search this blog

Wednesday, May 10, 2017

Ancient population shifts in western Iberia (Martiniano et al. 2017 preprint)


Update 27/07/2017: Yamnaya-related migrations into Iberia: infiltration rather than invasion.

...


Over at bioRxiv at this LINK:

Abstract: We analyse new genomic data (0.05-2.95x) from 14 ancient individuals from Portugal distributed from the Middle Neolithic (4200-3500 BC) to the Middle Bronze Age (1740-1430 BC) and impute genomewide diploid genotypes in these together with published ancient Eurasians. While discontinuity is evident in the transition to agriculture across the region, sensitive haplotype-based analyses suggest a significant degree of local hunter-gatherer contribution to later Iberian Neolithic populations. A more subtle genetic influx is also apparent in the Bronze Age, detectable from analyses including haplotype sharing with both ancient and modern genomes, D-statistics and Y-chromosome lineages. However, the limited nature of this introgression contrasts with the major Steppe migration turnovers within third Millennium northern Europe and echoes the survival of non-Indo-European language in Iberia. Changes in genomic estimates of individual height across Europe are also associated with these major cultural transitions, and ancestral components continue to correlate with modern differences in stature.

Martiniano et al., The Population Genomics Of Archaeological Transition In West Iberia, bioRxiv, Posted May 10, 2017, doi: https://doi.org/10.1101/134254

See also...

New resource: 67 diploid ancient genomes

46 comments:

André de Vasconcelos said...

Something struck me when I read it:

"A recurring feature of ADMIXTURE analyses of ancient northern Europeans is the
appearance and subsequent dissemination within the Bronze Age of a component (teal)
that is earliest identified in our dataset in HGs from the Caucasus (CHG). Unlike
contemporaries elsewhere (but similarly to earlier Hungarian BA), Portuguese BA
individuals show no signal of this component, although a slight but discernible increase in European HG ancestry (red component) is apparent. D-Statistic tests would suggest this increase is associated not with Western HG ancestry, but instead reveal significant introgression from several steppe populations into the Portuguese BA relative to the preceding LNCA (S4 Text, S6 Table)."

What's 'steppe ancestry' then, if it has no CGH?


Anyway, R1b1a2a1a2 (P312) is present on the three MBA male samples, so thereabouts should be when R1b M269 and steppe ancestry got into Iberia - or at least to its westernmost reaches - but as far as we know, besides Celtic/Lusitanian, the only languages spoke in Iberia were non-IE (I'm including Tartessian here). Does this mean that R1b carriers arrived in such a low frequency that they were not numerous enough to impose their own language, despite having successful lineages? Today R1b is higher in the East than in the West aswell, where non-IE was spoken 2000 ago

bellbeakerblogger said...

The MBA samples illuminate what's really in the LC/EBA in Iberia, not detected in megaliths and caves.

AWood said...

Admittedly I don't understand all the paper's analysis and at times it appears a little contradictory, (ie: MBA Portugese had no CHG, but had "steppe" ancestry.) but it seems clear the MBA had an introduction of non-Megalithic people who were R1b-P312, and had different burial customs to the previous I2 rich megalithic folks. The paper hints at them arriving from central Europe, but I am wondering if there are other alternatives, perhaps from the eastern Mediterranean considering they allegedly "lacked" CHG.

rozenfag said...

The questions may sound stupid, but: Why the hell they didn't included EHG samples into comparison?

Anthro Survey said...

@rozenfag

They did. Third column from the left. Top of it.

P Piranha said...

They had no 'teal' ancestry as inferred from ADMIXTURE; this does not mean they had no CHG ancestry. You can get runs where Corded Ware has no 'teal' ancestry, e.g. in Mathieson et al at some K.

The fact that they specifically mention BA input into these samples as the best formal stat comparison show that real steppic contribution is there.

Nirjhar007 said...

OM, comment man ..

Simon_W said...

I suppose even Maju will now have to accept that Basque R1b isn't from the Atlantic/Megalithic Neolithic. The sampling is really satisfactory now.

Matt said...

No proportions here. Measuring on screenshots, Finestructure PCA position is 33% roughly between Atlantic Neolithic and Central NW/BA-AS implying Portuguese Bronze Age 16% Steppe, or 19% between Atlantic Neolithic and Yamnaya implying 19% Steppe.

Imputed TVD distance to modern populations - http://i.imgur.com/tnTsTho.png.
Some oddities and interesting factors; e.g. how much of Irish Bronze Age donation to Germany is real vs an artefact of this graphic normalising the distance against Irish Neolithic? Poland / Ukraine has lower excess of TVD similarity to Yamnaya compared to Scotland, relative to TVD similarity to LBK_EN. PT Neolithic vs PT Bronze are very weak, even for Basques. Hard to interpret.

If anyone involved in the paper is reading this, in the supplement to the final print, would be great to just have the table of raw TVD distances from the ancients, as in Cassidy 2016 (e.g. does Poland / Ukraine have lower TVD distance imputed to Yamnaya than Scotland, or is this just lower haplotype distance from LBK_EN?).

Like others, pretty curious as to why samples which seem likely to have steppe ancestry (Hungarian and Portugal BA) don't show that ancestry via CHG in ADMIXTURE when it is much under 50%... This does not happen in recent people. E.g. Southeast and Southwest Europeans probably have less steppe ancestry than Hungary_BA, but they show CHG fraction in ADMIXTURE.

Matt said...

Quick comments on their genetic height analysis:

- General level of EuroHG ancestry correlated with height, and all EuroHG samples fairly tall compared to other samples. Not really so clear Yamnaya ancestry correlated with height.

- Iberian_Chal samples taller than early Neolithic, in contradiction to Mathieson's analysis.

- However, height not just correlated with Yamnaya vs MN fraction; tallest grouping of samples are Iron Age Britons and Anglo Saxons. We know males from these populations were about 168-172 cm tall (5'6" to 5'7") in actual realized male height, whatever the genetic height - https://www.economics.ox.ac.uk/materials/papers/14997/151-final.pdf).

- Bronze Age Central European samples (75%-50% steppe) overlap with Hungarian Bronze Age samples. Portuguese BA overlap with Iberia_Chalcolithic.

- Hard to get a sense of predicted magnitudes in height differentiation without modern panel for comparison.

I've taken a screenshot and added annotations to help identify populations/samples : http://i.imgur.com/GOPVVKh.png

Josep Coderch said...

"This does not happen in recent people. E.g. Southeast and Southwest Europeans probably have less steppe ancestry than Hungary_BA, but they show CHG fraction in ADMIXTURE."

Modern people have post-BA CHG admixture brought by yDNA J and E folks, specially in southern Europe.

Anthro Survey said...

@Andre:
Regarding the possible scenarios I offered in my last comment to you in the other thread---we'll just have to wait for the supplement to come out and see if they ran any relevant d-stats which would provide clues as to any NA or Levantine introgression.

If they did not, maybe Davidski could run a few?

P Piranha said...

David, could you simply ban the posters who contribute little to the discussion? They sucking up all the bandwidth away from stuff that should interest us more. E.g. the current BA genomes are less steppic than even Basques, meaning that continual gene flow from Central Europe must have occurred post-Bronze Age.

War Lord said...

"Anyway, R1b1a2a1a2 (P312) is present on the three MBA male samples, so thereabouts should be when R1b M269 and steppe ancestry got into Iberia"

...and reached 87% frequency in Basques via a miracle, despite that their proportion of the 'steppe' ancestry is minimal?

War Lord said...

Where did all the Magdalenian hunter-gatherers (El Miron et al.) disappear, by the way? We cannot find them past the Mesolithic somehow. Even the 'West European hunter-gatherers' are now regarded as migrants from the Epigravettian refugium in the Balkans(see Mathieson's paper).

P Piranha said...

David, just mentioning this so we can KIV: once the Balkan HG data are out, we should investigate BR1 again, who was quite EHG and WHG heavy but had no or little CHG. Perhaps this guy did not require steppic input, and just required admixture with local HGs, like SE EEFs such as Varna and Trypilia. Then again he may require input from Ukraine_Neolithic or Ukraine_Eneolithic, which will be equally intriguing.

If the Iberians do not contain CHG, and this is a big if, a connection between BR1 and Iberians would be fruitful to check as well.

War Lord said...

"all three Iberian BA males are R1b, the haplogroup that has been strongly associated with Steppe-related migrations"

Where? In Samara???

P Piranha said...

Thinking about post-Bronze Age processes: it appears that continued influx of steppic ancestry must have occurred for both non-IE (basque) and IE groups. The Iron age movements and post-Roman movements, e.g. Unetice, might both have contributed. Certainly NW African ancestry must be quite new, as it is ~1% in the Neolithic and Chalcolithic and there is little reason to expect it to increase during the Bronze Age. Increased CHG might be recent too.

We know for a fact that the Balkans saw massive turnovers throughout and after the Bronze Age as well, such that they moved from a French-like position to a more Northern one and then massive introgession from Southern populations dragged them into the triangle pointing at the middle east that they currently occupy.

This makes me doubt the hypothesis that the genetic composition of N Europe was static all through this time. How would you know, when all the migrating groups have the same ancestry? Perhaps Eurogenes K13, MDLP K36 or other analyses with high dimensionality, or haplotype or rarecoal based analyses, will show continuing turnover in N Europe, which formal stats on unlinked markers will obscure.

The other thing to think about is what social process produced an elite which is very, very low in steppic ancestry, but retain foreign Y HGs. Perhaps they just derive from a different population via a different route, e.g. from the Balkans from BR1 ('stelae' people anyone?). Or perhaps there is a legacy of social stratification here in Iberia from the preceding Chalcolithic citadels, which there wasn't in Northeast and Northwest Europe, such that NE and NW Europe had massive folk migration with relatively egalitarian social structure post-migration--even more egalitarian than the preceding Neolithics--but Iberia already had deeply differentiated and structured societies with roles that the incoming people could capitalise on. This is the case with the Amazonian tribes, for example--they have a hereditary chieftainship and an extremely formalised ritual life and artistic/visual language, and codes of etiquette, and grid arrangement of settlements, all remnants from the time when they populated huge city complexes in the middle of the Amazon down to the 15th century, but of which they don't even retain a folk memory.

Matt said...

P Piranha: We know for a fact that the Balkans saw massive turnovers throughout and after the Bronze Age as well, such that they moved from a French-like position to a more Northern one and then massive introgession from Southern populations dragged them into the triangle pointing at the middle east that they currently occupy.

Eh? I don't by any means say you will turn out to be wrong, but I just don't see any data in the paper that compares recent Balkan to ancient populations in a way that allows us to say that is true. If there is, I would like to read it, so if you could point it out...?

Only thing where there is modern to ancient comparison I can see is their Extended Data Figure 1, which is badly hit by projection when comparing ancient and modern to each other. See here, with labeling for modern clusters - http://i.imgur.com/zJb9Y8J.png(e.g. Balts are not that close to Ukraine Mesolithic and Sardinians not that close to Iberia-Chal), etc.. (Will be very cool to see Davidski use his DoHA method on these samples).

(There are some proportions in the qpAdm analysis. Balkans_Bronze_Age are at WHG:AN:Yamnaya of 11:59:30. Comparing Portugal_Bronze_Age would be tricky with no explicit qpAdm model, but I'll assume 19% Yamnaya, and an Iberia_MN ancestry matching proportions for the rest, so WHG:AN:Yamnaya 21:60:19.

Assuming Yamnaya 66% EHG, then WHG:EHG:AN:CHG goes PBA - 21:12:60:7, and BBA - 11:20:59:10. That's just back of envelope but overall projection of HG ancestry would be roughly equal in PBA as BBA in that scenario.).

Ariel said...

@ Piranha

"We know for a fact that the Balkans saw massive turnovers throughout and after the Bronze Age as well, such that they moved from a French-like position to a more Northern one and then massive introgession from Southern populations dragged them into the triangle pointing at the middle east that they currently occupy."

I don't think so, bronze age Balkans are a more neolithic version of today's Balkans, in fact bronze age Balkans don't cluster with France but with northern Italy. Look a the modern reference for that PCA.

https://www.researchgate.net/profile/Jean-Michel_Guinet/publication/259441354/figure/fig3/AS:271610733133833@1441768453301/Figure-2-Principal-Component-Analysis-PCA-on-all-present-day-west-Eurasians-with.png

Ariel said...

Matt

I actually think they made a decent PCA

Matt said...

@Ariel, I think it's likely at least OK for showing differences of ancients relative to each other (Fig 1B), as they're all projected, but I don't think you could take the position of non-projected moderns relative to projected ancients literally. E.g. NW Europe is not likely to have an excess of HG ancestry compared to CE Bell Beakers, when the qpAdm and ADMIXTURE say it's either the same or less. Extended Figure 1 shows some cool things (e.g. look at the position of Tepecik-Citflik).

An example of a PCA that looks better, check out the one in Olalde's Bell Beaker paper (modern samples more overlap the ancients they resemble and don't the ones they don't). No idea why two collaborator papers from the same lab do it differently but there you go.

I think Piranha might be right in the sense of some offset in the Balkans from a) CHG brought in by Anatolia Bronze Age related, followed by b) Slavic expansion brings in a more WHG / Narva rich offset to that. But not modern Balkans being straight up more "southern" than ancient Bronze Age Balkans though. That remains to be seen from more specific stats and PCA, I think.

Still, even if this plays some role I generally think the main difference between the post-Neolithic SE Europe and post-Neolithic SW Europe stands - SE compared to SW is more Steppe heavy, but (on average) got less HG admixture prior to Steppe expanision; Balkans more like a straight up Yamnaya+AN mixture (but not "more HG" overall). That difference looks to withstand changes from other populations mixing into Balkans populations. That's how it looks to me anyway atm; maybe it will look more discontinuous once all is said and done.

Rob said...

@ Matt
What did you think of their sex-biased analysis wrt SEE ?
At a zoomed out look, and without raw datantwt available, it doesn't seem many steppe Y haplogroups made it in. There's a Z2105 in Vucedol, and a curious Z93 later on in Bulgaria

Ariel said...

Tepecik-Citflik shouldn't be there, that's for sure. Maybe it had some Iran Neolithic ancestry, but yes there are problem with the PCA. But after all is similar to the Davidisky's one (http://eurogenes.blogspot.it/search?updated-max=2017-04-29T00:02:00-07:00&max-results=10). Some Hungary BA samples are similar in position to the Bronze Balkans that we have in this paper. BB are more or less in the same position. Also the extra WHG effect is there, at least a little bit, even in the Davidisky's one (still a mistery). I don't think that they are that different. If they din't account for projection bias it would be so much worst, like we saw in many other papers.

Richard Holtman said...
This comment has been removed by the author.
MaxT said...

@RichardHoltman

We already know what true "AR1an" haplogroup is lol

Grey said...

André de Vasconcelos said...

"What's 'steppe ancestry' then, if it has no CGH?"

1) very early steppe ancestry before mixing with CHG?
2) steppe ancestry from a region that didn't mix with CHG?

#

"Does this mean that R1b carriers arrived in such a low frequency that they were not numerous enough to impose their own language, despite having successful lineages?"

To me this implies a minority group integrating into a pre-existing culture (atlantic megalith) but bringing with them something that allowed them to gain power later.

My guess is metalworkers who settled along the Atlantic fringe and as a result got early access to bronze swords.

Grey said...

André de Vasconcelos said...
"Anyway, R1b1a2a1a2 (P312) is present on the three MBA male samples, so thereabouts should be when R1b M269 and steppe ancestry got into Iberia"

War Lord said...
"...and reached 87% frequency in Basques via a miracle, despite that their proportion of the 'steppe' ancestry is minimal?"

If they were miners/artisans who arrived early but in very small numbers then only finding lots of samples of their presence in the MBA may simply be a function of their numbers increasing enough to be noticeable by that time.

The ancestral proportions would fit them being mostly male as you might expect if they were miners - with local HG wives maybe.

Also a while back someone on here posted a map of Iberian mining over time which iirc showed mines in the Basque region being more copper than gold and silver - so maybe the best spots were already taken so the new arrivals were concentrated in the Basque region.

Grey said...

"so maybe the best spots were already taken so the new arrivals were concentrated in the Basque region."

assuming here the original atlantic megalith culture people were G2 and I2

mooreisbetter said...

Simon W, are you kidding me? The authors said the opposite of what you did:

"However, the limited nature of this introgression contrasts with the major Steppe migration turnovers within third Millennium northern Europe and echoes the survival of non-Indo-European language in Iberia. "

Matt said...

@Ariel: But after all is similar to the Davidisky's one Very similar, but look at the distances between EHG and CHG from moderns in Davidski's PCA vs the projected PCA. As well, the Yamnaya-Bronze Age cline cuts right through the moderns in Davidski's PCA, for NW Europe, and Balto-Slavic populations have a signature of additional Kunda/Narva related admixture. Some samples even in NW Europe do have more than the cline implied for Yamnaya-Bronze Age, but we do have very large sample size for moderns, so more outliers are expected.

I'm not saying the PCA with ancient+modern in Mathieson 2017's supplement is awful, but Davidski's or the one from Olalade would give a better picture.

P Piranha said...

@ Matt

Matt, if you look at their ADMIXTURE estimates, early Bronze Age Balkans have only ~9% Steppe ancestry on average. A more detailed breakdown can be found in this post:

Link

30% Steppe ancestry is for the Bronze Age as a whole, and this figure is arrived at only when three very steppe-rich samples from more than twelve hundred years later is included in the calculation.

In ADMIXTURE analyses with Yamnaya components, all modern Balkans populations have generally ~20% Yamnaya admixture. In qpAdm they generally have 20-25%.

I know that the PCA suffers from massive projection bias, which is why we can only deduce their position from where they score relative to other ancients. The Bronze Age BA overlap with Vatya, Maros, Varna and with Hungarian BA and CA with lower steppic proportions in the Mathieson PCA.

These populations cluster with Spanish, French and Basque in David's PCAs, which are non-projected. Therefore its reasonable to deduce that EBA Balkans will behave the same way.

So its quite clear to me at least that large turnovers continued after the initial Yamnaya intrusion, including further increase in steppe and CHG ancestry.

Josep Coderch said...

These 3 bronze age individuals with steppe ancestry from Portugal date from around 1800 BCE but in the Bell Beaker paper 2 of the samples from Brugos (~2400 BCE) also show steppe ancestry even though the rest doesn't.

I wonder if this shows the first people with steppe admixture reached the Iberian peninsula with boats via somewhere in atlantic northwestern Europe, they landed in Galicia and spread from there. The same route was probably done by the people bringing celtic languages in the peninsula.

Later on R1b folks crossed the Pyrenees and mixed with neolithic-like natives. They were less numerous but immigration was gradual over a longer period of time (plus there was probably a higher ratio of males), thus being able to change the yDNA of the population without changing their language.

Alogo said...

The Bronze Age and Iron Age Balkan samples seem to have the Yamnaya orange at some 20% taken all together and most are (omitting the 0 and 60%+) 5-30% steppe or so. Btw, I didn't see any actual breakdowns of the component percentages in the supplements, are there any?

That's comparable to Albanians and mainland Greeks, while the rest of the Balkans seems to have some 30-40%. Of course, the samples are generally from more Northern locations so we might expect to find lower average steppe input in the unsampled south in the Bronze - Iron age period (but we'll have to see). For the time being, it does seem like there has been somewhat of an increase in 'steppe' ancestry overall, perhaps due to the reason Rob mentioned.

As for the BA samples from Hungary, here's how they plot according to David's PCA: http://i.imgur.com/4kWs8AC.png

Some seem to reach all the way to Tuscany (the ones close to the Iron Age sample?). The Balkans seem a bit shifted towards the ANE side of things (extra CHG and EHG depending on the location?) compared to North Italians and Tuscans and the more Southernmost Europe-plotting Hungary BA and Balkan BA samples.

André de Vasconcelos said...

@Anthro

You're right, but these are just 3 guys anyway, there's a very good chance they do not represent the whole population living along the western fringes of Iberia at the time.
I'm also curious to see how these samples relate to modern Iberians, particularly those in the south, but the supplementary material isn't avaliable yet.

@Piranha
" The Iron age movements and post-Roman movements, e.g. Unetice, might both have contributed. Certainly NW African ancestry must be quite new, as it is ~1% in the Neolithic and Chalcolithic and there is little reason to expect it to increase during the Bronze Age. Increased CHG might be recent too."

I disagree, there are NW African archaeological findings for instance in northern Portugal dating from this period (circa 2000BC), plus you have modern NA ancestry peaking in the west and decreasing as you go east - Galicia has more than Valencia - which points to Atlantic contacts which could have happened during the Atlantic Bronze. There's also the apparent wave of CHG+EEF that brought greater amounts of yDNA J1/J2/etc into southern Europe (as Josep said) probably coming from the Levant or Caucasus. You also have Pheonicians traders along the Atlantic, or during the Roman Period during which moving from one place to the other was easier (within the empires' borders). There are many chances of it happening before recent times.
In fact my guess is that by the late roman period very little changed in Iberia


@Grey

How could it be very early ancestry if these samples are from ~1800BC and we find nothing like it anywhere else? I really have a hard time understanding this. Anyway, I agree with everything else you said

Grey said...

André de Vasconcelos said...
"How could it be very early ancestry if these samples are from ~1800BC and we find nothing like it anywhere else?"

I'm working from the premise (maybe wrong?) of them having steppe but not CHG and *if* that was true then the two options i can think of are:
- they came in a time before the ehg + chg mix
or
- they came from a different region that didn't have that mix
and while the second option is simplest taking the 1st option as a "what if?" *might* imply they maybe came in very low numbers initially and thus are hard to find.

It's a what if thing.

#

there's a map of neolithic mining sites in iberia which iirc showed two regions which had copper mines but not silver / gold mines (iirc Basque region and central meseta) which may be related to this if i'm not remembering wrong.

Matt said...

@P Piranha, OK, I have a bit more sympathy with where you're coming from and appreciate that you've gone looked at the samples in more depth to build your conclusions.

Though I would say, that still doesn't seem like good evidence for a southern influx to me necessarily. Even though the samples are late, there could still have been high AN populations about and heterogenity, so we could talk about admixture breaking down over time (as quite likely to have happened with Europes EN->MN transition).

I would also say as well, and perhaps more pressingly, I actually have a few more problems with the ideas and proportions you're thinking in.

Three of the four late samples you talk within the Balkans Bronze Age group they label about are from Croatia (I4331, I4332, I3313 - 1700-900 calBCE), and only one is from Bulgaria (I2163 - 1750-1625 calBCE).

Now, in ADMIXTURE, I4331, I4332 and I3313 all have about the same level of Yamnaya component - looks 30% - while I2163 is quite different - looks like 55%. I2163 looks to basically cluster with Corded Ware on the PCA. (My percentages from measuring Yamnaya component pixels vs the whole bar length.)

(By the way, looking at your link, I generally respect Ryu's analysis, but where is he getting the individual samples' levels of Steppe Ancestry from? I can't see them in the paper anywhere.)

Modern day Croats have about 38% Yamnaya per Davidski's "qpAdm tour", about the same as the French (36% with slightly more WHG), or in WHG and EHG terms Croats have 20% EHG, 11% WHG vs French 19% EHG, 13% WHG. At any rate, Croats are very similar to French in overall HG. Bulgarians are markedly different from both, with 33% Steppe EMBA according to Davidski's calculation, or or 17% EHG, 8% WHG. But this is not 20% steppe; in Greeks have about that at 22%.

(See these posts for Davidski's qpAdm tour - eurogenes.blogspot.com/2017/01/qpadm-tour-of-europe-mesolithic-to.html and eurogenes.blogspot.de/2017/01/qpadm-tour-of-europe-bronze-age-invasion.html. Pretty clear that Bulgaria, Croatia, Romania do not have 20% steppe, and are in the 30%s).

I've got a bit more confidence that three Croat female samples aren't outliers for their location (albeit they are from Southern Croatia) or populations living there later in history; less confidence on the Bulgarian ancient. I think it's possibly more likely that he was atypical for his location and time. I just don't see Bulgarian Late Bronze Age being like "and suddenly Corded Ware!" on a mass scale if sample sizes go up.

Another way to look at it which may be simpler as well, taking Davidski's PCA and this PCA: The Bronze Age Balkans as a whole, including the later Croatian samples and earlier Bulgarian samples, but ignoring the Corded Ware like outlier, basically run parallel on the North-South axis to Italy_CA (Remedello) to Iberia_Chalcolithic.

So to project an estimate on Davidski's graph - http://i.imgur.com/Cdscp5C.png.

That essentially looks to me like a bit like it overlaps the Croatia->Bulgaria cline. But we'll only know when for sure in future.

To be sure, I think these ancient Hungarian, Croatian and Bulgarian Bronze Age samples are *very* interesting, as they take on minority steppe components and have generally no R1, in contrast to the situation in Iberia, the Baltic and Central Europe. Suggesting a quite different dynamic of admixture and male founder effects. One that is perhaps also true of Italy.

@ Alogo, I think the only proportions are in Supplementary Table 4 and Supplementary Table 5 from their Supplements, done with qpAdm.

P Piranha said...

@ Andre

I admit there is no good explanation for why N African ancestry peaks in Galicia instead of Andalusia, but there is little evidence of North African ancestry in early Iberian BB nor late incoming C European BB elites, so that leaves rapid movement of people into Iberia via non-elites in the late BA if we want it to take place in the BA, since we need a 10% figure. Is there any evidence of such movement? Is there also any reason to think why this would avoid reaching the Basques?

P Piranha said...

@ Matt

I don't think they will fall there; the average of the entire Hungary_BA is 20% Steppe ancestry, and EBA Balkans have only 9% on average, so they certainly wouldn't. They would be as western, or even more western, than Basques. As for the LBA Balkans, they need to have extra ancestry from CHG, say from Armenia_EBA, to fall in that position you marked out, and its not clear that they will have it. Ultimately we should wait for David to process the data.

Grey said...

P Piranha

"I admit there is no good explanation for why N African ancestry peaks in Galicia instead of Andalusia"

a simple explanation would be a connection between NW Africa and the Atlantic Megalith culture. if the problem with that is

"but there is little evidence of North African ancestry in early Iberian BB nor late incoming C European BB elites, so that leaves rapid movement of people into Iberia via non-elites in the late BA if we want it to take place in the BA, since we need a 10% figure. Is there any evidence of such movement? Is there also any reason to think why this would avoid reaching the Basques?"

then an alternative possibility might be there wasn't a complete overlap between Atlantic Megalith culture sites in Iberia (containing the NA ancestry) and some later incomer sites e.g. regions that were suited to the neolitihic farming package and/or which had gold and silver deposits vs regions that only had copper.

Samuel Andrews said...

Several new Hungarian BBC genomes carry 50%+ Steppe ancestry. One has 74% and R1b Z2103 while a female from the same cemetary had 0%.

Point is there was plenty of diversity in LNBA Hungary.

Olympus Mons said...

Bad Timing. When paper came out was about to go to Chigado and only this morning arrived back in Lisbon. Will try to post several Points.
This is Point A
I could peep into comments. And totally agree withn Richard Rocca. If there was a sentence that kills my thesis is how Rui Martiniano wrote it:
"Unlike contemporaries elsewhere (but similarly to earlier Hungarian BA), Portuguese BA individuals show no signal of this component... but instead reveal significant introgression from several steppe populations into the Portuguese BA relative to the preceding LNCA

There is not coming back from it. Actually the opposite would prove me super right. But this is a killer. So I give up. There is no way one can argue on papers with this on the back ground. I won’t be the one arguing against others posting based on papers. So unless something really new comes out of further samples I won’t argue North Africa migration route to Iberia and those becoming bell beakers and so forth.

Now in the real world.
I happen to go and read Rui paper. How many of you realized that very cunningly Rui paper avoided THE ALL CHALCOLITHC and COPPER AGE and BELL BEAKER in Portugal? You else noticed the trick?
CA , CM and DA samples have nothing, nothing to do with all that we usually argue in this debates. No shit they came out as I2a and G2a. DA (dolmen ansiao) were the guys that in my thesis as the males being chased by the incoming population and put down (arrows in the back and huge DCP fractures) and buried in traditional caves in the mid of 4th millennia! Then they jump to middle Bronze age… in southern Portugal. And what else was happening in the southern Iberia? EL argar (from 1900bc) what some actually call “the greeks arrived “ (Mycenaean actually) . will try to research deeper into those samples to see how Algar they were …. Oh but wait! There is the Monte canelas (MC).... although not BB actually in last quarter of 3rd milenia. Yes this is going to be interesting because MC is going to be representative of Chalc and Copper age…Oh no, got no Mtdna, no Y-dna and a coverage of, gosh cant believe 0,05x! then why is even there? Because without it, there would have been a gap of almost 2000 years between earlier and bronze age samples. No copper age, no Bell beaker period, no 3300 bc incoming Iberia population.

But you cant really beat it. So, no more post from me in the future apart this my next Points in the incoming days.

Olympus Mons said...

...to be even more clear.
the paper avoids the all VNSP complex and jump from really neolithic people to bronze age Atalaia complex.

https://upload.wikimedia.org/wikipedia/commons/c/c4/Iberia_Bronze.gif

Olympus Mons said...

so much talk about copper...
And just a few days ago, in Perdigoes, this copper tool just got dug. And I asked A C valera and he says context is clearly first half 3rd millennia.

https://3.bp.blogspot.com/-9XOM16YOA8Y/WRId-s2dMLI/AAAAAAAAFoo/lmnKGeshkJU7DBgEMyl2G6239Le6FhCjACLcB/s400/IMG_4701.JPG


So many samples and inhumations in Perdigoes and many bell beakers and they keep on looking in caves!

Cossue said...

P Piranha:
"I admit there is no good explanation for why N African ancestry peaks in Galicia instead of Andalusia"

Actually N African ancestry doesn't peaks in Galicia proper, but in Leon, east of Galicia. In Galicia E-M81 is found at a 4% (N=292 Galician males, ref. goo.gl/nb2TfN), but if falls from a 8% in eastern Galicia, next to Leon, to 1% in the coastal areas. So, either the N African influx is very old, and the seashore population was "renewed" later during the Bronze Age or something, or the N African influx is recent (medieval records show the presence of Muslim slaves and prisoners in Galicia, León and N Portugal; eventually they converted and admixed with the locals). Also, most Muslims were expelled from Valencia, Aragon and Andalucia in the 17th century.

Rafs said...

"Modern people have post-BA CHG admixture brought by yDNA J and E folks, specially in southern Europe."

Basques possess teal admixture despite near zero levels of these two haplogroups.