Tuesday, September 8, 2015

New Northern Iberian Ancient DNA Doesn't Shake Basque Bell Beaker Hypothesis

On November 4, 2011, in a post at this blog, I laid out some thoughts about the origins of the Basque people in Europe that I've developed somewhat over the years, but still believe is the most plausible narrative to explain the facts. A newly released, open access PNAS paper with eight ancient DNA samples from the Copper Age and early Bronze Age in what is now Basque County may tweak this hypothesis, but doesn't seriously overcome it. (See also the Supplemental Materials).

In my view, the Bell Beaker culture was a linguistically Vasconic source of Y-DNA R1b in Europe. The archaeology reported in the new PNAS paper is consistent with the hypothesis that these people (including two males who were not Y-DNA R1b) are pre-Basque people who lived where the Basques live now, not actual culturally Basque individuals. The introduction of the PNAS paper explains the archaeological context (citations and references omitted):
We investigated the remains of eight individuals from the Chalcolithic and Bronze Age periods excavated from the cave of the El Portalón de Cueva Mayor, of the Sierra de Atapuerca—a site with a remarkably rich archaeological record, with human occupation from the Paleolithic to the historical period. The human remains were associated with offerings, such as domestic animals and pottery vessels corresponding to the pre-Bell Beaker culture, and were directly radiocarbon-dated to between ∼5,500 (Chalcolithic) and ∼3,500 cal yBP (Bronze Age). Seven of the burials contained fragmentary human remains whereas one burial was a near-complete skeleton of a male child showing signs of chronic malnutrition.
The harder question, however, is the strong autosomal similarity between these people and modern Basques, who have somewhat more European hunter-gatherer component, and somewhat less Early European farmer component than these individuals, but only modestly so, and don't seem to have any ancestral components not found in some proportion in Early European farmers. 

How could the Y-DNA landscape of Western Europe change so dramatically, while making such a negligible change to the autosomal and mtDNA mix?

The Basque ethnicity, in my Bell Beaker hypothesis, probably had their ethnogenesis around 2900 BCE in Southern Portugal, with a migrant population drawn to the area's rich resources of copper and tin from someplace far to the East of Iberia (probably ultimately from the general vicinity of the Southern Caucasus Mountains give or take 200 km or so), arriving either by land or by sea. Either initially, or as they expanded, they then married a succession of local women, some descendants of first wave Cardial Pottery farmers, some descendants of European hunter-gatherers, in an ever expanding pool of local women as the men expanded their horizons to the frontier of Atlantic and Western Europe. Socially, sons of Bell Beaker men were favored and assumed the reins of local megalithic farming communities, while daughters of Bell Beaker men were absorbed into the general gene pool.

In my view, the modern day Basques were relatively late on the scene in modern Basque country (ca. 2500 BCE) and probably arrived from France, but unlike follow members of their linguistic community in Western Europe, Central Europe, and Northern Europe were not overrun by Indo-Europeans around the time of Bronze Age collapse (starting ca. 1300 BCE), for a variety of reasons.

By the time that the Indo-Europeans arrived, 50 generations or so of introgression of local women into Bell Beaker patrilineages diluted the Bell Beaker autosomal contribution and mtDNA contribution to an imperceptible level, while causing Y-DNA R1b which was predominant (but probably not universal) in the founding population of Bell Beaker men, to become the most common Y-DNA type in Europe.  And, 14 generations or so would have passed between Basque ethnogenesis in Southern Portugal and their arrival in modern Basque Country.

The cultural and Y-DNA descendants of the Bell Beaker people who were the most important contributors to Vasconic ethnogenesis in Portugal ca. 2900 BCE, by the time that reached modern day Basque country, were so genetically diluted with contributions from local women. As a result, they were autosomally similar to other Iberians, except that they had less North African influence because more of their ancestors were from further North in Western Europe where there was no possibility for trace North African influences to penetrate the gene pool.

There is some hope that we may before too much longer have more data to explain the contradictions with evidence from the same site. The Supplemental Materials note that:
43 additional human bone fragments have been recovered in Middle Bronze Age levels at the space known as the Salón del Coro or Galería Principal, which is also part of the El Portalón site itself
The Supplemental Materials have this to say regarding historical linguistics:
The Neolithic cultures appear in the Iberian Peninsula around 7,500 cal BP as a result of dispersal of human groups along the Mediterranean coastal areas and (eventually) visible as the Cardial culture in the western part of the Mediterranean. According to [104] and others, the dispersal of the Neolithic communities was related to the spread of Indo-European languages to Europe. However, this model (the Anatolian Hypothesis) coexists with a number of competing models. In particular, other models have placed the origin of Indo-European languages (or Proto-IndoEuropean) in the East, North of the Caspian and Black Seas, and in a chronologically younger period (often termed the Steppe Hypothesis). The recently confirmed eastern migration of human groups (linked to the Yamnaya group) into Europe around 4500 cal BP has been interpreted as evidence for the Steppe Hypothesis, but the population movement is also consistent with a secondary expansion under the Anatolian Hypothesis.

The linguistic implications of this are still under debate. There are a number of different linguistic scenarios consistent with these genetic results, although as always it is difficult to associate genetic information confidently with archaeological groups or language families. The Basque language (Euskara) is a linguistic isolate, and is believed to be the last surviving pre-Indo-European language in Western Europe. The only known precursor to Basque is Aquitanian, reported during the Roman empire and spoken in southwest Gaul, the Pyrenees, and some adjoining areas. This language is clearly related to Basque, but is probably a relative rather than the direct ancestor. Basque has always been a magnet for extravagant linguistic speculation, but the hypothesis of Paleolithic roots of Basque is most wide accepted on the groups on the grounds of explanatory parsimony, and in the absence of adequate evidence for other hypotheses.

One intriguing suggestion is that the Basque language exhibits similarities to the pre-Roman language of Sardinia (Paleosardo) based on, for example, place-names on Sardinia. The number of linguistic forms is small, but this is particularly interesting given Sardinians and Basques are the two modern populations with the highest genetic proportion of early farmer ancestry. Contacts between Iberia and Sardinia in the Neolithic are indicated by recent studies of Obsidian artifacts, facilitated by maritime (and coastal) movement. This suggests the Basque might be the remnant of a much larger Vasconic speaking area, suggesting a the possibility that language family spread along with the first farmers. If so it would be tempting to suppose that it was the only language of the first farmers, which would support the Steppe Hypothesis of Indo-European origins over the Anatolian Hypothesis.

Language isolates are however not uncommon outside of Europe. Of the approximately 350 language families in the world, 121 of are isolates. The existence of such isolates is not really surprising given the highly skewed of linguistic diversity, and isolates are sporadically encountered embedded within the ranges of most large language families worldwide. Within the Indo-European languages, the Greek, Armenian and Albanian subgroups are also (near) isolates, consistent with the prediction of the Anatolian Hypothesis that the center of linguistic diversity in Europe would coincide with the entry points of the first farmers (and not contradicted by [68]). It is not implausible that Basque is an indigenous language that expanded in place after adoption of agriculture, or that Basque entered Europe alongside these other Indo-European languages. There is some hope that advances in Proto-Basque reconstruction will shed light onto these issues. Proposals of linguistic similarities between the Basque and other languages must however be evaluated with caution.
For what it is worth, the Anatolian Hypothesis isn't remotely credible given the facts as we know them today. And, it is almost certain that Basque is the remnant of a much larger Vasconic speaking area at some point in the past.

In my view, the real hard question is not between the Steppe Hypothesis and Anatolian Hypothesis of Indo-European language origins, which is largely resolved, or between a Paleolithic and post-Paleolithic origin for the Basque languages.

The hard question, one which the Supplemental Materials don't quite seem to grasp, is whether Basque is a language associated with the Early European Farmers of Europe, or the period of time starting with the Bell Beaker culture's appearance and ending with Bronze Age collapse. Neither position, by the way, is inconsistent with an apparent link between Paleosardo and Basque as Sardinia also had a Bell Beaker period during which there was maritime trade between the island and the nearby mainland.  Literate history post-dates the beginning of the Bell Beaker period in Sardinia, and all oral historical and naming conventions had to pass through the Bell Beaker period in Sardinia from the early Neolithic to reach the present.

Secondary Hunter-Gatherer Admixture

The increase in European hunter-gatherer proportion relative to the first wave of farmers (whose hunter-gatherer component shows more affinity to Hungary than to Western Europe), as these individuals indicate, may have come from new infusions of local hunter-gatherers into the gene pool during the collapse of the first wave of farming.

This new data strongly favors the idea that the first wave of European farmers were formed somewhere to the Southeast of Hungary and probably either in Anatolia or just beyond it, and fused at about a 60-40 ratio with local hunter-gatherers (with the farmer contribution disproportionately from men and the hunter-gatherer contribution disproportionately from women).

Support for a basically common first wave of hunter-gather admixture into European farmer populations before they went on to have additional local hunter-gatherer admixture is found in the following excerpt from the Supplemental Materials (citations and references omitted):
In this case we used – as others before - KO1 (individual found in a farming context) as a proxy for Hungarian hunter-gatherers since he grouped with Mesolithic individuals in all other analyses. The highest proportion of Mesolithic ancestry in the Portalón individuals seems to be related to central European hunter-gatherers (KO1, Loschbour) and not to the geographically close LaBrana (several |Z|>2). 
Central European farmers (CO1, Iceman, NE1) exclude only Mesolithic Scandinavians (Motala12) as a possible source so it seems likely that their admixture happened in Central Europe as well. 
Surprisingly, Mesolithic Scandinavians (Motala12) are excluded as a possible source of admixture into the Scandinavian farmer Gok2 whereas all other hunter gatherer groups (including the Neolithic Scandinavian Ajv58) are consistent with the data. This suggests multiple admixture events into Scandinavian farmers which happened in different parts of Europe. 
However, we note that the currently available data does not allow us to detect a strong population structure in Mesolithic Europe. Only Scandinavia seems to be an outlier from a relatively uniform Mesolithic population.
After this Early European Farmer ethnogenesis, these Early European Farmers probably migrated across Europe in a mostly endogamous basis until the first wave of farming collapsed at various times in various places, leading to a partial reversion to hunting and gathering and substantial introgression of hunter-gather populations into their communities (probably 15%-35% of total ancestry, with higher percentages on the frontiers and lower percentages in Central Europe), until this ceases when farming recovered in the Bronze Age. The model done in the new study suggests a 30% introgression of European hunter-gatherers similar to the ancestor of two hunter-gatherers from whom we have ancient DNA, one from Iberia and one from Luxembourg. Otzi the Iceman appears to have a 19% introgression from a Scandinavian hunter-gatherer, while a Scandinavian first wave farmer appears to have a 33% introgression from a Scandinavian hunter-gatherer.
The closest extant populations to hunter–gatherers from Iberia, Scandinavia, and Central Europe are Northern Europeans; however, the hunter–gatherers fell outside the range of modern-day European genetic variation. In contrast, early farmers from Iberia, Scandinavia, and Central Europe grouped with modern-day Southern Europeans, consistent with outgroup f3 statistics. These results demonstrate that early European farmers, including those in Iberia, emerged from a common group of people. This observation indicates that farming was brought to Iberia via migration, similar to the process in Scandinavia and Central Europe.

Chalcolithic farmers (Iberian ATP2 and the Tyrolean Iceman) and Scandinavian Neolithic farmers (Gok2) traced a substantial amount of their genetic ancestry to European HG groups, in contrast to the earliest farmers of Central Europe (NE1 and Stuttgart), and this increase in HG admixture across Europe was significant as a function of time (R^2 = 0.69, P = 0.001). The best fitting source for the HG admixture into the El Portalón individuals was the common ancestor of the nearby La Braña Mesolithic individual and a Mesolithic individual from Luxembourg (Loschbour) whereas contemporary farmers from Central Europe (Iceman) and Scandinavia received their (best-fit) HG admixture from Scandinavian hunter–gatherers. 
These inferred admixture events demonstrate that different farmer populations had different HG groups as the best proxy for the source of admixture (D-tests showed similar results of multiple admixture events in different parts of Europe). These analyses showed that, whereas early farmers—who were likely more numerous than the hunter–gatherers —spread across Europe, they assimilated HG populations, a process that continued locally for several millennia.
This discussion continues in the Supplemental Materials:
Chalcolithic and Scandinavian Neolithic farmers (ATP2, Gok2, Iceman, which are all dated to approx. 5000 BP) seem to harbor a higher proportion of hunter-gather related ancestry than the first Neolithic farmers of central and eastern Europe (Stuttgart, NE1).
This is, of course, perfectly consistent with the hypothesis that substantial new, local hunter-gatherer introgresion into farming populations occurs after the universally experienced first collapse of farming among first wave European farmers, although this doesn't happen at exactly the same time in each place.

The New Data Points

All eight of the ancient DNA samples produced mtDNA haplogroups, but only four produced autosomal genetic profiles and only two of those were men. The other two men had only 3% genome coverage so Y-DNA haplotyping was not possible. Three of the autosomal samples were Copper Age, while one was a woman from the early Bronze Age.

The Y-DNA Data

Of the two men for whom Y-DNA profiles could be determined one had Y-DNA I2a2a (dated to about 2,960-2829 BCE), and the other had Y-DNA H2 (dated to about 2,849-2,628 BCE).

Y-DNA I2 appears to have been predominant among European hunter-gatherers, but is not terribly uncommon in early European farmers. The Supplemental Materials have this to say about this individual's Y-DNA after discussing the specific genetic markers used to classify this Y-DNA sample (citations and references omitted):
While the almost European-specific haplogroup I arose approximately 20000 to 25000 years ago, haplogroup I2a2a may have diverged as a subclade, around 15000 years ago, possibly during the recolonization of Europe following the Last Glacial Maximum (LGM). Unlike the more common subclades of I1 and I2a1, haplogroup I2a2a appears at relatively low frequencies across much of Europe. Its highest levels (10-12%) are found in modern-day Germany and the Netherlands, with frequencies of around 5%, notably occurring in parts of modern-day France as well as Mordvin in the Volga region of central Eastern Europe. Other members of haplogroup I have been discovered previously in ancient individuals; e.g. I* in Mesolithic Scandinavians, I1 in Hungary, I2a in Neolithic individuals from Hungary and France, I2a1 in Neolithic Croatia and a late hunter-gatherer from Sweden, and I2a1b in Mesolithic individuals from Luxembourg and Sweden.

Other haplogroups found among ancient specimens include C* in Upper Paleolithic Russia and Mesolithic Spain, C6 in Neolithic Hungary, E1b1b1 in Neolithic Spain, F* in Neolithic Germany and Neolithic Hungary, G2 in Neolithic Hungary, G2a in Neolithic France, Neolithic Germany, Neolithic Hungary, Chalcolithic Italy, and Neolithic Spain, J2a1 in Bronze Age Hungary, K (xLT) in Upper Paleolithic western Siberia, N in Iron Age Hungary, R* in south central Upper Paleolithic Siberia, R1a in Neolithic Germany, and Neolithic R1b in Germany.
The citation for the "Neolithic R1b in Germany" is Lee EJ, Makarewicz C, Renneberg R, Harder M, Krause-Kyora B, Müller S, et al. Emerging genetic patterns of the european neolithic: Perspectives from a late neolithic bell beaker burial site in Germany. American Journal of Physical Anthropology. 2012;148: 571–579. doi:10.1002/ajpa.22074 The Bell Beaker context makes clear that it wasn't really Neolithic as opposed to Copper Age or Bronze Age.

The H2, as discussed further below, is rare but has a generally West Eurasian non-hunter-gatherer distribution.

The big story there is that Y-DNA R1b is absent from these two individuals despite the fact that modern Basque people have one of the highest percentages of R1b in Europe.

The mtDNA Data

On the mtDNA side:
The eight individuals, genetically inferred to be four males and four females, carried mtDNA haplogroups associated with early farmers of Europe (e.g., haplogroups K, J, and X), with hunter–gatherers (e.g., haplogroup U5), or with both groups (e.g., haplogroup H).
The Supplemental Materials have a nice further analysis of the mtDNA findings in the context of the larger literature (citations and references omitted):
All eight individuals from Atapuerca displayed unique haplotypes. The most abundant haplogroup, U5, was found in three temporally non-overlapping individuals. Two belonged to subtypes of U5b (U5b3 and U5b1b) and one belonged to U5a (U5a1c).... Two individuals belonged to H3. They were dated to within the same time-frame but were not maternally related as one of them carried a T to C transition at np 12957 classifying it to H3c. The remaining three individuals belonged to the haplogroups J, K and X (J1c1b1, K1a2b and X2c).

The mitochondrial lineages of the ATP individuals show a heterogeneous ancestry and can be traced back both to hunter-gatherer (HG) and subsequent farmer contexts. The most frequent haplogroup in ATP, U5, is commonly found in HG groups in Iberia and across Europe and Scandinavia. U5 subhaplogroups are also found in Neolithic farmer populations in Europe although at lower frequencies. The remaining four haplogroups found in ATP, H, J, K and X, are present in other farmer populations from the Neolithic and onwards. In southern Europe (e.g. Spain, Portugal and Italy), however, haplogroup H is also frequent in Paleolithic and Mesolithic HG populations.

Even though some haplogroups (U5b and H) are shared between ATP and HGs from Mesolithic Iberia (southern hunter-gatherers SHG), the general haplogroup composition between the groups differ, similar to the differences between other farmer and HG populations in Europe. None of the previously investigated Neolithic farmer populations from Iberia have similar haplogroup distribution as ATP. These farmer groups also differ from each other. Analysis of haplogroup frequency data have for example shown that early Neolithic north-eastern Iberian populations cluster with early- and middle Neolithic populations from central Europe while other Neolithic Iberian populations (from Basque Country and Navarre,NBQ and Portugal, NPO) share a closer affinity to HG populations. NBQ is the population that share the largest number of haplogroups with ATP (X, H, J, U5b and K although the frequencies differ and NBQ also display additional haplogroups (U, T, HV and I). The Chalcolithic individuals from El Mirador (MIR), a cave located in the same mountain system as ATP (Sierra de Atapuerca), present a somewhat different haplogroup distribution than ATP. MIR clusters with early Neolithic Iberians and early and middle Neolithic central European populations. They lack the U5 subhaplogroups found in ATP and instead display T2 and U3. RFLP data from another Chalcolithic population from the Basque Country show the same main haplogroups as found in ATP and MIR (38% H, 17% U, 13% J, 21% K and 9% T+X), although the lower resolution of the data cannot specify which population (ATP or MIR) that it is most similar to. It has further been suggested that the mt-haplogroup composition of Basque populations differs between Chalcolithic and historical times (600-700 AD) with increasing frequencies of H and V haplotypes and with increasing similarities to present-day western European populations.

The picture of the ancient farmers in Iberia remains unresolved and the limited level of information retrieved from mitochondrial DNA has not been able to go beyond the above described observations. Present-day European populations are genetically quite homogenous in terms of mitochondrial haplogroup distributions and it is mainly haplogroup frequency differences that separate different populations. It is therefore not a straightforward process to assess the potential connections between ATP and specific present-day populations. We note that the most abundant lineages in the ATP individuals are found in higher frequencies in some Basque-speaking populations (U5b and H3) than in other European populations. Further, several haplotypes have been suggested to be autochthonous to present-day Basque populations. Two of these are J1c1, a lineage ancestral to the J2c1b1 haplotype in ATP7, and H3c2a, a lineage that derives from the H3 and H3c haplotypes found in ATP17 and ATP12-1420.
The whole genome data

In autosomal genetics, these four individuals cluster together in both PCA and a ten population admixture analysis.

Like modern Basque persons, they have essentially no North African component and no Caucasian/Central Asian component. Other other modern Spanish populations have some trace North African component (perhaps 1-3%) (the time depth of this component isn't entirely clear and much of it could be from as late as the Moorish era in Spain). Modern French populations (except some people from Southern France) and modern Spanish populations (apart from the Basque) also all have small Causasian/Central Asian components (perhaps 2%-10%) (probably due to Indo-European migration into the region in the Iron Age).

These individuals, like modern Basque persons and Sardinians and ancient DNA from other first wave early European farmers are a mix of two admixture components: European hunter-gatherer (which is pretty much the sole component of ancient DNA from European hunter-gatherers) and Early European Farmer. Sardinians and first wave European farmers have the highest percentage of Early European farmer (more than half). Basque people have similar amounts of Early European farmer to other Spanish and French people. These four individuals have an intermediate amount of Early European farmer, suggesting hunter-gatherer introgression beyond the portion that went into the ethnogenesis of the LBK and Cardial Pottery farmers.

The PCA chart in the paper is a bit of a puzzle.  PC1 clearly represents a hunter-gatherer to early European farmer proportion continuum.  But, PC2 is harder to make sense of.  Most European population are at roughly the same spot on PC2 as the Sardinians and the ancient DNA from first wave farmers across Europe.

But, European hunter-gatherers and the Basque are significantly to the left and many Spanish people lean in that direction on PC2, while Cyprus and Malta are strongly to the right on PC2.  It might be some Middle Eastern contribution which was present in modest amounts in early European farmers, is present at higher levels in island populations near the Middle East, and is absent in European hunter-gatherers.

But, if the Basque simply diluted this "Middle Eastern" PC2 contribution with European hunter-gatherer contributions, then they would be much higher on PC1 than they are in fact.  This suggests some sort of "anti-Middle Eastern" farmer contribution in Basque and Spanish farmers that counteracts the pull of the Middle Eastern tendency of the early European farmer contribution, perhaps one too much like the early European farmer contribution to be distinct from it in a K=10 ancestry analysis, but which might pop out at a higher number of ancestral populations.

The Supplemental Materials have this to say about additional PCA Analysis done that was not discussed in the main paper (citations and references omitted):
We repeated this analysis including North African populations in order to look for any additional component of the Iberian farmers to modern North Africans. PC1 separates North Africans from Europeans while PC2 seems to be correlated with the amount of Near Eastern ancestry. All ancient samples line up along this gradient with one extreme in Druze and the other in Mesolithic Europeans. Farmers from El Portalon and Sweden are slightly shifted towards hunter-gatherers in comparison to central European farmers. The PCA suggests no additional North African ancestry in any of the ancient farmers.

An additional PCA including modern populations from the Caucasus was conducted since ADMIXTURE results suggest some Eastern ancestry in some samples. PC1 correlates largely with Near Eastern ancestry with Druze and Mesolithic Europeans as the two extremes. PC2 has Sardinians and Tajiks as extremes suggesting some correlation with longitude. Ancient farmers group around Sardinians and the Chalcolithic El Portalon individuals form a line between Sardinians and Basques whereas central European farmers are shifted towards Near Eastern populations. There is no specific affinity to modern-day Caucasian populations for any of the ancient individuals.
Another discussion of potential North African genetic tries from the Supplemental Materials is here (citations and references omitted):
The geographic proximity of Iberia to Northern Africa opens up possibilities to migrations across the Strait of Gibraltar. In fact, farming reached Northern Africa and Southern Spain long before Northern Iberia, and modern Iberian populations show a significant proportion of North African ancestry. Admixture estimates and outgroup f3 statistics do not support a strong contribution of North African populations to the individuals of El Portalón. 
Modern-day North Africans are highly admixed with contributions from Europe, sub-Saharan Africa, the Near East and Neandertals, and the level of admixture vary among groups. In order to avoid other components in reference populations from confounding the D-tests, we assume that all early European farmers contain the same Near Eastern component (which is also found in North Africa to some degree) and conduct D-test in the form of (Mbuti, modernday North African; ancient farmer 1, ancient farmer 2). 
We use Mozabite, Saharawi, Algerian, Tunisian and Burbur as representatives of modern-day North African populations since the a particular ancestry component (the ‘North African component’) is maximized in these groups in the admixture analysis. 
These analyses demonstrate that ATP2 and ATP12-1420 have similar genetic affinities to North Africans as Central European early farmers have. 
However, ATP16 shows higher affinities to North Africa than other ancient farmers, suggesting that there was at least some contribution from North Africa ~5,000 years ago (in one out of eight Portalón individuals). 
Surprisingly, ATP9 shows the lowest North African affinity of all ancient farmers. Since ATP9 also represents the youngest individual (Bronze Age) in the analysis, we suspect that this is the result of increased admixture with other European groups in the Bronze age, which contained less North African or Near Eastern ancestry. Generally, genomic data from Neolithic North Africans is needed to solve the question whether there was a strong Neolithic African contribution to the Iberian Neolithic population.
Phenotype and Inbreeding Data

The Supplementary Materials also discuss from phenotype conclusions that can be drawn from the genes of these individuals and about inbreeding in the entire sample of ancient DNA (citations omitted):
[T]he inhabitants of the El Portalon cave were probably all lactose intolerant in adulthood. This suggests a much later spread of this variant that has been the target of adaptation to a milk-rich diet in modern-day northwestern Europeans which also occurs at reasonably high frequencies in Northern Spain and Basques.

All sequence for the SLC24A5 (rs1426654) variant showed the derived state in the El Portalon individuals, which together with two derived variants at SLC45A2 (rs16891982), suggest that the pigmentation of the Chalcolithic Iberians was lighter than the Mesolithic LaBrana1 individual who carried the ancestral states at these major pigmentation loci. rs1805007 in MC1R which is associated with red hair and light skin is ancestral in all but one sequence (out of eleven in all El Portalon individuals) but that single derived base call might also be due to post-mortem damage. rs12913832, a SNP that explains more than 56% of the variation between blue and brown eyes, has been shown to be derived in the Mesolithic LaBrana1. Two of the El Portalon individuals show only ancestral alleles at this site whereas one individual shows both variants suggesting the individual is heterozygot at the site. These observations suggest some eye color variation but also a tendency towards brown eyes in the Chalcolithic Iberians.

To summarize, Chalcolithic Iberian farmers seem to be lactose intolerant as the Mesolithic inhabitants of the Peninsula. However, their pigmentation was fairer and their eyes were darker than in the hunter-gatherer LaBrana1. . . .

Diversity was estimated for all sites or cultures with two reasonably contemporary individuals and decent coverage: sites Ajvide (using Ajv58 and Ajv70), Motala (Motala12 and Motala1), Gökhem (Gok2 and Gok4), El Portalón (ATP2 and ATP12-1420) and the culture Alföld Linear Pottery (ALP; NE1 and NE5). This procedure was chosen to avoid the effects of potential inbreeding. The Scandinavian hunter-gatherers show the lowest diversity of all groups whereas the Scandinavian farmers from Gökhem are intermediate between those and the central European and Iberian farmers. Generally, farmers show a higher diversity than hunter-gatherers which is consistent with previous results and might be attributed to the increased carrying capacity of farming groups and/or the admixture with hunter-gatherers.
Boundaries In Time on Basque Origins

What we know about the genetics makes it increasingly unlikely that the Basque culture and language emerged from European hunter-gatherers of Western Europe (particularly the Franco-Cantabrian refugium).  These autochronous people may have had a substrate influence and may have made genetic contributions, particular maternally (although mtDNA H now looks like a likely Mesolithic contribution to Iberia that expanded with the Bell Beaker/Vasconic surge or with the early megalithic expansion).

But, the earliest plausible time that the Basque culture and the language could have emerged is from the first wave of European farmers, who reached Iberia relatively late (ca. 5,500 BCE).

The first farmers of Europe were apparently very similar to each other genetically all across Europe. We only have a couple of data points of first farmer Y-DNA in Western Europe, but not one of them is Y-DNA R1b, and the same is true of the first farmer data points we have from elsewhere in Europe, which are more numerous.

We know that the Basque culture and language had already come into being and was losing ground to the Indo-Europeans (first the Urnfield culture, ca. 1,300 BCE, and then the Celts and then the Romans), by the time of Bronze Age collapse and the Iron Age.

This is roughly a 4,200 year window.

Something happened to turn Western Europe which had almost no Y-DNA R1b, when the first farmers arrived, into a place where Y-DNA R1b was the dominant Y-DNA type of men in the region.

Yet, the fact that Basque men have one of the highest Y-DNA R1b percentages in Europe, rather than one of the lowest ones, almost surely implies that Indo-Europeans were not the source of Y-DNA R1b in Europe the way that Indo-Europeans were almost surely the source of Y-DNA R1a in Central and Eastern Europe (as we now know from ample direct ancient DNA evidence).

Likewise the high rate of lactose persistence in Basque persons, which this most recent paper strongly indicates arrived with the Bell Beaker culture or later, similarly can't have had Indo-European origins and had to have had fairly recent origins.  The LP dynamics are different than the R1b, which is merely ancestry informative, because LP clearly conferred some strong selective fitness advantage, even if the exact mechanism by which this happened isn't entirely clear.  So, a very low frequency introgression into the population combined with strong selective fitness effects could produce a dramatic change in the frequency of this genotype without having to hypothesize a major demographic event.  But, in all likelihood, given the timing involved, R1b and the LP gene probably entered the population of Northwest Spain at about the same time. (It would be interesting to know the RH negative or positive blood type of these individuals as the high rate of RH negative individuals is another distinctive aspect of the modern Basque gene pool and this might shed additional light on the extent to which they are ancestral to modern Basque people.)

The fact that the range of the Bell Beaker culture matches the Vasconic linguistic substrate in Europe and the approximately range where Y-DNA R1b is found in Europe now, and that I can't find any other plausible sources of this change, makes me conclude that the Bell Beaker culture was Vasconic linguistically and was the source of the transformation of Western Europe's Y-DNA mix and pinpoints the transition very precisely in time.

I just can't find any other explanation that can explain these singular data points.  While it could certainly be a coincidence, particularly taking this find in isolation, that these new four autosomal ancient genomes, two from men, don't contain Y-DNA R1b and are pre-Bell Beaker, I don't think that it is a coincidence.  I think it is much more likely that these individuals are some of the last people living in what is now Basque County who were not Basque and instead were pure first farmer descendants with an extra infusion of local hunter-gatherer ancestry picked up in the wake of the collapse of the first wave of farming that produced sad situations like the skeleton of the little boy who died of starvation in this most recent find.

A few centuries later or even contemporaneously a few valleys over, Vasconic people from France would arrive and put in place their culture in the only place where it would ultimately survive.  The extent to which this happened through the cultural influence of the thin Vasconic ruling class, and the extent to which this happened through population replacement, is hard to know, especially since the incoming Vasconic people, and the resident first wave farmer people who already lived there, may not have been that different from each other genetically at least in autosomal and mtDNA population genetics.

The need to explain the modern high proportion of Y-DNA R1b and what is increasingly clear was the near total absence of Y-DNA R1b in either European hunter-gatherers, or in the first wave of European farmers, compels some sort of solution, and makes otherwise less plausible narratives seem like the only possible explanations for the facts.

What About The Man From El Trocs Cave?

This said, there is one chink in this argument.  In a 2015 paper by Haak et al., that reports 69 ancient DNA results from Europe, there is a reference to a man from Neolithic Spain ca. 5100 BCE whose body was found in the El Trocs cave in the Pyrenees Mountains in Northern Aragon whose Y-DNA haplogroup was found to be R1b1* ancestral to all extant forms of Y-DNA R1b (both V-88 from Africa and the Eurasian haplogroups; he is not R1b-V88 as has been frequently misreported).  This long predates Bell Beaker and coincides with the very early Neolithic era in the region, and is also not far from modern Basque country.  Unfortunately, since so much ancient DNA was dumped on the world in this one paper all at once, there is essentially no detailed analysis of the context of theses El Trocs remains, even though they have the potential to be paradigm shaping.

My inclination is to think that this instance is a fluke outlier individual whose ancestors joined the wave of the expanding Neolithic revolution but ultimately left no modern living descendants (perhaps they died off in the bust that followed the first wave Neolithic), because the phylogeny of R1b and its distribution around Europe are not a good fit to this man or his kin being the an important source of Y-DNA R1b in Europe - it isn't a good fit for the pattern of Y-DNA R1b haplotype diversity, for example, or the apparent path from the Steppe to Western Europe that phylogeny analysis of modern R1b haplogroups supports. But, if more pre-Bell Beaker Y-DNA R1b turns up in Southern Europe, I might be persuaded otherwise.

A North African origin for this individual is even more unlikely than the narrative I suggest, since the El Trocs individual's autosomal genetics bear no similarity to North Africans.  He is autosomally pretty much identical to lots of other first wave European farmers.  Also Neolithic Spain in that vicinity at the time involves cereals and legumes with either wild caught or domesticated pigs or rabbits, while the Chadic people associated with R1b-V88 in Africa at approximately the same time, have at least sheep and goats (and possibly cattle) and seafood, but don't seem to have farmed cereals or legumes.

The oldest instance of R1b1* is found in Samara, Russia in a hunter-gatherer individual about four hundred years earlier (with autosomal genetics similar to other European hunter-gatherers) where the region is teaming with ancient Y-DNA R1b similar to that found in Europe two thousand years later in the Yamanya culture.

A Footnote Regarding Y-DNA H

Y-DNA H, like its parent clade, Y-DNA F, is centered around South Asia.

Y-DNA H1 and H3 in South Asia

Most Y-DNA H is found in South Asia and among South Asian expatriates, including the Romani of Europe.  But, this is predominantly Y-DNA H1 (aka H-M69) with a sprinkling of Y-DNA H3 (aka H-Z5857) in South Asia. Y-DNA H1 is present at the highest percentages in Southern India (25%-40%).  It is found in about 10% of upper caste men in South Asia, and in about 25%-35% of tribal men in India.  It's other sister clade, Y-DNA H3 (aka H-Z5857) is much rare than Y-DNA H1 and is found in some individuals in South Asia.

Outside of India, the most common clade of Y-DNA H is Y-DNA H1a1 (aka H-M82). This is found at rates of 13% to 50% in European Romani men (i.e. European Gypsies). But, a 2003 study of 20 Andaman Islanders found it absent there.

Y-DNA H1a1 Outside South Asia

In Southeast Asian, a 2006 study found one of six Cambodian men tested had Y-DNA H1a1, and a 2000 study found one of eighteen men in Cambodia and Laos had Y-DNA H1a1, as did only 2 of 1090 men in a 2012 study in Northeast India.

At the Northern fringe of South Asia, Y-DNA H1a1 was found in 8 of 188 men in Nepal in a 2007 study, in 7 of 204 men in Afghanistan in a 2012 study.  But, a 2007 study found none of 156 men in Tibet had Y-DNA H1a1, and a 2006 study of 26 men in Japan and 18 Siberian men found that it was absent.

Y-DNA H1a1 is found at more than trace levels in Iran and at trace levels in adjacent areas.  In West Asia, a 2004 study of 523 men from Turkey found one man with Y-DNA H1a1.  A 2009 study found 2 of 150 men with Y-DNA H1a1, and a 2012 study of Iran found 11 such men out of 938.  It was not found in a 2011 study of 1789 Caucasian men, or a 2009 study of 66 Georgian men, but was found in 1 of 38 Balkarian men (a Caucasian ethnicity) in the same 2009 study. In the Middle East, a study of 298 men in Yemen, UAE and Qatar found 3 men with Y-DNA H1a1. A 2009 study of 1891 men in Saudi Arabia, Oman, Egypt, Somalia, Lebanon, Jordan and Iraq found one man with Y-DNA H1a1 (in Saudi Arabia where 157 men were tested).

In Europe, Y-DNA H1a1 is found at only trace levels and only in populations that historically probably had contact with the Romani people.  In non-Romani European samples, Y-DNA H1a1 was found in 1 in 92 Ukrainian men in a 2009 study, in 1 of 113 Serbian men in a 2005 study (but none among 141 Herzegovians in the same study), and in 2 of 57 Macedonian Greeks in a 2008 study.  In the same 2009 study as the Ukrainian men, it found no Y-H1a1 in 92 Greeks, 55 Albanians, 324 Bozniaks, 75 Slovenians, 67 Northeastern Italians, 53 Hungarians, 75 Czechs, and 99 Poles.


Y-DNA H2 (aka H-P96, formerly known as H* and F3) is less familiar and has a very different distribution.  It is found in Western Europe in France, Switzerland, Germany, and the Netherlands, and also among Armenians, in Iran and in India.

The Supplemental Materials state with regard to Y-DNA H2 after clarifying the history of this label and identifying the loci used to make the classification that:
While only a few H2 individuals have ever been found, the haplogroup appears to have a west Eurasian distribution; with a low level Middle Eastern presence in modern-day Iran, Turkey, Bahrain, Kuwait and Qatar (Family Tree DNA), as well as minor occurrences in modern-day England, France, Sardinia, Sweden and the Netherlands (Family Tree DNA). H2 also seems to occur at low frequencies in Neolithic sample.*
* Citing for that sentence  Haak W, Lazaridis I, Patterson N, Rohland N, Mallick S, Llamas B, et al. Massive migration from the steppe was a source for Indo-European languages in Europe. Nature. 2015; doi:10.1038/nature14317

This is suggestive of the couple of possibilities. One is that Y-DNA H2 could have been a Cardial Pottery minor component that spread from Iberia to the megalithic first wave farmers further to the North as well.

Perhaps as few as just one Y-DNA H2 could have made his was from Southwestern Anatolia (once home to Armenians) who had ancestors who made it there from Iran, and then migrated further to wind up in the Cardial Pottery founding population, without leaving descendants whose descendants survived to the present along the way.

It would be interesting to see if there is any Y-DNA H2 in Tuscany where the Etruscans had their non-Indo-European civilization.

Given the dates of the Y-DNA H2 individual of 2,849-2,628 BCE, the pre-Bell Beaker archaeological context, and the location of the remains, any theory that the H2 was associate with Bell Beaker origins has to dramatically less likely.  The oldest Bell Beaker wares in Iberia are from ca. 2900 BCE, and the Bell Beaker culture reached Northwest Iberia much later than that.  In principle, a single man could have made this way from Southwestern Iberia to Northwestern Iberia and been incorporated into this community of pre-Bell Beaker Cardial Pottery farmers, but that does not seem very likely, and if Y-DNA H2 had more than a trace presence in a population that was a source of the Y-DNA R1b expansion in Western Europe, Y-DNA H2 would probably be much more common in Europe today than it is in fact, unless there were only one or two individuals with Y-DNA H2 in the entire founding population of the group that caused Y-DNA R1b to expand in Western Europe.


Maju said...

Andrew: nonsense. Too many instances of nonsense plaguing a prejudiced analysis in all aspects, that is ignoring the real data (did you ever even read what I wrote about Basque ancient mtDNA?) I would discuss them one by one but feel that it will be worthless and also I'm too tired to bother right now.

bellbeakerblogger said...

I think it's a good question whether it's possible for a y-chromosome to be completely disassociated from its original ancestry when it comprises the majority frequency of a derived population. (Beyond individuals)

I believe that this is the case in highly refined domesticates where you have super-studs and complex back breeding.

So the question is whether R1b could exist in the Atlantic at a decent frequency and be stripped of its Eastern components? I think it's possible but there's no way to tell with no Atlantic dna

The other thing is that the particular R1b of modern Western Europe may have reached it from another place 1) VFR direct from Syria to South Eastern Iberia in its Early Chalcolithic (culturally this had been proposed early on)
2) with a particular Western epi-Cardial as ancestral to Beaker as erroneously proposed by Bosch-Gimbura
3) North Africa in 3500 along both sides of Iberia

Time to catch more arrows for the last one.

Maju said...

BBB: Y-DNA lineages can become almost completely disassociated from their autosomal and mtDNA original associates indeed. Native American Q1 and Finno-Ugric N1 are good examples. Or probably also Chadic R1b-V88 in Central Africa, etc. Notice that I said "almost" and that's because even in those extreme cases some lesser trail can still be found in autosomal and mtDNA data. So not quite completely.

However what it does not become disassociated is with an archaeological and cultural trail, including probably language. It also seems to happen in peculiar founder effect or low population density situations and not by mere dripping. What we see in all those cases is incorporation of local mtDNA and, with it, autosomal genetics until a point when the original ones associated to the Y-DNA become almost invisible.

In the case of Western R1b:

It is not a mere matter of "frequency". S116/P312 has a geographic/phylogenetic internal structure ("geostructure" for short) that is exclusively Western with apparent center in modern France, rather to the South. This alone is crucial to understand. Probably its Nordic "brother" U106 also has a NW origin but the matter awaits improved research.

In this sense all the rest of R1b upstream of L11 is as important for West Europeans as is Iranian Q* relative to Native Americans or Yunnan's N* relative to Finnish. You take notice of that remote relation and move on because it is precisely that: remote and therefore not too relevant.

The fact that between the M269 and the Western nodes (S116 and U106) there is only a thin trail does not help the case of trying to mix these two glasses with the whole bottle over the shelf where they ultimately (but quite trivially) came from. The question is what happened to these two western "glasses" afterwards. And in the case of S116 the story is clear: the glass was originally in Southern France (roughly) and its content was mostly divided in three other glasses: one for the Brits/Irish, another for the Swiss/Italians and another for the Basques/Iberians. The small amount that remained in the original glass (unclassified S116*) seems to be shared between Basques and Irish.

How does this relate to West Asia, Eastern Europe, Africa or Uyghuristan? Well, there's a bottle over the shelf, but then you're not looking anymore at this glass and how we poured them in various other smaller glasses. You are just diverting the issue pointlessly: is that attention deficit disorder?

I'm really beginning to lose my patience with all this, honestly. S116 is sooo obvious but surely not if you're looking elsewhere. Someone hide that stupid bottle please!

Whatever happened between the bottle and the glass is just a thin trail of some M269 → L11 "asterisk" individuals between West Asia and Western Europe. It was fixated in no population (at least not one that survived), it was just another haplogroup among many. It is as relevant as G2a, I2a or E1b. ONLY after the coalescence of S116 (and U106 in the North) things became interesting.

So focus, please.

bellbeakerblogger said...

We'll I think this recent individual increases the probability that the Baalberg man was R something and the other one one as well (Rossen?)that was also ambiguous. So it does seem possible that by the MN these were established in some areas.
On the other hand though, if you look at the Isles the frequency in so high among natives that it must be either 1) reformed Native Mesolithic with trimmed mustaches 2) farmers with little foreign males 3) Beaker immigration. After weeding out Norman and Viking surnames, what percentage in Ireland or Wales isn't R1b, 4%?
But again, Globular Amphora, Baalberg, Rossen, Peterborough, Artenac, they may start popping up R1bs one after another. I will stand facing the corner.

andrew said...

FWIW, I've hypothesized that Artenac may be the immediate predecessor to Northern Iberian Vasconic people in the November 2011 post,as well as mentioning a potential role for the Rossen culture and the Chasseen culture. I have, for better or worse, conceptually lumped them together as Bell Beaker influenced.

"Native American Q1 and Finno-Ugric N1 are good examples. Or probably also Chadic R1b-V88 in Central Africa, etc."

Excellent examples.

"3) North Africa in 3500 along both sides of Iberia

Time to catch more arrows for the last one."

Yeah. I still think this theory is not supported by the evidence. At first look it has plausibility. But, the more you look at the detailed DNA evidence, the less plausible it becomes. The North African v. European PCA in this study's Supplemental Materials is something that is particularly damning for that hypothesis.

Maju said...

Artenac is related but it is a more northernly culture. AFAIK there is no Artenacian in Iberia, even if Artenacian is clearly somehow related to Iberia (Bell Beaker, "modest" Megalithism, bow & arrow weaponry, material exchanges...) The Iron Age Basque ethno-cultural area is actually defined in the negative by the Celtic "siege", first by the South and East (broken in the East by the Iberianization of Catalonia c. 550 BCE) and then in the South (Celtiberians) and North (La Tène Gauls). It's surely similar to what caused the Ligurian genesis: defensive need of unity.

Basque should not be confused with Vasconic anyhow. Vasconic in those times surely corresponds to all the various populations of the mainline Neolithic (EEF and derived) of Aegean roots, i.e. most Europeans in the late Neolithic and early Chalcolithic.

"... a potential role for the Rossen culture and the Chasseen culture".

Rossen was smashed by Michelsberg (the most dynamic "funnelbeaker" culture), what is as much as saying that ENs were replaced by MNs in West-Central Europe, something we already know.

As for Chassey-La Lagozza, it may well be at the Ligurian roots but does not seem to have any particular relation to Basques, ancient or modern. In fact we have a post-Chassey sample known as Treilles and it is pretty much EN/EEF and not yet MN at all.

Chris Davies said...

"The North African v. European PCA in this study's Supplemental Materials is something that is particularly damning for that hypothesis."

The reason I think is because the Sardinians and early European farmers aren't necessarily *Berber-shifted*, but are Sudanic-shifted [like Dinka] - this is the so-called 'Basal Eurasian'.

I think that Nilo-Saharan, Chadic, and Cushitic populations, [and potentially also Fulani], may have more of this type of shift - but many Berbers do not [some likely do].

andrew said...

As Razib also pointed out, any similarity is more likely mostly due to Fertile Crescent Neolithic peoples surging both North to Europe and West Asia, and South to Egypt and beyond.

Chris Davies said...

Present-day Sardinians are a close fit for early European farmers. Lazaridis and Skoglund papers show a migration edge from Dinka into Sardinia, Stuttgart, Gokhem. Present-day Sardinians have several different clades of haplogroup E and A which are not present in Fertile Crescent populations but occur in Africans. And HLA haplotypes in Sardinia and Corsica which clearly derive from African populations and are either low frequency or non-existent in the fertile crescent. While I agree that 'basal Eurasian' (which is likely African-derived) has diffused from the Fertile Crescent, I am certain that the high levels of EEF/basal Eurasian in Sardinia and Sicily (Europe's highest levels in fact) is only due in part to Fertile Crescent origins but is due in greater part to the genetic impact of admixture from a Saharan population in prehistory. Anyway, Razib believes that EEF = WHG + 'basal Eurasian'; and Berbers = 'paleo-North African' + 'basal Eurasian'. Which is probably correct.

Maju said...

"Razib believes that EEF = WHG + 'basal Eurasian'; and Berbers = 'paleo-North African' + 'basal Eurasian'. Which is probably correct".

Very clearly incorrect, IMO.

EEF = West Asian of "Arab", "southern" or "lowlander" type + lesser NE African (Basal Eurasian) + UHG (~WHG).

NW African = minor Aterian (OoA epoch) layer (more apparent in some isolate populations) + Paleoeuropean (proto-WHG, Oranian layer) + NE African with ancient West Asian admixture (Basal Eurasian, Capsian layer) + minor recent West African. Most probably because the studies are a bit limited but that's what Henn and myself have approximated in separate analyses.

On one side, there is a strong European, probably Paleo-Iberian, component in NW Africa that cannot be ignored. On the other, "lowlander" West Asian cannot be just described as "Basal Eurasian" (it goes against the grain of Lazaridis' definition).