By Roy Banks
The reader will find there is more uncertainty than certainty about the migrations that may have spread haplogroup G. The best hints come from examining the combinations of ancestral origins reported in each set of subgroups in the geographic distribution section of this site. Much of this section, however, will examine migrations that MAY have been involved in spreading haplogroup G and will indicate why some others are unlikely sources. In some cases, the principal spread to Europe may have occurred by small bands of men up to 3,000 yrs ago who have been lost to history and became part of later identified groups.
Haplogroup G seems to have expanded from southwestern Asia or the Middle East to Europe west of the Black Sea and to the western Mediterranean and to central Asia primarily about 1,700 to 3,000 years ago. The time period is based on time calculations derived by comparing the number of STR marker mutations among men in these three areas and the presumed region of origin. This region of origin for the major G groups of today was most likely somewhere in Turkey, the Caucasus Mountains region, Iran or the eastern Mediterranean where G is found both in significant numbers and with the diversity expected with a long presence.
The principal spread of G from the mentioned region was to Europe west of the Black Sea. As to the most common G groups there, their (a) presence in 4% or more of the European population plus (b) seeming ancestral splits (based on mutations in 67-marker STR samples) between persons in southern and northern Europe no more recently than the early Middle Ages make it possible the G ancestors came as part of identifiable groups (or a group) rather than as isolated movements of individuals. But a migration to Europe shortly after some key mutations took place perhaps 3000 yrs ago might also explain the large numbers in certain G groups. There are also Jewish components of the major G groups which require explanation when advocating any particular ancestral group for these major G groups now in Europe.
The likeliest geographical sources for spread of each of the following principal G subgroups is:
G1 -- Iran or eastern Middle East......
G2a1 -- the northern or southern Caucasus region.....
G2a3a -- Turkey or the eastern Mediterranean......
G2a3b1a (DYS388= 13 type) -- perhaps southern Caucasus region or developed in Europe.....
G2a3b1a (DYS568= 9 type) -- unknown.....can't rule out a development in Europe
G2a3b1a1 -- the northern or southern Caucasus.....
G2a3b1a1a -- Iran or southern Caucasus region......
G2c -- northern Middle East
North of the Caucasus Mountains
If G migrations originated in the area north of the Caucasus Mountains, surviving traces of an original G population in today's Ukraine and almost all of Russia are few. The exceptions today are in Russia along the north slope of the Caucasus where G2a3b1a1 is found among the Kabardinians, G2a1 is common among the Ossetians, and an unknown G2 type is found in Dagestan and among the Adyghe. Other G types are found in small numbers in these three areas. The variety of G types there could indicate a long G presence in the northern Caucasus if G did not arrive piecemeal from southern locations a long time after the formation of each of these G types.
There were two major groups immediately north of the Caucasus area in the period of interest who engaged in migrations: the Sarmatians and the Huns. Both groups have cultural similarities, such as manipulating infant head shapes. One after another, the two moved into the area north of the Caucasus from the east when a westward movement provided better options than resisting groups to the east. The Sarmatians subsumed an existing population of Scythians during their occupation of this area. Like the Scythians, the Sarmatians spoke an Iranian language. The more central portion of the Sarmatians was identified as Alan Sarmatians. Strabo claimed these groups could assemble hundreds of thousands of warriors. (S-2) Being nomadic with a history of moving en masse, it is conceivable that the Sarmatians when they began the Great Migration westward in the 1st century B.C.E. took with them a high percentage of their population. The Huns were also nomadic and had been as far east as central Asia. Their language is not well defined. Complicating an understanding of the genetic make-up of these groups, scholars have described both Sarmatians and Huns as confederations of groups rather than solitary ethnic groups.
Greek historian Diodorus Siculus related that the Scythians had carried Medes to the River Don north of the Caucasus presumably from today's northwestern Iran, this event giving rise to their name Sauromatians. (D-1) Scythians had occupied Media 653-625 B.C.E. Pliny also suggested Sarmatians descended from Medes. (P-1) But historians have difficulty explaining how the Sarmatians instead seemed to come from central Asia off to the east in their occupation of Scythian lands around the 5th century B.C.E. The Huns in their westward migration had reached today's Ukraine by the 2nd century C.E. replacing the Sarmatians who left, and they moved dramatically into more western lands in the 5th century. Under Hun rule, some Alans who had not migrated remained in the area north of the Caucasus. In the Middle Ages, invading Mongols reportedly forced them down to the Caucasus area. And there are Alan remnants which appeared in multiple locations. A group of 30,000 Alans, for example, formed the royal guard of the Yuan court in Dadu (Beijing, China), and a group of Alans arrived from unknown locations in Medieval Hungary.
While it is conceivable that many of the G persons in Europe proper today had ancestors part of the Great Migration that carried large populations out of southwestern Russia, they left few close relatives in the area they vacated if so. G2a1 which predominates among the North Ossetians is rare in Europe proper. The Caucasus G2a3b1a1 common among the Kabardinians (DYS390=23, DYS388=12, DYS385=13,14), for example, is seen only rarely in Europe proper where the DYS388=13 type (DYS390=22, DYS385= 14,15) predominates. In addition, none of the ancient DNA recovered so far from skeletons in the Scythian/Sarmatian area of occupation has belonged to haplogroup G.
While these mentioned northern Caucasus types of G are uncommon in lands to the north or in Europe proper, they have significant equivalents in the southeastern Caucasus, particularly in Georgia, including South Ossetia which is often included with Georgia.
If these persons in the north Caucasus were originally Medes who became Sauromatians (later Sarmatians and Alans) when carried to the north Caucasus by the Scythians, the G groups in the land of the Kingdom of the Medes in northwestern Iran, and also the rest of Iran, are different from what is seen in the northern Caucasus. G1, for example, is common in Iran, but rare in the north Caucasus.
South of the Caucasus Mountains
In contrast to the regional types of G seen north of the Caucasus Mountains, the three countries of Georgia, Armenia and Azerbaijan provide STR samples consistent with all types of haplogroup G seen today in Europe west of the Black Sea with the exception of G1 and G2c. And the regional STR types north of the Caucasus are seen in smaller numbers south of the Caucasus. Because of the rich diversity of these G types, this area of the world is a leading candidate for the starting point for most G migrations especially those to western and central Europe.
The STR marker samples in the YHRD database (Y-1) from these three countries obtained from the Nasidze study, (N-1) as well as samples in the Haplogroup G Project, contain examples of (1) G2a1, (2) G2a3a, (3) G2a3b1a (DYS388=13 type), (4) G2a3b1a1 and (5) G2a3b1a1a. These five G types represent over 90% of G types found in western and central Europe. However, the presence of the DYS388=13 type of G2a3b1a and G2a3b1a1a in this region have yet to be confirmed by SNP testing.
Both the Babylonian Empire and Persian Empire once included the southern Caucasus and stretched to the Mediterranean, but there is no seeming report of groups from the Caucasus being relocated to the western limits of these empires. But the Assyrians deported Jews to Armenia, and 10,000 Jews moved there on their own. The commonly accepted view is that the Mountain Jews of Azerbaijan were early Medieval immigrants from southwestern Persia and this immigration continued throughout the Middle Ages. More recently a high percentage of this community emigrated to Israel. The seemingly related non-Jewish Tats of Azerbaijan and southern Dagestan are descended from Persians who in the time of the dynasty of Sassanids (III-VII C.E.) built cities and founded military garrisons to strengthen their positions in the north. Kurds from Turkey and Iran settled in Georgia and Armenia beginning the late 19th century, but a Nasidze study did not find haplogroup G among them. (N-2) Arab persons from the Middle East are also known to have emigrated to the southern Caucasus with the spread of Islam.
Another possible source of a G migration from the Caucasus would be men recruited into military units during the Roman Empire. Armenia which encompassed nearly all the southern Caucasus region was occupied for a while in the 2nd century C.E. None of the Roman legions bore the name of locations in the southern Caucasus. Other possible sources of southern Caucasus G migrations were the Jewish communities. There are significant Jewish subgroups within G2a3a, G2a1 and certain types of G2a3b1a. The Jewish Radhanites had significant presences in western and central Europe in the early Medieval period with trade routes that included the eastern shore of the Caucasus together with some known Jewish communities in the Caucasus region.
There was also a significant outflow of persons�perhaps more than 100,000 persons�from Armenia during the Byzantine Empire due to various emperors deporting persons to parts of modern-day Turkey or Bulgaria or conscripting soldiers.G-1) (S-3) (D-2) (S-4) (T-1) (P-2) In addition, 10,000 Armenians were said to have settled in Cyprus in 578 C.E. (D-3)
One problem with identifying the southern Caucasus as the source of most of the G in western and central Europe is the minority status in the southern Caucasus of the DYS388=13 type of G2a3b1 and the inability so far to obtain SNP confirmation of its presence in the southern Caucasus. This DYS388=13 group, in contrast, is dominant in Europe but its rarity elsewhere actually makes no area the most logical location as a source for its migration.
In Iran G1 is found in significant numbers in all tested areas. The Persian Empire once extended to the Mediterranean, and Persian military forces may have carried G persons into the southern Caucasus as well as to the shores of the Mediterranean. In fact, the highest percentages of G1 outside of Iran are in the countries just to the west of southern Iran that were once part of the Persian Empire. If the G movement occurred instead before the Persian Empire, G1 could have originated either in the Middle East or Iran.
While G2a3a occurs in Iran, it is apparently too rare there to account for any source group for G2a3a persons who migrated in numbers elsewhere. The same situation applies to G2a1 there. Seeming G2a3b1a Iranian samples are available, especially G2a3b1a1a1, (though not confirmed by SNP testing) and several unclassified types of G2a and G2a3 are noted. Neither G2c nor G2a3b1a1 has surfaced so far among Iranian samples.
There are significant Jewish components to G1 groups. Main article: See Jews with Haplogroup G. The Jewish communities in the Holy Land contributed substantially to the Persian (now Iranian) population beginning in 727 and 721 B.C.E. with the deportation of Jews to Media and Persia. This was followed by the deportation of over 27,000 Jews to southwestern Persia. In 680 B.C.E. to escape another bad situation there was a Jewish emigration primarily into the Persian city ofIsfahan. In the 6th century B.C.E. at the end of the Babylonian Captivity,Cyrus allowed additional Jews into Persia. In 135 C.E., there was yet another wave of emigration into Persia by Jews fleeing the Roman persecutions. At the time of the original emigrations, Jews intermarried with the local population. (P-3) This practice soon stopped, but Jews were not forced into ghettos until Islamic times. The Jewish Radhanites could also have spread G samples to Europe in their travels in the early Middle Ages. They maintained a significant presence on the trade routes that connected Iran to the rest of the world. A major gap in genetic knowledge of Iran involves theKurdish areas. Jews were once commonly found there, but moved in large numbers toIsrael after its founding.
Turkey and the eastern Mediterranean
This region has unusual percentages of G2a3a. Beginning with the Phoenicians, any of the trading and conquering governments of the Mediterranean could have spread G2a3a to the western Mediterranean where it is more common than in Europe to the north. G2a3a also has a Jewish-Italian subgroup, and any of the Jewish diaspora movements may have contributed to the spread. The percentages of G2a1 and the DYS388=13 type of G2a3b1 are so tiny in this area, that it is unlikely Turkey was the starting point for a migration of these G groups.
G1 is common in the northern Middle East, and that area may have been the source of G movements to Iran or vice versa. G2c is found in small numbers in the northern Middle East, and the Middle East could have been the source from which European G2c and Pakistani G2c1 derived.