The Demise of Oriental Neolithic Admixture
The old idea that Europe is a result of Neolithic immigration from the Near East is ever harder to sustain. Certainly, a variety of grains jumped over from Southwest Asia to the virgin Neolithic world outside “as is”, but much of Neolithic culture discovered in Europe can’t be derived as easily and unequivocally from the same eastern source. Not even pottery, that developed pretty late and almost simultaneously on both sides of the Bosporus from what now increasingly emerges as a quite ‘backward’ PPNB-level of Neolithic civilization.
Until recently, the idea of a Neolithic ‘wave of advance’ could still count on the overwhelming support of geneticists, that interpreted the undeniable NW-SE cline in the distribution of several genetic markers as evidence of Neolithic expansion at the cost of Europes native population. Paleogenetic investigation already attested the low impact of Neolithic mtDNA, the Neolithic contribution of Y-DNA on current populations is ever more contested, and the autosomal DNA of Ötzi the Iceman rather suggests important post-Neolithic genetic shifts, that moreover appears to have been predominantly towards the north. The attested lack of an eastern component in Ötzi’s genome could be readily identified in comparison to modern populations, and raises sceptical questions about his purported “eastern” identity. The simple observation that Ötzi’s utter absence of any genetic Asia-shift necessarily implies that Ötzi does not represent the genetic heritage of Asiatic agriculturists is sound enough, notwithstanding the continuous stream of worthless articles that still demand otherwise. The same could be stated about his Y chromosome marked by haplogroup G2a. As I already explained elsewhere on this blog his DNA may have been different from Mesolithic ethnicities of northern origin, but closely related to the native population south of the Alps. Thus also the very proposition that Europe’s main Neolithic proponents, to be found in LBK and Cardium cultures, are genetically “eastern” at all is increasingly at odds with their sharing of their most common Y-DNA with Ötzi’s.
Now the study of Arenas et al. (2012) reveals that true Western Asian range expansions into European could only be considered reminiscent within the current pattern of genetic traces, if their age would be predominantly older than Neolithic. Europe has a clear SE-NW cline of genetic variation and right from the start this was always attributed to Neolithic expansion from the Near East. This interpretation, however, simply can’t stand up to careful scrutiny. Human genes are not like barley, emmer, lents and all other Neolithic crops being attributed an ultimate origin in the Near East, that all can be easily handed over between Neolithic societies and blurred from north to south and back again. Indeed, population geneticists rarely seem to realize that a range expansion is a two-dimensional process, of a population on an expansion wave that are fanning out slowly over a wider area rather than moving on a straight line. A population that expands from location A to equidistant locations B, C and D thus accumulate genetic distances AB, AC and AD that will be exceeded by lateral genetic distances BC, CD and DB. For the Principal Component Analysis (PCA) this should normally cause the first PC axis to be orthogonal to the expansion axis.
In a previous post I discussed the possible impact on genetic variance, that for this reason is most likely to increase within the circular ripples of the expansion wave (and hence the discussed increase of R1b-U106 variance towards the east, while the oldest and most varied haplotypes can be found in the west).
The authors realize that ‘PCA gradients can occur even when there is no expansion’, but only explore the possible impact of genetic differentiation over time and discard the scenario:
[…] Our simulations thus show that […] admixture between Neolithic and Paleolithic humans have drastic effects on PC1 gradients, and suggest that very large levels of Paleolithic ancestry are necessary to produce SE-NW PC1 gradients (Arenas et al., 2012)
Their model of paleolithic ancestry doesn’t take into account recent evidence on the continuity of archaic human genes, that might have strengthened the SE-NW gradient. Neither do the authors consider extremely quick range expansions without any orthogonal axis, that indeed weren’t feasible by Neolithic transport methods, but could have been a factor in Greek and Roman times up to nowadays. Instead, the autors chose to expand on an alternative range expansion from the hypothetized Iberian LGM (Last Glacial Maximum) refugium that was perpendicular to previous and subsequent range expansions from western Asia. Interestingly, the opposite effects on range expansion gradients didn’t simply phase out the Neolithic contribution, it was obliterated:
[…] our simulation results show that a PC1 SE-NW cline is not compatible with a major contribution of Neolithic populations into the gene pool of current Europeans, but with a major LGM refuge area for Paleolithic populations in the Iberian peninsula (Arenas et al., 2012)
Mol Biol Evol (2012) doi: 10.1093/molbev/mss203
Arenas et al. – Influence of admixture and Paleolithic range contractions on current European diversity gradients, 2012, link
Cavalli-Sforza and colleagues (1963) initiated the representation of genetic relationships among human populations with principal component analysis (PCA).Their study revealed the presence of a southeast–northwest (SE-NW) gradient of genetic variation in current European populations, which was interpreted as the result of the demic diffusion of early Neolithic farmers during their expansion from the Near East. However, this interpretation has been questioned, as PCA gradients can occur even when there is no expansion, and because the first PC axis is often orthogonal to the expansion axis. Here, we revisit PCA patterns obtained under realistic scenarios of the settlement of Europe, focusing on the effects of various levels of admixture between Paleolithic and Neolithic populations, and of range contractions during the Last Glacial Maximum (LGM). Using extensive simulations, we find that the first PC (PC1) gradients are orthogonal to the expansion axis, but only when the expansion is recent (Neolithic). More ancient (Paleolithic) expansions alter the orientation of the PC1 gradient due to a spatial homogenization of genetic diversity over time, and to the exact location of LGM refugia from which re-expansions proceeded. Overall we find that PC1 gradients consistently follow a SE-NW orientation if there is a large Paleolithic contribution to the current European gene pool, and if the main refuge area during the last ice age was in the Iberian Peninsula. Our study suggests that a SE-NW PC1 gradient is compatible with little genetic impact of Neolithic populations on the current European gene pool, and that range contractions have affected observed genetic patterns.
This study departs from the necessity of Last Glacial Maximum Refugiums, places where people survived during the last glacial period in the northern hemisphere. Nearly all ice sheets were at their Last Glacial Maximum (LGM) positions from 26.5 ka to 19 to 20 ka (Clark et al., 2009). According to the theory people disappeared – naturally – from the lands covered by glaciers, but also – and this is questionable – from a broad belt of steppe-tundra borderland in northern, central and eastern Europe down south to southwestern France, the Ligurian coast and the Adriatic Sea. A pattern of forest steppes emerged in southern Europe considered (more?) ‘benign’ to human habitation. Favorable places must have been Italy, by then connected through Tuscany by a land bridge with Elba, Corsica and Sardinia respectively, the Iberian peninsula and the southern Balkan, the latter being directly connected to Anatolia and the Middle East. Somehow ‘LGM refugionists’ considered LGM humanity trapped inside the southern forest steppes, effectively isolated by the southern limits of the extended steppe-tundras. Still, at Europe’s temperate latitudes intense sunlight and loess soils permitted a high level of bioproductivity; mosses, lichens, grasses, and low shrubs that fed mammoths, horses, bison, giant deer, aurochs and reindeer. It is hard to conceive why LGM Europeans would have left this northern paradise behind and contracted to southern refugia – and choose Iberia while during LGM the western Mediterranean basin was much stronger affected by climate change than the Balkan peninsula. Still the preferred LGM refugium, at least to scientific proponents and their mathematics, remained Iberia. This peninsula has a considerable overlap with the Franco-Cantabric region, that includes the southern half of France and the coastal area of northern Spain. In the prehistorical record this region was culturally homogeneous and possibly it was the most densely populated region of Europe in the Late Paleolithic. Highlights of artistic expression before and after LGM demostrate cultural continuity remained virtually unaffected even by the supposed upheavels during LGM. Well known highlights are the rock paintings of wild mammals and human hands in the cave of Altamira, Cantabria (~18,500 years ago, Upper Solutrean, and between ~16,500 and ~14,000 years ago, Lower Magdalenean); the famous Magdalenean paintings of Lascaux, Dordogne, estimated at 17,300 years ago; and the Chauvet Cave (Chauvet-Pont-d’Arc) in Ardèche, whose paintings were confirmed to be much older, between 30,000–32,000 years BP (Aurignacian).
The sudden appearence of real art of such high quality, defies all concepts of gradual evolution in artistic style and human mental capacities. However, the ultimate source of this art may be less visible, like the ancient cave paintings in Nerja, Andalusia (Spain), that emerged this year as possibly the oldest yet found. Organic remains at the spot indicated an even more incredible age than Chauvet Cave: being at least 42,000 years old, these must have been almost for sure the work of Neanderthals. Actually, there is not any reason to presume that knowledge of painting wasn’t native to the wider region. At least the use of paint, for whatever purpose, was already widely known among hominins about a quarter of a million years ago, from the Rhine to southern Africa:
Identification of the Maastricht-Belvédère finds as hematite pushes the use of red ochre by (early) Neandertals back in time significantly, to minimally 200–250 kya (i.e., to the same time range as the early ochre use in the African record) (Roebroeks et al., 2012)
Over this timedepth it would be more than amazing, even bizarre, that the most beautiful horses of Chauvet Cave have so much in common with the horses of the “nave” of Lascaux around the great black cow, almost 15,000 years younger! Indeed, on the basis of stylistic comparison, the Chauvet cave rock ornamentations were initially estimated as being Solutrean (22–17 ka BP) and Magdalenian (17–10 ka BP). This apparent attestation of cultural continuity over thousands of years, however, has a slight geographic component that I conceive as contradictory to the LGM concept. An Iberian LGM refugium would require the people of the Chauvet Cave cultural complex in Ardeche to have migrated to and fro Iberia before arriving in Lascaux, Dordogne. Since Altamira is located well inside the boundaries of the Iberian LGM refugium and dated only slightly after LGM, we might presume that according to the Iberian Refugium concept Altamira rock ornaments are transitionary between Chauvet and Lascaux. They are not. The Altamira hands are not sufficiently unique since painted hands are an ornament in rock art all over the world and the Altamira horses are of a different style. Even the Altamira crouching steppe bison does not have anything to do with Chauvet’s ‘The Venus and the Sorcerer’ having an extinct steppe bison crouching over a great black female pubic triangle in a sexual pose. If the crouching element in Altamira, devoid of all sexual implication, would nevertheless represent the survival of a technical style, preserved by travelling artists, it should be explained why this sexual implication was lost in Altamira while unique European mythological interpretations of the Sorcerer may be readily recognized in historic fertility gods like Crete’s Minotaur, the Frankish Quinotaur (Rhine) and the Celtic Cernunnos as depicted on the “Pillar of the Boatmen” (Seine). Another representation of the paleolithic myth may have survived even in the Sumerian god of creation Enki, sometimes depicted as a bull. Apparently, the Altamira representation attests a different tradition, with an equally problematic transition towards the art of Lascaux. Apparently, the argument in favor of a considerable detour of Paleolithic people quite north of the Pyrenees through-Spain-before-arriving-back-north in Lascaux, is not yet supported by compelling evidence and rather remains the product of grand speculation.
The LGM refugium boundaries of southern Europe, even the very concept of any LGM refugium at all from 26.5 ka to 19 to 20 ka, would be severely compromised if some kind of Franco-Cantabrian local continuity indeed persisted between the art of Chauvet and Lascaux. All scenarios investigated by Arenas et al. showed better results for simulations that ran with low Neolithic admixture. A best fit was warranted by a disproportionate role for the genetic component of a westernmost LGM refugium:
When southern Europe is considered as a single large refugium, PC1 maps show E-W gradients (Figures 2B and S3B), but when the LGM refugium is restricted to the Iberian Peninsula, PC1 maps show steeper NW-SE gradients (Figures 2C and S3C). (Arenas et al., 2012)
The scenarios described by Arenas et al. also investigated the option of mere (Upper) Paleolithic range expansions from oriental origin without subsequent LGM contraction, and indeed the results were very similar to the simulations of a refuge area restricted to the Iberian Peninsula with a history of expansion-range contraction-reexpansion for Paleolithic populations. However, a simple paleolithic range expansion may be insufficient for a true approximation of the NW-SE gradient if a longer history of genetic differentiation could compensate for the drift imposed by LGM refuge scenarios. Unwittingly Arenas et al. depart from the complete replacement of previous populations at the start of the Upper Paleolithic by modern man, and unfortunately this grand mistake already rendered the study obsolete before publication:
The onset of the initial settlement of Europe by Paleolithic populations was set to 1,600 generations ago, corresponding to 40,000 years ago (Mellars 2006) assuming a 25y generation time. In this initial range expansion, we assume that Paleolithic populations completely replaced archaic populations without any interbreeding. (Arenas et al., 2012)
Archaic admixture is a hot item nowadays and already shattered the once popular extinction scenarios attributed to the onset of the Upper Paleolithic. An older age of local genes would make the whole LGM refuge issue virtually irrelevant for understanding current genetic configurations, so I would call this a draw. The main difference of models without LGM refuges is they require continuous habitation everywhere south of the LGM glaciers for a much longer time – and actually there isn’t any reason why they didn’t. Humans just had to follow the game, and wasn’t deterred by the cold to follow their routes up to the food and water at the frontiers of iceage glaciers. Upper Paleolithic people were able enough to do so throughout LGM, and their inmediate predecessors had the advantage of a better climate: the last Ice Age in Europe (Weichselian) was nothing compared with the previous one (Saalien), that ended 130,000 years ago. Of course, simulations of Middle Paleolithic range expansions would account for a more pronounced genetic dichotomy on the European NW-SE axis and thus give better results than Upper Paleolithic range expansions, thus eliminating the need for incorporating LGM refugium related population contractions behind the Pyrenees in the model.
Having said the necessary on LGM refugia, I value the Arenas et al. study for debunking stale assumptions on the Asiatic character of Neolithic influence and the wrongfully implied impact on Europe’s NW-SE genetic gradient. However, the absence of strong mtDNA and autosomal DNA signals for Paleolithic SE-NW range expansions doesn’t necessarily imply consistency with a pre-Neolithic scenario – and indeed probably it doesn’t for Y-DNA. The source of inspiration for this study was mtDNA evidence (Pereira et al., 2005), whose European dichotomy and timeline is now mirrored by the geographic clines of autosomal DNA. Now, the Y-chromosome evidence features some intriguing dichotomies all by itself: ‘western’ R1b against ‘eastern’ R1a; R1b that features an internal dichotomy on SNP S127; and, more recently, R1a that apparently features another west-east dichotomy internally on SNP Z645. The latter dichotomy is continued further east by subclades of Z645, ie. European R1a SNP Z283 against Asiatic R1a SNP Z93. Hence it has all appearance the accumulation of various West/Central European Y-DNA haplogrouos are related and, despite a strong West to East gradient, are thus incompatible with results that previously suggested an important role of a Neolithic expansion:
[…] that R1b1b2 was carried as a rapidly expanding lineage from the Near East via Anatolia to the western fringe of Europe during the Neolithic. (Balaresque et al., 2010)
Naturally, this introduces a new inconsistency in the genetic evidence between Y-chromosome dating on one hand, and mtDNA and autosomal dating on the other hand. Klopfstein (2006) discussed Allele Frequency Clines (AFCs) were ‘mutations having arisen during Paleolithic range expansions should show larger absolute frequency differences than those having occurred during a pure Neolithic expansion’. Genetic differentiation perpendicular to the main direction of range expansions was not his concern for the very massive nature of his AFC-driven model, and still his model shared the preference for pre-Neolithic range expansions:
As expected from our previous results, the average final frequency of the mutation is found much higher after the Paleolithic expansion than after the Neolithic expansion (44% and 2% in the colonized area, respectively) – Klopfstein et al., 2006
Nor, indeed, was the Klopfstein study specifically meant to include Y chromosome genetics in his model. Still, only Y-chromosome substructure appears to be compatible with slow SE-NW range expansions (Balasques et al., 2010). The accepted YDNA dates for the European SE-NW gradiënt, however, are generally considered inconsistent with a pre-Neolithic or Paleolithic scenario. Between haplogroup dating of especially Y-chromosomes is based on obsolete assumptions on extremely large proportions of poorly understood ‘junk DNA’. Like already predicted in a previous post (Evolving Chimps are Messing Up Y-DNA Dating), the actual proportion of functional DNA is already accepted to be much higher, implying a more compromised viability at conception time of deleterious mutations, what in turn translates to actually lower mutation rates. It should be noted the Iberian refugium hypothesis isn’t compatible with Iberian Y-DNA, though maybe a pre-Neolithic development of some R1a and R1b in a wider European context is – especially now the accepted mutation rate keeps dropping.
At this moment it has all appearance that Neolithic DNA south of the Alps was rather much more ‘Sardinian’ and ‘Ötzi-like’ than anything else, ie. virtually devoid of any significant genetic shift to the east. Ethnical continuity of this type may have extended much further east than genetic analyses on current populations would allow us to consider without the evidence currently available in paleogenetic samples – even of an eastern location within Europe as remote as Bulgaria: ‘Strikingly, an analysis including novel ancient DNA data from an early Iron Age individual from Bulgaria also shows the strongest affinity of this individual with modern-day Sardinians’ (Sikora et al., 2012). Time will tell the publication of this find will deal the final blow to current scientific beliefs that concern important Neolithic immigration of Asiatic agriculturists. Much easier would it be to assume the remaining mtDNA and autosomal DNA gradients are simply the exaggerated, sex-biased result of quick range expansions from the east, that blurred all preceding slower range expansions beyond recognition – except for the Y-chromosome gradiënt mentioned above.****) After all, the lengthy debates and elaborate calculations about the origin of a considerable genetic east-shift, and the genetic SE-NW cline in Europe, may simply reduce to the ‘quick range expansion’ that was due to a previously unsuspected popularity in the Classic world of girls from across the Bosporus, as marriage partners.
****) The European clines for YDNA R1b and R1a are nowadays (2014) recognized as fairly recent (late-Neolithic) star-like expansions:
The Larmuseau et al. – Recent Radiation within Y-chromosomal Haplogroup R-M269 Resulted in High Y-STR Haplotype Resemblance (2014) study ‘reveals a strong Y-STR haplotype resemblance among West-European males belonging to haplogroup R-M269, which is most likely the result ofrapid population expansion. This expansion event should have been accompanied by an accumulation of allelic variance, such that the action of mutation and genetic drift had no chance to generate distinctive, subhaplogroup-specific haplotypes.’
Likewise, the Underhill et al. – The phylogenetic and geographic structure of Y-chromosome haplogroup R1a (2014) study admits that the ‘[w]hole Y-chromosome sequence analysis of eight R1a and five R1b individuals suggests a divergence time of ~25,000 (95% CI: 21,300–29,000) years ago and a coalescence time within R1a-M417 of ~5800 (95% CI: 4800–6800) years’.
In this light, it has become most unlikely any of these West-East clines of the YDNA R have anything to do with the Neolithic advance.
More Mesolithic origin of these clines have already been treated in The Mesolithic Blind Spot.
- Arenas et al. – Influence of admixture and Paleolithic range contractions on current European diversity gradients, 2012, link
- Balaresque et al. – A Predominantly Neolithic Origin for European Paternal Lineages, 2010, link
- Clark et al. – The Last Glacial Maximum, 2009, link
- Kuhlemann et al. – Last glaciation of the Šara Range (Balkan peninsula): Increasing dryness from the LGM to the Holocene, 2009, link
- Morelli et al. – A Comparison of Y-Chromosome Variation in Sardinia and Anatolia Is More Consistent with Cultural Rather than Demic Diffusion of Agriculture, 2010, link
- Pereira et al. – Highresolution mtDNA evidence for the late-glacial resettlement of Europe from an Iberian refugium, 2005, link
- Roebroeks et al. – Use of red ochre by early Neandertals, 2012, link, On the news
- Sadier et al. – Further constraints on the Chauvet cave artwork elaboration, 2012, link
- Sikora et al. – On the Sardinian ancestry of the Tyrolean Iceman. To be presented at the annual meeting of ASHG, 2012, link
- Züchner – Grotte Chauvet Archaeologically Dated, 2000, link
- The Paris Review Daily. The Spring Issue: Werner Herzog and Jan Simek on Caves
December 30, 2011 | by John Jeremiah Sullivan, link
- El Mundo – ¿La obra de arte más antigua de la Humanidad? 07/02/2012, link
- Mail online, 7th February 2012: ‘The oldest work of art ever’: 42,000-year-old paintings of seals found in Spanish cave, link