February 22, 2010

Global Genetic History of Homo sapiens Special Review Issue

Current Biology has several articles in its current issue (23 February, 2010 Volume 20, Issue 4). Will comment on them in this space. Just listing the titles for now:

46 comments:

Maju said...

FYI: the link to the article on East Asia is wrong, it leads to the general Current Biology issue. I could find it anyhow but you may want to fix that.

...

The articles are surely interesting in general but I found most interesting the one on East Asia and the one on selection.

The first one (Stoneking) emphasizes a lot how difficult is to make reliable age estimates, which should be a theme on its own right but is dealt in nice depth in this article anyhow. The opposite to this good style (educative and erudite at the same time) is the one on Europe (Soares), which falls all the time in the old vices, IMO.

The one on selection (Pritchard) is also very good, IMO, explaining how limited has been selective pressure in human evolution and how long is the road ahead for the proper detection of selective sweeps of any kind, if they exist at all.

The one on Oceania (Kayser) is also interesting, very specially (though not only) for the maps, always handy.

Average Joe said...

Here is an interesting paragraph from the Europe paper:

More surprising is the status of Y-chromosome haplogroup R1, which, unlike mtDNA haplogroup I, is not indigenous to West Eurasia but appears to have originated in South Asia, possibly in the early settlements associated with the southern route dispersal [64]. This appears better substantiated than the alternative suggestion of a Central Asian origin [65]. Two major subclades of R1 appear in Europe: R1b in the west and R1a in the north-east. It has been suggested that R1b mirrors mtDNA haplogroup H and the forerunner of V in arriving from the east shortly after the LGM. Then, with the Late Glacial, its main subclade R1b1b2 expanded into western and central Europe [66,67,68], with a possible expansion at the same time from Anatolia [69]. R1a might then represent an expansion from an eastern refuge, perhaps in the Ukraine, although it might also have been the result of more recent dispersals [62,66,70,71,72].

Gioiello said...

"Then, with the Late Glacial, its main subclade R1b1b2 expanded into western and central Europe, with a possible expansion at the same time from Anatolia": now I must go to work, but I dedicate this to Vizachero and Klyosov. To this afternoon.

Jean said...

The review by Soares et al is liable to come back to haunt them. It was out of date before it was even published. It can take such a long time from writing to publication.

It looks like it was written in late 2008. Other than the mtDNA dating paper by Soares himself, I spotted only one reference to a paper of 2009, and that one was actually available in late 2008.

It is just unfortunate for them that the old certainties of the early 2000s were overturned in paper after paper in 2009 and early 2010, to give a very different picture of the peopling of Europe.

On the plus side, they do make a number of sensible observations.
They reject Renfrew's thesis of IE languages spreading with agriculture, much to his annoyance, as we see from his guest piece.

Dienekes said...

Soares et al. argue that mtDNA and Y chromosome work has somehow cast doubt on the Neolithic origin theory and in the next sentence warn against conflating language with genetics!

Actually, the ancient DNA evidence, meagre as it is, argues strongly against the Pontic steppe theory, as the pre-Neolithic samples from Russia are uniformly in haplogroup U, just as in the rest of Europe, whereas the earliest Indo-European speakers of Siberia and Central Asia tested so far are light on U, suggesting that their origins are to be sought further west, certainly in Central Europe with likely Balkan and ultimately West Asian ancestors.

Maju said...

"... the pre-Neolithic samples from Russia are uniformly in haplogroup U..."

You should not forget the Sunghir burial.

"...just as in the rest of Europe..."

This a totally arbitrary, false, claim. And you know it.

Just check the nice page that Jean has and you'll see: Sunghir, Paglicci (three individuals, one of them within R0), Villabruna, Portugal (and Morocco, loads of likely H that must be of SW European ultimate origin).

...

Anyhow, dismissing Renfrew's hypothesis is one of the best things in Soares' article. But I have already explained why before and I would not like to be redundant.

I am more interested in other articles which do provide some good stuff, even if they don't deal with Europe, just a small corner of the world anyhow.

Notably, the article of Pritchard et al. on adaptation or lack of it is most educative and informative and I really recommend it to everybody.

Ponto said...

Age of a haplogroup does not mean it originated where it is found. So mtDNA U appears in some samples. U2 is common in India. It is likely all haplogroups found in Europeans of whatever age originated in Asia, which means east of Greece from Anatolia eastwards. Anatolia is not far from parts of the modern nation called Russia, nor is Russia far from India compared with most of Europe.

Nothing exists that proves any haplogroup formed in Europe. Where are the Paleolithic, Mesolithic R1b men? No proof at all. R1b is not European in origin, was not present in Europe at the LGM, was not in any refuge. It is due to the Bronze Age invaders of Europe with their cattle, milk drinking, wagons and horses. They just were a successful group of men, dominating completely the pre Bronze Age male inhabitants of Europe. Something like Bovines from the Middle East provided the main male line of European Taurine bovines. R1b=Wogs. As for haplogroup I, I* does not exist, but the problem exists that its parent IJ, is also the parent of haplogroup J. You have to explain how IJ went on to father I* but not J* in Europe, and vica versa in the Anatolia to Zagros mountain zone of the Middle East. I would say it is impossible for IJ to selectively found one population in one part of Anatolia and another population in Anatolia, one preferring Europe the other the Middle East. I know, an IJ man moved to Europe, drink the water there and hey presto, haplogroup I was created. Maybe God created it because the man looked like Adam?

mtDNA U* is so old it is older than Caucaosoids, Mongoloids and probably most Negroids since the Negroids seem to all descend from recent Bantu speaking farmers. So it is found in Europe. I am sure it will be found in bones of Middle Easterners, Indians, Central Asians and so on giving a wide range for that ancient haplogroup outside of Europe. It is just a minor and insignificant haplogroup in Europe anyways. Why the fuss? Oh yes, it is found in those ugly sandy haired people up in the outskirts of Europe everyone seems to love. Except me. There is a saying: You can't make a silk purse out of a sow's ear.

Maju said...

"It is likely all haplogroups found in Europeans of whatever age originated in Asia"...

Eventually yes, and if you follow that line, you end up in Africa as well. The issue is where each node and branch of the general phylogenetic tree originated.

"Nothing exists that proves any haplogroup formed in Europe".

I understand that certain haplogroups such as mtDNA H (at worst most of its subclades: H1, H3, H4, H7, etc.) or Y-DNA I or R1b1b2a1 are clearly original from Europe. These are just some examples.

"Where are the Paleolithic, Mesolithic R1b men?"

Where are the Paleolithic, Mesolithic anything else men? Is there any single Y-DNA sample from Europe before Chalcolithic? The answer is negative, of course, what makes your question a silly meaningless trap, Ponto.

terryt said...

"You have to explain how IJ went on to father I* but not J* in Europe, and vica versa in the Anatolia to Zagros mountain zone of the Middle East".

I presume haplogroup IJ* became spread right through the whole region but selection or drift led to two geographic varieties: I* in Europe and J* in Anatolia/Zagros. J* later, in turn, formed two geographic varieties: J1 in Arabia with J2 remaining in Anatolia/Zagros.

I found the Y-chromosome map in the East Asian article extremely interesting:

http://www.cell.com/current-biology/image/S0960-9822(09)02067-3?imageId=gr2&imageType=large

The map sorts Y-hap C into: C-M86 (C3c), C-M217 (presumably C3(xC3c)) and C-M130 (presumably C(xC3)).

I've always assumed that most Mongolian Y-hap C belonged to haplogroup C3c. But no. Well over 50% of all Mongolian Y-haps belong to the wider Y-hap C-M130, what we could perhaps regard as C*.

The map does not display any of the other C-M130-derived haplogroups such as C-M8 (C 1), which I've always assumed made a minor contribution to the Japanese and Ryukyu populations. (Ebizur has commented on the haplogroup at the leherensuge link below). But we might expect Y-hap C1 to appear on the map. So has Y-hap C1 been eliminated, or is it included in C M-130?

Similarly we might expect the C-M130-derived haplogroup C M-38 (C2) to appear somewhere in SE Asia. Has it too simply been included in C M-130?

The other C M-130-derived haplogroups (C M-347 (C4), C M-356 (C5) and C P-55 (C6)) don't occur in East Asia so they don't appear on the map. In order: they are found in Australia, India and New Guinea. Maju and I (with Ebizur providing much relevant information) are discussing the distribution of these Y-haps at this blog:

http://leherensuge.blogspot.com/2010/02/eurasian-y-dna-note.html

Let's go back to the map Dienekes put up some time ago:

http://4.bp.blogspot.com/_Ish7688voT0/SwWxkQGTZYI/AAAAAAAACCw/o3ls7yD1sig/s1600/maps-chiaroni.png

We notice in this map that it's very difficult to make the case that C* spread to SE Asia from India. Especially now we see that some sort of Y-hap C* (ancestral to Y-haps C1, C2, C3 and, presumably, Y-haps C4, C5 and C6) is so extremely common in Mogolia.

Central Asia? I know some are going to claim that it couldn't have happened. But to me that's not evidence that it didn't.

German Dziebel said...

"I've always assumed that most Mongolian Y-hap C belonged to haplogroup C3c. But no. Well over 50% of all Mongolian Y-haps belong to the wider Y-hap C-M130, what we could perhaps regard as C*."

Good observation, Terry. You probably also assumed that all American Indian C lineages are C3b. In fact, all C lineages observed in Siouan, Algonquian, Iroquoian and Muskogean populations are C* (M130).

See "Asymmetric Male and Female Genetic Histories among Native Americans from Eastern North America," by Bolnick et al.

Ebizur said...

terryt,

Practically all the haplogroup C-M130(xC3c-M86) Y-DNA in Mongols should belong to haplogroup C3-M217(xC3c-M86). I currently am preparing my next post for the thread at Maju's blog, but I will mention here that available data suggest that haplogroup C3(xC3c) is especially common in Asia among geographically northern and eastern Mongolic-speaking populations (i.e.Buryats, Khalkhs, and Měnggǔ-zú/Mongols and Dáwò'ěr-zú/Daurs of northeastern Inner Mongolia), geographically southern Tungusic-speaking (or formerly Tungusic-speaking) populations of Northeast China (Mǎn-zú/Manchus, Èlúnchūn-zú/Orochens, Èwēnkè-zú/PRC Evenks, and also Xībó-zú/Sibes, although this last population is now located mostly in Northwest China), and Koreans. It is also fairly common among other Mongolic- and Tungusic-speaking populations as well as among Palaeo-Siberians, though these populations usually exhibit C3c-M86 with greater frequency.

Among Mongolic-speaking populations, the ratio of C3(xC3c) to C3c appears to be positively correlated with longitude; the easterly Buryats, Khalkhs, and Daurs tend to belong to C3(xC3c) more often than C3c, whereas the westerly Oirats (including the geographically remote Kalmyks) tend to belong to C3c more often than C3(xC3c).

By the way, the so-called "peripheral Mongolic" populations of northwestern China, such as Dongxiang, Bonan, and Eastern Yugur, are practically devoid of haplogroup C-M130 Y-DNA altogether; instead, they exhibit high frequencies of haplogroup O3-M122 and various other haplogroups, notably R1a1a-M17 in the case of the Dongxiang and Bonan and P-M45(xR1a1a-M17) and F-M89(xJ-12f2, K-M9) in the case of the Eastern Yugurs. Haplogroup C-M130 does not account for even 5% of any of these populations' Y-DNA.

German Dziebel said,

"In fact, all C lineages observed in Siouan, Algonquian, Iroquoian and Muskogean populations are C* (M130)."

This is incorrect. Bolnick et al. (2006) have screened their samples only for M19, M3, M242, M173, M45, M130, YAP, and Tat. They have not tested for P39, which marks haplogroup C3b.

Zegura et al. (2004) have found all C-M130 individuals in their samples of Athabaskan (Tanana, Apache, Navajo), Algonquian (Cheyenne), and Siouan (Sioux) populations to belong to the subclade C3b-P39. However, they have found C3-M217(xC3b-P39) in two Wayus from South America.

German Dziebel said...

"Zegura et al. (2004) have found all C-M130 individuals in their samples of Athabaskan (Tanana, Apache, Navajo), Algonquian (Cheyenne), and Siouan (Sioux) populations to belong to the subclade C3b-P39. However, they have found C3-M217(xC3b-P39) in two Wayus from South America."

Yes, I checked and you're right about that. Strange that Bolnick et al. didn't test for a SNP. Their paper came out 2 years after the Zegura paper.

German Dziebel said...

"Among Mongolic-speaking populations, the ratio of C3(xC3c) to C3c appears to be positively correlated with longitude; the easterly Buryats, Khalkhs, and Daurs tend to belong to C3(xC3c) more often than C3c, whereas the westerly Oirats (including the geographically remote Kalmyks) tend to belong to C3c more often than C3(xC3c)."

On a second thought, all the C3(xC3c) lineages could be C3x lineages, i.e. lineages whose derived SNPs haven't been screened for. it seems unlikely that they would stay underived, while all others mutated in their own way.

terryt said...

"Practically all the haplogroup C-M130(xC3c-M86) Y-DNA in Mongols should belong to haplogroup C3-M217(xC3c-M86)".

Strange that they show it on the map as C(xC3) rather than C3 then.

"available data suggest that haplogroup C3(xC3c) is especially common in Asia among geographically northern and eastern Mongolic-speaking populations (i.e.Buryats, Khalkhs, and Měnggǔ-zú/Mongols and Dáwò'ěr-zú/Daurs of northeastern Inner Mongolia)"

The map supports that.

"it seems unlikely that they would stay underived, while all others mutated in their own way".

They could still be derived, but selection or drift hasn't reduced the number of haplogroups present. There are not enough widespread SNPs to justify defining a new haplogroup.

Ebizur said...

Here is a preliminary version of my compilation of published data regarding the distribution of haplogroup C3(xC3c) in Asia. The data points in italics do not represent confirmed C3(xC3c), but rather confirmed C or C3 that I have predicted to belong to C3(xC3c) in their entirety or at least in overwhelming majority.

C3-M217(xC3c-M48/M77/M86)
45/81 = 55.6% Buryat (Hammer et al. 2006)
17/45 = 37.8% Inner Mongolian (Xue et al. 2006)
51/149 = 34.2% Mongolia (Hammer et al. 2006)
22/65 = 33.8% Outer Mongolian (Xue et al. 2006)
4/12 = 33.3% Koryaks (Karafet et al. 2002; Pakendorf et al. 2007)
11/39 = 28.2% Daur (Xue et al. 2006)
7/26 = 26.9% Ewenki/PRC Evenks (Xue et al. 2006)
14/52 = 26.9% Manchu/Liaoning (Hammer et al. 2006)
1/4 = 25% Ainu (Hammer et al. 2006)
8/35 = 22.9% Manchu (Xue et al. 2006)
5/22 = 22.7% Oroqen (Hammer et al. 2006)
9/41 = 22.0% Xibe (Xue et al. 2006)
6/28 = 21.4% C-M130 Koreans (Karafet et al. 1999)
6/30 = 20.0% Han/Lanzhou (Xue et al. 2006)
6/31 = 19.4% Oroqen (Xue et al. 2006)
9/49 = 18.4% Tujia/Hunan (Hammer et al. 2006)
6/34 = 17.6% Hani (Xue et al. 2006)
17/98 = 17.3% Altai (Hammer et al. 2006)
1/6 = 17% C3-M217 Vietnam (Kayser et al. 2008)
13/78 = 16.7% Nyukzha Evenks (Pakendorf et al. 2007)
14/85 = 16.5% C(xC3c) Korean/Seoul (Katoh et al. 2004)
7/43 = 16.3% Korean/Korea (Xue et al. 2006)
7/45 = 15.6% C(xC3c) Korean (Wells et al. 2001)
5/35 = 14.3% Han/Harbin (Xue et al. 2006)
3/21 = 14.3% C3-M217 Korea (Kayser et al. 2008)
3/21 = 14.3% C3-M217 Han/Taiwan (Tajima et al. 2004)
1/7 = 14.3% Southern Altaians/Kosh-Agach (Kharkov et al. 2007)
13/95 = 13.7% Siberian Evenks (Hammer et al. 2006)
4/31 = 12.9% Eastern Evens (Karafet et al. 2002; Hammer et al. 2006; Pakendorf et al. 2007)
2/16 = 12.5% C3-M217 Ainu (Tajima et al. 2004)
3/25 = 12.0% Korean/PRC (Xue et al. 2006)
4/34 = 11.8% Han/Chengdu (Xue et al. 2006)
2/17 = 11.8% C3-M217 Malaysia (Kayser et al. 2008)
4/35 = 11.4% Hui (Xue et al. 2006)
6/54 = 11.1% C3-M217 Hui/Ningxia (Karafet et al. 2001)
5/45 = 11.1% Hezhe/PRC Nanai (Xue et al. 2006)
2/20 = 10.0% C-RPS4Y Yakuts (Karafet et al. 1999)
4/41 = 9.8% Manchurian/PRC Evenks (Hammer et al. 2006)
3/35 = 8.6% Han/Meixian (Xue et al. 2006)
4/49 = 8.2% C3-M217 Northern Han (Tajima et al. 2004)
1/13 = 7.7% Yukaghirs (Pakendorf et al. 2007)
6/84 = 7.1% Han/Taiwan (Hammer et al. 2006)
2/32 = 6.3% Han/Yili (Xue et al. 2006)
3/55 = 5.5% Tuvans (Pakendorf et al. 2007)
4/78 = 5.1% Cun/Hainan (Li et al. 2008)
2/40 = 5.0% Han/Guangdong (Hammer et al. 2006)
2/44 = 4.5% Han/Shaanxi (Hammer et al. 2006)
3/67 = 4.5% Uygurs/Xinjiang (Hammer et al. 2006)
3/70 = 4.3% Vietnamese (Hammer et al. 2006)
1/24 = 4.2% Central Evens (Pakendorf et al. 2007)
3/74 = 4.1% Ha/Hainan (Li et al. 2008)
2/53 = 3.8% Kyushu, Japan (Hammer et al. 2006)
2/58 = 3.4% Miao (Hammer et al. 2006)
1/32 = 3.1% NE Yakut (Pakendorf et al. 2006)
1/33 = 3.0% North Iran (Regueiro et al. 2006)
1/34 = 2.9% She (Xue et al. 2006)
1/35 = 2.9% Yao/Liannan, Guangdong (Xue et al. 2006)
2/77 = 2.6% Kathmandu, Nepal (Gayden et al. 2007)
4/156 = 2.6% Tibet (Gayden et al. 2007)
4/165 = 2.4% Japanese (Nonaka et al. 2007)
1/43 = 2.3% Yizu/Sichuan (Hammer et al. 2006)
1/46 = 2.2% Southern Altaians/Kulada (Kharkov et al. 2007)
1/47 = 2.1% Japanese (Xue et al. 2006)
1/47 = 2.1% Central Yakuts/"okayushie" dialect (Pakendorf et al. 2006)
2/96 = 2.1% Southern Altaians total (Kharkov et al. 2007)
1/48 = 2.1% Philippines (Hammer et al. 2006)
1/48 = 2.1% Taiwan Aborigines (Hammer et al. 2006)
1/49 = 2.0% Central Yakuts/"akayushie" dialect (Pakendorf et al. 2006)
1/50 = 2.0% Jiamao/Hainan (Li et al. 2008)
8/405 = 2.0% Hainan aborigines total (Li et al. 2008)
2/105 = 1.9% Tibet (Hammer et al. 2006)
1/61 = 1.6% Shizuoka, Japan (Hammer et al. 2006)
3/184 = 1.6% Yakuts total (Pakendorf et al. 2006)
4/255 = 1.6% Japan(xAinu) total (Hammer et al. 2006)
1/70 = 1.4% Tokushima, Japan (Hammer et al. 2006)

marnie said...

Thanks Ebizur for the exhaustive list. It's interesting in that it shows a relationship between Mongolian peoples and the Ainu, amongst other things.

Ebizur, any thoughts on early migration into Asia and around the Pacific Rim into the Americas?

I came across a fascinating paper published by a group from UC Davis a year or so ago:

Mitochondrial DNA and Prehistoric Settlements: Native Migrations on the Western Edge of North America

http://findarticles.com/p/
articles/mi_qa3659/is_200402/
ai_n9373568/?tag=content;col1

The ancient canoe cultures of many of the peoples mentioned in the paper are fascinating.

Just wondering if you know of any further research in this area.

South Central Haplo said...

Good Data Ebizur.

What is your thoughts on reference to O2 in South Asia Paper?.

marnie said...

High Resolution SNPs and Microsatellite Haplotypes Point to a single recent entry of Y Chromosomes into the Americas

http://mbe.oxfordjournals.org/
cgi/reprint/msh009v1.pdf

marnie said...

"Zegura et al. (2004) have found all C-M130 individuals in their samples of Athabaskan (Tanana, Apache, Navajo), Algonquian (Cheyenne), and Siouan (Sioux) populations to belong to the subclade C3b-P39.
"

There is a nice map of
the extent of Na-Dene Languages at:

http://en.wikipedia.org/wiki/
Na-Dene_languages

Ebizur, thanks for your work. Hopefully, we'll soon hear more about the North American side of C3. I'm originally from British Columbia. I've often been perplexed by the very Asian looking features of certain indigenous peoples such as the Sarcee and Tlingit.

Asian looking features are by no means universal in indigenous peoples of Western North America. It is engaging to see the science finally getting to the bottom of this very complex story.

terryt said...

"Here is a preliminary version of my compilation of published data regarding the distribution of haplogroup C3(xC3c) in Asia".

Certainly seems that most Mongolian Y-haps are C3 in fact, so I wonder where the authors get their Mongolian C(xC3) from. Do you have anything on that from that region? Or anywhere else? It would be C(xC1'6) in that case.

Andrew Oh-Willeke said...

The highlights in my view:

* Modern humans appear to have made their way to the Near East from Africa in the period from 130,000 to 75,000 years ago, then gone extinct or retreated to Africa and been replaced by Neanderthals from 75,000 to 45,000 years ago, when desert conditions no longer blocked modern human access to the Levant. (Europe paper)

* The retreat after 45kya and around the time of the LGM in Europe was also well explained. (Europe paper)

* The Nilo-Saharan languages probably have their roots in Sudan and subsequent migrations and subpopulation language switches to a Chadic are tracked. (Africa paper)

* There is remarkably low commonality between pre-Austronesean expansion New Guinea (i.e. pre-3,400 years ago) and Australia, despite the fact that they were geographically linked by land until 8,000 years ago, suggesting a possible separate migration from the West (i.e. from Indonesia) rather than the South (i.e. from Australia) or very early population segregation.

* Northeast Siberians retreated or went extinct during the LGM and was then repopulated as climate warmed again. It is possible that there was a Beringian refugia population (i.e. in Alaska and the exposed land bridge across the Bering strait) derived from the pre-ice age population of Northeast Siberia recolonized Northeastern Siberia from the East rather than from the South.

* Genetic evidence of a two wave American migration with Clovis wave from the North and Atlantic coast and a separate earlier Pacific coast migration looked pretty strong.

Ancient population retreats are an understudied/discussed subject.

terryt said...

"Modern humans appear to have made their way to the Near East from Africa in the period from 130,000 to 75,000 years ago, then gone extinct or retreated to Africa and been replaced by Neanderthals from 75,000 to 45,000 years ago, when desert conditions no longer blocked modern human access to the Levant".

There's a minor problem with that. Human occupation in India seems continuous through the Toba eruption of 75,000 years ago and the subsequent ice age of 70,000 years ago. So perhaps they'd made it all the way from the Near East to India before Neanderthal replaced them in the Near East.

"There is remarkably low commonality between pre-Austronesean expansion New Guinea (i.e. pre-3,400 years ago) and Australia, despite the fact that they were geographically linked by land until 8,000 years ago, suggesting a possible separate migration from the West"

Separate would be my guess, although both would have come from the west (Indonesia). And New Guinea/Australia may not actually have been one country even with lowered sea leval. The low-lying land between may have been dense, impassable mangrove swamp.

"Northeast Siberians retreated or went extinct during the LGM and was then repopulated as climate warmed again. It is possible that there was a Beringian refugia population"

Quite possibly further south than Beringia. Human occupation seems to be continuous across Central Asia from the Altai to the upper Amur for at least the last 100,000 years, although there is debate as to exactly what sort of human.

"Ancient population retreats are an understudied/discussed subject".

Difficult to study because, by definition, they will be small and isolated. However I'm sure that such retreats explain much about modern haplogroup distribution difficult to account for without them.

Andrew Oh-Willeke said...

"There's a minor problem with that. Human occupation in India seems continuous through the Toba eruption of 75,000 years ago and the subsequent ice age of 70,000 years ago. So perhaps they'd made it all the way from the Near East to India before Neanderthal replaced them in the Near East."

Any inconsistency is due to my sloppy use of the term "Near East."

The Europe paper is arguing that the coastal "Southern route" crossing over to Yemen along the coast to India and all parts East remained open in the period from 70,000 to 60,000 years ago, while the "Northern route" to the Levant was closed by deserts and enough of a Neanderthal population to stop the odd straggler that managed to cross the desert in the time period 75,000-45,000 years ago.

This keeps the ancestors of modern Europeans and Levantines all the way back in Africa until very shortly before they migrate widely across Europe, rather than in the Levant and Anatolia leaving much less time for the mutation clock to click for the part of the Out of Africa population that didn't take the Southern Route. This puts the divergence date for European and the most closely related Africans (probably the ancestors of Afro-Asiatic language speakers, about 55,000 to 85,000 years later than what we would have concluded from the earliest archeological evidence in the Levant. It also suggests a reason that we wouldn't expect to find many halotypes that appear to have arose in the Near East and Europe, but not Africa, in South Asia or all points east of there.

Maju said...

"The Europe paper is arguing that the coastal "Southern route" crossing over to Yemen along the coast to India and all parts East remained open in the period from 70,000 to 60,000 years ago, while the "Northern route" to the Levant was closed by deserts and enough of a Neanderthal population to stop the odd straggler that managed to cross the desert in the time period 75,000-45,000 years ago.

"This keeps the ancestors of modern Europeans and Levantines all the way back in Africa until very shortly before they migrate widely across Europe"...

Actually what keeps them is in South and East Asia.

The route was allegedly closed (only the northern route through the Fertile Crescent but not the one through Yemen-Oman) but H. sapiens was surely already in South Asia and maybe even as far east as SE Asia.

The genetic track does not leave room for a quick expansion from Africa to Europe, there must have been a Tropical Asian interlude in between.

Ronojoy said...

Any study done on aggressive behavior on western Europeans? Similar to the one done on Pakistanis.

terryt said...

"but H. sapiens was surely already in South Asia and maybe even as far east as SE Asia".

Agreed. I think we run into huge problems when we try to connect Europe directly with the OoA. The majority of European haplogroups derive from South Asia, if not even further east (perhaps much further east). That's why there is such a wide divergence between various dates offered for OoA. On the other hand I believe the genetic evidence is capable of supporting survival in some region to the northwest of India, somewhere on the Iranian Plateau or even as far north as the Altai.

Maju said...

"On the other hand I believe the genetic evidence is capable of supporting survival in some region to the northwest of India, somewhere on the Iranian Plateau or even as far north as the Altai".

We don't see any signature for that and all that region was Neanderthal anyhow. Yes, the Iranian Plateau too.

Ronojoy said...

Perhaps Europeans evolved in some other refuge we don't know about as yet..Maybe somewhere on the moon? They found water out there recently.

German Dziebel said...

"I think we run into huge problems when we try to connect Europe directly with the OoA."

We run into huge problems when we try to connect Asia to the OoA, too. The same ancient mtDNA N lineages that ended up giving rise to European haplotypes suddenly pop up in Southeast Asia without any OoA "trail." The same concerns mtDNA M haplotypes that suddenly pop up either in South Asia or in East Asia without any OoA trail. M lineages end up in Africa around 40-45K in the form of M1, precisely when N lineages show up in Europe.

So, we clearly see lineages moving westward from Asia to Europe and Africa, but not from Africa into Europe or Asia. This can be correlated with the spread of Upper Paleolithic technologies across Eurasia and into North Africa. Looks like we can't even connect Africa to the OoA.

Maju said...

German: shut up. L3 moved from Africa to Asia and that's a fact (you are merely in convenient stubborn denial about that because it fits your agenda).

And I know that you will reply to this fact with more "out of America" babbling and you should know that I won't reply to you. I am just making the point for the audience, among whom there may be one or two unsuspecting readers.

German Dziebel said...

"L3 moved from Africa to Asia and that's a fact"

It's not a fact. And you know this. M and N are derived from L3 "on paper" but there's no phylogeographic support for this derivation. There are no pre-M or pre-N lineages in Africa, there are no L lineages in India, Europe or Southeast Asia. And you gotta have this phylogeographic connection in order to argue for a "movement" of L3 from Africa to Asia.
In addition, L6 is the most divergent lineage of the L3/L4/L6 clade, and it apparently also originated outside of Africa, namely in Yemen.

Out of Africa is a hypothesis in need of a proof. And you better find it before you lose any credibility in this or other forums.

terryt said...

"and all that region was Neanderthal anyhow. Yes, the Iranian Plateau too".

That's a fairly brave statement. The link mentioned, 'This view has been extrapolated to much of the Iranian Middle Paleolithic, but recent research is starting to show that Middle Paleolithic technology was much more variable than previously believed, which is frankly unsurprising given the size of Iran and the topographical and environmental variability its modern boundaries encompass'. The 'much more variable' technology is probably relevant. Besides which Neanderthals and modern humans lived side by side (possibly separating the ecological niches) in Europe for at least 10,000 years. Besides which the 'view has been extrapolated', so there are many assumptions involved.

"The same ancient mtDNA N lineages that ended up giving rise to European haplotypes suddenly pop up in Southeast Asia without any OoA 'trail'".

I disagree about the lack of trail, but the trail I see is not accepted by most either.

"we clearly see lineages moving westward from Asia to Europe and Africa, but not from Africa into Europe or Asia. This can be correlated with the spread of Upper Paleolithic technologies across Eurasia and into North Africa. Looks like we can't even connect Africa to the OoA".

I'd agree totatlly that we can't correlate the Upper Paleolithic with any OoA movement. That particular expansion happened long after the OoA. A lot had already happened by the Upper Paleolithic.

"There are no pre-M or pre-N lineages in Africa"

There are actually. They are L1''6, L2''6, L2'3'4'6, L3'4'6 and L3'4. Lines M and N are simply regional variations of L3. Maju has done a magnificent series of detailed blogs on the subject starting here:

http://leherensuge.blogspot.com/2010/03/are-we-overlooking-signature-of-out-of.html

"In addition, L6 is the most divergent lineage of the L3/L4/L6 clade, and it apparently also originated outside of Africa, namely in Yemen".

L6 is hardly 'the most divergent lineage of the L3/L4/L6 clade'. It has been divided into just two subclades (as far as I know): L6a and L6b, neither of them found exclusively outside Africa. So I doubt very much that the clade as a whole originated outside Africa.

terryt said...

Oh. Another thing. I've just printed off the East Asian maps and noticed something interesting.

Y-hap NO (ancestral to N and O) is present in Borneo, Vietnam, Thailand, and further north into Taiwan, Korea, Mongolia and Tibet. But not in or east of Wallacea (counting the Philippines as Wallacean). And most of its closest relations are found around and east of Wallacea.

And what of another of its relations, P, ancestral to R and Q? Tenggara, Borneo, Java, Sumatra, Malaysia and the Philippines (the first and last actually in Wallacea, but the others immediately west of it). Perhaps the haplogroup Ps shown actually belong to R or Q, as those haplogroups are not listed on the map. P is also shown further north but in these northern regions it is very likely to be Q or R.

Maju said...

I said I would not reply but, German you're so extreme in your pseudologic that it's comical:

"Out of Africa is a hypothesis in need of a proof".

And a line before:

"L6 is the most divergent lineage of the L3/L4/L6 clade, and it apparently also originated outside of Africa, namely in Yemen".

Whoa! Far, far, far away from Africa, almost in... Seychelles!

No doubt that L6 in Yemen proofs Out-of America, we all know that Yemen is the 51st state of the USA and is somewhere around Nebraska. Obviousy it counters the OoA.

Yes, German you re right. M derives from D, L3 derives from M and I walk on my hands all the time. Let's shake feet in this upside down world of your imagination!

After these laughs I'll try to address serious mail non-sarcastically. It'll be difficult. :(

Maju said...

"That's a fairly brave statement. The link mentioned, 'This view has been extrapolated to much of the Iranian Middle Paleolithic, but recent research is starting to show that Middle Paleolithic technology was much more variable than previously believed, which is frankly unsurprising given the size of Iran and the topographical and environmental variability its modern boundaries encompass'. The 'much more variable' technology is probably relevant".

Did you check the two specific links within the thread. It's all Mousterian: variability within Mousterian, but Mousterian all the time. I've already screened the images in search of something that looks like MSA influenced or whatever but nope: it's all Mousterian.

You, Terry, tend to defend positions without first gathering any significant support for them. First you decide you like this option and then you try to support it somehow, what is desperating if you are at the other side of the argument, because whoever is discussing with you is forced to do all the critical work you never did and should have done first of all, in every single point.

At least, unlike German, you do yield sometimes when evidence is overwhelmingly against but, seriously, you should try to first analyze the evidence calmly and then decide what is what the evidence might be saying, not your desires. And not make others waste their time thinking critically for you and having to write down all those details only for them to be conveniently ignored while you jump to some other trench.

It's not a war: science it's critical thought. And that means self-critical to begin with.

So you have no evidence for H. sapiens in Iran before the backmigration from India, you have no evidence in general in support of any single of the aspects of your "India no, Siberia and Wallacea" conjecture. It's all "I wish this to be that way" and hence I will fight (in a military way) for it.

Well, enough said.

German Dziebel said...

"No doubt that L6 in Yemen proofs Out-of America, we all know that Yemen is the 51st state of the USA and is somewhere around Nebraska."

I'm not talking about out of America right now, but just going through the "evidence" for out of Africa. Yes, Yemen is outside of Africa and it harbors the greatest diversity and has the greatest frequency within a lineage called L6, which is the earliest offshoot of a clade that is pan-African and of which M and N are part. M shows signs of a population expansion starting in East Asia or South Asia. One branch, M1, ends up in North and East Africa. N "explodes" in Southeast Asia and never gets to Africa, until the U6 time.

So what I'm seeing is that there was a population movement westward into Europe and Africa. Even such a pan-African clade as L3-L4-L6 begin to form at the doorstep to Africa, namely in West Asia.

This is very consistent with Y-DNA, where pan-African E clade is derived from the non-African CFDE clade. The geographic signature of the Y-DNA C clade is very similar to the geographic signature of the mtDNA N clade, with early representatives found in places as far apart as America and Australia.

Yes, I do question the order of mutations because "out of Africa" doesn't make much sense. It doesn't make sense linguistically (Asia, Australasia and America are much more diverse than Africa and Europe) and it doesn't make much sense genetically. How in the world did N lineages end up in Southeast Asia? By helicopter?

"Yes, German you re right. M derives from D, L3 derives from M and I walk on my hands all the time."

You're probably right here. Glad we agree. Some humans evolved by descending from trees, others by turning upside down. If we assume a reverse order of the mutations "on paper," we'll end up with strong continuity in phylogeography: C and D lineages are found as far west as India, where we see a great basal diversity of M lineages, M is then found deep in Africa, as discussed.

"L6 is hardly 'the most divergent lineage of the L3/L4/L6 clade'. It has been divided into just two subclades (as far as I know): L6a and L6b, neither of them found exclusively outside Africa. So I doubt very much that the clade as a whole originated outside Africa."

Thanks, Terry, for pointing me to the amazing work Luis/Maju did on the L lineages. He writes at some point: "L6 shows 38 mutations, however the root of its very long stem is necessarily older than L3 (as it's the oldest branch of L3'4'6). Today it looks like Yemen harbors the highest diversity, with offshoots at Ethiopia and Egypt." It's possible that someone else is actually behind the human origins posts found at the Leherensuge blog.

"I'd agree totatlly that we can't correlate the Upper Paleolithic with any OoA movement. That particular expansion happened long after the OoA. A lot had already happened by the Upper Paleolithic."

We just need to get a proof that whatever happened archaeologically before the Upper Paleolithic happened to our direct ancestors. It's possible that African/West Asian "anatomically modern humans" were replaced by us, behaviorally modern humans, coming from the east (in the same way as Neanderthals were replaced by us). We know that the Skhul and Quafzeh folks were replaced by Neanderthals.

terryt said...

"It's all Mousterian: variability within Mousterian, but Mousterian all the time".

Yes, and early modern humans in the region used Mousterian technology. Further east they also used 'primitive' technology. Nowhere did the first modern humans outside Africa have the Upper Paleolithic.

"you should try to first analyze the evidence calmly and then decide what is what the evidence might be saying, not your desires".

So how did you analyze the evidence of Y-haps NO and P in Wallacea? Isn't that 'evidence in general in support of any single of the aspects of' my Wallacea conjecture'. Especially once we include M, S and the Ks. And there's more. In the East Asian Y-hap map we actually see that Y-hap F-M89 is easily common enough in both north and south China to explain the presence of all the mtDNA M haplogroups there, without any assistance from Y-hap C. And the mitochondrial DNA maps in the same article have Wallacean mtDNAs almost entirely R-derived: B4, B5, F and F1. The only non-R-derived haplogroup present to any extent at all is M7 (and still very much a minority haplogroup), presumably the old mtDNA E, a mainland SE Asian mtDNA haplogroup. Even on the ancient Sunda (in Borneo and Java) R-derived haplogroups easily form the majority. It's only once we get to Malaysia do we find M haplogroups beginning to make a significant contribution. Doesn't that suggest a Wallacean origin for mtDNA R? Or are you still going to ignore 'what the evidence might be saying' and still claim an Indian origin?

Maju said...

"So how did you analyze the evidence of Y-haps NO and P in Wallacea? Isn't that 'evidence in general in support of any single of the aspects of' my Wallacea conjecture'".

It would if anything support a "Melanesian conjecture", not a "Wallacean" one, right?

But what I say in that case is to wait and see what happens with MNOPS in the end. Because at this stage of total uncertainty drawing conclusions is premature.

I also say that mtDNA evidence has precedence, it's more solid, and mtDNA does not support such conjectures at all.

terryt said...

"It would if anything support a 'Melanesian conjecture', not a 'Wallacean' one, right?"

No. It's not actually found in Melanesia. Only along the western shore of Wallacea, the east Sunda coast. C2 and some Ks are Wallacean. Other Y-haps are confined to east of it. Melanesia has its own haplogroups, related to NO and P. So Wallacea seems to be a centre of expansion to both east and west.

"I also say that mtDNA evidence has precedence, it's more solid, and mtDNA does not support such conjectures at all".

It sort of does. We have B4, B5, F and F1 making up almost the entire Wallacean mtDNA spectrum, so thier ancestors were probably the first to reach the islands. We have the related mtDNA P beyond Wallacea in New Guinea and Australia. So it's possible to interpret the evidence as showing R* started from somewhere near Wallacea, even though R's greatest diversity is in India.

Maju said...

Karafet mentions "Lineages K2, K3, and K4
are found in Oceania, Indonesia, and/or Australia". Wikipedia says that K3 is found in Melanesia and Polynesia (not Indonesia nor Australia), datum which was confirmed in detail by Ebizur in our discussion here. Not even him seems to know any data on K2 and K4 and I'm starting to wonder if these lineages exist at all, at least with any entity (Australian K*?) or are just very rare "private" K* lineages which have received a name just on some caprice or confused appreciation (not the first time it happens).

The whole issue is confuse and I can't take any stand based on the current data. Wait and see is all I can do.

"It sort of does. We have B4, B5, F and F1 making up almost the entire Wallacean mtDNA spectrum, so thier ancestors were probably the first to reach the islands. We have the related mtDNA P beyond Wallacea in New Guinea and Australia. So it's possible to interpret the evidence as showing R* started from somewhere near Wallacea, even though R's greatest diversity is in India".

I don't see how. You're forcing that in order for your pre-conceived pet hypothesis to work but that's not what the evidence says.

The evidence says that mtDNA R spread from South Asia.

I think it happened about the time of the colonization of West Eurasia, maybe 50 Ka ago and that mtDNA R spread westward initially with mtDNA M1 and N1, and Y-DNA G, IJ and T, while it spread eastwards with Y-DNA MNOPS (if this lineage is confirmed).

I also suspect that N (eventually leading to R, N1'5 and N2) back-migrated to South Asia from SE Asia with C (leading to C5) at a previous stage.

And the only mystery I have left (if MNOPS is confirmed, and just within the frame of the Great Eurasian Expansion) is the process of back-migration of Y-DNA P (or R and some Q) westwards. Maybe the small scattered remnants of mtDNA B and F in India are related? I really can't say.

terryt said...

"The evidence says that mtDNA R spread from South Asia".

The only 'evidence' is that R's diversity is greatest in India. But we know very well that diversity can arise for a number of reasond other than indicating origin. In densely forested regions communities tend to become isolated, as they do in mountains. Mountainous regions tend to have more diversity than most other regions because each little valley develops its own version of any particular haplogroup, or because ancient haplogroups survive the arrival of newer ones. Another reason for diversity is if closely related haplogroups arrive in a region from different directions, or at different times.

"And the only mystery I have left (if MNOPS is confirmed, and just within the frame of the Great Eurasian Expansion) is the process of back-migration of Y-DNA P (or R and some Q) westwards. Maybe the small scattered remnants of mtDNA B and F in India are related? I really can't say".

I'd say that's correct. But it's not a 'back-migration' of Y-hap P. It's the original one from Wallacea after it had formed from MNOPS*. And it presumably included other members of mtDNA R as well as B and F. So no mystery.

Maju said...

"But we know very well that diversity can arise for a number of reasond other than indicating origin".

This argument is used here without the appropriate reasoning as of why this case would be an exception. Exceptions do exist but are not the rule.

This you say can happen when it gets immigration from various origins aboundant in the same haplogroup and only in that case. For example the USA or Brazil are likely to have more R1b diversity than Portugal or Britain, surely representing the various European origins of their populations.

But what we find in South Asia is not variations of R sublineages found elsewhere (with the exception of U maybe and some minor B, F and maybe even H) but totally original SA-only sublineages, directly spawn from the R node and found nowhere else.

So I say that your feeble counter-argument is a mere fallacy that you have decided to be "true" only because it favors your pet hypothesis and that can't hold against common sense scrutiny.

We are not talking of mere chaotic diversity here but of basal diversity within a phylogeny, more exactly within the earliest branches of a star-like node, a signature of rapid expansion that must have spread from South Asia, both because of diversity and geography (it's almost in the middle of the scatter).

"But it's not a 'back-migration' of Y-hap P. It's the original one from Wallacea after it had formed from MNOPS*".

It's a backmigration of some of the descendants of the K guys who migrated eastward, became there MNOPS (if the proposed haplogroup holds) and then back-migrated at a further stage as P or pre-P.

I don't think that the MNOPS (or K) leading to NO or P ever made it to Wallacea, only the one leading to M and S did (maybe as such M and S already). The reason is Wallace line, which was a major obstacle before sails were invented (so back-migration could never be massive and without massive numbers they would have never overcome the ones left behind: pure fluids' dynamics: you can't overcome a pressure unless you have previously accumulated larger pressure). This is what seemingly Karafet and co. have just found in Indonesian Y-DNA: Wallace line is a neat divide.

terryt said...

"The reason is Wallace line, which was a major obstacle before sails were invented"

No Maju. It was a major obstacle before boats were invented.

"I don't think that the MNOPS (or K) leading to NO or P ever made it to Wallacea, only the one leading to M and S did"

Doesn't make sense. M and S are not on the same line. They are as separatte from each other as NO and P are. If you're going to claim they craossed Wallacea as already established M and S you have to explain why they don't survive west of the line. So, MNOPS at Wallacea, perhaps some MNOPS* individuals not actually enetering it, but returning west as P and north as NO.

"and some minor B, F"

It would be really difficult to make the case that B and F arose in South Asia.

Maju said...

"It was a major obstacle before boats were invented".

No, my stubborn but misled preacher of the Holy Wallacean Dogmatic Faith of the Blind Ones, before boats were invented even a regular river was a major obstacle... and Wallace Line was an almost absolute barrier, only passable surfing with loads of luck on a tsunami wave or something like that.

With boats, the large strait was still a major barrier and it seems that the new genetic evidence confirms it (logically). If I'd got an euro each time I was right on this kind of stuff, I'd be rich by now. But as the song goes, "the truth matters to nobody, what for to say the truth then?"

Con artists and salesmen know better: they tell you what you want to hear. Yes sir, sure madam, of course, how could it be otherwise, you are right indeed. That's why they are rich and I'm poor like rats.

Enfin.

"M and S are not on the same line".

As far as current research has managed to discover so far... They are not the most intensely researched lineages, as you probably know...

Anyhow, nothing absolutely nothing impedes them from having migrated in two slightly different moments and places or even have migrated together and then parted ways only to meet each other a little later... because Melanesia is not that big anyhow, is it?

"It would be really difficult to make the case that B and F arose in South Asia".

Never was my point. I don't think that.

terryt said...

"With boats, the large strait was still a major barrier"

And absolutely impassable before boats were invented. The claim for four crossings suggests boating improved progressively over time so it gradually became less of a barrier.

"Anyhow, nothing absolutely nothing impedes them from having migrated in two slightly different moments and places"

But obviously mtDNA L6 couldn't have done that! What was that about 'the truth matters to nobody, what for to say the truth then?'

"If I'd got an euro each time I was right on this kind of stuff, I'd be rich by now".

Obviously so, because you never even consider any evidence that suggests you could be wrong. You still don't seem to realise that Sunda, Wallacea and Sahul are three separate geographic regions. The same with parts of Africa. Sudan is not Ethiopia, even if you like to call it East Africa (which is stretching things).

Maju said...

"But obviously mtDNA L6 couldn't have done that!"

Egypt and Ethiopia (the other locations of L6 besides Yemen, according to Behar'07) are two very different places. And we are anyhow talking of a very rare lineage so far located in just a handful of people, mostly Yemenis, who are the only ones that have both L6 sublineages.

Y-DNA M and S are dominant in Melanesia! Its relatives are largely dominant in many parts of Eurasia. L6 is not dominant anywhere. It's very existence today is almost a "miracle" (how many other such small and old lineages have not made it?) You are comparing Jupiter with a tiny asteroid.

"Obviously so, because you never even consider any evidence that suggests you could be wrong".

Believe it or not, I do. All the time. I just get heated sometimes when discussing but I do consider all the evidence I can manage.

"You still don't seem to realise that Sunda, Wallacea and Sahul are three separate geographic regions".

I do. It's you who likes to include Borneo in Wallacea and other extreme fancies of the like.

"The same with parts of Africa. Sudan is not Ethiopia, even if you like to call it East Africa (which is stretching things)".

I use "East Africa" and "the Nile" because they are more precise terms, because of their lack of extremist precision, precisely.

Anyhow, the different regions of Africa are not and have never been as strictly divided as the regions of Indonesia/Sahul you just mentioned: there's no such sea inside Africa. The only imperfect exception may be North Africa (because of the Sahara).