Haplogroup G from the Levant (early farmer), R1b & J (herders) from East Anatolia

[Maciamo]

Here is a new hypothesis. I, like most of the population genetics community, have assumed until now that haplogroup G originated in Anatolia or in the Caucasus, because this haplogroup is the most common there. A few elements make me think that haplogroup G could have expanded from the Levant instead.

1) Agriculture expanded from the Levant to Greece, then to the Balkans and the Mediterranean. Most of the early farmers tested so far being members of hg G2a, it would make sense that they originated where agriculture started, i.e. the Levant.

2) Haplogroup G has a fairly high frequency in Egypt (9%). There is so far no explanation as to how it got there. Either hg G was already in Egypt since the late Paleolithic, or it came with the spread of agriculture from the Levant. If Anatolian G people learned to farm from their Levantine neighbours (so far assumed to be E1b1b or J2 people) then expanded to Europe, there is no reason to find any G in Egypt. If G was absorbed by E1b1b an/or J2 farmers, then these two haplogroups should be found among early farmers in Europe, but so far it hasn't been the case.

3) What's more, hg G is found (at low frequencies) as far south as East Africa, Yemen and Oman, and as far east as India and Central Asia. So I think that the most reasonable assumption is that agriculture started with G people in the Levant, then expanded to Egypt, East Africa, all the Middle East, South Asia, Central Asia, and of course Europe. However they would probably have been outnumbered by local hunter-gatherers as G is a minority haplogroup almost everywhere.


If farming started with G in the Levant, then domestication started in East Anatolia/North Mesopotamia, and the prime candidates for spreading the new herder/stockbreeder lifestyle must have been haplogroups J and R1b1b, which both seem to have originated in that region in the Mesolithic. I see a stronger association of R1b1b with cattle and sheep, and J with goats.

As I have stated many times before, I believe that R1b1b crossed the Caucasus and settled in the Pontic Steppes, becoming one branch of the Proto-Indo-European speakers, a language that they probably spread to their R1a1a neighbours to the north and east.

As for haplogroup J, both J1 and J2 and found all over the Middle East, North Africa, Europe, and Central Asia. I really don't think that the medieval Arabic expansion is responsible for spreading all of it in North Africa and Central Asia. J2 being found among Hindus in India, was almost certainly in South Asia since the Neolithic. Dry, mountainous regions like the Maghreb, the Caucasus, most of Iran and Central Asia are also better adapted to herding than to agriculture, which may be why there is more haplogroup J than G in these regions. The same is true within Greece; Thessaly is better suited to agriculture and has a lot of hg G, while Crete is better suited to (goat) herding and has a lot of hg J2. In Iberia, hg G is more common in the wetter Northwest, while J1 and J2 are both more common in dryer Andalusia and Castilla-La Mancha.

[Dorianfinder]

This is a step in the right direction, I can envisage haplogroup G as an early farming community with certain locations more suggestive of stock-breeding due to altitude and poor soil types. I like where you are heading with J and R1b, they fit better when we put the clock back a bit as opposed to the Pax Romana hypothesis. J and G are Caucasian elements in Egypt's genetic history, the Phoenicians et al. are evidence throughout the early Mediterranean civilization of a significant levantine R1b history. Crete remains an interesting case as E is relatively rare and in its place there is significantly more G. The Aegean islands may prove to have similar levels of G as on Crete, time will tell.

[Knovas]

In Spain, Murcia appears to have a lot of G2a according to your map, wich is quite in the South but also very in the East. I supose you used the figures from Iberian roots, giving 14% G2a. Not sure if it's much reliable such figure, it seems there's lack of Murcian data to say something, but if it really is, it comes as the exception in Iberia. I'd like to see, if it's true, wich is the most common G2a subclade between Murcians, or if there's significant presence of an especific one. Nothing clear for the moment.

[Dorianfinder]

In Spain, Murcia appears to have a lot of G2a according to your map, wich is quite in the South but also very in the East. I supose you used the figures from Iberian roots, giving 14% G2a. Not sure if it's much reliable such figure, it seems there's lack of Murcian data to say something, but if it really is, it comes as the exception in Iberia. I'd like to see, if it's true, wich is the most common G2a subclade between Murcians, or if there's significant presence of an especific one. Nothing clear for the moment.

The Southeastern/W-Asian/Caucasus components in Iberia may be reminiscent of this G2a marker found in some isolated locations on the peninsula. The evident relationship between S-Eastern and Caucasian within the Dodecad euro7 results for Iberia may reflect the prehistoric migratory route taken from the Caucasus to Iberia through the Levant, showing up as Southeastern instead of Caucasian/W-Asian.

[Goga]

According to me there's very much hg. G in Egypt because of the Mamluks. Mamluks were slave children (sold to the Ottomans by their parents) from the Caucasus who became the greatest warriors in Egypt who defeated Saladin. The Mamluks were Caucasians (Georgians etc.) who were coverted into Islam and they fought also against the Crusaders.

Later the Mamluks were defeated by the Mongols, but that's another story...

[Hakan67]

According to me there's very much hg. G in Egypt because of the Mamluks. Mamluks were slave children (sold to the Ottomans by their parents) from the Caucasus who became the greatest warriors in Egypt who defeated Saladin. The Mamluks were Caucasians (Georgians etc.) who were coverted into Islam and they fought also against the Crusaders.

Later the Mamluks were defeated by the Mongols, but that's another story...

Thats Wrong.
Mamluks were slaves from Central Asia or Caucasus with turkish origin and later more Adyghe origin. They defeat the progeny of Saladin and became the Rulers of Agypt. They defeated the Mongols. Later they were defeated by the Ottomans.

[Goga]

Thats Wrong.
Mamluks were slaves from Central Asia or Caucasus with turkish origin and later more Adyghe origin. They defeat the progeny of Saladin and became the Rulers of Agypt. They defeated the Mongols. Later they were defeated by the Ottomans.

Yes, you're RIGHT. Mamluks defeated the Mongols! but the Ottomans (formal leaders of the Mamluks) defeated the Mamluks later. I made a mistake.

But Mamluks were Caucasians of Circassian & GEORGIAN origin and not Turkish! They were sold by their parents and became high ranked officers in the Ottoman armies. And later rulers of Egypt.

[Hakan67]

Yes, you're RIGHT. Mamluks defeated the Mongols! but the Ottomans (formal leaders of the Mamluks) defeated the Mamluks later. I made a mistake.

But Mamluks were Caucasians of Circassian & GEORGIAN origin and not Turkish! They were sold by their parents and became high ranked officers in the Ottoman armies.

?? Where did you read that?
At first the Mamluks where from turkish origin, but later more and more Mamluks were from Adyghe origin (not georgian!).
Because of the rules of Islam, moslems cant be sold as slaves.
Later, almost all Turks were Muslims, so the Mamluks went on to train Adyghe slaves.

One of the most famous Mamluk ruler, Baibar, is from turkish origin.

Now stop talking offtopic

[Goga]

?? Where did you read that?
At first the Mamluks where from turkish origin, but later more and more Mamluks were from Adyghe origin (not georgian!).
Because of the rules of Islam, moslems cant be sold as slaves.
Later, almost all Turks were Muslims, so the Mamluks went on to train Adyghe slaves.

One of the most famous Mamluk ruler, Baibar, is from turkish origin.

Now stop talking offtopic

? Mamluks were 100% not Turkish!

Adyghe were Circassian; and Circassians & Georgians are the same, all Caucasians in origin!

These children were sold by their parents and later these children (Mamluks) got the Islamic schooling from the Ottomans.

This is not offtopic, because according to me the Mamluks carried hg. G into Egypt!

[RayBanks]

Goga,

Please refer to another of my postings today which referred in detail to the absence of certain G Caucasus marker values in the available Egyptian G samples. For example, there is a high percentage of G2a1a (all with DYS392=10) in the Ossetias and Georgia, but none in the Egyptian samples. The G2a3b1a1 marker values common in the western Caucasus are also absent in the available Egyptian samples. Half of the Turkish and Greek G samples have DYS390=21 which is typical of G2a3a, but only one such sample so far from Egypt.

Ray Banks, Administrator, Haplogroup G Project

[Goga]

Goga,

Please refer to another of my postings today which referred in detail to the absence of certain G Caucasus marker values in the available Egyptian G samples. For example, there is a high percentage of G2a1a (all with DYS392=10) in the Ossetias and Georgia, but none in the Egyptian samples. The G2a3b1a1 marker values common in the western Caucasus are also absent in the available Egyptian samples. Half of the Turkish and Greek G samples have DYS390=21 which is typical of G2a3a, but only one such sample so far from Egypt.

Ray Banks, Administrator, Haplogroup G Project

Ok, you made your point. Hg. G is an ancient and very old haplogroup, so it seems that G could arrive in Egypt from West Asia in the prehistoric times!

Thank you for the correction!

[Ivan]

Ivan,

In reply to your comments about haplo G and mountainous areas

There seems a fascination unique to this site linking haplogroup G to high mountain areas.

I see the same fascination in your theory linking G in Europe almost exclusively with Jewish communities.

I never said they were exclusively anything.
All I said is that there is a strange correlation between mountainous regions and today's substantial G numbers.

You are right, they probably went there to find shelter from invaders in the first place, and that is probably why there is lack of G in today's genetics, as they were probably not involved so much in spreading genes (they were absent in a way). Scattered cases of G in Europe, I believe were from merchants who were Jewish (probably Khazarian), and christian Armenian Jugaites who were known merchants all over the world. There was 30% of Jugaites among all merchants in some 200 years ago in Venice and there is a street with their name there. Armenian Hamsheni (lots of G1a),who live in mountains like Trabzon are probably closely connected to Assyrian Christians from Nakhchivan or perhaps the same one genetically. Also Jugaites are from this area.

It is 15% among the nomadic Bakhtiari who cross the high mountains traditionally, but it is found in about the same percentage among the Mazandarani farmers and fishermen on the lowland south end of the Caspian sea.

What you forgot to say is of no less importance. Mazandarani province is the highest region with 2nd highest peak in allready mountainous Iran. It is mostly consisted of high mountains, and a Caspian sea plane. The province would not include so much mountains if in history nobody lived there. If they are fisherman now, there is nothing strange there, and it is quite normal to use all the resources as possible. It does not change a fact it is the highest part of Iran. I believe much of them came down gradually to become fisherman etc. Also here they could just follow the traditions of the rest 85%. In Serbia and in my family, most of people were shepherds, but in last 300 years most of them settled in plains, or in my case on border of Austro-Hungarian empire.

But the best known postal stamp of Mazandarani is also suggestive. A typical Mazandarani man is holding an unusual fishing rod.



As you can see the highest mountains are where Bahktiari and Mazandarani live.
Name Elevation Latitude/Longitude
1 Kūh-e Kalū Farangī, Lorestān 5,642 m 33.576 / 48.822
2 Kūh-e Damāvand, Māzandarān 5,628 m 35.955 / 52.109
3 Kūh-e Zelzeleh Khīz, Razavi Khorasan 5,150 m 35.683 / 58.484
4 Kūh-e Alborz, Qazvīn 5,025 m 36.315 / 50.753
5 ‘Alam Kūh, Māzandarān 4,822 m 36.376 / 50.962
6 Kūh-e Khersān, Māzandarān 4,659 m 36.365 / 50.952
7 Kūh-e Sabalān, Ardabīl 4,597 m 38.256 / 47.817
8 Kūh-e Haft Khvān, Māzandarān 4,537 m 36.371 / 50.939
9 Kūh-e Hezār, Kermān 4,465 m 29.512 / 57.272
10 Kūh-e Langarī, Māzandarān 4,457 m 36.379 / 50.928

We have one G2a4 man among the Bakthiari of Iran.

I thought I read most of G among Bakhtiari were G1, but not sure about that, nevertheless there is also G2a4 as you said. This is quite indicative of the way of life Otzi showed, as they are shepherds crossing mountains twice a year.
I really liked watching the movie "Grass" done by authors of King Kong in 1925. It is quite interesting as they seek better pastures in valleys.

So following your logic one can call them valley men since they graze their stock there on better pasture lands, just like Circassians did in the steppe, and where Georgians collectively became men from valleys in your interpretation.

In Switzerland I found only about 4% G in the general Swiss samples in YHRD

You noticed lack of Jews among G in Mountain Jews society in the last Yunusbayev, Caucasus revisited study, compared to that of Nasidze. From 15% to 0%. Also G1 in Armenians rose from 1% up to 10%, and half of Georgia became G.


I must say I value your work and from time to time read statistical data and peoples thoughts about G, but, I am not that versed in modern statistical and scientific tools like you are. I am just an old fashioned guy. I consider your modern tools and knowledge far more advanced.

I read carefully everything on the web considering G, and just gave my humble opinion regarding an overall G character, using my perhaps outdated and somewhat amateurish tools, but with an artistic twist, maybe just like a police sketch artist would do. I expect from you and other more intelligent and qualified people to use this sketch or disregard it.

That is why I also, chose to capture the essence of G character much more in those communities that thrived in some place, and where are still in high numbers in comparison to other groups.

Thank you for the time and effort you put in this altruistic endeavor called G haplogroup project.

[Knovas]

Possibly. It's obvious West Asian autosomes in the Peninsula have been quite replaced, and the Southeastern could reflect different migratory ways. The problem is that the Southeastern needs to be focussed more in Europe, since includes a lot of Southwest Asian in the last run. Probably you Greeks have one of the highest amounts of real Southeastern, rather than Armenians as the spreadsheet shows.

And well, I'd really like to know more information about Murcians. Paternal grandfather was from there, but I reflect very low Southeastern, and no West Asian or Southwest Asian. Perhaps G2a is really not that huge there, or the answer is in the subclades. Who knows.

[Yetos]

hmmmm

So diversity shows G around Levant?

I was thinking around minor Asia

[Dorianfinder]

hmmmm

So diversity shows G around Levant?

I was thinking around minor Asia

Hi Yetos and welcome to the forum. It is not the Levant, it is to the north in the Caucasus and through a process of diffusion in eastern Anatolia. If an individual however shows more Southwestern or Southwest Asian on a Dodecad admixture run and they are known to be haplogroup G2a, it suggests their ancestors possibly traveled along the Levantine/Mediterranean route as opposed to the Central/Alpine corridor, into Europe. But for all intents and purposes haplogroup G is Caucasian (Caucasus).

[Yetos]

Hi Yetos and welcome to the forum. It is not the Levant, it is to the north in the Caucasus and through a process of diffusion in eastern Anatolia. If an individual however shows more Southwestern or Southwest Asian on a Dodecad admixture run and they are known to be haplogroup G2a, it suggests their ancestors possibly traveled along the Levantine/Mediterranean route as opposed to the Central/Alpine corridor, into Europe. But for all intents and purposes haplogroup G is Caucasian (Caucasus).

ty that is also my believe

I'll do a quick diversity analysis for dummies using the Haplogroup G FTDNA Project, to see what subclades are in the Levant. The fuller the G tree is there, the more likely this theory is correct:

G1: Yes, Lebanon and Palestine
G2*: No, but few have tested as such
G2a*: No, but few have tested as such
G2a1: No
G2a2: No, but appears quite young
G2a3: Yes, and quite diverse in Lebanon, also present in Palestine
G2a4: No, but does include members that surround the Levant, in North Africa and the Caucasus
G2a5: No, but appears quite young
G2a6: Yes, in Lebanon
G2a7: No, but appears quite young
G2a8: No, but appears quite young
G2a9: No, but appears quite young
G2c: Yes, in Lebanon and Syria
G2d: No, but appears quite young

So, my preliminary thought is that the diversity of G in the Levant holds well enough to suppose that levels of G in the Levant could have been higher then than they are today (which is already pretty high), especially since the common J2 subclade in the Levant that makes J2 particularly frequent in the region, J2a4h, seems to be a more recent expansion (although its presence outside of the Levant may indicate otherwise... I'm not that familiar with it). I'm less clear on J1 and E in the region.

So far so good, but it needs more analysis.

the G2a3 hmmmm
and also next to J2a

maybe do we find similar ratio in any other areas? G2a3/J2a = c (almost stable ratio)
that is interesting cause maybe, if it can be proved maybe explains a lot in pre Hettit minor asia

[razor]

Here is a new hypothesis. ...
If farming started with G in the Levant, then domestication started in East Anatolia/North Mesopotamia, and the prime candidates for spreading the new herder/stockbreeder lifestyle must have been haplogroups J and R1b1b, which both seem to have originated in that region in the Mesolithic. I see a stronger association of R1b1b with cattle and sheep, and J with goats.

As I have stated many times before, I believe that R1b1b crossed the Caucasus and settled in the Pontic Steppes, becoming one branch of the Proto-Indo-European speakers, a language that they probably spread to their R1a1a neighbours to the north and east.

We are still waiting for proof of the latter mantra (:=)). What contemporary archaeology has established is that (except for the earlier Bug-Dnister culture) the hunter-forager communities of the Pontic-Caspian steppes had no domesticated cattle or sheep prior to ca. 5.200/5.000 BCE [Cf. generally Anthony,2007, pp. 134-159], and that the breakthrough for innovation came from the farming communities of southeast Europe (esp. from the Crish and subsequently from the Cucuteni-Trypilia peoples). We have no evidence for "transCaucasian" arrivals in the steppes at that time. And even if we were to assume that many Crish and Trypolian people were R1b1b, there is even less evidence of their settlement in the steppes. If anything, this latest hypothesis would support the notion that R1b were assimilated fellow-travellers of the IE's (after Yamna) rather than originators of IE speech and culture.

[sparkey]

I'll do a quick diversity analysis for dummies using the Haplogroup G FTDNA Project, to see what subclades are in the Levant. The fuller the G tree is there, the more likely this theory is correct:

G1: Yes, Lebanon and Palestine
G2*: No, but few have tested as such
G2a*: No, but few have tested as such
G2a1: No
G2a2: No, but appears quite young
G2a3: Yes, and quite diverse in Lebanon, also present in Palestine
G2a4: No, but does include members that surround the Levant, in North Africa and the Caucasus
G2a5: No, but appears quite young
G2a6: Yes, in Lebanon
G2a7: No, but appears quite young
G2a8: No, but appears quite young
G2a9: No, but appears quite young
G2c: Yes, in Lebanon and Syria
G2d: No, but appears quite young

So, my preliminary thought is that the diversity of G in the Levant holds well enough to suppose that levels of G in the Levant could have been higher then than they are today (which is already pretty high), especially since the common J2 subclade in the Levant that makes J2 particularly frequent in the region, J2a4h, seems to be a more recent expansion (although its presence outside of the Levant may indicate otherwise... I'm not that familiar with it). I'm less clear on J1 and E in the region.

So far so good, but it needs more analysis.

[Hammonia III]

I'll do a quick diversity analysis for dummies using the Haplogroup G FTDNA Project, to see what subclades are in the Levant. The fuller the G tree is there, the more likely this theory is correct:

G1: Yes, Lebanon and Palestine
G2*: No, but few have tested as such
G2a*: No, but few have tested as such
G2a1: No
G2a2: No, but appears quite young
G2a3: Yes, and quite diverse in Lebanon, also present in Palestine
G2a4: No, but does include members that surround the Levant, in North Africa and the Caucasus
G2a5: No, but appears quite young
G2a6: Yes, in Lebanon
G2a7: No, but appears quite young
G2a8: No, but appears quite young
G2a9: No, but appears quite young
G2c: Yes, in Lebanon and Syria
G2d: No, but appears quite young

So, my preliminary thought is that the diversity of G in the Levant holds well enough to suppose that levels of G in the Levant could have been higher then than they are today (which is already pretty high), especially since the common J2 subclade in the Levant that makes J2 particularly frequent in the region, J2a4h, seems to be a more recent expansion (although its presence outside of the Levant may indicate otherwise... I'm not that familiar with it). I'm less clear on J1 and E in the region.

So far so good, but it needs more analysis.

As a confirmed "dummie" I guess my question would be: Can the diversity of the Levant be shown to be greater than that of the Caucasus?

[sparkey]

As a confirmed "dummie" I guess my question would be: Can the diversity of the Levant be shown to be greater than that of the Caucasus?

No, the Caucasus seems equally diverse as the Levant, going by the same method. The only major subclade the Caucasus seems to be missing is G2a6, which probably proves a point about the G2a6 in Europe, but little about other subclades.

So I'd say that Maciamo's theory has proven to be plausible, but hasn't been proven.

[RayBanks]

No, the Caucasus seems equally diverse as the Levant, going by the same method. The only major subclade the Caucasus seems to be missing is G2a6, which probably proves a point about the G2a6 in Europe, but little about other subclades.

So I'd say that Maciamo's theory has proven to be plausible, but hasn't been proven.

The question of diversity is a valid one. Most of the areas immediately abutting the Caucasus are not very diverse, with one type of G dominating the area. Armenia so far seems the most diverse containing a high percentage of the G subgroups.

If one is looking for the ultimate area of origin, then instead Iran would attract attention because it is about the only country where there is a good mixture of both G1 and G2. But G1 is uncommon among European G persons. If one is looking at diversity to explain origins, Iran does not seem to have been a major source of European G because of the lack of G1 in Europe.

The Iceman (Italian-Austrian border) over 5000 yrs old was recently revealed as belonging to the rare for anywhere G2a4 subgroup. We have one G2a4 man among the Bakthiari of Iran.

It would seem that one can argue quite a few scenarios about the signifance of diversity depending of which G groups are included in the mix.

Ray Banks, Administrator, Haplogroup G Project.

[sparkey]

The question of diversity is a valid one. Most of the areas immediately abutting the Caucasus are not very diverse, with one type of G dominating the area. Armenia so far seems the most diverse containing a high percentage of the G subgroups.

I had assumed that the apparent lower diversity was due to lower samples sizes outside of Armenia (basically, the diversity is there, we just haven't taken samples to prove it yet). Am I wrong? If I am wrong, this is an intriguing pattern.

If one is looking for the ultimate area of origin, then instead Iran would attract attention because it is about the only country where there is a good mixture of both G1 and G2. But G1 is uncommon among European G persons. If one is looking at diversity to explain origins, Iran does not seem to have been a major source of European G because of the lack of G1 in Europe.

I agree. In the context of Europe, which is the focus here, we should look for a location where diversity is good enough to match the Neolithic expansion on Europe (doesn't need to be at the root of the G tree). I'm less concerned with frequency distributions of subclades, since those can expand and contract in both the region of origin and the region of expansion, so we shouldn't really expect them to match in both places.

[Asturrulumbo]

The Levant is a quite likely candidate in my opinion, but I must say that agriculture did not come directly to Greece from the Levant, but rather arrived through migrations from southwest Anatolia (or so I have read).

[Yetos]

The Levant is a quite likely candidate in my opinion, but I must say that agriculture did not come directly to Greece from the Levant, but rather arrived through migrations from southwest Anatolia (or so I have read).

that is an interesting, but following ancient Greek systems of agriculture, and plantation, it is rather another way,
Greeks follow that system from west minor asia and maybe came to them y messopotamia

the connection of Greece with Levant shares 2 branches,

1 is the Arcado-Cypriot Branch and the trade of Copper, remember Cyprus in Greek means copper,
2 is a colony system that creates trading posts in another area, remember that Greeks and phoenician had such colonies all over mediterrenean, changing goods, Levantines were famous for obsidian and perfums while greek for olive oil and wine, and in mid Cyprus with copper

according Greek historians Cadmus was from that areas, but we don't know if we classify Cypriot or Phoenician, or phillistine etc, most possible is a Crete-Akkadian who colonize central greece and went to ilyria,

[rms2]

The fact that the Chadic-speaking Fulani of Mali possess both a high frequency of R1b-V88 and the European T-13910 lactase persistence allele (somewhat uniquely among Africans) can be seen as supporting Maciamo's idea that R1b has a pastoralist origin, or was pastoralist when it spread to Europe.

The Fulani are nomadic pastoralists.

http://www.malariajournal.com/content/10/1/9