J-L283 The Case For a North of the Black Sea Migration from W Asia to Europe

Some recent discussion suggests a Phoenician-mediated migration of J-L283 to Italy from the Levant, countering the prevailing view that it migrated to Europe from north of the Black Sea / Pontic Steppes.

This misconception arises from:

  1. Inaccurate equivalence in a cited study of a Bronze Age Sidonian with J-L283 which is only confirmed J-M12
  2. Failure to recognize greater European diversity within the only two subclades of J-L283 now found in the Levant
  3. A misunderstanding of the modern presence in Sardinia of many of the oldest J-L283 subclades
  4. Incorrectly assumed origin of sibling J-Z2432 as in or near Levant
  5. The discounting of ancient J-L283 samples found in the North Caucasus and Croatia and Italy, predating the Phoenicians and/or well outside their sphere of influence
    1. 1800 BC Bronze Age Kabardino-Balkaria basal/near basal L283* [xZ627,YP91 Veseli]
    2. 1600 BC Bronze Age Croatian I4331 basal/near basal Y15058* [xZ40053,Z38241 Veseli]
    3. 1200 BC Sardinian unknown subclade
    4. 1209-1009 BC Late Bronze Age Norabak Culture in Armenia (sample RISE408) Z600(xZ627)
    5. 600-700 BC Etruscan CTS6190 [Veseli]

1. The study mistakenly equating a M12 sample from Sidon as L283 is quoted below:

"The observation in Punic sites (South-West [Sardinia] area) of haplogroups J1-L862 and J2-L283, described within Levantine Bronze Age individuals [69] and very common in modern North African [70] and Balkan populations [71], respectively, may represent traces of migrations from Levant/North African following the conquest of the island by the growing Carthaginians" - Y-chromosome and Surname Analyses for Reconstructing Past Population Structures: The Sardinian Population as a Test Case

This quote from the above study suggests that the 1200 BC ancient Nuragic samples from Sardinia who were J-L283 could plausibly have migrated there from the Levant, on the basis of it being found in Levantine Bronze Age individuals. However looking at the study linked, the sample found in Sidon dated 1630 BC is not confirmed L283, but only M12. In fact the authors themselves classify him as J2b(xJ2b2a) which meant negative for L283. In fact the sample was analyzed by Ted Kandell and found to be M205.

Since the sample is not confirmed to be J-L283 it makes more sense that it may be a different subclade of J-M12 that contains more established ancient diversity in the Levant, specifically J-M205 or to a lesser extent J-Z2453. Additionally, L283 is almost non-existent in North Africa [one Sephardim from Morocco] while J-M205 has been found in ancient Egyptians.

2. Let's examine the confirmed J-L283 from the Levant. So far no ancient samples have been found there, but we have on YFull a PH1602* from Syria and an FGC64029* from the Lebanon.

Given no presence in upstream or sibling clades in the Levant, but given diversity in Europe, the most likely conclusion, based on YFull TMRCA estimates for modern Levantine J-L283 is:

  1. One PH1602* man migrated from Europe to Syria sometime after 1100 BC
  2. One FGC64029* man migrated from Europe to Lebanon around 70 AD

3. Some of the most ancient branches of L283 have living descendants found only in Sardinia. However this is the case with many haplogroups on Sardinia and is attributed to a refugium effect compounded with an oversampling of Sardinia due to numerous studies.

The theory is that the ancestor of these branches did not actually live in Sardinia, but the ones who migrated there ended up not dying out as much. In fact, we have actually just discovered a cluster of Alsatian/Baden-Württemberg origin that splits YP29, the oldest branch of L283, near its MRCA 3400 BC. The Alsatian's Big Y is probably still processing and the Badener has submitted for YFull analysis so we should see them any day now.

Additionally a Norwegian from Rogaland and German from Voltlage form a subclade below YP29 with MRCA who lived about 1000 years ago (STR estimate) - the Norwegian is already on the YFull tree. On YFull this branch had previously been represented by a single geolocated sample, a Sardinian from Francalacci's 2013 study. In addition to these northerners, I/we predict an Albanian (Kurvelesh), Russian, German (Silesia), another German and an Armenian who traces descent to an Armenian community within Devely, Kayseri, in central Anatolia historically considered Cappadocia, to join these men under YP29.

The oldest branch of J-L283, known as YP29 on YFull and YP61 on FTDNA is about to split. Both subclades have modern presence in Germany, only one is present in Sardinia. So rather than being indicative of an ancient migration from Sardinia, it makes more sense that only one of these men migrated to found a lineage on Sardinia (possibly during Roman times) - the others stayed on the continent.

4. Incorrectly Assumed Origin of J-Z2432 in or Near Levant

Of the two main branches of Z2432, one is more well understood, Z2433. This has deep diversity in South Asia and the Persian Gulf starting 7500 years ago. Recently an Afghani was found positive.

The other branch, Y28235, has fewer samples. The earliest reliable placement of this branch in South Asia starts with Y28241 in the Bronze Age, 1600 BC.  The only other people in Y28235 on YFull are two samples - one from Azerbaijan and the other from Syria.

One sample from the Levant found in one poorly understood branch of Z2432 is simply not enough evidence to assume that it originated there. Especially given that its sibling J-L283 has no diversity or ancient samples in the Levant either.

Similarly I have seen an argument raised that recently discovered sibling to J-M241, Y167175, found in Scotland, Saudi Arabia and Yemen had a Levantine origin. This subclade is known only by three people and the Yemeni's sample is still in analysis on YFull. Regardless of the YFull analysis, this isn't evidence for a Levantine origin because these samples are not from the Levant.


5. Ancient Samples Point to Steppe Migration

We know that Bell Beakers entered Sardinia in two waves prior to the formation of the Nuragic culture around 1800 BC. One from Iberia/France and the other from Italy.

Because the ancient Sardinian L283 must predate 1200 BC, they may have entered during that time. This would fit with a steppe migration theory as the Central European Bell Beakers, which was the source of the Italian Bell Beakers who migrated to Sardinia in the second wave, would have had steppe DNA. By the time they got to Sardinia, or shortly thereafter, they may have been admixed to the point that their autosomal DNA would have lost any traces of steppe ancestry. The Y-chromosome however, immune to admixture, remains as a possible indicator of steppe ancestry.

Indeed the I4331 boy from Croatia, near basal Y15058, had steppe admixture in his autosomal DNA. He lived only 300 years after the MRCA of Y15058 so his location is especially significant for tracing this haplogroup. The sibling clade of Y15058, Z638, while widely distributed notably has Albanians or other Balkanites in almost every subclade. Further upstream, J-Z597 is the oldest subclade in J-L283 which has Balkan presence in all of its children subclades. The simplest explanation is that L283 had, at the very latest, by 2400 BC established itself in the Balkans.

The most ancient J-L283 sample ever found was a L283* (or near basal, as he has no call for BY55372 and is YP91- and Z627-) who died in Kabardino-Balkaria, just north of the Caucasus Mountains, in 1800 BC. The proximity of this sample to the Terek River, the single corridor connecting North and South Caucasus, and the fact that J-M241 is believed to have originated in W Iran (the only ancient sample of ancestor J-M102 so far been found in Tepe Abdul Hosein) suggests this man's ancestors may have migrated from W Iran via the Terek River corridor. West Iran makes sense as the origin of J-M241 because this location facilitates the diversity of its other child Z2432 between the Caspian, Persian Gulf and South Asia.

There is also an ancient Armenian from around 1200 BC Norabak Culture who is Z600(xZ627). There are other modern Armenian samples who are yet unclassified J-L283.

While we have known for some time that there are modern Armenians in J-L283, until recently there was no evidence that any were positive for sibling branch BY55372 (formerly known as YP29 on YFull). Based on an STR analysis I now predict an Armenian from Kayseri (central Anatolia) to be BY55372>YP61* because he shares rare STRs with men in YP61, today found in Albania, Sardinia, Lower Saxony, Silesia, Norway and Russia. I predict basal because some values he shares with one cluster, some with the other and some with both, and his genetic distance is about the same to both clusters (absolute GD 22-24/67). STR Match Finder generated the output below taking as input the FTDNA classic STRs along with YSEQ samples in J-M241 public projects.

Output from STR Match Finder for an Armenian J-L283 who appears to be basal J-YP61 due to shared rare STRs with known clusters and relatively equal genetic distance to both

Because Armenians are found / predicted to be in both of the oldest branches of L283, it may make sense to view them as "remnants" who stayed near the ancestral location of L283 in the South Caucasus. This is not far from the assumed ancestral location of M241 in West Iran.

If the ancient North Caucasian is not in fact Z600(xZ627) like the ancient Armenian, the Caucasus homeland argument is further strengthened by the addition of either a third L283 lineage or an offshoot of BY55372(xYP91) found in the Caucasus.

If this Armenian man from Kayseri's male ancestors were Armenians, then they had likely migrated to Cappadocia from Greater Armenia around between 900-1100. However it is possible that the Armenian from Kayseri's male ancestor was Hittite, Persian, Greek, Roman or other peoples who lived or held power in ancient Cappadocia, and only later did one of the descendants assume an Armenian identity. Before it became known as Cappadocia the region was part of Hatti, and Develi is just 240 km from the Hittite capital of Hattusa.

It is also possible that some or all of these Armenians may be the result of back migrations from the North Caucasus, as some Indo European groups, for example R-Z2103, are theorized to have migrated to the South Caucasus via the Terek river.

If we can identify exclusively Caucasus/Anatolian branches within Z600 or BY55372 we will be able to more definitively answer these questions. Getting a more accurate estimate of the TMRCA of YP61 will also help regarding the theory of Armenian/Cappadocian remnant vs back migration. As it currently stands, the MRCA of YP61 is estimated by YFull to have lived 400 BC. I don't think this estimate is reliable because it is based on a single Big Y compared against a low coverage scientific sample. If this date stands then a Greek, Balkan or Roman mediated back migration would best explain the YP61 in Armenia/Cappadocia.

So there are now four ancient samples found in many key regions pertaining to the theorized Indo European migration to Europe from the steppes, linking J-L283 to this migration due to timing and autosomal admixture after having migrated through the Tergi River corridor from the South Caucasus sometime after 3400 BC.

The Horse, the Wheel, and Language, the revised Kurgan Theory by David Anthony spells out the Indo-European migration theory in greater detail, fusing archaeological, linguistic and cultural traits of ancient and successor cultures.

Perhaps there were L283 originally in the Kura Araxes culture of the South Caucasus who migrated north and became integrated in the Maykop Culture, some of whom later went on to integrate with steppe cultures described by Anthony. Notably the tribes that they fused with did not go on to displace large portions of the British Isles as some other haplogroups did because we don't see huge founding effects in J-L283 during this time. Instead it appears more likely based on the ancient sample with steppe admixture and diversity in Italy, that these tribes primarily concentrated in the Balkans and Italy.

Leave a Reply

Your email address will not be published. Required fields are marked *