search this blog

Tuesday, May 19, 2015

Large-scale recent expansion of European patrilineages


Open access at Nature Communications today: Large-scale recent expansion of European patrilineages shown by population resequencing by Batini et al.

It's a shame the authors failed to sample any Eastern European populations, but it's still a very useful effort which moves us in the right direction in an area of study that has floundered from the start, largely due to the widespread use of bikini STR haplotypes and faulty methodology. From the paper:

Here, we use targeted NGS of European and Middle Eastern populations to show that Europe was affected by a major continent-wide expansion in patrilineages that post-dates the Neolithic transition. Resequencing at high coverage of 3.7 Mb of MSY DNA, in each of 334 males comprising 17 population samples, defines an unbiased phylogeny containing 5,996 high-confidence single-nucleotide polymorphisms (SNPs). Dating indicates that three major lineages (I1, R1a and R1b), accounting for 64% of the sampled chromosomes, have very recent coalescent times, ranging between 3.5 and 7.3 KYA. In demographic reconstructions (17) a continuous swathe of 13/17 populations from the Balkans to the British and Irish Isles share similar histories featuring a minimum effective population size ~2.1–4.2 KYA, followed by expansion to the present. Together with other data on maternally inherited mtDNA (16, 18) and autosomal DNA (19), our results indicate a recent widespread male-specific phenomenon that may point to social selection, and refocuses interest on the social and population structure of Bronze Age Europe.

...

The shapes of different clades within the tree (Fig. 1a) vary greatly. Haplogroups E1b-M35, G2a-L31, I2-P215, J2-M172, L-M11 and T-M70 contain long branches with deep-rooting nodes, whereas I1-M253, N1c-M178, R1a-M198 and R1b-M269 show much shallower genealogies.


The recent and rapid continent-wide demographic changes we observe suggest a remarkably widespread transition affecting paternal lineages. This picture is confirmed in an independent analysis of MSY diversity in the pooled HGDP CEPH panel European samples (16), and is compatible with current (n=98) ancient DNA data for MSY (Fig. 3; Supplementary Table 8), in which hgs R1a, R1b and I1 are absent or rare in sites dating before 5 KYA, whereas hgs G2a and I2 are prevalent.


The period 4–5 KYA (the Early Bronze Age) is characterized by rapid and widespread change, involving changes in burial practices that might signify an emphasis on individuals or kin groups, the spread of horse riding, and the emergence of elites and developments in weaponry (35). In principle male-driven social selection (36) associated with these changes could have led to rapid local increases in the frequencies of introgressing haplogroups (34), and subsequent spread, as has been suggested for Asia (37). However, cultures across Europe remain diverse during this period; clarifying the temporal and geographical pattern of the shift will rely heavily on additional ancient DNA data.

Batini, C. et al. Large-scale recent expansion of European patrilineages shown by population resequencing. Nat. Commun. 6:7152 doi: 10.1038/ncomms8152 (2015).

See also...

R1a1a from an Early Bronze Age warrior grave in Poland

Massive migration from the steppe is a source for Indo-European languages in Europe (Haak et al. 2015 preprint)

159 comments:

Krefter said...

I'm very confused how certain paternal lineages became so successful in Bronze age Europe. It was intended by human minds we know that much. It didn't just happen randomly.

Steppe-maternal lineages have a presence to, so Steppe-intrusion wasn't solely male. Also, plenty of R1a and R1b lineages from the Steppe must have gone extinct(not completely for all) like native lineages. R1a-M417* from Corded Ware is evidence of this.

Davidski said...

How do you know the Corded Ware M417* isn't ancestral to modern Z282*?

Mike Thomas said...

Krefter

"it didn't happen randomly".

Maybe not entirely, but chance mihjt have a greater role than you think

(Btw it's "too" not "to")

Krefter said...

He was negative for R1a1a1b-S224(ancestral to Z93, and Z283) and R1a1a1a-CT27083(Same clade as L664).

S224 didn't originate in Germany. So his lineage is probably extinct to close to extinct. But his cousin Z283 is dominating East Europe today.

Since Bronze age R1a-rich intruders who went into Europe had lineages which died out, we shouldn't be surprised the Tarim mummies had an R1a lineage which died out in Asia.

@Mike,
"Maybe not entirely, but chance mihjt have a greater role than you think"

I agree. I doubt anyone was conscious of what was happening. The Corded Ware genomes sampled by Haak may have thought of themselves a native to the region. Human memory of times before they lived probably loses its accuracy after you get before their grandparents lived.

Davidski said...

Yes, I know he was negative for S224, but wasn't that because S224 didn't exist at the time?

Yuri said...

Do Basques consistently stand out as a G2a-rich island in the west or was this a fluke? Who knows, maybe this could be evidence that they preserved the "Oetzi" (early E Neolithic) language.

capra internetensis said...

@Yuri

It's a fluke (only 19 samples). Martinez_Cruz et al (2012) sampled 835 Basques - they had 0.7% G.

Krefter said...

@Yuri,

I've also read the Basque have a high amount of I2a1a1-M26.

http://www.eupedia.com/europe/european_y-dna_haplogroups_by_region.shtml

0% G is listed for Basque by Eupedia. BTW, all the Neolithic and Mesolithic I2as negative for I2a1b so far are positive for I2a1a(father of M26 and others). Their DNA files are online that's how I know.

Y DNA isn't the place to look for what the autosomal DNA of people is. Just look at Sardinians. It's obvious mainland Italy is the source of a lot of their paternal lineages but didn't make a big impact autosomally.

Maju said...

Which is the calibration point or points? Using age(CT)=110Ka (convenient because the fig.1 tree measures 10-11 cm in my screen but also approximately correct), I get some 10-11 Ka for the R1b-S116 gray rectangle. Considering that the M269 and M198 (R1a) branches appear here to be parallel and that, following Underhill's 2014 data, R1a expansion is also Epipaleolithic to Neolithic.

It's a bit fuzzy because not all branches are of equal length anyhow but it is crucial to know what calibration reference they are using. It's not the same imagining a most implausible Out-of-Africa migration c. 60 Ka ago (against all the material evidence) or accepting a realistic and well documented OoA c. 125-100 Ka ago. It has major implications.

Maju said...

Well, I just made a quick calculation based on the E-M35 age estimate of 18 Ka provided in table 1. If we would have to accept those chronologies at face value, then the E/F (=DE-CF) divergence node would be just 28 Ka old. Not 60 nor 70 nor anything logical: barely 30 Ka!!!

This means that you have to multiply x4 all age "estimates", sorry. Intra-disciplinary circular thinking at its worst but very illuminating, I must say.

capra internetensis said...

@Maju

They are using an mutation rate estimate from a study of Icelandic families with known pedigrees. I agree it is too fast, but I think your calculation is wrong.

Looking at the same table of dates, I figure the average is 268 years per mutation. Then measuring by the bar marked 50 mutations in the phylogenetic tree, the age of the CDEF node would be about 45 000 years old.

Also the K node would be about 30 000 years old, when Ust' Ishim man was K at 45 000 years ago.

Maju said...

@Capra: I don't know what's the problem with the "pedigree rate" but there is a serious problem with it, quite obviously. Most probably selection is acting in a conservative way: mutations are in general deleterious, surely affecting fertility and other male fitness aspects, and they get selected against, so long-term evolution is much slower than observed pedigree rate, which does not account for that.

With the two references you provide, surely more precise than my visual estimates, I'd make that:

1. For a realistic age(CDEF)=125-100 Ka (archaeological OoA), the correction rate to apply is: x2.8 to x2.2.

2. For a realistic age(K)=75-70 Ka (Toba episode), the correction age to apply is: x2.5 to x2.3

So I'd say that the realistic corrections for the estimates provided in this study are x2.4 or x2.5. Would you agree with that?

Krefter said...

@Maju,

Ust' Ishim man lived 45,000 years ago and isn't significantly more related to WHG/ANE than to East Asians and vice versa. Another 55,000 years is a long time, and for most of that time the ancestors of Eurasians must have been in the same general population and region.

Ryan said...

Can anyone get supplementary figure 1 to display properly? I can't get it resolved well enough to read the population names.

Maju said...

@Krefter there are K people in all Greater Eurasia, from Australasia to the Mediterranean, etc. K is the largest single macro-haplogroup of Y-DNA at that phylogenetic level. Right?

Furthermore we know that K2 (its main top level subhaplogroup) spread from SE Asia (Karafet) associated to mtDNA R (very high correlation) in a quite impressive demographic expansion that must have culminated in the West Eurasian colonization (early Upper Paleolithic, Ust-Ishim for example). That's why I associate K (K2 to be precise) and mtDNA R (maybe even N itself) with the Toba episode, which must have created a major crisis that triggered new founder effects, hard to explain otherwise.

It is also notable that this inference, arguable on itself, I reckon, but still very plausible, produces a an almost identical correction factor as the other calibration (the OoA one), which is much less arguable because it is much better studied by 21st century archaeology, with several researchers (Petraglia, Armitage, Rose, etc.) doing a great job in expanding our knowledge of the prehistory of Arabia, India and even Australia and America, in all cases pushing back the H. sapiens first settlement dates quite a bit relative to obsolete 20th century paradigms, which still burden scholastically the population genetics field.

So for me it is a very consistent correction factor, take or give a decimal figure, relative to the dates pushed by this study. And when 5000 years ago becomes 12,500 years ago, it is critically important for paleohistorical reconstruction. So it's a major issue of systematic erring what is going on with this calibration problem that plagues the "molecular clock" pseudoscience (or "wannabe science" if we are generous).

As Capra unederlined with the issue of Ust Ishim, the figures are off by at least 50% but nothing impedes that they are off by 100% or, as I suggest with good arguments 150%.

"Ust' Ishim man lived 45,000 years ago and isn't significantly more related to WHG/ANE than to East Asians and vice versa".

Ust Ishim is roughly "crown" to West Eurasians. Those dates 50-45 Ka BP are when the proto-WEA founders expanded. As demonstrated by Karafet with Y-DNA K2 (and also common sense in relation to mtDNA), they had an origin (at least partial but quite important anyhow) in SE Asia, origin related to Papuans (K2b) and East Asians (NO). That migration from SE Asia to West Eurasia (via North India, as can be tracked in both sides of the genetic legacy) must have taken place between the Toba catastrophe (74 Ka BP) and the first UP, which we can roughly date to c. 50 Ka BP. So no wonder that he is relatively indistinct Eurasian, IMO (never mind later E-W readmixture via Siberia, which can't be ignored: proto-Amerindians migrating East first, proto-Uralics migrating West later, etc.)

"Another 55,000 years is a long time"...

I'm only arguing for another 25 Ka more (between 50 Ka of first UP and 74 Ka of Toba), maybe a bit less. Before this period most early Eurasians seem to have been concentrated in the Tropical and Subtropical belt, i.e. in South and SE Asia, maybe with a few early offshoots to Australia and Central-NE China of little consequence and also surely a remnant of "Basal Eurasian" in Arabia also of little consequence. Migrations forth and back, notably this one of Y-DNA K2 plus mtDNA R people, kept the genetic pool in that relatively homogeneous state. However since that period we see greater regionalization, it seems to me, probably helped by greater densities, leading to the modern "races" or distinctive subcontinental populations, or at least their ultimate foundations.

Davidski said...

Ryan, the same and other samples can be seen here.

http://mbe.oxfordjournals.org/content/suppl/2014/11/26/msu327.DC1/FigureS1_TreeWithSampleNames.pdf

Grey said...

1. LP imo - mainly cos it would be fun if it was true.

2. mutation rate - wouldn't it vary over time with effective population so the rate at the early end would be slower than the more recent end?

terryt said...

"Also the K node would be about 30 000 years old, when Ust' Ishim man was K at 45 000 years ago".

A huge problem for the dating calibration.

"Furthermore we know that K2 (its main top level subhaplogroup) spread from SE Asia (Karafet) associated to mtDNA R (very high correlation) in a quite impressive demographic expansion"

Something I have been suggesting for years, in spite of vehement opposition.

"That's why I associate K (K2 to be precise) and mtDNA R (maybe even N itself) with the Toba episode, which must have created a major crisis that triggered new founder effects, hard to explain otherwise".

I have long been convinced there is another explanation. To reach New Guinea/Australia humans would have had to develop a reasonably sophisticated boating technology. Such technology would have enabled a very rapid expansion along some coastal regions and major rivers. Such an expansion through South Asia via the Ganges would explain the absence of any real level of Y-DNA P in the region. And also explain the reasonably high level of mt-DNA R of various sorts.

Mike Thomas said...

Terry& Maju

Re K- maybe that study is based in modern phylogeny; and is not necessarily incorrect. Ie all modern K descends from c. 3o kya
Indeed, ustishm is an extinct line

But terry, aborigines needn't have had "sophisticated" boating technology to island hop during favourable tides and water levels. All that's needed is a few logs strung together, literally

Shaikorth said...

U-I is an extinct line, but derived enough that it seems to be somewhere in the pre-NO branch. Whether that means pre-NO originally moved from Siberia/Central Asia to east is another matter, but it does imply the ancestor of all modern K's can't be just 30k years old, and that's what NGS of modern Y-DNA says as well.

Carlos said...

In this light how can we explain the existence of ancestral to 7700 old Swedish I2c2 in Georgia, the Caucasus

http://yfull.com/tree/I2/

Maju said...

@Saikorth: Just to be precise, IMO all those lineages at high phylogenetic levels like Ust Ishim or Mal'ta, as well as comparable more recent erratics like La Braña's C or the occasional F* or P* found here and there, seem to imply something that for me is rather obvious but maybe not for all: initially there was greater basal diversity in the WEA macro-population that was gradually fixated, mostly by drift but surely also by founder effects.

So I wouldn't see Ust Ishim as ancestral to NO but rather as a capricious variant of K2 moving West and I wouldn't see Mal'ta as ancestral to R1 (it is too drifted in its own R* line for that, much more recent than the R1 node in the time-drift equivalence) but rather as a capricious variant of R, which was already scattered in at least parts of West and South Eurasia. In other words: I don't see in them "ancestors" but rather horizontal relatives, "cousins", of other people (probably living further south) who were indeed true ancestors but whose remains we may never find, let alone sequence.

But one thing is clear: at least in Siberia there were "rare variants" that did not succeed in the long run.

Maju said...

@Carlos: no idea of the details but it seems apparent to me that the Swedish genetic pool has notable Eastern European influence, be it from Pitted Ware (related to Dniepr-Don) or from Kurgans.

Carlos said...

Thank you, Maju

Mike Thomas said...

Carlos
My take is different to Maju
Rather, I'd see I2c in Georgians and Armenians as the easternmost extent of WHG type groups, which miraculously survived there.

Shaikorth said...

@Maju

Ust-Ishim isn't really part of the West Eurasian macropopulation (unlike the C-V20 Kostenki) as it's equally close to "pure" East Eurasians as it's to ANE-influenced ones such as Amerindians or Siberians.

To me finding C or in modern context eastern forms of K in paleolithic West Eurasia or Siberia isn't something that should be considered an outlier *in those times* if they're the only samples found, which is the case. For all we know the regional distribution of at least mainland Eurasian K lineages is a more recent thing, a large-scale version of something like the near-total replacement of Scandinavian HG's I2.

Carlos said...

@ Mike Thomas

Interesting... In that case would I2c reach the Caucasus before 7700 ybp?

Mike Thomas said...

Carlos
Yes- must be. All of I2 dates to c. 18 ky BP; after which point various subgroups of I2 split, undoubtedly as they were repopulating much of Europe; and apparently parts of the caucasus, following the climactic amelioration.

Carlos said...

@ Mike Thomas

Fascinating, thanks

Gaspar said...

Interesting is that they found all the subclades for these 98 ancients

Krefter said...

@Gasper,
"Interesting is that they found all the subclades for these 98 ancients"

With ancient DNA data put in 23andme format by Felix's(his blog), you can further define their Y DNA haplogroup.

I did for some hg Is and older HGs.

https://docs.google.com/document/d/1YMr0G7ccyPtv8X9O4byfz09aH9XmjGXoRFjlm7Y5dhg/edit?usp=sharing

NE7 could be broken down pretty far to I2a1a2-L1286. There's a poster at Anthrogencia who belongs to that clade. I've heard it is found scattered around Europe, and is especially popular in Ireland and Britain.

terryt said...

"Re K- maybe that study is based in modern phylogeny; and is not necessarily incorrect. Ie all modern K descends from c. 3o kya
Indeed, ustishm is an extinct line"

But as Shaikorth said it is an extinct line within K2a. And Maju has also elaborated on the matter. That means K2a's ancestor must have split from the other K2s by that time. I leave for now the problem of whether the ancestors of K2b, K2c and K2d had also already formed separately.

"But terry, aborigines needn't have had "sophisticated" boating technology to island hop during favourable tides and water levels. All that's needed is a few logs strung together, literally"

It needs to be reasonably sophisticated as it was not simply island hopping. There are considerable stretches of open water. I remember something like 150 kms for the largest stretch. And the individuals who made the original crossing almost certainly had to return in order to gather up their women, although it is possible they were a family group moving from island to island. But even that requires more than 'a few logs strung together'.

rozenfag said...

Ok, this may be not related directly to the topic of this blog, but another ancient wolf-dog DNA:
http://www.cell.com/current-biology/abstract/S0960-9822%2815%2900432-7

An ancient Siberian wolf yields a first draft genome sequence of a Pleistocene carnivore
•The 35,000-year-old wolf genome allowed recalibration of the lupine mutation rate
•Dog ancestors diverged from modern wolf ancestors at least 27,000 years ago
•Ancient Siberian wolves contributed to the ancestry of high-latitude dog breeds

Maju said...

@Saikorth: I'm arguing that UI is "founder" of a yet not differentiated WEA population. Not only founder of West Eurasians but also partly of Amerinds and therefore "his people" influencing genetically NE Asia with the spread of the UP since c. 30 Ka ago. You should compare with SE Asians, Negritos, Papuans, where you may find that incipient divergence more apparent. Also remember that Northern Europeans have some lesser NE Asian admixture via Uralics and whatever admixture Ma1-like Paleosiberians had, so maybe you prefer to use a more select WEA sample like Sardinians or Basques, which lacks those Siberian influences or has them only at low values.

Maju said...

@Terry: it's less than 150 Km, rather 80 Km or maybe less. Still it's clear that they must have got some seafaring skills, as the diversity of lineages indicates it was not any random "tsunami accident".

There is some evidence from a later date that people in Timor practiced fishing with hooks. Sadly organic materials like those used for rafts or boats are all gone long ago.

I recall that legends suggest that they followed the sea crocodile, what makes some sense, as they should have been familiar with the habits of this animal, which does not actually live in the sea most of the time. So they plausibly inferred that crocs knew where land (and freshwater) was and followed them.

Shaikorth said...

@Maju

Formal testing shows affinity to Ust-Ishim in SE-Asians including Melanesians and Papuans is more on the level of Siberians or Amerindians, and clearly above any West Eurasians. Here's a graph of f-stats:

http://s28.postimg.org/wuj44klpp/Kostenki14_Ust_Ishim_MA1ls.png

If there is a trend, it's that affinity to Ust-Ishim is increased by lack of Near Eastern/SSA mixture (which today is minimum in East Asians and Native Americans), making it more like ancestral generic "Eurasian". Basques don't stand out among Europeans while Spanish and Sardinians have less relation to it, and low-EEF NE-Euros such as Finns and Estonians are a bit closer than Erzya or Kargopol Russians who have similar or higher Siberian affinity (in fact they are about as close as Uzbeks and Hazaras who have much more ENA but also higher Near Eastern contribution). If European hunter-gatherers were in that graph, they'd be closer to U-I than any modern European population.

Looking at what we know of the genome, Ust-Ishim being an ancestral node of "West Eurasians" is possible if that means just WHG/ANE types and not Near Easterners. Kostenki14 affinity shown on the horizontal axis of the graph definitely looks more like ancestral West Eurasian - Sardinians, Armenians and Jews for instance are much closer to it than Chinese are.

terryt said...

Thanks for the clarification Maju.

Lank said...

Based on Ust'-Ishim's Y-DNA, the authors estimated the TMRCA of the K node at ~50 kya, and ~70 kya for CT. See Table S9.1: http://www.nature.com/nature/journal/v514/n7523/extref/nature13810-s1.pdf

terryt said...

Makes sense Lank. Humans may have arrived in Australia by 50 kya. Although I actually suspect that Y-DNA C arrived there some time before just the single branch of K2b1a did. That branch of K2 is shared with New Guinea and the other K2b1a branches are basically confined to that island.

capra internetensis said...

@Terry

You might want to have a look at the recent study by Karmin et al. By then (shortly after 50 000 years ago) K2b had already split into P and MS, P had split into P1 and P2, MS had split into pre-M and pre-S, and pre-S had split into 3 different surviving lineages. This was from sequencing a couple of dozen men from Borneo and a handful of Philippine Negritos.

No K2c or K2d, as they didn't test anywhere else in Indonesia.

C1b also split up at the same time (at multiple levels) and shows up in the same populations. Quite possibly K and C arrived simultaneously.

Krefter said...

@rozenfag.

"Ok, this may be not related directly to the topic of this blog, but another ancient wolf-dog DNA:"

I'm telling my dogs about this right now.....

They said they'll post their notes on it later today.

terryt said...

"By then (shortly after 50 000 years ago) K2b had already split into P and MS, P had split into P1 and P2, MS had split into pre-M and pre-S, and pre-S had split into 3 different surviving lineages. This was from sequencing a couple of dozen men from Borneo and a handful of Philippine Negritos".

Thanks. I was aware of that but I thought that would be a distraction from the main discussion.

"MS had split into pre-M and pre-S, and pre-S had split into 3 different surviving lineages".

The information I have is specifically that MS (or K2b1 if you prefer) split into K2b1b found in Borneo and East Indonesia, K2b1c found in The Aeta of the Philippines, M, sort of widespread in SE Asia/Melanesia, and a fourth branch that included S along with several other branches, basically all confined to New Guinea.

"No K2c or K2d, as they didn't test anywhere else in Indonesia".

And interestingly the whole K2 groups can be placed into a geographic sequence: K2a in East Asia, K2* in Sumatra, K2d in Java, K2c in Bali and K2b in: one branch in island SE Asia/New Guinea and the other right through Eurasia (P).

"C1b also split up at the same time (at multiple levels) and shows up in the same populations. Quite possibly K and C arrived simultaneously".

But Australian C1b2b is confined to that continent (not present in New Guinea at all) whereas Australia shares its K2b1a with New Guinea. C1b2a's arrival in New Guinea seems to considerably postdate its arrival in Eastern Indonesia and so we have two totally separate Y-DNAs in New Guinea and Australia. And virtually no mt-DNA common to both either. To me that indicates a separate arrival. But whose ancestor was first to cross Wallace's Line? The Australian of New Guinea haplogroup. In other words was Australia or New Guinea settled first?

Maju said...

@Saikorth: Alright, I stand corrected. Maybe he was some sort of early East Asian and I figured it all wrong. Strange to say the least but whatever.

"If there is a trend, it's that affinity to Ust-Ishim is increased by lack of Near Eastern/SSA mixture"...

Maybe there was a true Basal Eurasian genetic influence in West Eurasians (but not Paleosiberians) at the Upper Paleolithic genesis c. 50 Ka ago? I mean: people "left behind" in Arabia and nearby areas in the eastward migration (second phase of the OoA c. 100 Ka BP) and then reincorporated to the WEA pool, but not to Ust Ishim or other Paleosiberians. The Westward early UP migration probably branched in southern Central Asia (there's a cave in Uzbekistan that may have the oldest UP chronology), with one branch heading to Siberia and the other, the main one, heading West via Iran.

Maju said...

@Lank: Alright, let's assume that the Ust Ishim calibrated ages are correct (they pose some problems to me but will deal with them on another occasion), it still means that all the dates of this study need to be multiplied x1.7 (at least). So where it reads 5-6000 years ago, it should say 9-10 Ka ago, which is totally different.

Anyhow, to be precise, in that link you provide is clearly stated (table S9) that age(K2)=47-55 Ka BP, and age(IJK)=49-59 Ka, so we are rather getting towards >50 Ka for K.

Also age(CF)=65-80 Ka and age(CT)=63-81 Ka. Again tending to >70 Ka.

All these are 95% CI (what always leaves a 5% chance of error outside the brackets, assuming everything else is correct).

Maju said...

Re. the colonization of Australia, there are now at least two solid dates pointing to slightly before 60 Ka ago.

→ http://www.sciencedirect.com/science/article/pii/S0047248499903056

Another date of c. 80 Ka has a wide error margin however, so it could be near 60 Ka as well.

This could be contradictory with nearly all age estimates but then of course those early settlers could have carried lineages now extinct like A* (not too likely but who knows?).

terryt said...

Thanks Maju, but I think the Mungo Man date has been hotly disputed. It was dated as part of the study that showed the mt-DNA did not belong to any modern group, including African ones. However I am quite prepared to accept dates for Australia's first settlement considerably older than 50 kya.

Kristiina said...

Shaikorth, I refer to your f-stats you posted and ask if my understanding is correct if I gather that by contrast to the modern variation where we have an increased affinity between West Eurasians, i.e. Europeans and Near Easterners, at the time of Ust Ishim, there was a higher affinity between all Eurasians and South Pacific populations to the exclusion of Near Easteners. I am wondering what could have caused this pattern. Do you have ideas?

terryt said...

"at the time of Ust Ishim, there was a higher affinity between all Eurasians and South Pacific populations to the exclusion of Near Easteners. I am wondering what could have caused this pattern. Do you have ideas?"

Ust-Ishim's Y-DNA is K2a of some sort. That means it is part of the K2 clade which is centred on East and Southeast Asia, especially the latter. K2a in particular is the East Asian representative of the clade, the others being specifically island SE Asian. In other words the most likely interpretation of all the evidence is that the Ust-Ishim population moved into southern Siberia from East Asia, thus offering no possibility of contact with the Near East.

Shaikorth said...

WHG's, Amerindians, Onge and East Asians are roughly equally close to Ust-Ishim. The options I see here are that Ust-Ishim-like population contributed to all W/E Eurasians but not to Near Easterners, or that something that wasn't shared with other Eurasians happened to Near Easterners, that would be things like extra gene flow with Africa.

Kristiina said...

The first explanation should require that the Aurignacian culture came to Europe from Central Asia. In that case, the Aurignacian y-line could well be C.

Without the modern sofisticated genetic methods, I remember that Cavalli-Sforza showed that Europeans are in the midway between East Eurasians and Africans. However, this affinity is surely not only due to African geneflow to Western Eurasia but also to Eurasian geneflow to Africa.

However, it is intriguing to think about the amount and effect on our looks of this African-related ancestry that should be important in Near-Easterners and also in (southern) Europeans, as well as from where in Africa it comes (all north Africa or only northeast).

Mike Thomas said...

I thought UstIshm position is already established, within the limits of our data we currently possess.

He was we might call part of a "crown eurasian" pop that split off after Basal Eurasian. He is equidistant to both ENA and a "basal NW Eurasian" (the latter in turn giving off ancestors of WHG and ANE). Ultimately, however, UI is an extinct line.

Mike Thomas said...

Kristiina
The Aurignacian didn't come from Central Asia. Well it might have, but there is zero evidence for it (even in considering that countries like Iran and Afghanistan are lagging in digs).

Kristiina said...

My observation was only based on Shaikorth's comment as an attempt to explain the lack of affinity of 40 kya Eurasians with the modern neareasterners, and not on archaeology, but when you look at the map of the Aurignacian culture on Wikipedia, it is difficult to say where it originated because also Turkey is blank. There is only an anomaly in the Levant. Is Wikipedia map inaccurate, and there are finds in Turkey?

Kristiina said...

When you look at the figure Shaikorth posted, the distance between for example Yemenese and Egyptians compared to almost all others is huge when measured against Ust Ishim and somewhat smaller when measured against Kostenki.

Mike Thomas said...

According to current state of research, the origin of "classic aurignacian" lies within Europe itself- in sites in Austria, Bulgaria, Bohemia etc.

It's precursor the" Emirian-Bacho Kirian" arose in the East balkans and the near East. This is beyond dispute.

In Turkey, most notable is Uglaci cave; but yes there are hardly any other sites. Maybe due to state of research, but there is always the possibility that it mived from near East toward Central Asia, then back to Balkans. But that is less parasiminous, but my mind is open to future evidence.

Kristiina said...

Why do you take it for granted that Aurignacian has originated in the Near East? Couldn't it have moved from the Balkans to the Near East along the coast?

In any case, I do not have any opinion on this, because I am not really a specialist in archaeology.

Kristiina said...

Thank you! That was a good explanation! However, it looks like the current research is unable to clarify whether Aurignacian people came from Central Asia or Near East or if there was a mixture of both.

Mike Thomas said...

Im not taking anything for granted, kristiina

And if your read what I wrote above carefully, you'll note that I didn't say what you iterated.

What I actually wrote was that the "classical aurignacian" of europe- as a technocomplex- actually arose in Europe. This was an internal development from preceding industries at sites like like Bacho Kiro (Bulgaria), Bohunice (Bohemia), etc; who in turn have undeniable technological and chronological links to near eastern sites of the so called Emirian industry. This has been confirmed time and again, and now confirm dated with modern methods like thermoluscence and AMS.

This "Emerian-Bacho Kirian" horizon is what palaeoarchaeologists link as the earliest undeniably human assemblages in eruope (barring other "transitional industries" whose modern human provenance remain unproven)

Now, Emirian -descended sites are also found in Central Asia, and even Siberia. But these date later than the near East.

So that is the current state of research. I have no reason to doubt it's veracity.

Mike Thomas said...

Sorry had to correct spelling

Well, as stated, it's clear that it came from near East, and not Central Asia, on the balance of current evidence.
But i highlight the poor state of research in Iran, etc; and admit that there might have been multiple waves not just one.

Perhaps Krause et al's upcoming paper will clarify these issues.

Maju said...

@Kristiina: "Without the modern sofisticated genetic methods, I remember that Cavalli-Sforza showed that Europeans are in the midway between East Eurasians and Africans".

He didn't show anything: he committed an error of reading a global K=2 as something meaningful, when it is not. In fact that same global K=2, depending on sampling strategies (notably more or less Africans in the samples, more or less West Eurasians) sometimes shows an Africa vs East Asia polarity (in which case West Eurasians appear as "mixed" - but it's a mirage) or a West-East Eurasian polarity (in which Africans appear as "West Eurasians"). All that implies that minor African admixture in West Eurasians (of the kind or E1b or Lazaridis' "Basal Eurasian") does exist and brings WEA slightly closer to Tropical Africa than some would have expected but not at all meaning what Cavalli-Sforza claimed (~ 67% Asian + 33% African mixture). He read too much in the global K=2.

Maju said...

PS- with smaller Asian samples and larger WEA and African ones the K2 should be "correct", showing an African-Eurasian polarity, with clinality only in the contact zone. That's because in that case the Eurasian component would be defined by WEA peoples, with EA trailing them. Anyhow a minimally meaningful K depth for global analysis is K=3, and a serious one is surely quite deeper. Statistical component analysis has limited inference power in the phylogenetic sense.

Maju said...

@Mike: Obi Rakhmat (Uzbekistan, NE of Tashkent) has some of the oldest dates for the earliest Upper Paleolithic ("Aurignacoid"), bordering 50Ka. A South Central Asian origin of the UP is therefore at least plausible but it's also possible that this Uzbekistan group is just a branch of an undocumented more southern origin somewhere between NW India and Iran. If so it should be precursor of the Altai branch, which is only slightly more recent. The Central Asian dates in general are contemporaneous of the European dates (or even more ancient, depending how you judge some of the oldest European datings).

What is less clear as of now is the relationship with the LSA (NE Africa), which is quite apparently the same phenomenon and roughly contemporary too but it is less studied in the fine detail (less clear chronologies, etc.) My impression based on genetics (which show much more Eurasian influence, and seemingly quite old, in NE Africa than vice versa) is that LSA is derived from Eurasian (West Asian) early UP, but I could be wrong.

Maju said...

@Mike again: "The origin of "classic aurignacian" lies within Europe"...

The origin of Aurignacian proper is in Istallosko (NE Hungary) but Aurignacian senso stricto (which only affects Europe and since c. 41 Ka BP in most regions) is just "the end" of the wider "Aurignacoid" or early UP migratory phenomenon that saw H. sapiens taking over nearly all of the Neanderthal regions of Europe, Central Asia and Highland West Asia. Recently it has been demonstrated that North Italian proto-Aurignacian were H. sapiens, similarly S. Italian Uluzzian also seem to be H. sapiens (they apparently crossed from Greece). And there are many other early UP cultures and groups in Europe, West and Central Asia, all with similar features and similar chronology of before 40 Ka BP.

You may "read only peer-reviewed papers" but obviously you are not reading enough of them. Anyhow you're clearly confusing Aurignacian senso stricto with Aurignacian senso lato, aka "Aurignacoid", aka initial Upper Paleolithic (or most of it).

For those who can read Spanish, here there is quite good synthesis by Prehistorian David Sánchez: http://prehistorialdia.blogspot.com/2014/03/la-expansion-de-homo-sapiens-hacia.html

Maju said...

@Kristiina: "Why do you take it for granted that Aurignacian has originated in the Near East?"

That's an excellent question. The known chronologies for anywhere in West Asia are too recent. And that's why David Sánchez (the maps and graphs are accessible for all) suggests a steppe expansion from Central Asia. However I'd contest it because the chronologies of Central Asia, Europe (not just Central Europe but also some Pyrenean sites like Isturitz) and the Caucasus are roughly the same, hence it's probable that they all had some West Asian origin. The problem is that very few sites outside Palestine (Israel) have been studied in that area at least with sufficient rigor, also some areas like Central Turkey are highly erosive and don't retain the archaeological evidence in a context that can be properly studied.

So there are big question marks in West Asia, South Asia and NE Africa regarding this issue, with West Asia being pivotal. Otherwise West and Central Europe, the Caucasus and Central Asia show very similar chronologies, suggesting a radial expansion from West Asia. But where exactly from? An interesting potential region highlighted by recent archaeological research in Arabia is the so-called Persian Gulf Oasis, i.e. the Persian Gulf itself, which was an emerged marshy area in the Ice Age and probably acted as refugium in the arid periods that dominate the late Middle Paleolithic and all the Upper Paleolithic, excepted the Mousterian Pluvial c. 50 Ka BP (which is precisely when the UP formed and expanded). However as of yet I'm not aware of any site that could shed light on first UP development in that region.

Another potentially promising region are the Zagros. We can't forget the issue of the Neanderthal man (Zewi Chemi) with a demonstrated projectile injury suggestive of West Eurasian H. sapiens favored weaponry (atlatls, only demonstrated in Gravettian however because human remains in Aurignacian, senso lato, are very rare). I personally lean for somewhere in the Zagros area without excluding Iran-Pakistan-Afghanistan-NW India in general because in that Pluvial period the deserts of today would be much more hospitable (and anyhow a relation is demonstrated in the presence of microblades in India since c. 38 Ka BP).

Kristiina said...

Maju, thank you for sharing your ideas! I appreciate your knowledge in archaeology.

I still think that we need an explanation to the huge difference between ancient Eurasians and Near Easterners. It appears that the Neolithic or post-Neolithic gene flow from Africa into Near Easterners should be massive to explain the difference. At least, I do not believe that the ever repeated slave trade is enough to explain this difference.

Mike Thomas said...

Maju

True to your style, you've come out swinging with Straw man arguments - mostly due to your own confusion. Im getting worried about you - is your eyesight going ??

I've clearly differentiated between aurignacian propper (or narrow) which is a european phenomenon, and the broad pre-Aurignacian phenomenon which precedes it- sites like Bacho Kiro, Bohunice and Emirian sites in the near East, as well as similar sites which stretch to Siberia as early as 45 kya (which I clearly stated!)

The earliest phase of AMH technology * in Europe* is called the Emirian-Bohunician- Bacho-Kirian horizon (EBB); or "Aurignacoid" to use your typically sloppy and self-invented terminology.

Now, the "proto- Aurignacian" sites you refer to are in Spain, Italy, etc. I wholly agree they are o
the earliest manifestation of AMHs via a Mediterranean / coastal rout, but they nevertheless seem anchored in SE Europe (where earliest PA finds are found in Temneta Dupka). I merely left discussion of PA out becuase it post-dates the Emirian-Bacho Kirian, and its modern human affinitiy is not doubted, esp in light of the recent paper which we all know about. But in terms of techno-complex; it differs from Classical Aurignacian; so much that Mellars suggests it be called "Fumanian"- after the eponymous Italian site.

So when I say "Aurignacian' - I distinctly mean aurignacian propper of Europe; when I want to anything else, I'll say so. I'm merely following the terminologies of academics, and not Maju's world.

Also, Contrary to what you pontificate, the Ulizzian has not be "shown" to be AMH. I'm aware there is a big drive by italian scholars to claim it as such, but in reality it's lithic technology is clearly MP dervived, and on the whole restricted compared to aurignacian sites. Moreover, there have been not confirmatory finds by either DNA or even classical morphology (apart from a tooth or two dubiously classed as "modern"). So that's why I left this out.

Maju said...

@Kristiina: the difference seems to be polarized to Yemen (and Egypt, which is already Africa) rather than West Asians in general, who are much closer to the WEA normal. So that's quite clearly the clue and surely it has some relation with Lazaridis' "Basal Eurasian" as contrasting with mainline Eurasian. West Eurasians (even Finns!) probably have some "Basal Eurasian" admixture in general that Ust Ishim did not. Another thing is what is this BEA thing: leftovers of the OoA migration or African admxiture or a mix of both.

Maju said...

@Mike: Not in agreement at all. There is a wild diversity of "Aurignacoid", some more similar to "true Aurignacian" other less so but it'd be extremely complex to even understand enough to be able to debate those degrees of similitude/difference, so I'll let it to prehistorians, who often disagree among them in this aspect. We cannot in any case limit it to Emirian and Bohunician-Bachokirian because Swabian "Aurignacian" (c. 49 Ka calBP), Pyrenean Proto-Aurignacian (c. 48 Ka calBP), Uluzzian (demonstrated to be H. sapiens indeed, Lincombian, aka early British Aurignacian, also, similar very old dates) and even Aurignacian I of Istallosko (c. 48 Ka calBP) are older than all these. And that's only in Europe: in Central Asian, as mentioned before there are also two regions (Uzbekistan and Altai) that are also older than Emirian itself (both are c. 47 Ka BP).

Your dates for Europe are obsolete. Please download this database, introduced here and recommended by several important prehistorians as John Hawks and CSIC researcher Millán Mozota. Bohunician-Bachokirian is clearly more recent than several other "Aurignacoid" cultures, although Bohunician is anyhow one of the oldest such cultures in Europe, it is not the oldest one.

As for the Levant early UP, in Ksar Akil (Lebanon) there is a Neanderthal jaw and the evolution seems rooted in Mousterian. The dates don't seem older than other initial UP industries. I can't say for sure but I'd suspect Neanderthal presence in the Levant until Ahmarian. Earlier Mousterian penetration to as far south as Yemen is documented also. So we can perfectly imagine a Neanderthal "buffer" zone between the African and Eurasian branches of Homo sapiens population after the OoA.

It wasn't probably until our kin developed ranged weapons and domesticated dogs that they could face the much stronger (and similarly smart) Neanderthals with advantage. Our kin's advantage was not in brawn, nor probably brains either, but in longer legs and lighter bodies, allowing for longer marches (Neanderthals were quite fast in sprint anyhow), what allowed us to reach farther than them into Asia (early boating maybe helped too, as the biggest Neanderthal naval feat documented, reaching a coastal island in Western Greece is ridiculously short, practically within swimming distance) and also to exploit larger areas (everything else equal), as it has been documented in Europe.

So in my view (add Mezhmayskaya to all the other data), the most probable origin of "Aurignacoid" (or first UP) should be around Iran-Pakistan. However it has yet to be documented.

Mike Thomas said...

Maju

Thanks for the links, i'll have read.

For now, as I stated Im open to more data from central asian UP, which is under-researched. I have suspicions based on some (? perhaps forthcoming aDNA) that it might have played a role.

But as for now, Im not going hold too much weight on your opinion, which you self-admittedly rests on unsubstantiated finds. Rather, I am hapy with the recent perspectives of actual experts (eg Hoffecker and Hublin (http://www.pnas.org/content/109/34/13471.full)(http://www.pnas.org/content/106/38/16040.full).)

Furthermore, i am happy to fall back to what actual palaeoanthropologsts and archaeologists define as Aurignacian strictu sensi, rather than what you claim to be so.

Moreover, you seem to be unaware of the dubiousness of old C14 datings (and hence the outdated summary you linked to), and that it is now conisdeered as gold standard to rely on calibrated chronology based on techniques such as thermoluscence, when it comes to dating over 40 kya.

All in all, it seems you're a few years behind the 8 ball on this issue- which is otherwise a no-brainer anyway unless you wish to claim that AMHs somehow parachuted into Swabia or Britain.

As for your Ulizzian teeth, Ive already informed you that their modern taxonomy, and in fact the entire context / stratigraphy of their finds remain disputed. In today's age of aDNA, I guess the onus lies on proving their 'modernity' with properly excavated skeletal remains and genetic analysis. Its a mere baseline standard i'd have thought. But Ill be the first person to celebrate if they are confirmed to be AMH, as it will show that there was diversely adapted pioneer AMHs entering Europe.

Maju said...

@Mike: my substantiated opinion is that the initial UP or "Aurignacoid" complex is nearly synchronous (50-45 Ka calBP) in the following areas: Central Europe ("Swabian Aurignacian" and Aurignacian I of Istallosko, as well as Bohunician soon aftewards), Pyrenees (proto-Aurignacian of Isturitz and possibly other sites in Catalonia and Cantabria), North Italy (proto-Aurignacian), South Italy (Uluzzian), Balcans, Caucasus (Mezhmaiskaya), Uzbekistan (Obi Rakhmat) and Altai. Instead the Eastern European steppe was yet Neanderthal Mousterian (only later becomes Aurignacian, true Aurignacian, but that's after the HE4), so, unlike David Sánchez, I think that the migration went through the South of the Caspian and Black Seas. Earlier all those regions were inhabited by Neanderthals, so it is a mass migration of H. sapiens from some other place, most probably from somewhere near India because that's where the genetic trail leads us in the case of F (→G), IJK (→IJ) and P1 (→Q and R1), which are ancestral to all relevant Western haplogroups. But I don't mean to impose my opinion (we can agree to disagree) but to exchange information and substantiated opinions (theories).

Your link for the Hoffecker paper is broken. I did however for some time follow Hoffecker 2009 (I believe that's what you meant to link to) but there have been other studies since then (very early dates for Swabian Aurignacian, confirmation of very old dates for Isturitz, etc.) and also Hoffecker does not deal with the issue of Central Asian or Caucasus "Aurignacoid", limiting himself to offer a good attempt at interpreting the European initial UP on 2009 data.

A critical issue is that, if effectively Bacho-Kiro dates have been revised as the linked database suggests (and none the actual prehistorians I communicate with has ever objected to that), then Bachokirian-Bohunician is not anymore the oldest European UP and actually Bachokirian appears as derived from Bohunician, which now seems older (but who knows: we don't have all the data, there will always be blanks and doubts for fine detail such as this).

"As for your Ulizzian teeth, Ive already informed you that their modern taxonomy remains and the stratigraphy of their finds remain disputed".

Believe what you will: if even fanatic Neanderthalists like Millán Mozota seem to reluctantly accept this as a fact, not contesting it without even a line (neither in blog, correspondence or Google group), I have to think that it is pretty solid.

"In today's age of aDNA, I guess the onus lies on proving their 'modernity' with properly excavated skeletal remains and genetic analysis. Its a mere baseline standard i'd have thought".

Not always is possible to obtain aDNA. Also today's understanding of teeth morphological differences between Neanderthals and us seems to be quite solid. I punctually follow other specialists on the matter and they seem to be very strongly persuaded of teeth morphology being very clearly definitory of the species.

But I agree to disagree with you, sure. No problem.

terryt said...

"WHG's, Amerindians, Onge and East Asians are roughly equally close to Ust-Ishim. The options I see here are that Ust-Ishim-like population contributed to all W/E Eurasians but not to Near Easterners, or that something that wasn't shared with other Eurasians happened to Near Easterners, that would be things like extra gene flow with Africa".

There is no mystery here. As I tried to explain before Ust-Ishim is from East Asia therefore it has no connection to the Middle East at all. Even it s mt-DNA is apparently related to that of Tianyuan, some sort of B which is also and eastern haplogroup. But it does more than adequately explain its relationship with "WHG's, Amerindians, Onge and East Asians.

Mike Thomas said...

Maju

In skeptical of the dates you provide, not least because they're based on C14 dates, which are unreliable (as I've mentioned). I'm also at a loss as to how you come up with hyper-inflated dates for west european sites when it is crystal clear that the aurignacian in the Swabia you mention isn't a day over 38ky BP (recent summary Conard, Bolus). The proto-aurignacian in Italy dates from only 42 ky BP.
Bacho Kiro does *not* dervive from Bihunice, nor is it "younger".

Whilst I disagree with much of what you say, your views are nevertheless interesting as they provide insight into some of the fringe theories which exist in the blogosphere

Maju said...

So ultrafiltered C14 dates are unreliable but "molecular clock" rocks? Ahem! C14 is pretty good, scientifically very solid since the very beginning, and has been improved a lot in the last decades.

"The proto-aurignacian in Italy dates from only 42 ky BP".

Yes, that's probably correct more or less (43 Ka from memory). But Isturitz is much older and has been recently confirmed in its antiquity.

"I know you "talk" to a lot of "academics""...

That's not true. Only a few - but I tend to trust their opinion (unless I have to disagree, that also happens).

Mike Thomas said...

Who's talking about molecular clocks, Maju ?

it is well known that there are major problems with C14 dating beyond c. 40 kya. That is why we need thermoluninescence dating (a "new" technique discovered 20 years ago).

Anyhow, I share your interest on Central Asian aurignacian . Know of any good papers ?

Maju said...

I already mentioned the Obi Rakhmat study. Otherwise I'm only really familiar with Altai:

Derevianko's "The Middle to Upper Paleolithic transition in the Altai" is very important to understand the early UP (and previous Mousterian) of the region. Sadly it seems to have vanished from the Internet except in citations.

2. Vasiliev on the C14 prehistory of Siberia and relevance for the colonization of America. This one is pretty good too, and more recent, but it's rather focused on the Eastern expansion of early UP.

I'm also missing another bookmark (404 error) with similar content as above but more focused on Mongolia and North China rather than Siberia (Russian side of the border). Can't recall the author.

Mike Thomas said...

Ok thanks
I know certainly that marcel Otte's numerous papers afire for centrala asian origin of Aurignacian. But I'm taking his conclusions with a touch of salt because of limited sites

Kristiina said...

Thankyou, now I have a good explanation that will hopefully be confirmed or modified in the future: modern humans were expanding from East to West somewhere North of the Caspian Sea and the Black Sea. In fact, yDNA K is India/Central Asia-centered and also yDNA C was frequent in Europe and not in the Near East(?). As for mtDNA, R probably developed in the same area.

There were many Neanderthals living in the Near East, Turkey and Europe and these Neanderthals separated North Eurasians from Near Easterners which caused this huge genetic difference between them 40 000 kya. When Neanderthals disappeared and tighter contacts were established, this distance started to shrink.

Mike Thomas said...

Kristiina

I don't know where u keep imagining this North of the Black Sea scenario ??


At present, the only solid and confirmed data we have undeniably points to southeastern europe- namely Bulgaria as the entry point of AMHs into europe, in turn, with links to sites in southern turkey, Israel and Syria.

Yes a handful of early dates also exits in Southern Russia- but these in anyway are **later** (40 vs 48 kya) and became extinct after the CI. Quite clearly humans were not yet adapted to Northern latitudes and must have arrived via more temperate land bridges (Asia Minor- Balkans).

(Bulgaria is *south* (& west) of the Black Sea FYI)

At present there is *no evidence* that Central Asian -> russian sites are earlier than than near eastern -> Balkan. If anything, all the evidence points to the very opposite.

Kristiina said...

You are better in archaeology, so you may be right in what you say, but my point is that there was a kind of Neanderthal barrier between northern ancient modern humans and Near Easterners blocking/ hindering gene flow at Ust Ishim's time depth.

I still think that if Eurasians entered Europe through the Near East during the Aurignacian and the so called North African Caucasoid populations are this old (30-40 kya ???), there could not be such a huge difference between Ust Ishim and modern Near Easterners such as Egyptians.

In any case, I may be wrong but it really doesn't matter as the change of ideas is more important.

terryt said...

"In fact, yDNA K is India/Central Asia-centered"

Depends on how far down the line you consider K. K2, the ancestor of both NO and P, is almost certainly Southeast Asian although it presumably arrived there via India. And P most probably also moved back west through India.

"yDNA C was frequent in Europe and not in the Near East(?). As for mtDNA, R probably developed in the same area".

mt-DNA also almost certainly developed in Southeast Asia. The Kostenki Y-DNA C is some sort of C1b, presumably having arrived via India but now the specific line is extinct. Surviving European Y-DNA C is a separate line C1a2 which has its closest relationship with Japanese C1a1. The pattern of movement is far more complicated than just a sojourn in the Middle East before expanding from there in one major movement in all directions.

"my point is that there was a kind of Neanderthal barrier between northern ancient modern humans and Near Easterners blocking/ hindering gene flow at Ust Ishim's time depth".

I don't think the Neanderthal presence had anything to do with the separation. The haplogroup evidence shows the first North Eurasians we have genetic evidence for (represented by Ust-Ishim) entered from the east, not from the Near East. They had separate origins going back to from soon after humans first emerged from Africa.

Mike Thomas said...

Kristiina

I agree with terry
I don't think that the divergence of UI has anything to do with Neanderthals ; and moreover we don't actually have many UP samples to make any definitive conclusions

In fact, if you look at UI and K14; they both possess very basal comonenents , just in different proportions. And also remember that UI is just as divergent to East asians as he is to WHG.

Back to Neanderthals; the latest Neanderthal sites in EE are from the Adriatic coast and Greece, on the one hand; and east cetral european plain on the other (Szeletian culture etc) -ie exactly where your proposed path lies

Mike Thomas said...

Terry

"mt-DNA also almost certainly developed in Southeast Asia. The Kostenki Y-DNA C is some sort of C1b, presumably having arrived via India but now the specific line is extinct. Surviving European Y-DNA C is a separate line C1a2 which has its closest relationship with Japanese C1a1. The pattern of movement is far more complicated than just a sojourn in the Middle East before expanding from there in one major movement in all directions. "

Interesting points. HHmm . I guess that makes central Asia a strong possibility.

I can only wait with eargerness for more aDNA from near East and central asia to clarify these intriguing debates.

Kristiina said...

Terry, you do not see my point, so I repeat it: if there was a flow of modern humans from east to west 40 kya (as seen in yDNA K and C and mtDNA R) and they entered Europe through the Near east, why Near Easterners remain so much more deviant from Ust Ishim like humans than Europeans do. Of course, you can say that it is all due to African gene flow into Near Easterners or even due to the non-Eurasianness of Egyptians, but I do not know if it is enough. May be.

I still find it intriguing that 40 kya the biggest genetic difference was not between East Asians and West Eurasians/Near East but between Eurasians/Pacific Islanders and North Africans/ Arabians.

Maju said...

@Kristiina: as far as I can discern the flow of humans from the East you visualize, happened via North India (and from SE Asia) and it was quite older than 40 Ka BP. We can track it very well in the case of K2, which seems to have an origin in SE Asia (possibly Sundaland) and the following branching:

1. K2a=NO to North
2. K2b stays put (or moves to Sundaland if the coalescence of K2 took place in Indochina)
2a. P stays put but P1 already (back-)migrates to India.
2b. K2b1 (M, S and several Australasian K subclades) migrates towards Papua.

Based on Karafet 2014, I synthesized the findings here. The whole K→P process takes almost 20 Ka (according to Karafet's estimates).

K as such only has another sublineage (K1=LT), which is very clearly centered in Pakistan. So the origin of K as such must be somewhere between Pakistan and Malaysia, most probably still in South Asia.

As for Ust Ishim my main theory is that his lineage is an erratic of a wider K2-derived pool dominated by P1, which eventually gained an even stronger dominance (by reason of drift and replacements). However its similitude with K2a=NO (as well as autosomal greater affinity with modern East Asians) may indeed suggest that there was some minor E→W flow via Siberia, otherwise completely undocumented: not at all archaeologically and also inconsistent with the pattern we can discern in modern K2. The case of European C (La Braña and very few modern relatives) also fits with the same scenario, as undoubtedly early C and early K2 intermingled to some extent in the area of SE Asia principally. Some C back-migration to India is also apparent (as I'd argue that the coalescence of C seems SE Asian in geography, just as that of K2 but maybe older).

Kristiina said...

Maju, I know your South East Asian theory, although some recent results have weakened it: the frequency of C in whole Eurasia, finding of XNO in Ust Ishim and the younger coalescence age of Chinese N branch compared to European/Siberian N branch. In any case, let's see what future research will show us.

I checked the Eurogenes K12 results of Yemenites, Egyptians, Ust Ishim, Kostenki and Ma1. On the one hand, a salient feature is the lack/paucity of Transcaucasian in Europe (Karelian/ Hungarian/ Motala hunter gatherers) and Ust Ishim and its higher frequency in Egyptians and Yemenites. On the other hand, Egyptians and Yemenites lack Euro-HG which is high in Ma1 and found also in Kostenki and Ust Ishim. Then of course, Ust Ishim, Kostenki and Ma1 all have a lot of South Asian which is lacking in Egyptians. Instead, Egyptians carry 0.26-0.32 of Northeast African, but I do not know if this is enough to put Egyptians in such an extreme position in the figure.

Maju said...

I wouldn't put too much trust into zombie-based calculators. It is clear that North Africans, Egyptians included, have some strong Euro-affinities when you let them run freely, as found by Henn and myself. I doubt it can be all attributed to the EEF component, which is anyhow like 50% WHG-like (UHG).

Mike Thomas said...

Kristiina
Didn't the recent study by Kang supposedly confirm the Chinese origin of Hg N (http://dienekes.blogspot.com.au/2015/05/structure-of-y-haplogroup-n.html?m=1);
Or are you skeptical of their conclusions ?

terryt said...

"I guess that makes central Asia a strong possibility".

And a long-term one. I strongly suspect that humans reached Eastern Eurasia via Central Asia quite some time before they moved in South Asia. The mt-DNA haplogroup N has long been noted as having separate eastern and western branches. My guess is that branches through Central Asia were exterminated soon after the original movement although it seems now that A may be some sort of Central Asian remnant. My further guess is that Y-DNA C accompanied N east resulting in C being the eastern representative of CF while F is the western representative.

"I can only wait with eargerness for more aDNA from near East and central asia to clarify these intriguing debates".

I could not agree more.

"Terry, you do not see my point, so I repeat it"

I still don't see your problem. None of the early haplogroups appear to have entered Europe via the Near East.

"you can say that it is all due to African gene flow into Near Easterners"

I don't see that as necessary. Ust-Ishim certainly never went anywhere near the Near East and so it is obvious why he is so different from Near Easterners. The only K in Europe then remains R, with a little bit of Q, much more recent than 40 kya. It also seems doubtful that R entered Europe via the Near East.

"I still find it intriguing that 40 kya the biggest genetic difference was not between East Asians and West Eurasians/Near East but between Eurasians/Pacific Islanders and North Africans/ Arabians".

Yes, but not really a surprise.

"As for Ust Ishim my main theory is that his lineage is an erratic of a wider K2-derived pool dominated by P1"

I don't think so. Ust-Ishim K is K2a of some sort. In other words related to NO. A completely separate branch from K2b, source of P. I agree Ust-Ishim was probably an erratic, but associated with just a wider K2a pool.

"may indeed suggest that there was some minor E→W flow via Siberia, otherwise completely undocumented: not at all archaeologically and also inconsistent with the pattern we can discern in modern K2".

Probably a very short-lived E-W flow that rapidly became extinct. Modern western K2, in the form of P, back-migrated through India and later into Siberia. A separate population movement. I agree early Y-DNA C in Europe (Kostenki C1b*) probably moved there through India, but La Brana C1a2 connects with Japanese C1a1 and shows no South Asian connection.

"I'd argue that the coalescence of C seems SE Asian in geography, just as that of K2 but maybe older"

Agreed. Although possibly somewhat north of SE Asia in the case of C as a whole.

"I know your South East Asian theory, although some recent results have weakened it"

I haven't seen anything that weakens a SE Asian origin for K2 as a whole.

"XNO in Ust Ishim"

That supports the above concept of a widespread K2a, present through East Asia from Northeast India to Ust-Ishim.

"the younger coalescence age of Chinese N branch compared to European/Siberian N branch".

I agree with Mike.

terryt said...

I've just noticed a mistake in my post yesterday:

"mt-DNA also almost certainly developed in Southeast Asia".

I left out the 'R'. I meant to say: 'mt-DNA R also almost certainly developed in Southeast Asia'. R is a basal N haplogroup and we have no basal N between Australia and the Malay Peninsula. Yet Australia contains four basal N haplogroups plus a highly diversified basal R haplogroup in the form of P. And the gap between Australia and the Malay Peninsula is filled with basal R haplogroups: R12/21, R14, R22, R23 and the greatly diversified B.

Kristiina said...

Mike, yes, that was their conclusion, but their coalescence ages are the following:
N-M231 mainly Han and Mongol 9.8 kya, linearly calibrated 15.8 kya

N1-F2130 mainly Han and Mongol 8.4 kya, linearly calibrated 13.5 kya

N2-F2930 Mainly Han 6.7 kya, linearly calibrated 10.8 kya.

In their calculations they used mainly Han and Mongol samples, and, as a result, the Northern/Siberian branch was older than the southern branch. However, it should be borne in mind that in their age estimations they disregarded the western branches that split first, i.e. N-P189.2 and N-L732.

I am waiting for new results, in particular ancient ydna, before making any definite conclusions.

Mike Thomas said...

Ok interesting
I have no expertise in Hg N; but two things to keep in mind

1) the phylogeny of NO within K looks like rooted in SE Asia rather than anywhere west of India at most

2) is the older coalescence of Hg N you refer (presumably calculated by various genealogists etc) a true calculation ? Ie did they remember to disentangle the various subclades so as not to arrive at a falsely high coalesce of N in west eurasia when it in fact represents high diversity due to multiple gene flows ?

Terry / Kratina / Maju

Could not Hg C be simply an early off shoot (the earliest) coming to europe straight from near East ? I suspect mega-group CT began to diversify in the near / Mid East as the population grew and expanded East. This dovetails with the centering of F & G in MEast, H in west India, IJ somewhere in west asia also; with K being the sole truly eastern lineage. On the other hand, C might have been ubiquitous although the different sublineages in Kistenki vs La Brana do not allow a straightforward narrative

Kristiina said...

In this new paper

http://genome.cshlp.org/content/suppl/2015/02/18/gr.186684.114.DC1/Supplemental_Tables_S1-S7.pdf

http://genome.cshlp.org/content/suppl/2015/02/18/gr.186684.114.DC1/Supplemental_Text.pdf

the age estimate for N1-F2130 is 16 kya, and 12.5 - 22.5 kya considering uncertainty due to the confidence intervals of
the mutation rate 0.63-0.95x 10-9.

So, using non-chinese samples the coalescence age is considerably older; and even in this paper the Balcanic branches are omitted.

NB, I do not know how comparable these two papers are.

The Chinese paper says that:
The age of each clade was calculated as the average STR-based time between the mean STR values (supposed as the ancestor node) and each modern member using Slatkin’s method and father-son pair Y-STR mutation rate, and the dates were linearly calibrated to the divergence time of N1-F1206 and N2-F2930 at 15.8 kya (thousand years ago), which was achieved from the high-throughput sequencing of Y chromosomes18.

The latter paper says that:
Table S7. Age estimates of Y chromosome clades reported in Figure S3
No - node number, Name -node name, Post - posterior support to the clade, Age - average age estimate (years), using 0.74x10-9 mutations/base/year mutation rate (see SI3 for further details), Lower - lower 95% boundary, Upper -upper 95% boundary of the age estimate, a -considering only the variance of branch length estimation in BEAST, b - considering also the uncertainty due to the confidence intervals of the mutation rate 0.63-0.95x 10-9.



Kristiina said...

IMO, a problem with all of us is that we underestimate the amount of extinct y lineages that have however contributed autosomally to surviving populations. We have very old haplogroups, such as C, P, E, NO, that had plenty of time to cover huge distances and become autosomally very varied. A successful younger y line could pop up in a place that was very far from its point of origin, and as a result, it probably developed different autosomal traits from its point of origin.

terryt said...

"I am waiting for new results, in particular ancient ydna, before making any definite conclusions".

Kristiina, check out ISOGG. Your N-P189.2 and N-L732 are now N1a and N1b respectively. They also show a N1c-L729. I can't work out where these fit with your N1 and N2 as the F2130 and F2930 mutations are not listed. Seems as though both the studies leave branches out.

"the phylogeny of NO within K looks like rooted in SE Asia rather than anywhere west of India at most"

Agreed.

"Could not Hg C be simply an early off shoot (the earliest) coming to europe straight from near East ?"

Possibly, but C is primarily and eastern haplogroup and western C belongs to two separate branches within C1. Those two separate branches have separate distributions in the east. One is now Australian/Indian and points between and the other Japan/Europe and nowhere in between. Interesting situation.

"I suspect mega-group CT began to diversify in the near / Mid East as the population grew and expanded East".

I think that, in general and especially in small populations, diversity is a product of geographic expansion. In other words haplotypes confined within a small geographic region do not diversify. Descendant haplogroups do tend to occasionally replace parent haplogroups but probably in a somewhat random manner. This also has the effect of making any sort of molecular clock pretty unreliable. I would think that CT diversified as it spread out and formed geographically separate branches rather than diversifying and then spreading.

"This dovetails with the centering of F & G in MEast, H in west India, IJ somewhere in west asia also; with K being the sole truly eastern lineage".

Somewhat correct. G is Middle East, yes, but under recent phylogenies F1 and F2 are specifically South Asian and F3 specifically East Asian. Of the H Y-DNAs just H1 is south Asian with H2 being Middle East and H3 Southeast Asian. So K is not ' the sole truly eastern lineage'. Otherwise: yes.

And C doesn't fit that geographic spread at all. C2 forms six branches and is very much East Asian, with two branches specifically Japanese (C2f and C2a) and C2b1a American. C1 forms two branches, C1a in Japan and Europe, as does C1b with C1b1 forming separate branches in South Asia and Borneo, and C1b2 forming separate branches in Eastern Indonesia and Australia.

terryt said...

A further thought. The geographically specialised F haplogroups argues very much against your idea that 'mega-group CT began to diversify in the near / Mid East as the population grew and expanded East'. If that were so the haplogroups would not be so geography specific but their distribution would be patchy all the way from the Middle East.

terryt said...

"we underestimate the amount of extinct y lineages that have however contributed autosomally to surviving populations".

Agreed. And we know of three definite examples. Many will have been lost leaving no evidence.

"A successful younger y line could pop up in a place that was very far from its point of origin"

Unlikely I think. Especially in the Paleolithic when survival itself depended on being part of a tribe. In general I think populations containing specific haplogroups tended to expand rather like ink through blotting paper, but being diverted by geography at times of course. After initial expansion and exhaustion of resources extinction would be the lot of much of the regional diversity that had developed during the expansion.

"it probably developed different autosomal traits from its point of origin".

I think the main driver of autosomal diversity was population mixing. Of course populations that became geographically isolated would sometimes undergo severe selection if survival was tight. Thosegenes that developed would then be spread when another population reached them and mixed once more.

capra internetensis said...

@Terry

C1a2 was also found in Nepal by Hallast et al. There is a good deal of untested C* in China and India which might end up falling into C1a as well (or not).

IMO the old primary branches of C2 are no longer very useful. C2a-M93 has only ever been reported in one man in one old study that I know of, so should probably be considered a private mutation; also, it has not been tested for any upstream SNPs of the main C2 branches. So it has no weight as a primary branch and should probably be ignored altogether. The case of C2d-P62 is similar. C2c is defined by P53, which is an unstable mutation; I have read somewhere that it may fall under C2b.

C2b, C2e, and C2f on the other hand appear to be solid: C2e and C2f form a clade, but with a very ancient split. A C2e'f* has been reported in a southern Han man.

But I agree with your overall conclusion - on the present evidence, C2 looks more-or-less Northeast Asian, though of course that could change with further study.

Kristiina said...

Now that we are getting ancient yDNA, things are not any more what they used to be. Most of us expected that Neolithic farmers should have been J, E or R1b and instead, samples have been mostly G, F and I2. According to the old school thinking La Braña should have been R1b but he was C which came as a complete shock to everybody. When Kostenki came out, I thought it should have been IJ and not again that ”crazy” C. When Ust Ishim was coming, many were hoping for P and he was instead XNO. In Samara, most of us were sure that R1a must be found and all were R1b. Personally, I thought that the Karelian hunter gatherer should have been I*, N or Q and he was R1a. Anzick is among the rare ancient samples that met our expectations.

I do not know if you share my feelings about this but this have been often hilarious.

Davidski said...

C in Paleolithic Europe wasn't a shock to me. I was expecting it there after this paper came out.

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0049170

See the section “Out of Africa” haplogroups.

Mike Thomas said...

Terry

Yes, I meant CT diversified as it expanded East, not diversified within NE, then expanded.

As for C, sure today it's "East asian", but what about 50 kya ??

Kristiina

Your right about Y haplgroups, they're just patrilineage, and represent only 12.5% of overall diversity from as near as 3 generations worth of inheritance. Often forgotten. They are also liable to rapid and wide "replacement".

Kristiina said...

"Haplogroups tended to expand rather like ink through blotting paper"

Terry, initially it was probably like you say, but when we look at the situation from 30 or 40 kya perspective, the picture is different. It looks like nearly all y haplotypes get extinct at a certain time depth and a successful y-line may be the only one that survived on its long path before its own extinction sooner or later.

terryt said...

"Yes, I meant CT diversified as it expanded East, not diversified within NE, then expanded".

Sorry. I misunderstood.

"As for C, sure today it's "East asian", but what about 50 kya ??"

I am reasonably certain C1 had arrived in Australia by 50 kya. That's fairly 'eastern'.

"Your right about Y haplgroups, they're just patrilineage, and represent only 12.5% of overall diversity"

I agree, but any level of rapid genetic expansion is more often associated with population expansion rather than just selection. Population expansions are usually led by males rather than by females. Consequently male haplotypes serve as a pretty good indicator of population expansions.

"They are also liable to rapid and wide 'replacement'".

By population expansion even if only by male lineages.

"It looks like nearly all y haplotypes get extinct at a certain time depth and a successful y-line may be the only one that survived on its long path before its own extinction sooner or later".

But surviving haplotypes are presumably the survivors of previous expansions.

"I do not know if you share my feelings about this but this have been often hilarious".

All the ancient discoveries have made sense to me although I was a little surprised that K2a was as old as 50,000.

"C2a-M93 has only ever been reported in one man in one old study that I know of, so should probably be considered a private mutation; also, it has not been tested for any upstream SNPs of the main C2 branches. So it has no weight as a primary branch and should probably be ignored altogether. The case of C2d-P62 is similar. C2c is defined by P53, which is an unstable mutation; I have read somewhere that it may fall under C2b".

Thanks.

"C in Paleolithic Europe wasn't a shock to me. I was expecting it there after this paper came out".

But European C is hardly basal. It is at least post the split between the ancestors of C1 and C2.

Krefter said...

@Krist,
"It looks like nearly all y haplotypes get extinct at a certain time depth and a successful y-line may be the only one that survived on its long path before its own extinction sooner or later."

I've noticed that lineages that became popular founder effects and had a time period of rapid expansion came from pops with lots of their brothers who died out.

Think of it this way: There's a mommy R1b in Mesolithic Russia, and she has a large pen of R1b puppies. From the Mesolithic-Bronze age those R1b puppies fight each other for survival. In the end only one survives: R1b-L23.

It's not that no one was R1b-rich till the Bronze age. It's that there was a R1b-rich pop who had a son(puppy) who killed off all of his siblings.

In Mesolithic Russia we find two R1 lineages which today are extinct, but were brothers of other R1 lineages which are popular today.

K14 had an extinct C1* who was killed off(not literally) by the C1 ancestors of La Brana-1 who would become C1a2-V20.

MA1's R* was killed off other Rs.

I'm sure we'll eventually find an I* or IJ* in Upper Palaeolithic Europe.

terryt said...

Sorry. A bit more:

"C1a2 was also found in Nepal by Hallast et al. There is a good deal of untested C* in China and India which might end up falling into C1a as well (or not)".

My guess is that C1a was a widespread haplogroup at some time and what we have left is the survivors of that wide expansion. But that expansion most likely originated in the East.

But further to Y-DNA K2. I see ISOGG has incorporated the Karafet et. al. study as well as other information. The highlight for me was the resurrection of the old mysterious Indian K1-M147 haplogroup. It is now placed on an equal footing with K2b, K2c and K2d. It is a bit ambiguous but K2b2 of P has been promoted to equal footing with that lot too although still labelled as a member of K2b. That leaves open the possibility that K2b2 is actually Indian as well. Also somewhat ambiguously all the K2s apart from K2a (or NO) have been united in one group. That makes sense to me as I have assumed that K2 emerged from India into East Asia via the Burma/NE India/SW China border region. From there K2a went north and K2b went south.

terryt said...

" From the Mesolithic-Bronze age those R1b puppies fight each other for survival. In the end only one survives: R1b-L23".

I don't think fighting between them is the reason for extinction within clades. As far as I'm aware it is usually considered that diversity develops, and new haplogroup branches form, on the advancing margin of a population as groups become isolated from the core. Therefore it is unlikely that close 'relations' would meet each other until descendant haplotypes began their own expansion.

Krefter said...

I didn't mean literal fighting. It took thousands of years for the Y mutations to pop up. No one back then was aware what was going on. Many of the mega-lineages in humanity were probably on the winner's side in conflicts though that helped lead to their popularity. But it's not like they knew what Y DNA haplogroup they had, and intentionally killed off other haplogroups.

Mike Thomas said...

Krefter

I think they main "killing off" back then was by the environment, hunting availability; and indirect competition between groups. Sure there might have been competition over prized hunting grounds etc, but the main killer was Mother Nature, and not fellow man

Mike Thomas said...

Terry

Are you suggesting that the genetic evidence points to a colonisation of Europe from Asia, perhaps Central asia, rather than Near East ?

Thereby implying there is a wealth of yet unknown or invisible UP finds in that region which should date to between 80-50 kya ??

Maju said...

@Krefter: First of all R1b (Y-DNA) is a "daddy" not a "mommy". Secondly, neither humans nor wolves/dogs fight each other the way you say: both species are social and that means that they tend to cooperate above anything else (and that's a big key to our success, incidentally, be it as whole species or as distinctive populations). The wolf-eats-wolf reality you imagine is not real, at least not so primarily and overwhelmingly real as you fantasize.

Your fantasies rather belong to imaginary orcs or goblins than real people (or canids).

Maju said...

Actually I read recently about sibling cooperation, particularly among males, being important in anthropological reality. So, if anything, a lineage may succeed because the brothers cooperate (at least for some generations) rather than because the brothers fight each other to death (totally unreal).

Additionally, since Bronze Age onwards (or maybe Chalcolithic even), the increasing specialization and hierarchization of societies implies even more distinctive dynamics: dynamics of power (oppression and slavery/serfdom), quasi-feudal dynamics. In these elitist dynamics, the >90% of the peasant masses are not exterminated or displaced in conquest but usually enslaved or otherwise put to service of the conquering elites. They survive well therefore. Additionally there are many historical societies in which even this "feudalization" was minimal, for example Basques but also Nordics or even Athenians, etc., and where the freeman (farmer-warrior) predominated. Not everywhere was like Sparta, in fact Sparta is the oddball one (although it does reflect the oligarchic-militarist trend to its extreme).

You are putting the cart before the horses: your ideas about how the reality "should be" before the facts. That's called wishful thinking.

Maju said...

@Mike: I'm with Terry in the general aspects of what he's saying. And the archaeology is not that invisible: we have African typology (MSA and Aterian like) tools in India since c. 100 Ka BP and H. sapiens remains in East Asia since c. 100 Ka BP. We have also the Jurreru valley sites that show that African style (generic MSA-like) MP toolkits since 80 Ka BP and surviving through the Toba ash layer, rather surprisingly. The evidence grows ever thicker after Toba.

It is true that some sites, old findings, are poorly dated and that there is a huge void of research especially in SE Asia. But overall the evidence is rather supportive of an MP (MSA-like) migration to Tropical Asia (Arabia/Palestine c. 125 Ka BP, South and SE Asia c. 100 Ka BP and finally Australia c. 60 Ka BP).

Genetically it is pretty much unmistakable as well (you may appreciate it better in the mtDNA than the Y-DNA but both are quite "parallel" anyhow, see here). It is clear that West Eurasia was colonized from Asia and not the other way around (Arabia Peninsula and nearby areas like the Persian Gulf "oasis" excepted).

terryt said...

"You are putting the cart before the horses"

I totally agree.

"Are you suggesting that the genetic evidence points to a colonisation of Europe from Asia, perhaps Central asia, rather than Near East ?"

Possibly.

"Thereby implying there is a wealth of yet unknown or invisible UP finds in that region which should date to between 80-50 kya ??"

Through much of that time Neanderthals were an obstruction to movement into Europe. However it seems Neanderthals did not expand into Central Asia until at least later than 70 kya. There appears to have been modern human movement eastward before that time as demonstrated by the high level of Neanderthal and Denisova genetic element in East and Southeast Asia. On the other hand I would expect to find 'Upper Paleolithic' remains in the region as the UP did not develop until around 50 kya at the earliest. The first modern humans into Australia, for example, did not have an Upper Paleolithic culture.

Krefter said...

Whatever, we can't make up stories about why certain Y DNA haplogroups are so popular. All there is evidence of when and where they originated and when they expanded. How is unknown and may never be known. Ancient Y DNA results has thrown out lots of surprises, so we should keep an open mind to all the possibilities.

Kristiina said...

”[Ust Ishim] may indeed suggest that there was some minor E→W flow via Siberia, otherwise completely undocumented: not at all archaeologically and also inconsistent with the pattern we can discern in modern K2".

”Probably a very short-lived E-W flow that rapidly became extinct.”

”As I tried to explain before Ust-Ishim is from East Asia”

Maju and Terry, I do not think that Ust Ishim’s yDNa came from China via Siberia. When I look at Ust Ishim’s Eurogenes K12 results, he seems to have originated in India:
South Asian: 35%
Oceanian: 11.4%
Northeast African: 11.3%
Southeast Asian: 11.3%
Euro hunter gatherer: 7.5%
EEF WHG: 6.9%
Sub Saharan 4.5%
Pygmy: 4,3%
East Siberian 3%
Siberian: 2%
Amerindian: 1.4%
Transcaucasian: 0,5%

If his y line had came from China through Siberia, he would have much more Siberian and Amerindian and less African and Indian. However, I agree that his line went extinct and his population seems to have mixed with Altaian Neanderthals later on (which was showed by the admixture graph), so his population did not necessarily pass any/much genes to surviving human populations.

Krefter, I think that, apart from the Mother Nature that Mike mentioned and the fact that half of ancestry is passed through females who do not pass their father’s y line (a man with 3 daughters loses his y line but his autosomal ancestry is however passed on), a third way in which y lines go extinct is the loss of reproduction success, with which I mean that there are many males that cannot get offspring even if they wanted to, because women are more interested in successful or at least moderately successful men. And even if less successful males get offspring, the same battle goes on in the next generation and his line easily goes extinct just for the lack resources, i.e. inability to provide a favourable start/protection for his offspring. As a woman I also have to add that I really do not believe that many y lines thrive because of rape, because children have to be maintained and cared for. I think that children of raped women usually had a short and miserable life.

Mike Thomas said...

Terry, Maju

Thanks for those perspectives.

Yes Id heard about the fossils eg in China which date to 80 kya, but I know they remained disputed / controversial

Mike Thomas said...

Kristina
I think davidski mentioned before you can't take too seriously K15s on such old samples.
Eg "west asian" doesn't appear to even exist; neither does "Red Sea".

Kristiina said...

Mike, I agree with you that there are plenty of uncertainties everywhere, but what is the evidence that Ust Ishim came from China through Siberia?

According to Fig S10.5 of Ust Ishim paper, Ust Ishim is not on the same branch with Han/Dai and Mixe/ Karitiana but on the Eurasian main branch.

http://genetics.med.harvard.edu/reich/Reich_Lab/Publications_files/Supplementary_Q_Fu_Genome%20sequence%20of%20a%2045,000-year-old%20modern%20human%20from%20western%20Siberia-s1.pdf

Maju said...

@Mike: The Southern China fossils date, in at least one case, to 100 Ka and they are pretty much uncontroversial today, at least the Zhirendong jaw, which has a very clear chin, a Sapiens-only trait. The other remain is controversial only on the dating (sapienness is even more unmistakable), with brackets, if I recall correctly, between 120 and 70 Ka. Some people are just entrenched in old paradigms and find difficult to modify them according to new evidence but that's not "controversy", just "knee-jerking".

AFAIK you also know about the African related industries of India dated to 100 Ka BP (98 Ka?), including unmistakable Aterian-style blades. These are most important archaeological references that doubtlessly indicate H. sapiens presence in India in those dates.

And you may also know, unsure, about the many findings of MSA-like industries in Arabia (and Palestine) in the 125-90 Ka bracket, which is the Abbassia Pluvial and maybe about the 70 Ka Mousterian migration southwards of (presumably) Neanderthals to Yemen, what seems to rather strictly detach for some time (some 20 Ka) the African and Asian branches of Humankind, limiting the bracket for the OoA to the 125-70 Ka period, although I'd say that it's quite clear that 125 Ka is the date for Arabia-Palestine and 100 Ka for South and SE Asia.

If you lack of references about some of these events, please ask or make a search in my blog.

Maju said...

@Kristiina: "Maju and Terry, I do not think that Ust Ishim’s yDNa came from China via Siberia. When I look at Ust Ishim’s Eurogenes K12 results, he seems to have originated in India".

Makes sense to me. As I believe I said before, my main theory re. Ust Ishim is that modern West Eurasians are somewhat admixed with true "Basal Eurasian" (remnants of the OoA in Arabia and the Persian Gulf) and maybe also with African blood (false "Basal Eurasian"). These traits should be more prominent in people like Yemenis, so it does makes sense. This places them more distant of a crown ancestor as could be Ust Ishim than East Asians, also because NE Asians have some "crown Western" admixture via proto-Amerinds.

"[Ust Ishim's population] seems to have mixed with Altaian Neanderthals later on"...

Why do you think that? I know of no such evidence (we should see it someone like Ma1, for example, but we do not). Also Altai Neanderthals seem to have been cornered in a northern cave c. 47 Ka BP, where they survived with distinctive Mousterian industry until c. 35 or 33 Ka BP.

terryt said...

If his y line had came from China through Siberia, he would have much more Siberian and Amerindian and less African and Indian".

Not necessarily. Remember Y-DNA is only part of the story. Haploid DNA, especially Y-DNA, tend to introgress to a considerable extent.

"his population did not necessarily pass any/much genes to surviving human populations".

The Y-DNA line became extinct. The population may not have.

"what is the evidence that Ust Ishim came from China through Siberia?"

His Y-DNA is K2a, basically an early branch of NO. K2b-e are all SE Asian while K2a is East Asian. ThHe NO related X is South Asian but is northeast Indian. And India seems to be very accommodating of haplogroup variety and so it is unlikely XNO has become extinct through the remainder of India. That makes it unlikely that Ust-Ishim's Y-DNA came from there.

"Ust Ishim is not on the same branch with Han/Dai and Mixe/ Karitiana but on the Eurasian main branch".

But it probably predates the formation of the Mongoloid phenotype which would skew that comparison.

Maju said...

@Terry:

"And India seems to be very accommodating of haplogroup variety and so it is unlikely XNO has become extinct through the remainder of India".

We had back in the day endless discussions about this kind of stuff relative of mtDNA N, particularly X, which does not have Indian precursors but should, as other N, ultimately derive from SE Asian ancestors. It seems clear that, as someone said above, minor lineages do tend to vanish in some cases at least, while also having a founder effect further away.

I believe that in these "invisible trail" rather exceptional cases we have to look at the wider picture and fit them inside it rather than imagine whole alternative (and unsupported) scenarios specifically for the oddball one.

Kristiina said...

Maju, look at Fig S10.5 of Ust Ishim paper, p. 60. http://genetics.med.harvard.edu/reich/Reich_Lab/Publications_files/Supplementary_Q_Fu_Genome%20sequence%20of%20a%2045,000-year-old%20modern%20human%20from%20western%20Siberia-s1.pdf

There is an arrow going from Ust Ishim to Altai Neanderthals. I think that this interaction/conflict with Neanderthals is exciting and more genetic papers on this encounter seem to be coming up (like Dienekes' post on May 13, 2015).

Kristiina said...

In addition to the fact that Ust Ishim's population seems to have mixed with Neanderthals in Altai later on, he already had a big chunk of Neanderthal in his genome that seems not have been passed on to existing modern humans. (Figure S17.1. p. 98)

http://genetics.med.harvard.edu/reich/Reich_Lab/Publications_files/Supplementary_Q_Fu_Genome%20sequence%20of%20a%2045,000-year-old%20modern%20human%20from%20western%20Siberia-s1.pdf

Kristiina said...

"Thhe NO related X is South Asian but is northeast Indian."

Terry, X has been found in a Telugu and Telugu is a Dravidian language spoken in southeast India. It is not a Tibeto-Burman or an Austro-Asiatic language, so it has no connection with Southeast Asia.

http://www.chrismary.com/languagekeyboard/resource/dravidian/telugu.htm

Maju said...

@Kristiina: Thanks for the explanations, most interesting. It does seem that "Altai Aurignacian" people mixed bidirectionally with Neanderthals. It also seems true that the UI (or first Siberian) population went extinct and that Ma1 is less related to UI than I would have expected (more apparent in fig. 10.6B). This may suggest a more dramatic influence of the Gravettian wave more than 10 millennia later, not just in Ma1 (expected) but also in proto-Amerinds (not really expected). A bit perplex but guess that fair enough.

I won't knee-jerk: I'll adapt my theoretical frame. It would seem that the Gravettian wave therefore was the one most closely associated with Y-DNA R1 and Q (the Western subclades of P1), although it may have included other lineages (we don't know that much yet).

Maju said...

"Terry, Telugu is a Dravidian language spoken in southeast India".

Well said. I thought it was Dravidian but did not bother checking. Also in the sup. materials that you linked to the South Asian affinity of Ust Ishim is quite outstanding, very particularly in the PCAs.

In general we can infer that Indian C and this "X" (K2a*), as well as European C and Ust Ishim's K2a* should be somehow related to the overall westward flow of K2, whose most notable legacy is P1-derived subclades (R and Q, as well as the P1* found mostly in South Asia). I would say that, given the evidence we have from both ancient and modern DNA, the conclusion is unavoidable. I wish there was more archaeological research, especially for the MP-UP transition, in Pakistan, Iran and nearby areas, because I think they are critical to understand the UP genesis itself and the associated westward flow(s) of Asian H. sapiens that displaced Neanderthals and established our species in most of the Western section of the Eurasian continent.

terryt said...

"Terry, X has been found in a Telugu and Telugu is a Dravidian language spoken in southeast India. It is not a Tibeto-Burman or an Austro-Asiatic language, so it has no connection with Southeast Asia".

But it does have a strong connection with East Asia. Its closest relations are there. Which brings us to this:

"When I look at Ust Ishim’s Eurogenes K12 results, he seems to have originated in India:
South Asian: 35%"

What does 'South Asia' mean in this context? Check out Razib Khan's musings on the subject:

http://www.unz.com/gnxp/agriculture-came-with-men-to-the-indian-subcontinent/

Quote:

"More or less all South Asian populations are a fusion between a deeply indigenous strain which distant affinities to the peoples of eastern Eurasia (ASI), and a group very close to the ones typically found in Western Eurasia (ANI). There are no pure indigenes".

It is quite possible that, rather than being a source population, the South Asian population is a sink from both east and west. That leaves open the possibility that the genetic connection between Ust-Ishim and South Asia is a product of an Ust-Ishim-like population moving into South Asia.

"We had back in the day endless discussions about this kind of stuff relative of mtDNA N, particularly X, which does not have Indian precursors but should, as other N, ultimately derive from SE Asian ancestors. It seems clear that, as someone said above, minor lineages do tend to vanish in some cases at least, while also having a founder effect further away".

Mitochondrial lineages are more long-lasting than are Y-DNA lineages, minor or not. I doubt you can point to a single possible Y-DNA line that has become extinct in India. There is absolutely no evidence that basal N haplogroups were ever present in South Asia. On the contrary a recent paper showed A originated somewhere in Central Asia (sorry, can't remember where I saw it). The same is probably true of X. The split between eastern and western mt-DNA N is almost certainly a product of extinction of intermediates through Central Asia rather than extinction through India.

terryt said...

" a recent paper showed A originated somewhere in Central Asia (sorry, can't remember where I saw it)".

Found it:

http://www.plosone.org/article/fetchObject.action?uri=info:doi/10.1371/journal.pone.0127182&representation=PDF

It deals specifically with mt-DNA A10, and with the Neolithic, but A10 happens to be the most westerly A haplogroup. And it is a basal division within the hplogroup. Quote:

"In the light of our ancient DNA data, we suggest that the mtDNA haplogroup A10 is an autochthonous component of the gene pool of West Siberian indigenous populations, and that this component underwent a long evolution in West Siberia before the arrival in the region of
genetically contrasting western Eurasian and eastern Eurasian groups".

And there's more:

"the evolution of A2 cluster is related to early settlement of North-East Asia in the Late Paleolithic, and the subsequent settlement of America [39, 40]; the appearance of cluster A11 in Tibet dates back to the post-LGM period [41, 42]. Subsequent genetic isolation of human populations in these regions has led to the preservation of these components with high frequencies and diversity"

These two branches are usually shown as belonging to the same basal group as each other, separate from A10.

Kristiina said...

"This may suggest a more dramatic influence of the Gravettian wave more than 10 millennia later, not just in Ma1 (expected) but also in proto-Amerinds (not really expected). A bit perplex but guess that fair enough."

Maju, the above did not occur to me as I am not so well acquainted with the archaeology, but now that I look at Figure S10.6, it is very fascinating if it really means that Ust Ishim came from India and did not have much Western/Northern admixture even if he may have mixed with people (yDNA C) that developed Aurignacian in Europe and Ma1 instead was mixed between populations originating in Central Asia and Europe. His mtDNA U could have developed from R within the Aurignacian culture. Hopefully we get some Gravettian yDNA some day.

Maju said...

Burning nails, Terry.

First, it's clear that you were wrong in the NE Asia claim, yet you fail to make even the minimal self-criticism. This is the kind of issue that really irks me about you: you always have to be right and insist beyond reason.

Second, ASI is not anymore "East Asian" than it is "West Eurasian" or any other generic Asian or Eurasian or Australasian. It's quite apparently the aboriginal South Asian pool of UP roots.

Third, the Central-South Asian sample used in the study (and in so many others) is a Pakistan sample (plus some Hazaras and Uyghurs, easy to spot because they do tend towards East Asia). Pakistan is in South Asia the most West-Eurasian-like region of all, having lots of ANI (a West-originated component, surely Neolithic). So South Asian affinity here means both ANI and WEA affinity, of the Highland West Asian kind to be specific (Caucaso-Baloch component splits into Baloch and Caucasus, with "Baloch" being equal to ANI).

This means that if South and West Eurasians were just before Ust Ishim's time a single population (which was never surely the case but may have been to some extent at least in the North of South Asia), Pakistanis make a good proxy for that reconstructed (and somewhat conjectural) ancient population.

Fourth, it is not as simple as saying that the "closest relations" of K2a* are in East Asia (i.e. NO). Once we accept that K2 originated in SE Asia that becomes trivial: K2 originated in SE Asia and some branches went West via India. The most notable is P1 but it seems that some K2a* also migrated with them (and maybe was more numerically important back in the day). Other branches (NO, K2a1) went in other directions. We also see that in C, with most branches staying in the East (East Asia, Oceania) but some heading to India (C*, C5 - again among Dravidians mostly, particularly tribals) and even Europe (C1a2).

Fifth, "mitochondrial lineages are more long-lasting than are Y-DNA lineages, minor or not" is a spurious argument. All lineages are subject in the end to the logic of drift and founder effects. Even if you'd be right in you assumption it amounts to zero proof because there are always exceptions, being a matter of statistics and chance.

"I doubt you can point to a single possible Y-DNA line that has become extinct in India".

What about pre-IJ, pre-G, pre-C and pre-D? They all must have been at some point in South Asia but are nowhere to be found today (as their direct descendants that have not undergone migration through some other region). Even C2a1 must have migrated through South Asia to the West (inside a K2a- or rather P1-dominated population) but go find it now in India!

"There is absolutely no evidence that basal [mtDNA] N haplogroups were ever present in South Asia".

N2 (incl. W)? Again it is a matter of what lineages probably dominated the migration, which in this case was R. R did indeed leave a host of legacies in South Asia.

"The split between eastern and western mt-DNA N is almost certainly a product of extinction of intermediates through Central Asia rather than extinction through India".

There's no evidence of any H. sapiens migration via Central Asia before the time of Ust Ishim, sorry. The area was inhabited by Heidelbergensis (Denisovans) and then Neanderthals. There's no evidence of any H. sapiens presence in NE Asia either before around those dates.

Add to that some genetic and geographic common sense.

And I close this conversation here. I don't want to be trapped again in one of your endless "debates".

Maju said...

@Kristiina: "Hopefully we get some Gravettian yDNA some day".

By the moment all we have is Ma1 (which is a bit peripheral but Gravettian anyhow).

Anyhow, I'd dare say that Epipaleolithic European DNA should be at least partly Gravettian-derived. Problem is discerning what part.

Mike Thomas said...

Maju

Not denying eastern links; but I think yours is an overly "flexible" understanding of the Gravettian- which was european; but the nature orbits transition remains to be determined. Whateber the case, it appears to be a North alike phenomenon rather than a southern one (greece, Italy, iberia).

Its origin might be a case of internal evolution; but if I had to make a simplistic guess; Id bet Hg C was aurignacian, and IJ* Gravettian

Kristiina said...

Mike, I would also bet IJ* for Gravettian, but the ancient DNA has proved me wrong so many times that I am not eager to play with high stakes in this game. :-)

Mike Thomas said...

Don't know what happened with my keyboard there: but it meant to say Gravettian was concentrated in North Alpine europe.

Kristiina - you have nothing to lose ! :)

terryt said...

"First, it's clear that you were wrong in the NE Asia claim"

What 'NE Asia claim'? It seems that, as usual, you have adopted a belief and will not change that belief until the evidence becomes overwhelming. And I remind you of the heated disagreements we had over the possibility that YDA K arose in SE Asia. Any self-criticism or apologies?

"ASI is not anymore 'East Asian' than it is 'West Eurasian' or any other generic Asian or Eurasian or Australasian. It's quite apparently the aboriginal South Asian pool of UP roots".

That is not what Razib suggested.

"Pakistan is in South Asia the most West-Eurasian-like region of all, having lots of ANI"

But which is the direction of movement for that ANI? Once again Razib suggests it is from further west.

"K2 originated in SE Asia and some branches went West via India".

Some. Yes. But K2a? Unlikely.

"it seems that some K2a* also migrated with them (and maybe was more numerically important back in the day)".

What evidence do you have for that?

"Even if you'd be right in you assumption it amounts to zero proof because there are always exceptions, being a matter of statistics and chance".

Mitochondrial lineages are almost always older than surviving Y-DNA lines.

"What about pre-IJ, pre-G, pre-C and pre-D?"

You have evidence that all were present in South Asia at some time? Please provide. That is another example of your you having adopted a belief and sticking with it.

"They all must have been at some point in South Asia"

On what grounds do you claim that? A belief you have adopted again?

"Even C2a1 must have migrated through South Asia to the West"

I presume you mean C1a1 but once more, on what grounds do you claim that to be the case?

"but go find it now in India!"

Exactly. You won't find it because C1a was never there. C1b1a1 on the other hand definitely has a presence there.

"N2 (incl. W)?"

As far as I'm aware there is no actual N2 in South Asia. And W is primarily a West Eurasian with only a minor presence in South Asia.

"R did indeed leave a host of legacies in South Asia".

Yes, But R is very unlikely to have originated in South Asia. In fact it almost certainly 'migrated through South Asia to the West (inside a K2a- or rather P1-dominated population)"

Mike Thomas said...

Terry

Are you suggesting R* arose in SEA ? I'd put it in South- Central Asia

Maju said...

@Mike: I don't feel we have the capacity here to make a throughout discussion on Gravettian, more so when prehistorians themselves not always agree; for example some have recently argued for a West Asian "Ahmarian" origin and a radical population replacement at least in Central Europe (core area of both Aurignacian and Gravettian in Europe) and others however argue for some sort of Eastern European origin that I don't see plausible at all.

But in any case Gravettian is, like Aurignacian before it, universal in Europe (very few and quite dubious pockets of pre-Gravettian persistence have been argued for, all them in Western Europe). However it is also complex in its implementation: for example in many Franco-Cantabrian sites (but not all) it seems "intrusive" in the sense of causing a gradual techno-cultural transition soon cut short by the Solutrean. However in South Iberia instead it seems so consolidated that it actually modified the Solutrean in unique an peculiar ways (Iberian Gravetto-Solutrean, which is associated to North African Oranian or Iberomaurusian).

It is also the techno-cultural identity of Ma'lta in Central Siberia.

The Gravettian was also decisive in Italy and Eastern Europe, persisting in Epi-Gravettian form until Neolithic.

You make a lot of emphasis in N-S distinction but in those times the North was an ice sheet and anyhow there are a number or regions which can be capriciously organized in N-S axes or W-E axes or neither. After Gravettian, i.e. since the LGM and Solutrean, the main cultural axis is W-E, excepted Italy, which, as I just said, remains Epi-Gravettian (the iced Alps were a formidable barrier to human interaction by the west and north of the peninsula - there's some Holocene Epi-Gravettian in the area of Provence anyhow, indicating migration from Italy).

Maju said...

"What 'NE Asia claim'?"

*NE India (was a typo or lapsus).

Mike Thomas said...

Nice comments Maju

But North was not an ice aheet; at least not until 25 kya when peak glacial approached (20-18 kya).

But I agree as to the doubts on the "origins" of the Gravetian - with notable scholars (eg Svoboda) changing their own theory.

But my point was "Gravettian" is at best very loosely applied outside europe. In fact the mal'ta site of often classified as its own "culture", not Gravettian.

Moreover, my "N-S distinction" was not one at all. Rather it was a remark pointing out that the Gravettian concentrates on North-central Europe over the South in absolute number terms. Sites along the Danube of Europe outweigh by far those in Italy or Balkans.

Whatever, the case, the sharpest transition appears to be that of the Gravettian- Epigravettian. Not in Italy or southern balkans- where a true EpiGravettian continued, but certainly in East Central Europe and Russia, where there was marked changes in settlement topography, hunting economy and possibly population.

Mike Thomas said...

Further on North South, c. 30-25 ky; in. balkans you have sites where an Aurignacian persists till 25 kya ! (Istria -Sandalja etc); Klisoura Greece). So maybe Gravettian influences arrived from East, and perhaps northeast ?

terryt said...

"NE India (was a typo or lapsus)".

OK. Returning to Kristiina's comment regarding Y-DNA X:

"Terry, X has been found in a Telugu and Telugu is a Dravidian language spoken in southeast India".

Thanks for the clarification. I had presumed it would have moved overland and therefore would be in Northeast India. Its presence further south in Andhra Pradesh suggests an arrival by sea.

It is apparent there was considerable movement around the Bay of Bengal at some early period, including the settlement of the Andaman Islands around 35,000 years ago, presumably from the Burmese mainland. We have Y-DNA D* in the islands accompanied by mt-DNA M31a1 (with other M31 branches found from Orissa north to Tibet and Nepal) along with mt-DNA M32a with the connecting branch M56 also in India. Y-DNA D2 probably formed in the Philippines around the same time, and we have some very interesting connections between Borneo and Eastern India. On the Y-DNA side we have C1b1a2 in Borneo and C1b1a1 in Bangladesh and Gujarat. On the mt-DNA side we have M19'58, with M19 in Palawan and M58 in Orissa, and M24'41 with M24 also in Palawan and M41 in Eastern India.

Supporting the dating of around 35 kya for the movement is the now extinct C1b* found at Kotenski at 37 kya. It is actually quite possible that the expansion of Y-DNA P and mt-DNA R also was part of this major movement around, and from, the Bay of Bengal.

I seem to remember Maju made some comment as though he believed the eastern movement into India was confined to speakers of Austo-Asiatic and Tibeto-Burman languages. That would apply to just Y-DNAsO2 and D respectively. Much later arrivals than the above.

"Are you suggesting R* arose in SEA ? I'd put it in South- Central Asia"

I am fairly certain of SEA for reasons I spelt out above. Basically N is absent in island SE Asia but diverse in Australia whereas R is present and diverse in island SE Asia. Although R is also present and diverse in South Asia it is most so along the east coast. You may find this interesting, and as far as I'm aware Maju agrees with it totally:

http://ourorigins.wikia.com/wiki/Mt_R_east_to_west

"Further on North South, c. 30-25 ky; in. balkans you have sites where an Aurignacian persists till 25 kya ! (Istria -Sandalja etc); Klisoura Greece). So maybe Gravettian influences arrived from East, and perhaps northeast ?"

Quite possibly. And remember that most of the action around the Bay of Bengal took place before the dates you mention here. Interesting connotations if P was involved in the Gravettian.

Grey said...

"Central Europe (core area of both Aurignacian and Gravettian in Europe)"

I know nothing about this but a comment from another site sounds like it might be relevant if you assume (like I do) that innovations are likely to appear in places with unusually high population densities for the time period.

originally from bicicleur
"also Moravia in Europe was a good place before the ice age, when northern Europe was a cold steppe ; Moravia was a corridor between the northern European plain (northern Germany & Poland) and the Carpathian basin ; every spring and every automn herds of animals would pass through this corridor between their winter and summer grazing fields ; there were permanent HG settlements in Moravia ; it was allready densely inhabited by Neanderthals ; 48000 years ago the Bohunicians came to this place ; it is the oldest European culture that is assigned to homo sapiens sapiens, they settled here before the Balkans were settled"

If correct it would be a good example of how geographical/bioregion flukes - like a narrow physical corridor for herd migration - might have had a dramatic effect on human history.

Maju said...

"But North was not an ice aheet; at least not until 25 kya when peak glacial approached (20-18 kya)".

I don't know of any good maps for before the LGM but the overall Ice Age shows much of Northern Europe iced, including Poland, much of Britain, all the Baltic, etc. There is also lack of evidence regarding any population, other than Neanderthals, before the icing. Also there were several very cold episodes resembling the LGM, such as the Heinrich Event 4 (coincident with the expansion of Aurignacian proper) or the Younger Dryas more recently. I would like to emphasize the HE4 because it is that period when Europe is first colonized by our species.

What happened in the LGM is that even Central Europe became pretty much impossible to live in: permafrost reaching as far south as Budapest, lack of wood, extreme temperatures... Archaeologically speaking there are only pockets of survivors in Moravia and maybe (unsure) the Low Countries, and in general it is very apparent the displacement of the demographic centrality towards the south, particularly the Ocean-blessed Southwest (and even into North Africa).

"But my point was "Gravettian" is at best very loosely applied outside europe. In fact the mal'ta site of often classified as its own "culture", not Gravettian".

Sure. It is distinctive but it also has many elements of connection with not just Gravettian but in general Europe (the Mal'ta venuses are totally European-like and AFAIK there are no precursors anywhere else of this kind of iconography).

"Moreover, my "N-S distinction" was not one at all. Rather it was a remark pointing out that the Gravettian concentrates on North-central Europe over the South in absolute number terms. Sites along the Danube of Europe outweigh by far those in Italy or Balkans".

Well, Italy and the Balcans are not too important in the overall European Upper Paleolithic. In most of the periods Central Europe and even Eastern Europe (Dniepr and Don basins) seem to be much more important. However the most important region of all, particularly since the LGM but also to a large extent before it was the Southwest, particularly the Franco-Cantabrian region. Not sure if you place it in "Northern Europe", as it is North of the mountains, but for me it's South (or West if you wish).

"Whatever, the case, the sharpest transition appears to be that of the Gravettian- Epigravettian. Not in Italy or southern balkans- where a true EpiGravettian continued, but certainly in East Central Europe and Russia, where there was marked changes in settlement topography, hunting economy and possibly population".

Intriguing observation. I don't know enough to judge. Maybe you can point me to some key resources?

Mike Thomas said...

Maju

"I don't know of any good maps for before the LGM but the overall Ice Age shows much of Northern Europe iced, including Poland, much of Britain, all the Baltic, etc. There is also lack of evidence regarding any population, other than Neanderthals, before the icing. Also there were several very cold episodes resembling the LGM, such as the Heinrich Event 4 (coincident with the expansion of Aurignacian proper) or the Younger Dryas more recently. I would like to emphasize the HE4 because it is that period when Europe is first colonized by our species."

Yes, I agree. But settlement in WCE continues until 25 kya at least. In ECE, this continued well beyond that, albeit diminished till the Epipalaeolithic, althugh these might have been temporary camp sites (by WCE I mean sth Germany, Switzerland, etc; ECE - Austria, Czech lands, Hungary, Sth Poland, etc)

https://www.academia.edu/3164654/The_Human_Presence_in_Europe_during_the_Last_Glacial_Period_I_Human_Migrations_and_the_Changing_Climate

There is also Alex Verpoorte - Pleniglacial in central Europe.


"Well, Italy and the Balcans are not too important in the overall European Upper Paleolithic. In most of the periods Central Europe and even Eastern Europe (Dniepr and Don basins) seem to be much more important. However the most important region of all, particularly since the LGM but also to a large extent before it was the Southwest, particularly the Franco-Cantabrian region. Not sure if you place it in "Northern Europe", as it is North of the mountains, but for me it's South (or West if you wish)."


Of course Iberia in in SWE. The number of sites in the F-C refuge is indeed marked, and
they double going into the LGM. But part of this might be due to better preserved conditions in Iberia, less cryo-perturbation, erosional processes, c.f. East (ie the Balkans). But the importance of Iberia cannot be doubted, not the least because of the extend of Magdalenian expansion impact on the rest of Europe. That's why Im frustrated with studies from SPain coming up with nothing but mtDNA. Cant really construct much of a story

"Intriguing observation. I don't know enough to judge. Maybe you can point me to some key resources?"

Well that Svoboda paper in your own log. Also Hoffecker on "the East Gravettian -Kostenki culture as a cold adaptation.."

Mike Thomas said...

Maju
Slightly off topic- but what's your view on the early (>4o ky BP) vs late Neanderthal extinction in Iberia (& East adriatic) debate ?

Maju said...

@Mike: On the contrary I think that the Iberian province (i.e. excluding the Cantabrian strip, mostly from Valencia to Gibraltar but also offshoots in some other areas like Lisbon, Salamanca, Catalonia and even Asturias) is not too important, except regarding the Oranian genesis and some higher densities in the LGM (Gravetto-Solutrean). Unlike many people I keep quite strictly separated the Franco-Cantabrian and the Iberian provinces (they are related but mostly by repeated flow from the FC region to the Iberian one, also the relation is not much more intense than that of the FC region with Central Europe probably).

As for the Magdalenian (and previously Solutrean) expansion, it did not originate in Iberia but in the Franco-Cantabrian region, surely in the most densely populated area of it: Dordogne (aka Perigord), which is some 350 km north of the Pyrenees. Nothing originated in Iberia in the Upper Paleolithic except very possibly the Iberomaurusian/Oranian of North Africa. France, Occitania or Aquitania... is not Iberia.

Anyhow a criticism that can be done to the van Andel paper is that he seems to ignore the important penetration of Gravettian in the Iberian province. There are some other inconsistencies with other data I have in mind but guess it depends on what database you rely on.

I'm not sure I can agree with the argument about cryo-perturbation or erosional process to explain the low densities of population in most of the UP in the Balcans. Even today the Balcans have rather low densities but, regardless, I have observed other theories on this issue: that the dry steppe regions (North France and the Balcans for most of the Pleistocene) were disdained. Instead loess steppe regions like the Rhine, Upper Danube or the Dniepr-Don areas were preferred. This does not apply well to the more complex (and surely richer) ecosystem of the Franco-Cantabrian region but the same sources suggest that the Mediterranean was also of relative low densities because of its dense forest cover. In brief: that there are ecological reasons to the selection or preference of inhabitation areas. For example Moravia was favored not just because of being a migration corridors but also because it had a microclimate more favorable for inhabitation, including even some woodlands in the LGM.

"Well that Svoboda paper in your own log".

Is it from Svoboda? I don't recall that part. I'll re-read it.

Maju said...

"what's your view on the early (>4o ky BP) vs late Neanderthal extinction in Iberia (& East adriatic) debate ?"

Not sure but IMO there were several pockets of Neanderthal persistence, not just in Iberia and Croatia but also in Russia (Komi and Altai regions) and very possibly others. An intriguing issue is a date of c. 22 Ka BP in a remote mountain valley of Cantabria for Mousterian tools. It's the kind of isolated find that has to be taken with a pinch of salt, not because it's poorly done (doesn't seem so: five different datings support it) but because there's nothing else to back it up. But it is possible that the very last Neanderthals lived in the mountains of SW Cantabria.

Mike Thomas said...

Sorry my sloppy terminology. I mean franco - Cantabria which should rightly be separated from Iberia

And I think you're right about the dense forests cover in the Balkans making it unfavourable (unsuitable) for game, esp. big game, hunters.

Mike Thomas said...

Although Gamble arguea that it was iberia rather than France that was the major G refugium

Maju said...

@Mike: I have no idea why would Gamble argue such thing, really. The archaeological density is at all times much larger in the Franco-Cantabrian region than in the Iberian region. The Cantabrian strip, the only area that can overlap both geographical definitions was not that important, although it was no doubt much more densely populated than the rest of Iberia or any single Iberian district south of the mountains I can think of.

We can even go through examining the three main sub-regions of the Franco-Cantabrian region: the the Dordogne-Garonne one, the Cantabro-Pyrenean one and the Rhône-Mediterranean one, and I would think that the order of primacy in terms of overall population in all periods is in the stated order, i.e. Dordogne and nearby areas were the true #1 center. I actually have used the term "Paleolithic metropolis" to refer to Dordogne, whose densities were apparently like four times larger than any other populated district for reasons I cannot really explain with data - but I suspect that fishing was important as has happened with other high density HG populations from Lepenski-Vir to the Chinooks or the Jomon People of ancient Japan. If this is true, then coastal regions now submerged further west might have got even higher densities but we can't document it (just get some indirect hints at best).

Maju said...

@Mike: Hold on! What do you mean by "the major G refugium"? Y-DNA G? Or are you talking of Gravettian? If the former, the lineage seems to have arrived only with Neolithic (but you tell me if you suspect otherwise), if the latter, Gravettian did not experience any secondary re-expansion so the term "refugium" can only be understood as "last stand". If so, I'd agree in the context of Western Europe only.

Mike Thomas said...

Ha ha no

I should have defined: "G refugium" refers to Geogrphic refuge, vs M refugium which is the metpopulation in question. So, no, nothing to do with Y Hg G- which we all know was a Neilithi arrival.

Maju said...

Well, I think that the study of Bocquet-Appel 2005 leaves quite clear that the main density in the LGM and afterwards was in the Franco-Cantabrian region (FCR). The Iberian region had a mild spike in population density in the LGM but nothing in comparison to further North. Everything else I know on the matter is coincidental.

In fact, fig. 5, leaves clear that the FCR was all the time the most populated region of Europe in the Upper Paleolithic (possibly also before but who cares!): it had 2x the est. population than Central Europe in the Aurignacian and Gravettian periods, 6x the est. population of Central Europe and 8x that of the Iberian region in the LGM (Solutrean), and an undescribed but clearly much larger amount in the Late UP (Magdalenian) phase. It was not just a refugium: it was all the time the most important region or "country" in terms of inhabitants and density.

Mike Thomas said...

Maju
Those maps are very revealing

Maju said...

Yeah, they are. I've already mentioned in the past about that study in several discussions in this blog, so I thought you were familiar with it.

Simon_W said...

I was intrigued by Figure 3, showing the presence of J2 in a sample from 7250 years ago. I started puzzling where that may have been found and thought it must be from a new, perhaps not yet published paper. But it's just a stupid error. It's good old BR2, Late Bronze Age Hungary. In Table S8 he figures under "Late Bronze Age, Hungary, 7100 YBP", lol.

Maju said...

@Mike: regarding our early discussion in this thread on the beginnings of the UP (and the associated westward migration of H. sapiens from SE and South Asia), I have just stumbled on this PNAS paper on Ksar Akil and other earliest Levant UP sites (via a Spanish-language nice quality blog).

The study is largely on Ksar Akil (Lebanon), whose H. sapiens remains, associated to "Aurignacoid" tools, are as old as Oase, but what intrigues me the most are the very early dates of Manot cave (Galilee), a site I was oblivious about (was discovered in 2008) and that seems to have the oldest early UP (and associated H. sapiens remains) in all West Eurasia. The dates are U-Th rather than radiocarbon but they seem to demand a minimal age of 49 Ka ago for this site (max. 60 Ka). I imagine that you and maybe others would be interested.

terryt said...

Maju, thanks for the link.

"they seem to demand a minimal age of 49 Ka ago for this site (max. 60 Ka)".

Almost certainly quite some time after any OoA event.

MfA said...

bav-13 sample from Bavaria Germany is E-FGC18401+
Ycalls from the VCF file

Y 16437358 FGC18393 T C 166 PASS . GT:DP:GQ:PL:SP 1/1:17:21:255,51,0:0
Y 15372926 FGC18389 T G 166 PASS . GT:DP:GQ:PL:SP 1/1:58:99:255,175,0:0
Y 14641492 PH1393 G A 166 PASS . GT:DP:GQ:PL:SP 1/1:85:0:255,255,0:0