Sardinian Population Genetics

Why Care About Sardinian Genetics?

The population genetics of Sardinia are notable mostly because it is the population that is the closest modern match to the population genetics of the Neolithic era first farmers of Europe. 

In large part because it was an island in the relatively warm Mediterranean sea, it was basically unpopulated prior to the Neolithic Revolution, and experienced far less Middle Neolithic to Early Bronze Age change in its populations than Continental Europe, due to migration from the Steppe. 

This was from both Indo-European initially in the form of the Corded Ware Culture, from the Bell Beaker culture (of uncertain linguistic affiliation) and from later linguistically Indo-European and Uralic migrations into Europe. 

Findings Of The New Large Sample Size Study

A new study with a larger sample size than its predecessors has found that the mountainous regions of the island have been particular stable population genetically since it was settled by first wave farmers of the Neolithic era. 

As in most cases in Europe, new migration to Sardinia from the mainland appears to have been male biased, even before the Bronze Age.

Why Is There A Basque Affinity In Sardinian Population Genetics?

It isn't entirely clear if the stronger genetic affinity of this Sardinian highlands population to the Basque population is due to a greater proportion of hunter-gatherer and Neolithic ancestry, or is due to demic Bell Beaker impacts on Sardinia. 

One way to model the population genetics of Western Europe is to see four primary layers: a Mesolithic layer of Western European hunter-gatherers, a first wave Neolithic farmer layers, a Bell Beaker layer (dating to the early Bronze Age), and an Italo-Celtic-Germanic Indo-European layer (dating to the early Iron Age). 

In this model, the Basque have contributions from the first three layers, while the rest of Western Europe has contributions from all four layers. A reduced "Bronze Age Steppe" component in Sardinians and Basque people could reflect the absence of the fourth layer but not the third, or could reflect a reduced third and fourth layer contribution.

This study, however, does not appear to really consider this possible model and instead assumes three layer model without distinct Bronze Age and Iron Age Steppe migrations, even though the Y-DNA data seems to support such a model.

The body of the paper states regarding the Basque affinity:

Due to its smaller long-term effective population size (Figure 5A), Sardinia is expected to have undergone accelerated rates of genetic drift. To correct for this when measuring similarity to other mainland populations, we used “shared drift” outgroup-f3 statistics (Raghavan et al. 2014), which are robust to population-specific drift. Using this metric, we find the Basque are the most similar to Sardinia, even more so than neighboring mainland Italian populations such as Tuscany and Bergamo (Figure S6A, S6B). 

This relationship is corroborated by identityby-descent (“IBD”) tract length sharing, where among mainland European populations, French Basque showed the highest median length of shared segments (1.525 cM) with Arzana (Figure S7). 

We also tested the affinity between Sardinians and Basque with the D-statistics of the form D(Outgroup, Sardinia; Bergamo or Tuscan, Basque). In this formulation, significant allele sharing between Sardinia and Basque, relative to sharing between Sardinia and Italian populations, will result in positive values for the D-statistic. We find that Sardinia consistently showed increased sharing with the Basque populations compared to mainland Italians (|Z|> 4; Figure S6C), and the result was stronger when using the Arzana than Cagliari sample (DARZ = 0.008 and 0.0096, DCAG = 0.0072 and 0.0087 for French Basque and Spanish Basque, respectively). In contrast, sharing with other Spanish samples in our dataset was generally weaker and not significant ( |Z| < 3.5; Figure S6C), suggesting the shared drift with the Basque is not mediated through Spanish ancestry. . . . 

We generally find that Sardinians have the highest observed levels of shared drift with early Neolithic farming cultures and low levels of shared drift with earlier hunter-gather samples (Figure 6A, Figure S9, Table S5), consistent with previous reports. Surprisingly though, when examining allele sharing within Sardinia, we found that both ancient Neolithic farmer ancestry and pre-Neolithic ancestry are enriched in the Gennargentu-region. 

First, we find that shared drift with Neolithic farmers and with pre-Neolithic hunter-gatherers is significantly correlated with the proportion of “Gennargentu-region” ancestral component estimated from ADMIXTURE analysis, while shared drift with Steppe pastoralists has a weak negative correlation with Gennargentu-region ancestry (Figure 6B). 

Second, using supervised estimation of ancestry proportion based on aDNA (Haak et al. 2015), we estimate higher levels of Neolithic and preNeolithic ancestries in the Gennargentu region and higher levels of Steppe Pastoralist ancestry outside the region (Figure S10). Finally, calculations with Patterson’s D-statistics of the form D(Outgroup, Ancient, Ogliastra, Non-Ogliastra) also support increased sharing with Neolithic and pre-Neolithic individuals, but not post-Neolithic individuals from the Steppe, in the Ogliastra samples (D = -0.0037 and -0.0042, |Z| = 7.4 and 7.9 when aDNA sample = Stuttgart and Loschbour, respectively; D = -0.0009, |Z| is not significant, when aDNA sample = Yamnaya) . 

Together, these results suggest that while at the regional-level Sardinia appears to harbor the highest amounts of Neolithic farmer ancestry and very little of the pre-Neolithic hunter-gatherer or Bronze Age pastoralists ancestries, there exists within-island variation. Specifically, we find that with increasing level of isolation (represented by increasing level of the Gennargentu ancestry), there are greater Neolithic farmer and pre-Neolithic hunter-gatherer ancestry, while the less isolated Sardinians have a stronger signal of ancestry from the Steppe pastoralist source. . . . 

We also found Sardinians show an impressive signal of shared ancestry with the Basque, in terms of identity-by-descent tracts and the outgroup f3 shared-drift metric. Such a connection is consistent with long-held arguments of a connection between the two populations, including claims of Basque-like non-Indo-European language words among Sardinian placenames (Blasco Ferrer 2010). More recently the Basque have been shown to be enriched for Neolithic farmer ancestry (Lazaridis et al. 2014, Gunther et al. 2015) and Indo-European languages have been associated with Steppe population expansions in the Bronze Age (Allentoft et al. 2015, Haak et al. 2015). These results support a model in which Sardinians and the Basque may both retain a legacy of pre-Indo-European, Neolithic ancestry (Gunther et al. 2015). . . . 

The high frequency of particular Y-chromosome haplogroups (particularly I2a1a2 and R1b1a2) that are not commonly affiliated with Neolithic ancestry is one challenge to a model in which Sardinian principally has Neolithic ancestry. Whether such haplogroup frequencies are due to simple genetic drift and/or a signal of sex-biased demographic processes has been an open question. By carrying out X versus autosome comparisons we uncovered evidence of sex-biased patterns of ancestry in Sardinia, and found an enrichment of Neolithic and preNeolithic ancestry on the X-chromosome across all groups on the island. 

The Abstract and Citation For The Paper

The population of the Mediterranean island of Sardinia has made important contributions to genome-wide association studies of traits and diseases. The history of the Sardinian population has also been the focus of much research, and in recent ancient DNA (aDNA) studies, Sardinia has provided unique insight into the peopling of Europe and the spread of agriculture. 

In this study, we analyze whole-genome sequences of 3,514 Sardinians to address hypotheses regarding the founding of Sardinia and its relation to the peopling of Europe, including examining fine-scale substructure, population size history, and signals of admixture. 

We find the population of the mountainous Gennargentu region shows elevated genetic isolation with higher levels of ancestry associated with mainland Neolithic farmers and depleted ancestry associated with more recent Bronze Age Steppe migrations on the mainland. Notably, the Gennargentu region also has elevated levels of pre-Neolithic hunter-gatherer ancestry and increased affinity to Basque populations. 

Further, allele sharing with pre-Neolithic and Neolithic mainland populations is larger on the X chromosome compared to the autosome, providing evidence for a sex-biased demographic history in Sardinia. These results give new insight to the demography of ancestral Sardinians and help further the understanding of sharing of disease risk alleles between Sardinia and mainland populations.

Charleston et al., Population history of the Sardinian people inferred from whole-genome sequencing, bioRxiv, Posted December 7, 2016, doi: http://dx.doi.org/10.1101/092148 via Eurogenes.

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

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.

Scattered Data and Preliminary Musings About Basque Prehistory

I've been digging around the archaeological cultures and archaeology of Basque prehistory and the immediate vicinity.

Particularly interesting have been the links between Artenacian culture and the closely related Aquitani people ca. 2600 BCE (also here and here), geographical extents suggested by toponymns (also here), megalithic monuments (and also here), nearby and potentially relevant cultural complexes such as the Beaker culture, the Rossen culture and the Chasseen culture, and the culture of the Stele people (also here, and here), the role of RH negative genetics, lactase persistance genetics, Y-DNA genetics for R1b1b2, linguistic landscape considerations, outstanding academic issues in the Upper Paleolithic transition, and more.

Either in one of the links above, or somewhere else that I'll have to track down, was some discussion of the timing of the arrival of cattle herding and megalithic structures in Basque country (which was quite a bit later, at least in the case of the megalithic structures, than the rest of Iberia and close in time to the rise in cattle herding and the end of the Artenacian). The Artenacian also is quite a good fit to some guesses I'd made based on latase persistance genetics before I learned that it even existed, in terms of timing, geographical extent, and direction of migration, which is just the kind of thing you like to see happen.

The Stele people story unfolds close in time, and importantly pre-Bell Beaker, but post Cardial Pottery in Iberia. But, given that the oldest evidence of the Bell Beaker people is ca. 2900 BCE in Iberia, about three centuries after the Stele people are seen in Sardinia, it is hard to make out if they are proto-Bell Beaker people (and hence arguably Indo-Europeans, which would be consistent with their place of origin), or are a layer that comes earlier and integrates with the later megalithic culture.

Then come the complexities and question marks.

We have basically no pre-Neolithic Y-DNA samples from Europe that are R1b, and indeed, basically no pre-Neolithic Y-DNA samples anywhere in Europe. We do have lots of R1a from ca. 2000 BCE and a bit older in Eastern Europe and Central Asia, but that doesn't resolve the R1b timeline much at all.

To answer the question of Basque origins, you have to say what you mean by the Basque people (which in turn flows from why you want to know the answer to that question) and entertain the possibility of phenomena like language shift due to elite dominance, demic migration accompanied by admixture with existing populations (possibly on a sex biased basis), Iberian demic interactions with North Africa in each direction, transfers with technology and cultures with neighboring cultures (e.g. bow wave models in which cultures neighboring Indo-Europeans that have adopted IE technologies and cultural features migrate with expanding Indo-Europeans at their heels), the relationship of the Corded Ware (which the Artenacian held at bay in Western Europe) to proto-Indo-European culture, and the deeper roots of the immediate antecedents to the Basque people.

I haven't digested the information to my satisfaction enough to write a post reaching strong conclusions. My preliminary avenues of inquiry are (1) to see the time period from about 3500 BCE to 2500 BCE (and especially 3000 BCE to 2600 BCE) as central to the ethnogenesis of the Basque people in the region where they are found today, (2) to see them as arriving by land from what is now France with the Aternican which is probably an important source of the Y-DNA R1b and lactose persistence genes in modern Basque, (3) to see significant amounts Paleolithic era mtDNA as the product of admixture between proto-Basque and indigeneous hunter-gatherer or Cardial Pottery populations, (4) to see Basque society at a remote level going back to the early Neolithic as probably more culturally derived from the LBK than the Cardial Pottery Neolithic, (5) to see the Basque people as either derived from or significantly influenced by upon contact with the Atlantic megalithic culture in France, (6) to conclude that anything the Basque and Indo-European populations of Iberia differ in is probably largely attributable to the later waves of Indo-European migration (Bell Beaker, Celt, Roman and Reconquest), (7) to see strong suggestive connections between the Basque and Sardinians that nevertheless remain cryptic, and (8) to have continuing uncertainty regarding the linguistic affiliations of the Bell Beaker people.

Scenarios in which the Bell Beaker people are culturally Pre-Vasconic, with the Celts representing the first Indo-European migration wave, and those in which the Bell Beaker people are culturally Indo-European and are the first Indo-European migration wave both have some plausibility. Bell Beaker integration into megalithic culture by 2500BCE-2400 BCE argues for a Pre-Vasconic reading, but there are arguments that Celtic origins are too young to account for the spread of Indo-European culture. The Bell Beaker geographic distribution is a decent fit to the Old European toponym area. Bell Beaker is sometimes seen as having Funnelbeaker culture antecedents (TRB) which was a clearly megalithic culture, and at other times seen as having antecedents in Corded Ware to whom the Aternicians were opposed. Neither of these theories, however, is a fit to the oldest Bell Beaker wares being found in Iberia ca. 2900 BCE.

The Bell Beaker physical anthropology, a discipline with a good track record of predicting genetic links in these kinds of situations carries its own implications:

Historical craniometric studies found that the Beaker people appeared to be of a different physical type than those earlier populations in the same geographic areas. They were described as tall, heavy boned and brachycephalic. The early studies on the Beakers which were based on the analysis of their skeletal remains, were craniometric. This apparent evidence of migration was in line with archaeological discoveries linking Beaker culture to new farming techniques, mortuary practices, copper-working skills, and other cultural innovations. . . . Margaret Cox and Simon Mays sum up the position: "Although it can hardly be said that craniometric data provide an unequivocal answer to the problem of the Beaker folk, the balance of the evidence would at present seem to favour a migration hypothesis."

Non-metrical research concerning the Beaker people in Britain also cautiously pointed in the direction of immigration. Subsequent studies, such as one concerning the Carpathian Basin, and a non-metrical analysis of skeletons in central-southern Germany, have also identified marked typological differences with the pre-Beaker inhabitants.

Jocelyne Desideri examined the teeth in skeletons from Bell Beaker sites in Northern Spain, Southern France, Switzerland, the Czech Republic and Hungary for her thesis. Looking at inherited dental traits, she found that only in Northern Spain and the Czech Republic were there demonstrable genetic links between immediately previous populations and Bell Beaker populations. Elsewhere there was a genetic discontinuity.

These data points are suggestive of a possible identity in time and space between the Stele people and the proto-Bell Beaker people and with an Iberian origination of the Bell Beaker people. An Iberian origin Bell Beaker people also puts it in opposition to Corded Ware and aligned with the megalithic people, in accord with the alignment of the later Aternicians. From the same source: "A review of radiocarbon dates for Bell Beaker across Europe found that some of the earliest were found in Portugal, where the range from Zambujal and Cerro de la Virgen ran between 2900 BC and 2500 BC, in contrast to the rather later range for Andalusia (between 2500 BC to 2200 BC)."

My preliminary inclination is to see early megalithic culture (with separate Cardial Pottery and LBK branches), intermediate copper to bronze age Bell Beaker culture (recently found to have extended as far as the Baltics at its peak, which is the furthest extent of Old European topomyns) that was fusional with local neolithic cultures and has a meaningful demographic component, and Indo-European culture (which was culturally disruptive, but perhaps not as strongly disruptive demographically) as three separate migration waves, with Bell Beaker as a non-Indo-European copper age culture more prone to integrate with the megalithic culture, and the Urnfield culture as the first Indo-European culture that fully displaced the prior cultures of the region in Western Europe.

In the North, I am inclined to see Indo-European culture date not to the Funnelbeaker culture or the Bell Beaker, but to the Nordic Bronze Age (starting sometime 1700 BCE-1200 BCE) which is seen as synonymous with proto-Germanic in this frame, and the Urnfield Culture (as well as the Hallstat, La Tene, and Celtic cultures that grew out of it) would likewise have been an Indo-European one. This would also suggest that the Central European Tumulus culture that preceded the Urnfield Culture, and perhaps the Unetice culture before it, 2300-1600 BC, may also have been Indo-European, but casts doubt on the Unetice's culture's origins in the Bell Beaker culture. But, a Unetice culture source in the Southern Corded Ware culture seems like a better geographic fit and a more fitting to an Indo-European source culture.

But, this is problematic as well in a non-Indo-European read of Bell Beaker, because there is a good argument that the Stele people have roots in the Kemi oba culture very close to the kurgan proto-Indo-European urheimat.

On the other hand the Kemi oba culture had strong trade and cultural ties with, was contemporaneous with, and was adjacent to the Maykop culture, which eminent Indo-Europeanist Mallory identifies as probably non-Indo-European in unity with the Northwest Caucasus mountain region. In this view, the Kemi oba people would have had very little genetic contribution from the Northwest Caucasus but would have experienced language shift arising from cultural dominance arising from more advanced Northwest Caucasian technologies in areas like copper working.

In this scenario, then, we have Northwest Caucasian languages, with genetic and culture infusions from the Steppe, and ties to the some of the earliest centers of the copper industry. The Stele people migration was from about 3200 BCE to 3000 BCE, so the link that this would imply between Basque and Northwest Caucasian languages (influenced by existing substrates for a relatively fusional wave of migration), would be about as thin as the link between the most divergent branches of the Indo-European language family (e.g. Hindi and English) without the extant and documented intermediate branches to reveal the connections, particularly as the Basque people may have had origins from Southern Iberia to France and then back to Northern Iberia, rather than directly from Southern Iberia to Northern Iberia.

Rather than suggesting a common source of Indo-European and Basque languages, the two cultures might have represented similar technological complexes based on technological exchanges between two cultural and linguistic groups, the Indo-Europeans to the North, and the Caucasians to the South.

While it seems presumptuous, given that we know that there was a substantial demographic component to the Bell Beaker migration, it wouldn't be beyond plausibility to imagine that R1b, rather than being a Paleolithic or Indo-European marker (and how could R1b be predominant in the Basque if it was an Indo-European marker?), R1b might be a proto-Vasconic Bell Beaker marker, as its distribution fits the megalithic cultural region into which it integrated and spreads further at lower frequencies into areas were the Bell Beaker culture was in place for a shorter time period. This seems even less presumptuous when one digs down into accounts of just how primitive an existence pre-Bell Beaker Neolithic farmers in Western Europe endured. It wouldn't have taken much of an edge to become leading figures in this simple small villages.

It also provides a path for R1b with a likely origin around the Caspian Sea or Central Asia to make its way to Western Europe. This interpretation would leave, by process of elimination Y-DNA haplogroups I and T as the most plausible candidates for pre-Neolithic Y-DNA haplogroups, perhaps with T in place pre-LGM in Europe, and I in the Epipaleolithic or early Neolithic.

This could also help explain some of the cultural ambiguity of the Bell Beaker culture. Given that both Bell Beaker and Indo-Europeans are directly borrowing metallurgy and farming technologies and some other parts of the same cultural complex from each other before their respective expansions, it makes sense that artifacts found by archaeologists alone would have a hard time revealing an Indo-European or Vasconic affinity.

In this view, R1a really is an Indo-European marker and its relative absence from Western Europe is a direct consequence of the fact that the Vasconic R1b carrying Bell Beaker people beat them to it and delivered an equivalent technological package that allowed them to hold off the Indo-Europeans for a millenium. This scenario is also a good fit to the absence of early Neolithic R1b in Western Europe.

One down side of this interpretation of course, is that it suggests that the combined forces of the Indo-European and Vasconic expansions have utterly obliterated any distinguishable linguistic traces of either Upper Paleolithic Europeans or of early Neolithic Europeans from either the LBK or Cardial Pottery branches of that Neolithic expansion. Any trace of Upper Paleolithic languages would have become substrates of early Neolithic languages, which in turn would have become substrates in Vasconic and Indo-European languages.

In Basque, the lack of any way to parse substrate from superstate (except perhaps by remote comparisons to proto-Northwest Caucasian and purging Indo-European borrowings, that would still be highly conjectural) makes the task impossible. In Celtic and Germanic languages, the intervening Vasconic substrate would muddy the early Neolithic and hunter-gather substrates. In areas where Indo-European languages directly replaced early Neolithic and hunter-gatherer languages leaving perhaps identifiable substrates, subsequent Slavic language expansions would probably have obscured those substrates in the previous probably Indo-European languages of that region.

Basque, in this narrative, would be essentially equal in age to Indo-European, although it would have been the more ancient language in Western Europe and might be more directly descended from Northwest Caucasian than Indo-European is from its progenitors. My own inclination is to see Proto-Indo-European itself as a likely creole of an ancient variety of the Uralic language and the probably very different language of the LBK farmers.

This would also tend to place the origins of African R1b-V88, associated most strongly with the Chadic people, which is quite basal in the R1b clade, somewhat earlier in time than the Stele people migration of ca. 3200 BCE to 3000 BCE and to associate its likely place of origin as the Southern Steppe. Conveniently, this would put this migration before the invention of writing in Egypt and Sumeria, which would help to explain why there is no historical record of this migration.

It is also tempting to associate the G2a rich people of the Caucasus with the G2a rich source populations of early Cardial Potery and LBK Neolithic populations. Thus, the Bell Beaker people may have been linguistically separated linguistically from the older Neolithic populations of the region by only 1500-2500 years, little more than the separation of the Romance languages or the Sanskrit derived languages of India are from each other. Greater similarity in language and cultural roots between the Bell Beaker people and their Neolithic predecessors may have faciliated their greater fusion with the local people, as opposed to the far more complete lack of cultural continuity between Indo-European culture and its predecessor.

Y-DNA Expansions And Polygamy

Both Europe and Africa have dominant Y-DNA lineages that are probably associated with populations that had major expansions sometime in the Holocene era.

In principle, Y-DNA lineages can expand much more rapidly than mtDNA lineages. The maximum number of children that a group of women with identical mtDNA can have in a lifetime on average is much lower than the maximum number of children that a group of men with identical Y-DNA can have in a lifetime.

But, how often does this happen and does the historic anthropology support the suggestion from the genetic evidence.

Commentators like Razib Khan (who has migrated Gene Expressions to a new website as of yesterday) speak of androcide - the near complete exclusion of substrate population men from having male children who go on to reproduce themselves as at the heart of this process, while substrate population women are more often appropriate by men of incoming superstrate populations via polygyny of the equivalent (multiple women per man).

Both history and ethnography document mass population collapse in concert with an androcide of the Amerindians. By this, I mean that European males took Amerindian women as concubines, and engaged in de facto polygyny in the New World.

Polygyny remains common even today in much of Africa.

The evidence is sketchier in prehistoric populations like the Bell Beaker and Corded Ware peoples whose civilizations were superseded by later populations before it was possible for those civilizations to be documented in writing.

The alternative possibility that there was merely segregation of people into a rapidly expanding largely monogamous superstrate population and a stagnant largely monogamous substrate population, with some level of bride exchange between the two populations, is also possible.  In that scenario, bride exchange brings substrate women into the rapidly expanding population mix, while bride exchange in the other direction is hard to discern from genetic evidence in modern populations since it is hard to distinguish from superstrate women who remained in their home population and because exchanged brides would make up a small part of a stagnant population whose descendants are becoming an increasingly small proportion of the total population.

If the growth rate for the expanding population is much greater than for the segregated stagnant population, it is hard to distinguish a polygyny and androcide scenario from a scenario with monogamy and moderate levels of equitable bride exchange from each other on the basis of genetic evidence alone.

It ought, in principle, to be easier to distinguish a scenario in which migration and conquest was overwhelmingly by men (e.g. an invading army) from one that is a mixed gender migration - as there would be no non-indigenous mtDNA lines introduced in great frequencies while there was a great change involving a newly introduced group of Y-DNA lineages.  A male dominated conquering population could support a scenario of androcide (often literally with indigeneous men of reproductive age killed in war and their widows appropriated by the conquerors) without significant polygyny.

There is evidence of polygny and the equivalent in European cultures at the dawn of the historic era like the Celts (alluded to in several comments here and also inferred here and here).  This may have also been present in early Bell Beaker and Indo-European populations, and on the to do list is an effort to track down better authority on the extent to which it is likely that there was much polygyny in those prehistoric or ancient cultures.  There is also legendary history evidence of androcide in wars of conquest.

But, it is hard to tell.  Polygny leaves a far fainter trail of artifacts in archaeological digs than technologies like metal use or food production technologies.

A cursory search reveals that polygyny was practiced, at least among elites, in almost all of the cultures known at the dawn of the historic era in Eurasia, although sometimes secondary wives or concubines were permitted only if they held slave status.  The Mycenean Greeks and Trojans, the ancient Macedonians, the Hittites, the Celts, early Nordic pagans, the first century Merovingians and Carolingians, the early Jews, the Roman, the pre-Christian Slavs, Muslims and pre-Muslim pagans, the pre-Communist Chinese, the early Aryans (a controversial point to some), the pre-Zoroastrian religion of Iran, and the elites of ancient Egypt.  There is genetic evidence of prior polygamy in French Basques and many other modern populations.  Also, monogamous societies are a decided minority among anthropologically attested cultures (only about one in six fit that description).  The status of Avesta era Iran is unclear and may have been monogamous and polygamy was expressly forbidden by the legal code of the Visigoths as of the 1st century according to Tactius (although not by any other the other Germanic tribes of that time).  There is evidence of religious fertility rights involving annual divine marriages of Sumerian kings to temple priestesses in addition to their ordinary wives, but evidence regarding polygamy per se in ancient Sumeria is sketchy.

Still, as a Bayesian prior, we would strongly expect Chalcolithic and Bronze Age cultures of Europe to have been polygynous rather than monogamous.  Monogamy is a practice that seems to have been exceedingly rare as a universal rule of any culture prior to the Iron Age.  An academic case has been made for Proto-Indo-European monogamy, but I personally find it implausible given the prevalence of polygyny in the earliest attested Indo-European cultures.

However, if polygny was a practice largely restricted to kings and chiefs, rather than being widespread, the number of wives was not too immense, and the scale of chiefdoms was fairly large in this time period, the cumulative impact of this practice on the gene pool may have been fairly modest and this may not have been an important factor in explaining the expansion of Y-DNA lineages.  The considerable diversity of polygynous practices makes it hard to model.

Polygny isn't absolutely necessary to estimates of Y-DNA R1b expansion rates in Europe which, while high, were not so high that they couldn't be consistent with a sustained period of largely monogamous expansion by a population without the male selective androcide and without the polygyny that necessarily goes hand in hand with that pattern.

The expansion of mtDNA H in Europe apparently coincident to the R1b expansion (in both cases from a source in Iberia whatever prior sources those expansion had, perhaps the Bell Beaker population*) did not bring mtDNA H to quite as dominant a position a Y-DNA R1b in Western Europe, but without knowing what the mtDNA breakdown of the source population for the R1b expansion in Western Europe was and having detailed mtDNA haplogroup subtype data for non-mtDNA H haplogroups in Western Europe (both available to some extent but beyond the scope of this post), it is hard to know how much of the Y-DNA R1b expansion involved displacement of indigenous men and how much involved expansion of multiple mtDNA lineages in the source population complemented by bride exchange between growing and stagnant communities on an equitable basis.

UPDATED:

The 2008 analysis by Hammer et al. of sex differences in breeding ratio in six populations including the French Basque is most on point (internal references to figures and other papers of the paper omitted without acknowledgement of the omissions):

Our method uses a population genetic model (i.e., the coalescent) to account for the inherent uncertainty in estimating diversity and divergence rates from sequence data. For three out of six populations (Basque, Melanesians and Mandenka), the 95% confidence intervals for the ratio of X-linked and autosomal effective population sizes does not include 0.75 [Ed. the expected value in a monogamous population model.] (p = 0.001, 0.005 and 0.030 for the Basque, Melanesians and Mandenka, respectively). One interpretation of these results is that there is strong evidence for an unequal female and male Ne in at least three of our six populations, with estimates of the breeding sex ratio (i.e., the effective size of females to males) ranging from 2.1 in the San to 12.5 in the Basque. If the observed differences in nucleotide variability on the X chromosome and autosomes are caused by long-term (demographic) processes, then the estimates of N_x/N_a presented will be highly correlated due to shared population history. When we use the intersection of all six confidence intervals (0.87–1.02) to estimate the range of N_x/N_a values that are consistent with the data from all six populations, we estimate the range of the breeding sex ratio to be 2.4–8.7. We also note that even with a conservative Bonferroni correction, a 1:1 breeding sex ratio is rejected in two out of six populations.

We also employ a separate method for estimating the breeding sex ratio in each population that does not allow for intra-locus recombination but does permit independent mutation rates across loci. This method produces similar results to those described above, with estimates of the ratio of female to male effective population size ranging from 1.8 in the San to 14.0 in the Basque. We interpret this as additional evidence that the unusual patterns observed in our data are real and require explanation.

Naively, this suggests that the ancestors of the French Basque were, at some point, massively polygnyous.  There is an alternative explanation, however.

This is that while the pairings in some ancestral era of the French Basque were mostly of French Basque men and indigeneous women, that rather than being massively polygamous, the French Basque men were simply highly inbred and closely related. In other words, so many of the men were genetically identical or nearly so that they look in effective population size genetic statistics, like a single man.

This could happen, for example, if a Basque founding population were composed almost entirely of a group of migrating men who were all part of a patrilocal extended family group that had expanded fairly rapidly elsewhere in their insular community of origin, prior to migrating together to Iberia, while the indigeneous women of the pre-existing megalithic culture were highly diverse due to many generations of fairly long distance bride exchanges prior to the Basque male founding population arrived.

For example, suppose the early copper working in Central Europe or West Asia were organized on a patrilocal clan basis, so that only sons of existing (and prosperous and expanding) copper workers could become copper workers.  Then, suppose that word of rich unexploited copper resources in Southern Iberia reached them.  Large numbers of these men from the copper worker extended family might leave, en masse, in a Gold rush style, male dominated copper rush to Iberia that in this case proved to be not a myth, but a reality that led to multiple generations of migration to Iberia from the same patrilocal clan as those who arrived there prospered and married local women.

While the founding Bell Beaker population which I claim is proto-Basque probably was polygamous to some extent like other pre-Iron Age cultures, it seems highly unlikely to me that Basque men had, on average, for several hundred years, a dozen or so wives each.

After all, while the growth rate of the Western European specific clades of Y-DNA are phenomenal, they are merely at the very high end of what could be attained with a very fertile monogamous scenario.  Mildly polygamous family patterns (e.g. 1.5 or so wives per Basque man on average) fit the observed rates of R1b expansion much better than massively polygamous family patterns in which each man could conceivably have had as many as a dozen or more wives on average for ten or twelve generations in a row.

The patrilocal clan originating founding population scenario that I suggest seems like a more likely source for most of the disparity in effective male and female population sizes that is evident in French Basque DNA.

END UPDATE

* The old conventional wisdom was that the Bell Beaker people had predominantly a cultural impact on Western Europe and was largely a trading civilization with only modest demographic impact, while the subsequent Indo-Europeans may have had a major demographic impact.  Maju (who is blogging again after taking a hiatus for a while) has argued this case multiple times.  Ancient DNA evidence and modern population genetic studies seem to be favoring the reverse scenario - one with a major Bell Beaker demographic impact and a fairly shallow additional Indo-European wave demographic impact - although the ancient DNA data are still too fragmentary and inconclusive to definitively favor either scenario or some third scenario.  There is also some debate over the question of whether the Indo-Europeans may have themselves been Indo-European linguistically and culturally, a position that I consider a minority and disfavored view.  In my view, the Bell Beaker people were probably Vasconic (i.e. proto-Basque), and were themselves probably a second rather than an initial wave of farming people to expand in Western Europe.

The Latest Basque Genetics Paper

A new paper on Basque genetics is out with both ancient and modern samples, honing in on the geographically localized heterogeneity of Basque genetics even within Basque country in Northern Spain and Southeastern France. Razib blogs it here without any real commentary except from his readers. 

The paper shows that the Basques who speak the Basque language are indeed genetically distinct from their neighbors in Iberia and Southern France, with their immediately adjacent close neighbors, called "Peri-Basque" in the paper, being intermediate). 

The Basque lack the genetic contributions of the Iron Age Romans and from the Moors in the Middle Ages is largely found in other Iberians. It also shows strongly small scale geographic variation in Basque genetics from subregion to subregion of Basque country.

It does not appear to show any real Basque distinctiveness attributable to earlier eras, although the paper slightly hedges its bets on that score. The lack of pre-Iron Age genetic differentiation is a finding that I am skeptical of given prior publications and the historical and linguistic context, although I don't rule it out, out of hand. 

At least some Indo-European influences arrived in Iberia in the Bronze Age, not the Iron Age, and the Basque people obvious avoided cultural domination at that time, even though this paper's genetic analysis doesn't real reveal any genetic evidence in modern Basque people that this happened.

Certainly, the Romans are the source of all of the Romance languages spoken in Iberia today. But the pre-Roman linguistic character of Iberia is muddy, with Celtic languages present in or near parts of Iberia in pre-Roman times as well, for example, and the Bell Beaker phenomena originating there.

My ambivalence is, in part, because there is good reason based upon historical and linguistic evidence, to think that much more of Iberia was Vasconic in the pre-Roman era than it is today, with sister languages of the Basque language's going extinct in the face of Roman influence. It could be that influences distinct to Bronze Age Vasconic people are invisible because, apart from the Basque people, Iberian Vasconic people may have been thoroughly integrated into the Iberian general population in 85 or so generation that followed from Roman conquest until the modern samples were taken.

Another important conclusion reached, contrary to some of my conjectures in the past on the matter, is that there is no discernible Caucasian/Iranian farmer or Caucasian/Iranian hunter-gatherer ancestry found in the Basque population (or in the general Iberian population).

Image from Broushaki et al. (2016).

While it is well known that language shift can occur in a population whose population genetics are unchanged, the lack of a Caucasian/Iranian farmer genetic signal in Iberians still disfavors the possibility that Basque is a language that arrived in Iberian in the Copper Age/early Bronze Age from a Minoan/Hattic/Hurrian/Caucasian, pre-Indo-European Iranian, or Harappan source (all of which would be expected to have some of this genetic signal), or at least, that any migrants from these societies were few in number, making the likelihood that they brought about language shift in the Neolithic societies in which they arrived smaller. This component isn't entirely absent from Copper Age and Bronze Age Pontic Caspian steppe peoples (who did leave a discernible genetic impact on Iberians), but as Davidski notes in  recent post at Eurogenes, it was a quite dilute and minor component.

Instead, this data point tends to favor (by process of elimination) the hypothesis that that Vasconic languages are derived from the first farmers of Iberia and its vicinity, which in turn would have been derived from the languages of the Western Anatolian first farmers who expanded into Europe in the European Neolithic Revolution, primarily in a Northern Linear Pottery Culture and cultures derived from it, and in a Mediterranean coastal Cardial Pottery culture and cultures derived from it, which would be the most plausible source for most Iberian farmers. 

The paper is thinner on analysis and context that might be hoped given the depth of the literature on this particular matter, and does less to leverage its ancient DNA samples than it could, so I'll provide some context in this post to supplement it.

Iberia's Neolithic Revolution showed more internal regional variation than most of Europe, where most regions saw only a single wave of first farmers, did:

In the 6th millennium BC, Andalusia experiences the arrival of the first agriculturalists. Their origin is uncertain (though North Africa is a serious candidate) but they arrive with already developed crops (cereals and legumes). The presence of domestic animals instead is unlikely, as only pig and rabbit remains have been found and these could belong to wild animals. They also consumed large amounts of olives but it's uncertain too whether this tree was cultivated or merely harvested in its wild form. Their typical artifact is the La Almagra style pottery, quite variegated.

The Andalusian Neolithic also influenced other areas, notably Southern Portugal, where, soon after the arrival of agriculture, the first dolmen tombs begin to be built c. 4800 BC, being possibly the oldest of their kind anywhere.

C. 4700 BC Cardium pottery Neolithic culture (also known as Mediterranean Neolithic) arrives to Eastern Iberia. While some remains of this culture have been found as far west as Portugal, its distribution is basically Mediterranean (Catalonia, Valencian region, Ebro valley, Balearic islands).

The interior and the northern coastal areas remain largely marginal in this process of spread of agriculture. In most cases it would only arrive in a very late phase or even already in the Chalcolithic age, together with Megalithism. 

The location of Perdigões, in Reguengos de Monsaraz, is thought to have been an important location. Twenty small ivory statues dating to 4,500 years BP have been discovered there since 2011. It has constructions dating back to about 5,500 years. It has a necropolis. Outside the location there is a cromlech. The Almendres Cromlech site, in Évora, has megaliths from the late 6th to the early 3rd millennium BC. The Anta Grande do Zambujeiro, also in Évora, is dated between the early 4th and the mid 3rd millennium BC. The Dolmen of Cunha Baixa, in Mangualde Municipality, is dated between 3000 and 2500 BC. The Cave of Salemas was used as a burial ground during the Neolithic.

What was going on in Iberia in the Copper Age?

The Chalcolithic or Copper Age is the earliest phase of metallurgy. Copper, silver and gold started to be worked then, though these soft metals could hardly replace stone tools for most purposes. The Chalcolithic is also a period of increased social complexity and stratification and, in the case of Iberia, that of the rise of the first civilizations and of extensive exchange networks that would reach to the Baltic and Africa. The conventional date for the beginning of Chalcolithic in Iberia is c. 3000 BC. In the following centuries, especially in the south of the peninsula, metal goods, often decorative or ritual, become increasingly common. Additionally there is an increased evidence of exchanges with areas far away: amber from the Baltic and ivory and ostrich-egg products from Northern Africa.

The Beaker culture was present in Iberia during the Chalcolithic. Gordon Childe interpreted the presence of its characteristic artefact as the intrusion of "missionaries" expanding from Iberia along the Atlantic coast, spreading knowledge of Mediterranean copper metallurgy. Stephen Shennan interpreted their artefacts as belonging to a mobile cultural elite imposing itself over the indigenous substrate populations. Similarly, Sangmeister (1972) interpreted the "Beaker folk" (Glockenbecherleute) as small groups of highly mobile traders and artisans. Christian Strahm (1995) used the term "Bell Beaker phenomenon" (Glockenbecher-Phänomen) as a compromise in order to avoid the term "culture".

The Bell Beaker artefacts at least in their early phase are not distributed across a contiguous areal as is usual for archaeological cultures, but are found in insular concentrations scattered across Europe. Their presence is not associated with a characteristic type of architecture or of burial customs. However, the Bell Beaker culture does appear to coalesce into a coherent archaeological culture in its later phase.

More recent analyses of the "Beaker phenomenon", published since the 2000s, have persisted in describing the origin of the "Beaker phenomenon" as arising from a synthesis of elements, representing "an idea and style uniting different regions with different cultural traditions and background. "Archaeogenetics studies of the 2010s have been able to resolve the "migrationist vs. diffusionist" question to some extent. The study by Olalde et al. (2017) found only "limited genetic affinity" between individuals associated with the Beaker complex in Iberia and in Central Europe, suggesting that migration played a limited role in its early spread from Iberia. However, the same study found that the further dissemination of the mature Beaker complex was very strongly linked to migration. The spread and fluidity of the Beaker culture back and forth between the Rhine and its origin source in the peninsula may have introduced high levels of steppe-related ancestry, resulting in a near-complete transformation of the local gene pool within a few centuries, to the point of replacement of about 90% of the local Mesolithic-Neolithic patrilineal lineages.

The origin of the "Bell Beaker" artefact itself has been traced to the early 3rd millennium. The earliest examples of the "maritime" Bell Beaker design have been found at the Tagus estuary in Portugal, radiocarbon dated to c. the 28th century BC. The inspiration for the Maritime Bell Beaker is argued to have been the small and earlier Copoz beakers that have impressed decoration and which are found widely around the Tagus estuary in Portugal. Turek has recorded late Neolithic precursors in northern Africa, arguing the Maritime style emerged as a result of seaborne contacts between Iberia and Morocco in the first half of the third millennium BCE. In only a few centuries of their maritime spread, by 2600 BC. they had reached the rich lower Rhine estuary and further upstream into Bohemia and beyond the Elbe where they merged with Corded Ware culture, as also in the French coast of Provence and upstream the Rhone into the Alps and Danube.

A significant Chalcolithic archeological site in Portugal is the Castro of Vila Nova de São Pedro. Other settlements from this period include Pedra do Ouro and the Castro of Zambujal. Megaliths were created during this period, having started earlier, during the late 5th, and lasting until the early 2nd millennium BC. The Castelo Velho de Freixo de Numão, in Vila Nova de Foz Côa Municipality, was populated from about 3000 to 1300 BC. The Cerro do Castelo de Santa Justa, in Alcoutim, is dated to the 3rd millennium BC, between 2400 and 1900 BC.

It is also the period of the great expansion of megalithism, with its associated collective burial practices. In the early Chalcolithic period this cultural phenomenon, maybe of religious undertones, expands along the Atlantic regions and also through the south of the peninsula (additionally it's also found in virtually all European Atlantic regions). In contrast, most of the interior and the Mediterranean regions remain refractary to this phenomenon.

Another phenomenon found in the early chalcolithic is the development of new types of funerary monuments: tholoi and artificial caves. These are only found in the more developed areas: southern Iberia, from the Tagus estuary to Almería, and SE France.

Eventually, c. 2600 BC, urban communities began to appear, again especially in the south. The most important ones are Los Millares in SE Spain and Zambujal (belonging to Vila Nova de São Pedro culture) in Portuguese Estremadura, that can well be called civilizations, even if they lack of the literary component.

Extent of the Beaker culture

It is very unclear if any cultural influence originated in the Eastern Mediterranean (Cyprus?) could have sparked these civilizations. On one side the tholos does have a precedent in that area (even if not used yet as tomb) but on the other there is no material evidence of any exchange between the Eastern and Western Mediterranean, in contrast with the abundance of goods imported from Northern Europe and Africa.

Since c. 2150 BC, the Bell Beaker culture intrudes in Chalcolithic Iberia. After the early Corded style beaker, of quite clear Central European origin, the peninsula begins producing its own types of Bell Beaker pottery. Most important is the Maritime or International style that, associated especially with Megalithism, is for some centuries abundant in all the peninsula and southern France.

Since c. 1900 BC, the Bell Beaker phenomenon in Iberia shows a regionalization, with different styles being produced in the various regions: Palmela type in Portugal, Continental type in the plateau and Almerian type in Los Millares, among others. 

Our knowledge of the cultures present in Iberia by the Bronze Age is patchy.

The picture is clarified somewhat, however, by the eve of Bronze Age collapse.

Eventually, before Roman influence arrived in Iberia, however, most of the region was Celtic (and associated with this Indo-European linguistic family). This was an early Iron Age event:

The Iron Age in the Iberian peninsula has two focuses: the Hallstatt-related Iron Age Urnfields of the North-East and the Phoenician colonies of the South.

During the Iron Age, considered the protohistory of the territory, the Celts came, in several waves, starting possibly before 600 BC.

The Southwest Paleohispanic script, also called Tartessian, present in the Algarve and Lower Alentejo from about the late 8th to the 5th century BC, is possible the oldest script in Western Europe and it could have come from the Eastern Mediterranean, perhaps from Anatolia or Greece.

Since the late 8th century BC, the Urnfield culture of North-East Iberia began to develop Iron metallurgy and, eventually, elements of the Hallstatt culture. The earliest elements of this culture were found along the lower Ebro river, then gradually expanded upstream to La Rioja and in a hybrid local form to Alava. There was also expansion southwards into Castelló, with less marked influences reaching further south. Additionally, some offshoots have been detected along the Iberian Mountains, possibly a prelude to the formation of the Celtiberi.

During this period, the social differentiation became more visible with evidence of local chiefdoms and a horse-riding elite. It is possible that these transformations represent the arrival of a new wave of cultures from central Europe. From these outposts in the Upper Ebro and the Iberian mountains, Celtic culture expanded into the plateau and the Atlantic coast. Several groups can be described: 

* The Bernorio-Miraveche group (northern Burgos and Palencia provinces), that would influence the peoples of the northern fringe. 

* The north-west Castro culture, in today's Galicia and northern Portugal, a Celtic culture with peculiarities, due to the persistence of aspects of an earlier Atlantic Bronze Age culture. 

* The Duero group, possibly the precursor of the Celtic Vaccei. 

* The Cogotas II culture, likely precursor of the Celtic or Celtiberian Vettones (or a pre-Celtic culture with substantial Celtic influences), a markedly cattle-herder culture that gradually expanded southwards into what is today's Extremadura. 

* The Lusitanian culture, the precursor of the Lusitani tribe, located in what is today's central Portugal and Extremadura in western Spain, is generally not considered Celtic since the Lusitanian language does not meet some the accepted definitions of a Celtic language. Its relationship with the surrounding Celtic culture is unclear. Some believe it was essentially a pre-Celtic Iberian culture with substantial Celtic influences, while others argue that it was an essentially Celtic culture with strong indigenous pre-Celtic influences. There have been arguments for classifying its language as either Italic, a form of archaic Celtic, or proto-Celtic.

All these Indo-European groups have some common elements, like combed pottery since the 6th century and uniform weaponry.

After c. 600 BC, the Urnfields of the North-East were replaced by the Iberian culture, in a process that wasn't completed until the 4th century BC.

Approximate extent of the Celts c. 400 BCE

However: 

After c. 600 BC, the Urnfields of the North-East were replaced by the Iberian culture, in a process that wasn't completed until the 4th century BC. This physical separation from their continental relatives would mean that the Celts of the Iberian peninsula never received the cultural influences of La Tène culture, including Druidism.

The linguistic picture isn't really clearly documented until the eve of Roman conquest around 300 BCE, with the Iberian language probably sharing a linguistic family relationship with the Basque language. 

The linguistic classification of the Tartessian language is controversial:

Tartessian is generally left unclassified for lack of data or proposed to be a language isolate for lack of connections to the Indo-European languages. Some Tartessian names have been interpreted as Indo-European, more specifically as Celtic. However, the language as a whole remains inexplicable from the Celtic or Indo-European point of view; the structure of Tartessian syllables appears to be incompatible with Celtic or even Indo-European phonetics and more compatible with Iberian or Basque; all Celtic elements are thought to be borrowings by some scholars.

Since 2009, John T. Koch has argued that Tartessian is a Celtic language and that the texts can be translated. Koch's thesis has been popularised by the BBC TV series The Celts: Blood, Iron and Sacrifice and the associated book by Alice Roberts.

However, his proposals have been regarded with scepticism by academic linguists and the script, which is "hardly suitable for the denotation of an Indo-European language[,] leaves ample room for interpretation". In 2015, Terrence Kaufman published a book that suggested that Tartessian was a Celtic language but written using a script devised initially for a Vasconic "Hipponic" language (numerous SW placenames in -i(p)po(n)) although there are no extant inscriptions in such a language using the Tartessian script.

The Tartessian culture appears to be the first Iberian culture in which there is major use of cattle:

The name Tartessian, when applied in archaeology and linguistics does not necessarily correlate with the semi-mythical city of Tartessos but only roughly with the area where it is typically assumed it should have been located.

The Tartessian culture of southern Iberia actually is the local culture as modified by the increasing influence of eastern Mediterranean elements, especially Phoenician. Its core area is Western Andalusia, but soon extends to Eastern Andalusia, Extremadura and the Lands of Murcia and Valencia, where a Tartessian complex, rooted in the local Bronze cultures, is in the last stages of the Bronze Age (ninth-eighth centuries BC) before Phoenician influences can be seen clearly.

The full Tartessian culture, beginning c.720 BC, also extends to southern Portugal, where is eventually replaced by Lusitanian culture. One of the most significant elements of this culture is the introduction of the potter's wheel, that, along with other related technical developments, causes a major improvement in the quality of the pottery produced. There are other major advances in craftsmanship, affecting jewelry, weaving and architecture. This latter aspects is especially important, as the traditional circular huts were then gradually replaced by well finished rectangular buildings. It also allowed for the construction of the tower-like burial monuments that are so typical of this culture.

Agriculture also seems to have experienced major advances with the introduction of steel tools and, presumably, of the yoke and animal traction for the plough. In this period it's noticeable the increase of cattle accompanied by some decrease of sheep and goat types.

Another noticeable element is the major increase in economical specialization and social stratification. This is very noticeable in burials, with some showing off great wealth (chariots, gold, ivory), while the vast majority are much more modest. There is much diversity in burial rituals in this period but the elites seem to converge in one single style: a chambered mound. Some of the most affluent burials are generally attributed to local monarchs.

Some of the key illustrations from the new Basque genetics paper follow:

The paper says this in its discussion section:

[O]ur analyses support the notion that the genetic uniqueness of Basques cannot be attributed to a different origin relative to other Iberian populations but instead to a reduced and irregular external gene flow since the Iron Age[.] The observed clines of postIron Age gene flow in the region suggest that the specific genetic profile of Basques might be explained by the lack of recent gene flow received. Our analyses confirm that Basques were influenced by the major migration waves in Europe until the Iron Age, in a similar pattern as their surrounding populations. 

At that time, Basques experienced a process of isolation, characterized by an extremely low admixture with the posterior population movements that affected the Iberian Peninsula, such as the Romanization or the Islamic rule, as observed in the present genetic landscape. 

This does not exclude plausible previous periods of isolation, as attested by the presence of short ROHs and small Ne values that support signals of ancient inbreeding in the region, even higher than in Sardinia, which is suggested to be isolated after Neolithic times. Thus, the increase of the Ne observed only in the external groups about 1,000 generations ago might be potentially linked to the role of the Franco-Cantabrian region as glacial refugium during LGM periods and the subsequent expansion. 

Although our results support the genetic continuity from the Iron Age in most of the present day Basques, those located in the periphery of the Basque core area show signals of contacts compatible with the Roman Empire presence in the Iberian Peninsula. These results are in agreement with archaeological and historical records. An important presence of the Roman Empire has been reported in the whole Franco-Cantabrian region, but the scholars suggest a much higher impact in the peripheral areas of the southern side, specially Nafarroa and Araba. Otherwise, North African influence only fit the models where southern and northwestern Iberians are included. This confirms the reduced gene flow between the eastern and northern areas of the Iberian Peninsula with the North African incomers during the Islamic rule, as already reported by using uniparental markers and more recently through genome-wide data and haplotype-based methods.

The paper and its abstract are as follows:

Basques have historically lived along the Western Pyrenees, in the Franco-Cantabrian region, straddling the current Spanish and French territories. Over the last decades, they have been the focus of intense research due to their singular cultural and biological traits that, with high controversy, placed them as a heterogeneous, isolated, and unique population. Their non-Indo-European language, Euskara, is thought to be a major factor shaping the genetic landscape of the Basques. Yet there is still a lively debate about their history and assumed singularity due to the limitations of previous studies. 

Here, we analyze genome-wide data of Basque and surrounding groups that do not speak Euskara at a micro-geographical level. A total of
629,000 genome-wide variants were analyzed in 1,970 modern and ancient samples, including 190 new individuals from 18 sampling locations in the Basque area. For the first time, local- and wide-scale analyses from genome-wide data have been performed covering the whole Franco-Cantabrian region, combining allele frequency and haplotype-based methods.

Our results show a clear differentiation of Basques from the surrounding populations, with the non-Euskara-speaking Franco-Cantabrians located in an intermediate position. Moreover, a sharp genetic heterogeneity within Basques is observed with significant correlation with geography. 

Finally, the detected Basque differentiation cannot be attributed to an external origin compared to other Iberian and surrounding populations. Instead, we show that such differentiation results from genetic continuity since the Iron Age, characterized by periods of isolation and lack of recent gene flow that might have been reinforced by the language barrier.

Frederic Bauduer, et al., "Genetic origins, singularity, and heterogeneity of Basques" 31 Current Biology 1-11 (May 24 2021) (online ahead of publication). doi.org/ 10.1016/j.cub.2021.03.010