search this blog


Tuesday, April 22, 2014

Tuscan-like farmer from late Neolithic Iberia

I found this Principal Component Analysis (PCA) in a thesis published recently by Uppsala University (see here). It features an ancient human sample from the archeological site of El Portalón, near the city of Burgos in northern Spain. The thesis refers to this individual as a late Neolithic farmer, which is probably another way of saying that his remains date back to the Chalcolithic or Copper Age. Indeed, I'm pretty sure that El Portalón is a Chalcolithic site.

Interestingly, he doesn't cluster with modern Iberians, like the Swedish Neolithic farmer Gok4, nor with modern Sardinians, like Oetzi the Iceman. He's actually closest to modern Tuscans from Italy. This might well be an artefact of the low resolution of the data (only 66,476,944 bp of DNA sequences), but if not, then it could be a signal of population movements to Iberia from somewhere in the east during the Copper Age.

Needless to say, it's a shame we don't know what this guy's Y-haplogroup was, because it's now generally believed that haplogroup R1b made its appearance in Western European during the Copper Age, and that it arrived there from somewhere in the east. By the way, his mtDNA belonged to haplogroup U5b1b, which is actually a marker typical of Western and Central European hunter-gatherers.


Daskalaki, E. 2014. Archaeological Genetics - Approaching Human History throughDNA Analysis. Digital Comprehensive Summaries of Uppsala Dissertations from theFaculty of Science and Technology 1101. 61 pp. Uppsala: Acta Universitatis Upsaliensis

See also...

The story of R1b: it's complicated


Maju said...

Interesting finding. Thanks for sharing, David.

... "it could be a signal of population movements to Iberia from somewhere in the east during the Copper Age".

What about flow from the Western margins? I mean modern Iberians can probably be described (largely at least) as a mix of La Braña and El Portalón, right? The Ebro basin is generally considered, judging on craniometry, as an area with large (gracile-)Mediterranean influence in Neolithic times, some of which is still present, no doubt. Instead for example most of the Basque Country is considered less or not influenced by this Mediterranean inflow - although admittedly we know better now thanks to genetics. If this phenotype evidence indicates actually various degrees of admixture with the Mediterranean Neolithic inflows, just a bit further North (or probably to the West too) we should have populations with greater WHG affinity, although surely not "pure" anymore, of course.

Then of course there should be at least some "Eastern" input among non-Basque iberians from the East but not really (or very minor) among Basques, judging on ANE affinity scores. Northern Iberians in general (as southern French too) are less ANE-akin than mainstream Iberians but these are also less ANE-akin than French or Italians. Naturally this PC graph does not evaluate ANE affinity directly but, judging on the PCA, El Portalón should be more ANE-like than modern Iberians (with caution, of course) because Tuscans are too (about the same as French and unlike Sardinians or Basques).

The issue looks interesting because we know that Motala (Scandinavian HGs) were already more ANE-like (i.e. more "Eastern European") than WHGs and therefore it's possible that some of that extra ANE affinity is pre-IE (certainly in NW Europe, with partly similar Paleolithic non-Magdalenian roots as Scandinavia).

Another possibility would be that the influence could be Northwestern (British-like) but it still seems more likely to me something in the line of ancient Portugal and/or Brittany instead.

Maju said...

PD- 4000 BP is certainly Chalcolithic in Iberia, almost Bronze Age, although there's no reason to doubt the farmer's Neolithic ancestry in general terms. No clear intrusions are known before the Bronze Age (first proto-Tartessian "horizons" in SW Iberia, later Urnfields in NE Iberia), unless you follow the "colonial hypothesis" for the rise of Iberian civilizations, which anyhow are very distant from this particular area.

About Time said...

Maybe Paleo-Basque territories were originally mainly North European, and the Pyrenees are just a southern refugium where they held on longer than say, Germany/France/Denmark/England.

Maju said...

@AT: We can't judge that without considering the archaeology (material prehistory) nor the exact relations between the samples. A PCA may suggest things but alone it cannot show anything too clear, not in the context of West Eurasia, where genetic differences are almost like splitting hairs (way too subtle).

And the archaeology does not suggest that at all, exactly as it does not suggest that Etruscans come from Iberia or that Sardinians come from Austria. In the case of Sardinians, they seem to have the kind of mix that is very similar to Alpine Neolithic samples (Lazaridi's EEFs).

However other Neolithic samples from further West are different. Gök is in the way we would expect (i.e. more Westernized than Lazaridi's EEFs) but El Portalón is actually "more Eastern" in fact, what is anomalous and raises some interesting questions.

Questions we won't probably be able to answer without some deeper analysis.

barakobama said...

This study is nothing compared to Laz, but it's better than nothing. I wish they included other Iberians besides Basque. Basque, Sardinians, Otzi, and Gokhem may be at the far left because they have little to no ANE ancestry. Maybe Portalon is shifted a little more to the right, so with Tuscans, because he has some ANE ancestry. I doubt he had a high amount of south-west asian(modern like) ancestry like Italians do, and possibly if Spanish and Portuguese were put in this PCA they would cluster with Portalon.

The positioning on this PCA mirrors Davidski's PCAs. Spanish fill a gap between Basque and Tuscans, so in a similar position as 4,000BP Spaniard Portalon.

It is defintley possibly that R1b(primarlly Df27) and the autosomal alleles brought with the source population had already arrived in Iberia by 4,000BP. That could be why Portalon does not cluster with Neolithic European farmers, and i think would cluster with modern Spanish and Portuguese.

barakobama said...

" Interestingly, he doesn't cluster with modern Basques, like the Swedish Neolithic farmer Gok4, nor with modern Sardinians, like Oetzi the Iceman. He's actually closest to modern Tuscans from Italy. This might be an artefact of the low resolution of the data (only 66,476,944 bp of DNA sequences), but if not, then it could be a signal of population movements to Iberia from somewhere in the east during the Copper Age.

I tend to agree. Probably nothing from Mesolithic and Neolithic west Europe could explain Portalon's shift towards the east(ANE). He seems similar to modern Iberians(Basque-), who are certainly not pure Cardiel+La Brana-1. If R1b(primarily Df27) had already reached Iberia by 4,000BP, i would expect Portalon to be more WHG and ANE like since the admixture between Neolithic descended Iberians and the mysterious R1b L11 source population(Insular Celtic aka north west European-like?) had just begun. French are very intermediate between north-west Europeans and Iberians, so maybe they became more and more Neolithic west European like as they moved through France.

Davidski said...

Iberians cluster between Basques and North Italians, so it's unlikely that this Portalon farmer was a good fit for a modern Iberian.

Unfortunately, it's difficult to figure out what his ancestry was precisely from that PCA, because the same result, more or less, can be produced in at least four ways. For example:

1) Almost 100% Tuscan-like

2) 70% Sardinian, 30% North Caucasian (Adygei-like but more northern)

3) 60% Basque, 40% Levantine

4) 50% Balkan, 50% Sardinian

So take your pick for now. In any case, I won't be one bit surprised if higher resolution tests of remains from the Portalon site uncover several sequences belonging to R1b.

truth said...

@ barakobama

I think you got it wrong, iberians wouldn't cluster with El Portalon, since they cluster between Basques and North Italians. Here is a same PCA but with more populations :

Heraus said...

So, we've got a farmer from the late Neolithic age in the Ebro valley and he happens to be similar to modern-day Tuscans.

The question as Polako hints to is indeed the following one : did people from an already stabilized Tuscany genetically wise migrate into the Ebro valley or is this farmer's Tuscan similarity the consequence of a similar melting-pot ?

My opinion is that ths farmer illustrates a non-stabilized melting pot of autochtonous Stuttgart-like Iberians (read : more westernized Sardinians) and a newest wave from the East which brought some well-identified components (the ones associated with ANE).

To get modern North Iberians with this Tuscan-like farmer, you have to imagine further admixture with Basque-like people, which is probably what happened.

I presume it's impossible to detect the links of this farmer community in El Portalon with more rural nearby districts which probably were more akin to modern-day Basque people.

Maju said...


How was Atapuerca (where El Portalón site is) not "rural", i.e. implicitly "urban" in any way? It was probably a quite typical Cardium-derived site, if anything cosmopolitan is in that spot is that Atapuerca pass acts as the most natural "gate" between the Ebro and Duero basins and therefore sits on a traditional route of Megalithic and Bell Beaker times, nearly identical in that section to the historical St. James Road. Are you thinking of this: i.e. a bit like the difference between Ebro basin gallery dolmens vs. mountain/coast simple dolmens without gallery?

The typology of the pottery is totally Epicardial and even a notch or needle is tentatively compared with typical Cardial sites from the Mediterranean (see here - in Spanish).

You may well be right about a "non-stabilized" genetic pool (and therefore hardly representative of the general population of the peninsula), however the person probably approximates well the more intrusive Cardium Pottery original element, because it's likely that most of the Ebro basin was heavily colonized by those new arrivals (earlier there was not much population density that we know of and anthropometry suggests that kind of Mediterranean colonization in all the Ebro). On the other hand other regions like Murcia, Andalucía, Portugal or the mountainous North was probably more mixed or aboriginal. The plateau is a bit of a mystery in this aspect, at least to me.

barakobama said...

Is Portalon male or female.

barakobama said...

Maju do you think Portalon was purely descended Cardiel farmers or had some extra ancestry from the east that came after the Cardiel culture?

barakobama said...

Portalon is in a similar position as are southern and western Iberians in this PCA.

I doubt he had a significant amount of modern south-west Asian ancestry, probably just more early European farmer ancestry than most modern Iberians have. His ANE ancestry like with modern southern and western Iberians causes him cluster with Tuscans. Modern Iberians are more WHG, more EEF, and less ANE than Portalon. I don't think it's very possible that just 4,000 years ago a very Tuscan-like population lived in Iberia.

Davidski said...

The only Spaniards who are close to Tuscans on West Eurasian PCA appear to have inflated North African admixture, which pulls them south. But this is due to historic events which obviously couldn't have influenced the Copper Age genome from El Portalon.

The interesting thing to note is that Oetzi the Iceman was living in the Italian Alps over 1,000 years before the El Portalon farmer's time, and he was Sardinian-like. What this means is that if the El Portalon farmer came from somewhere around Tuscany, then there had to have been a major genetic shift in northern Italy during the Copper Age. This is actually very likely.

So perhaps a useful question is, who migrated to northern Italy during the Copper Age that made the Sardinian-like ancient Tuscans into modern Tuscans? Based on a test I did just now I'd say it was a Balkan-like population, perhaps from the Balkans, but perhaps not.

Maju said...

"But this is due to historic events which obviously couldn't have influenced the Copper Age genome from El Portalon".

That's not correct: the "Basal Eurasian" (African-like tendency) is present in both Iberian hunter-gatherers (La Braña) and all the Early Neolithic Farmers of Lazaridis (EEF).

You may be on the right track only on the North African affinity, which is in fact typical of Western Iberia since probably Solutrean times (it shows up everywhere: La Braña, Portuguese ancient mtDNA, modern West Iberian genetics...) and also of EEFs (not quite the same source but similar "Basal effect" when compared with Eurasians).

"if the El Portalon farmer came from somewhere around Tuscany"...

No, he did not. The site is totally Iberian Epicardial (with "decoración de mamelones" and all that) and there's absolutely nothing in the archaeological record suggesting migrations from Italy to Iberia between the Early Neolithic and the defeat of Carthage.

In the Chalcolithic one can imagine ("colonial hypothesis", which has no real material support) maybe some tenuous influences from... Cyprus??? but not Italy. In the Bronze Age influences from Mycenaean Greece are quite more clear but they did not leave any obvious legacy in language nor genetics. But in any case Italy was not sending migrants around in all that period I say (between Neolithic and Romans), not to Iberia, not to anywhere else.

Italy was in the Chalcolithic a rather isolated region, receiving influences from East and West but certainly not emitting anything that we can track. Even in the Heraklean legend it is very apparent that the route from Greece to Iberia was then via North Africa, not Italy (only later the Romans added the apocryphal episode of Caco, precisely because otherwise Italy was not mentioned at all). This explains why West and East did not meet more often in late prehistory: the in-between area (Italy in essence) was a passive receptor but not any dynamizer (Malta excepted probably). If Italy would have been a more proactive region back then, our Prehistory would have been quite different but it only acted as trampoline, if at all.

"who migrated to northern Italy during the Copper Age that made the Sardinian-like ancient Tuscans into modern Tuscans?"

Nobody yet. Only Western ("Occitanian", possibly proto-Ligurian) influences are apparent in North Italy by that time. In the South and Center however there was some influence from the Aegean but not enough to trigger a Bronze Age yet.

The genetic change in Italy (overall) was probably a sum of these Aegean influences (which persisted later as Etruscan and Greek) plus the Indoeuropean ones. While Inodoeuropean conquest was surely important, we must not dismiss the Etruscan colonization of Padania either, because it relocated people with a more SE affinity over there.

barakobama said...

Your forgetting about Tuscan's proto-Italic speaking ancestors who descended from the bronze age Urnfield culture, and were northwest European-like genetically. R1b S28 peaks around Tuscany, evidence they are important ancestors of Tuscans.

There is recent southwest Asian ancestry in Italy and it gradually becomes stronger as you move south. Maybe it came in the copper age.

I think Portalon has similar WHG/NE/ANE percentages as Tuscans, that doesn't mean he has recent common ancestry with Tuscans. Hopefully you or someone else will be able to get Portalon's raw data put him into admixtures. They will show if he has recent southwest Asian ancestry like Tuscans or was just very Stuttgart-like plus some ANE.

Not all north African ancestry in Iberia is historical, mtDNA L samples from Neolithic Iberia prove it. Also I have heard that there is evidence in mtDNA and Y DNA that alot of the north African ancestry in Iberia is very ancient.

barakobama said...

La Brana-1 had a small amount of farmer ancestry which could explain his estimated 17% basal Eurasian ancestry.

Anima Mortel said...

What does it matter where he came from to Spain; Modern Spanish people have nothing to do with that person either way in the same manner as with all the other Neolithic and Mesolithic Europeans.

And this person is from the early Bronze Age not Neolithic (after Bell Beaker culture). R1b might be possible because it was found before (in Bell Beaker site) in Germany.

Lazaridis first PCA chart was different form Skoglund and this new paper in that all Neolithic europeans clustered next to Sardinians and not with any other modern Europeans.

Davidski said...

No, not really.

If you mix Basques with Tuscans you'll basically get a Spanish-like result on a West Eurasian PCA. So it's very likely that the population that the El Portalon farmer derived from contributed in a major way to the modern Spanish gene pool.

Also, different PCA will often show different things, and that's because they're dependent on the samples that are used. But all of the PCA I've seen of these samples generally agree with each other, and with other types of analyses.

Gok4 is Sardinian-like, but clearly has more northern European admixture than Oetzi and Stuttgart, so it's no wonder she clusters closer to Basques and Spaniards on some plots.

Indeed, even Stuttgart and Oetzi don't always cluster exactly with Sardinians.

You need to interpret these results with some insight into how they're produced and also be aware of all of the factors that might affect them. Otherwise you won't know what you're looking at.

Anima Mortel said...

So your saying that modern Spanish are a mix of Basques and Tuscany?

I think you mentioned it in the text and Tuscany is not actually that close to this person from the Bronze Age (but of a neolithic backround) as Gok4 was not actually close to the Basques either.

Davidski said...

Yes, I think modern Spaniards are essentially a mixture of a Neolithic/Mesolithic Basque-like population, and a Tuscan-like group that arrived in Iberia during the Copper Age. That's what they look like in all my analyses.

But I don't know where this Tuscan-like group came from. It might have actually been from Tuscany or another part of Italy. Then again, it might have been from the same place that many of the ancestors of modern Tuscans came from, like maybe France, the Carpathian Basin and/or the Aegean region?

Also, Gok4 does appear Basque-like to me, but this affinity is probably indirect, due to similar ratios of Mediterranean and North European ancestry, plus a lack of ANE admixture.

Anima Mortel said...

But if 'Portalon farmer' is representative for this Tuscan like group than modern Tuscany did not change much according to that paper. Because according to it the 'Portalon farmer' is still closest to the modern Tuscany.

There has to be (according to this paper) a common ground between modern Tuscany and the 4,000 BP 'Portalon farmer'. But as you pointed out Historically that does not make much sense since there was no substantial movement from or to either way. The Cardial/Epicardial phase at best but than that was thousands of years earlier.

Maju said...

Not all PCAs are the same; even if you use similar populations different loading of each group can cause important variation. For example the typical Euro PCA approximating the European map is different from Lazaridi's PCA (fig S10-5), which looks like a T with the pivot in the Balcans and more ambiguously the Alps (and this happens also for Europe in the all West Eurasia graph). This is caused mostly because Lazaridis et al. used more Eastern European samples.

But regardless, Basque+Italian=Spaniard is an over-simplistic fallacy. It's true that in many PCAs Spaniards sit right in between Basques and Italians but also between Sardinians and French (or other West Europeans). You can imagine them as a mix of either combo or all them or none but the fact is that a mere PCA doesn't explain on its own the genesis of the various populations.

Judging on ancient DNA the modern Italian genetic pool is not directly involved in the genesis of Iberians (except at negligible levels). Mainline Iberians are very close to be considered a simple EEF-WHG mix, although they deviate enough for some extra continental input to be likely (IE?). After Sardinians, Basques, and near-Basque populations, they are the next Europeans closest to the EEF-WHG simple mix baseline (→ FineStructure PCA).

Is this extra-deviation towards Italy? It all depends on the baseline. And if El Portalón is the baseline, then the deviation is rather towards France and towards Basques.

In other words: if we factor modern day mainline Spaniards as ES=Portalón+X, then that X can only be France and the Basque Country, not Italy. If we factor ES=Stuttgart+X, then X=France (alone). Only if we factor ES=Basques+X, then the X becomes Italy, but it is precisely Portalón what tells us something about the actual Neolithic baseline, not Basques.

Similarly if we have to factor Basques as a mix of EH=Portalón+Y, then Y is WHG and if we do EH=Stuttgart+Y, then again Y is WHG (plus some minor deviation towards France).

So Basques are much more WHG than EEFs, including Portalón, and Spaniards are more French (and possibly also some more WHG) than EEFs (Portalón included). Q.E.D.

Davidski said...


Tuscany probably hasn't changed much genetically since the Copper Age, but it probably changed a lot during the Copper Age, because Oetzi the Iceman came from the Alps just to the northeast of Tuscany, and yet he's significantly more Mediterranean genetically than modern Tuscans.

See, the El Portalon farmer is Tuscan-like, but he's not from Tuscany. He's from Iberia and perhaps partly Iberian.

So what would be really interesting is to see who was actually in Tuscany at around this time or a bit earlier. Perhaps Balkan-like individuals, shifted genetically a few hundred kilometers to the east compared to modern Tuscans, just like the El Portalon farmer is relative to modern Iberians?

Maju said...

"Gok4 is Sardinian-like, but clearly has more northern European admixture than Oetzi and Stuttgart, so it's no wonder she clusters closer to Basques and Spaniards on some plots".

More "Northern European" or just more WHG? It may look the same thing but it is not actually the same thing, because North and Eastern European appearence of WHG likeness is not directly derived from WHGs but from their Eastern relatives (EHG). That is apparent in two facts: (1) the ANE element, which is absent in true WHGs but not among North and Eastern Europeans at all, and (2) the fact that French have much longer Lochsbour IBD segments than North and East Europeans, what implies a more direct relatedness (logical because Lochsbour lived in Luxembourg, not Ukraine).

But how does Gok4 score in these aspects? We do not know for sure but the fact that he tends to cluster with low ANE populations like Basques and Spaniards strongly suggests that his pseudo-Nordishness is actually low ANE and therefore an artifact of WHG (and not EHG) admixture.

That is very interesting because Gok 4 is the only sequenced Atlantic Megalithic individual I know of, and Megalithism must have played a major role in the Chalcolithic reorganization of the Western European genetic pool.

... "even Stuttgart and Oetzi don't always cluster exactly with Sardinians".

True, they can also be pereceived as more like modern Italians. If so, maybe there has not been such a radical refurbishing of Italian ancestry. However modern Italians have much more ANE tendency than EEFs. So if we factor IT=EEF+X (using fig. S19.2 of Lazaridis, as before), X is a coordinate between Russia (or Ukraine) and Turkey, but much more closer to Russia than to Turkey. Only for Sicilians and Maltese the position is truly intermediate between Ukraine and Turkey.

That probably confirms what I said before about two main sources of post-Neolithic ancestry in Italy: Aegean and Indoeuropean (some of which was also via Aegean).

Maju said...

"Tuscany probably hasn't changed much genetically since the Copper Age, but it probably changed a lot during the Copper Age"...

Copper Age yes (Aegean influences, also something from the West) but later periods too. The Etruscans only arrived in the Bronze Age and the Italics did the same (North), with expansion to the peninsula already in the Iron Age (contemporary to the main Celtic expansion in Iberia). The languages spoken in Italy at the beginning of history arrived all (except probably Ligurian and some possible but unknown pocket elsewhere) in the Bronze and Iron ages. I'm quite certain that if we had ancient DNA from Italy to compare, we'd get that kind of results: the most radical change was from the Bronze Age onwards.

"See, the El Portalon farmer is Tuscan-like, but he's not from Tuscany. He's from Iberia and perhaps partly Iberian".

He's Tuscan-like in this PCA, what is not rocket science, really. Whatever it means he was surely, like all known EEFs low in ANE, unlike modern Tuscans. That means, I understand and I believe that you do too, that modern Tuscans had extra inputs from the East and North (Indoeuropean inputs particularly). Those IE inputs are from the Bronze Age onwards, earlier IEs stayed North of the Alps and East of the Ionian and Adriatic Seas.

Davidski said...

Maju, why are you so sure El Portalon had lower ANE ancestry than modern Tuscans?

PC2 in that analysis seems to be heavily influenced by ANE admixture, which appears to be the rule for West Eurasian PCA. So I don't see why this guy wouldn't have a decent level of ANE?

And yes, Gok4 was probably significantly WHG admixed. I'd say her ancestors picked that up in Western Europe somewhere, like France or Germany.

Maju said...

"why are you so sure El Portalon had lower ANE ancestry than modern Tuscans?"

We don't yet know empirically but being clearly part of the wider Cardium (and maybe Atlantic-Megalithic) Neolithic it should. There was yet no flow from Eastern Europe that could act as vector for increased ANE.

Even if you consider Bell Beaker as such vector (what I do not), there's no evidence of BB in El Portalón (a single pottery fragment) until some time after the layer of this burial.

Also BB or any other continental influence should not drag samples to Italy but to France, Germany, etc. To me it looks rather a variant EEF, which just does not cluster so strongly with Sardinians or modern Iberians (although these are not plotted, as pointed above, and they should be closer to Italy/France than to Basques).

Maju said...

Consider this please: barring Ötzi, whose ascription to the ancient Italian ethnography is ambiguous, there has not been a single Mediterranean Neolithic autosomal sequence before this one of El Portalón. What happens if Lazaridis' EEF dataset actually only properly represents Danubian Neolithic genetics (or something like that) and Portalón represents more or less Mediterranean (or at least Iberian) Neolithic genetics?

The whole analysis may need recalibration.

Maybe is not as clear-cut as I put it above for simplicity but there was some serious Neolithic complexity and Lazaridis' EEF set only represents well the Alpine area, from where all the samples are from. North Italy was neolithized mostly by a secondary wave from the area of Bosnia-Croatia which arrived by land, not by sea, and although they are considered within the Impressed-Cardium cultural area, they were a different subgroup, more "Bosnio-Dalmatian" and not at all "Epiro-Albanian". So even if Ötzi is properly classified as ancient North Italian, he may well be from a particular distinctive group not too closely related to the main Cardium one (although incidentally yes to the Sardinian high I2a founder effect).

I'm really looking more and more at Portalón as an alternative Mediterranean Neolithic reference and possibly the most genuine one. If that would be the case, then maybe many of the conclusions from Lazaridis et al. need to be revised (they should be alright grosso modo but probably not too exact in the fine detail).

Of course it's hypothetically possible that this alternate EEF population would have greater ANE score. But why would they? It would be interesting to know more details like this one indeed.

Shaikorth said...

I'm not sure if PC2 is an ANE-representation. It seems that they use HGDP and 1kGenomes populations on the PCA, meaning the "Levantines" are Palestinians (with marginal ANE) and Druze (with about 11% ANE according to Lazaridis paper), yet they are as far from Sardinians as the northern Russians and more so than Orcadians. The ones more distant from hunter-gatherers (Palestinians?) are also more distant from Sardinians even though they should be the least ANE population on the plot.

While PC1 seems to have a WHG connection, with Levantines on one end and actual WHG's (or Scando-HG's) on the other, PC2 might just represent relatively recent Sardinian or Southwest European specific drift.

Maju said...

You're right Shaikorth: it's a generic SW-East axis where Levantines (low ANE) score in the same pole as East Europeans (high ANE). If anything it represents the SW European genetic distinctiveness, whichever it is (Sardinianness?, Basqueness? EEF-ness?).

Heraus said...

What does archeology say respecting that late Neolithic/Copper Age wave into Iberia ?

By sea directly from somewhere in the Aegean Sea ? Through France and the Eastern Pyrenees ?

Davidski said...


It seems there were two different cultural groups in northern Spain during the Copper Age, and a lot of people were dying from projectile wounds.

So the question is, were they both local, and just different from each other in terms of culture, or did one or both sail in from somewhere far away?

Maju said...

"What does archeology say respecting that late Neolithic/Copper Age wave into Iberia?"

What wave? Nothing other than some innovation in elite burials (tholoi and artificial caves), first fortified towns, etc. There are some instances of evidence of trade with the Levant (some glass beads and ivory, which was recently shown to be from that part of the world, not Africa). That's about it. It's a similar evolution to Southern France, which was largely integrated into the Iberian cultural and exchange networks.

Many decades ago some of these innovations were automatically believed to come from the Eastern Mediterranean, where tholoi for example had been used long before as common buildings (not for burial). But there is a lapse of 1000 years between the quite vertical Levantine tholoi of for example Khirokitia and the rather flat Iberian ones used for burial exclusively. A Greek origin is also impossible because these are more recent (probably an Iberian concept loan in the context of Bronze Age long distance interactions). Today the consensus strongly leans towards a local evolution, much as the Mexican pyramids have nothing to do with the Egyptian ones: just same civilizational stage reached, so let's build monuments for the glory of whatever (early elites' propaganda).

Maju said...

"It seems there were two different cultural groups in northern Spain during the Copper Age, and a lot of people were dying from projectile wounds".

Not sure where you get that there were specifically two cultural groups. Maybe this sentence:

"there are essentially two contemporary types of burial found in the Upper Ebro Valley: caves or rock shelters and megalithic graves".

This variation is common through much of Iberia. In some areas (center and east) predominates burial in caves, while in others dolmens are dominant but seldom a strict rule. Both are almost invariably "collective" (clan or family tombs). No other cultural differences are apparent, except ill-defined and fluctuating regional variations of pottery style, all of which seem loosely Epicardial until BB shows up. In Biscay for example you can find both types of burials, although dolmens (without gallery most of them) are more common.

As for "lots of people" dying from war:

"Only 12 of the 338 individuals have injuries by arrow wounds".

There was some increase in violence, possibly a milder reflex of what happened further north (France, Germany, etc.), where violence also seems to increase in the Chalcolithic but largely without any apparent cultural change related. Probably much of that violence is just clannish internecine violence and stuff like that, although it's also possible that there were full fledged wars among local groups, for which you don't need marked cultural differences (nation-states did not exist yet, inter-tribal and inter-polity wars are known everywhere). We need more than just an apparent (and highly arguable) increase in violence - 3% weapon injuries, less than half causing death is not that much in a society in which everybody carried weapons and there was no police, some may be even hunt accidents ("a boar, damn what was doing Pete here today?!, why doesn't someone invents orange jackets already?!"), we need a clear cultural change to be able to talk of new waves, demic change and all that. And we have nothing of relevance yet.

There is some cultural change in Chalcolithic Iberia but it seems all locally rooted and not too abrupt. The first civilizations worth that name (fortified towns) arise in the SE and SW, the already mentioned new burial monuments show up (also in the civilized south, and some parts of SE France), by the end of the age Bell Beaker shows up but mostly as randomly located fashion items (pottery, conic buttons) and only very rarely as true BB burials like those found elsewhere, while the rest doesn't really changes or just very gradually.

So there was an increase in conectedness (long distance "trade" and fashion flow) and civilization but nothing that indicates any major cultural change, much less of intrusive origin.

The first really notable intrusive phenomenon happened in the Bronze Age in SW Iberia and was probably at the origin of the Tartessian language (same area more or less) and may be related to possible rivalry between "pro-Greek" El Argar (SE) and the independent power of Zambujal/VNSP (Portugal) for the control of tin routes from the NW. The second one were the Urnfields, in the late Bronze Age, which conquered the NE (roughly Catalonia). These would really expand up the Ebro River sparking much more serious evidence of conflict at the beginning of the Iron Age. With the arrival of Hallstatt influences, they expanded to the central plateau and the Western coast but, after the founding of Marseilles, Catalonia was rapidly absorbed into the Iberian (post-Argar) cultural area and that way early Iberian Celts and other IEs (Lusitanians) were detached from their continental relatives, never adopting druidism nor anything La Tène.

Those are the real very apparent intrusions into Iberia: from much later periods only.

Maju said...

PS- Said that, there may well have been some general "ethnic" differences between the south-west-north (mostly) Megalithic area and the central-east (mostly) non-megalithic one, which may have been closer to the original Cardium wave and less influenced by aboriginal Epipaleolithic blood and ideas, much like there was some of that in other parts of Europe (change Cardium for Danubian further North).

However it's very difficult to imagine some sort of "clash of civilizations" between them, mostly because the central-east area lacked any civilizational centers that could lead their ethnicity and also because almost certainly the Megalithic area was also of Cardium influence one way or another.

I mean: there was almost no Megalithism in Murcia, never mind Valencia, but Iberian language and Basque language seem quite related anyhow.

Daniel Szelkey said...

I think he was between from 30-70% hunger gatherer, and he probably came was a population that was dominated by U5b. He had U5b1b and I would be suprised if he was a singleton. If he looks like he is 0% hunter gatherer on any analysis something is off.

Daniel Szelkey said...

I am prettyy sure he had R1b if you go to he is P3.

barakobama said...


Shaikorth said...

Most unfortunately page 26 in that new Skoglund paper's supplementary data says they went and projected all the ancient genomes on their PCA again, so it (posted on Dienekes' blog) is less informative than it could be.

Seinundzeit said...

The tree graph is very illustrative. The ancient European farmers seem to be strongly shifted towards Africa, and the ancient European hunter gatherers seem to be strongly shifted towards East Eurasia. I guess the evidence is now unequivocal, ancient European farmers simply weren't people who came straight from the Near East (by this, I mean they weren't unadmixed Near Easterners). Rather, they seem to have been part of complex, "synthetic" populations, the end product of admixture between Pan-Eurasian like European hunter gatherers (by this, I mean that these hunter gatherers, and MA1, seem to show strong affinities to rather diverse portions of contemporary Eurasia, not just West Eurasia. In fact, I think the concept of "West Eurasia" is simply inapplicable to these ancient populations. These hunter gatherers were just much too close to East Eurasian populations in comparison to living West Eurasians, and just much too distant from Sub-Saharan Africans in comparison to living West Eurasians), and actual Near Eastern populations which were deeply diverged from the rest of Eurasia.

I'm sure all of this was already fairly obvious to everyone, but it's nice to see that Skoglund et al. verifies and substantiates Lazaridis et al., and that too with the addition of some new data.

Seinundzeit said...

I should have mentioned this the first time. The HGDP Pashtuns share the most drift with Avjide58 out of all South Central Asians, and share more drift with this ancient individual than Iranians do. In fact, the Pashtuns aren't too far behind Northern Caucasians. This is despite the fact that they probably don't share any ancestry with this ancient individual. I guess Avjide58's 15% ANE admixture must play a role in this, and heightens the affinity, since the HGDP Pashtuns are surely very rich in ANE admixture.

But there is one odd detail. Namely, the shared drift between the HGDP Pashtuns, and Gökhem2. In this case, Pashtuns are the closest population in all of South Central Asia to this individual, as close as Iranians. They are even closer than populations from Balochistan, which all have very strong Southwest Asian admixture. I'm not really sure what this could possibly imply.

Heraus said...

Thank you Maju for your summary.

Still, there is something I don't get : who is that Tuscan-like farmer ?

That Tuscan-like farmer differs from aboriginal Basque-like Iberians of the time. The fact he clusters with modern-day Tuscans on autosomal maps must mean his "West Asian" ancestry is higher (I suppose so though an Admix run would be ideal to precisely know).

In a word, where did such influence come from ? Step by step ?

I don't doubt that the fusion of those people with aboriginal Basque-like Iberians will then create modern Iberians (with maybe subsequent migrations as well).

Maju said...

You suppose it "must" mean that. But where's the evidence? Tuscans are, just like all Europeans (Sicilians excepted), modeled as simple function x=EEF+WHG+ANE. Only Sicilians and Maltese need extra West Asianness to be explained.

As explained by David above, different PCAs will show the EEFs aligning more towards Sardinia or more towards Italy, and there is small but meaningful variation among them in any case, what may be due to greater or smaller "Basal Eurasian" ancestry, for example.

This ancient Iberian must be just another EEF with maybe less "Basal Eurasianness" in the ancestry or whatever other variation. Please don't read too much on a single graph and await for more data.

Anyhow, I would expect populations more like Görkheim (more WHG, less EEF) in the Megalithic zone of not just Iberia but all the European Atlantic facade. The expansion of these populations in the context of Megalithism must have set the basic layer for modern Western (and even to some extent Central) Europe, although IE expansions carrying East European genetics (initially EHG=WHG+ANE) have also significantly altered this too later on.

Heraus said...

Of course, I'm only making assumptions according to what I understand. I eagerly wait for more data.

I'm opened to the idea of all Europeans being explained through the function x=EEF+WHG+ANE but I have then a hard time matching that equation with some autosomal results, albeit imprecise (to sum up : what "West Asian" is about in Admixture runs, the ones western "Neolithic" people such as Sardinians and Basque people lack).

I fail to explain most Italian results without an extra migration from the East which brought an extra boost in some components (which were also brought by Indo-European migrations alongside more northern components through ANE ancestry).

You're probably right that this ancient Iberian clustering with Tuscans might be a simple artefact and that it "must be just another EEF with maybe less "Basal Eurasianness"". I actually don't know and I'm a bit clueless about that.

Still, in any case, that ancient Iberian is not Görkheim-like. He might be Sardinian-like, in that case, everything will be easily explained : just a less WHG-influenced farmer. But as far as I understand things, he's Tuscan-like, a modern-day population whose autosomal results differ from modern-day Sardinians because of subsequent migrations affecting Italy.

That's what I'd like to understand : who is that ancient Iberian from the Copper Age ? Is he representative of a population which would make Basque-like Iberian modern-day Spanish ? Who were those people ?

Maju said...

"That's what I'd like to understand : who is that ancient Iberian from the Copper Age ? Is he representative of a population which would make Basque-like Iberian modern-day Spanish ? Who were those people ?"

I'd say he's representative of the older Cardium layer and that this layer was somehow less important that the more Atlantic-like peoples such as the "hyper-Basques" of Neolithic Portugal. But possibly it's a matter of mere demographic weight: other than the (southern) Valencian Country, which does not show archaeological signs of replacement and must have got a serious Cardium colonization, the rest of the Cardium area was either tenuously inhabited (plateau) or suffered a lot with Urnfield invasions (Catalonia, Ebro river) and later Basque-Celtic wars (remember La Hoya), which probably were not less violent Ibero-Celtic wars, according to what Roman chronicles say about Celtiberian origins. Similarly Portugal was also ravaged by the Celtic invasions leaving a country that Romans marveled how it could be so underdeveloped with such great natural resources (of course, it was much more advanced before the Celtic invasion).

So the whole demographic landscape (except Eastern Andalusia, Murcia and Valencia, as well as the Northern mountain strip) probably changed a lot in the Bronze and especially the Iron age. Not nearly as much as Chalcolithic Germany but enough. Most of the forces involved were in any case local. So on one side we see some "orientalization" of Western Iberia and on the other a "westernization" (or rather south-easternization, i.e. Iberization) of NE Iberia.

What is clear to me is that this El Portalón sequence should be considered as representative of central-east Iberia (Cardium, Italy-tending EEF) and not of the Megalithic belt (Basque/Spain-tending Atlantic farmers like Görkheim).

Davidski said...


The West Asian component is just a signal of EEF mixing with ANE. Without that ANE admixture, we'd just have the Southwest Asian and Mediterranean components, which are closely related.

Moreover, the West Asian component peaks in the Caucasus because of recent endogamy and genetic drift among the various ethnic groups there, not because it expanded from the Caucasus. If, say, Balkan Europeans were more heavily drifted, then it would peak in the Balkans.

The reason it's missing among Basques and Sardinians is because both of these groups can be fitted as 0% ANE, and it's very likely that their non-Indo-European origins are the reason for this.

By the way, similar things can be said about many other components that appear in ADMIXTURE runs, like the Kalash one. It's just a heavily drifted mix of ASI, ANE and EEF, and that's why it spills outside the Hindu Kush, even as far as Western Europe.