Origins & History
The first J1 men lived in the Late Upper Paleolithic, shortly before the end of the last Ice Age. Like many other successful lineages from the Middle East, J1 is thought to have undergone a major population expansion during the Neolithic period.
Chiaroni et al. (2010) found that the greatest genetic diversity of J1 haplotypes was found in eastern Anatolia, near Lake Van in central Kurdistan. Eastern Anatolia and the Zagros mountains are the region where goats and sheep were first domesticated, some 11,000 years ago. Chiaroni et al. estimated that J1-P58 started expanding 9,000 to 10,000 years ago as pastoralists from the Fertile Crescent. Although they did not analyze the other branches, it is most likely that all surviving J1 lineages share the same origin as goat and sheep herders from the Taurus and Zagros mountains.
The mountainous terrain of the Caucasus, Anatolia and modern Iran, which wasn't suitable for early cereal farming, was an ideal ground for goat and sheep herding and catalyzed the propagation of J1 pastoralists. Having colonised most of Anatolia, J1 herders would have settled the mountainous regions of Europe, including the southern Balkans, the Carpathians, central and southern Italy (Apennines, Sicily, Sardinia), southern France (especially Auvergne), and most of the Iberian peninsula. Hotspots of J1 in northern Spain (Cantabria, Asturias) appear to be essentially lineages descended from these Southwest Asian Neolithic herders.
Most J1 Europeans belong to the J1-Z1828 branch, which is also found in Anatolia and the Caucasus, but not in Arabic countries. The Z1842 subclade of Z1828 is the most common variety of J1 in Armenia and Georgia. There are also two other minor European branches: J1-Z2223, which has been found in Anatolia, Germany, Belgium, Ireland and Spain, and J1-M365.1, identified only in England and Spain at the moment. Their very upstream position in the phylogenetic tree and their scarcity in the Middle East suggests that these were among the earliest J1 lineages to leave the Middle East, probably in the Early Neolithic, or possibly even as Late Paleolithic hunter-gatherers that wandered outside Anatolia and ended up in western Europe.
Within the Middle East, SNP analysis shows that the J1-L136 branch migrated south from eastern Anatolia and split in three directions: the Levant, the southern Zagros (and southern Mesopotamia ?), and the mountainous south-western corner of the Arabian peninsula (mostly in Yemen), bypassing the Arabian Desert. That latter group, consisting essentially of J1-P56 lineages, crossed the Red Sea to settle Sudan, Eritrea, Djibouti and northern Somalia. The climate would have been considerably less arid than today during the Neolithic period, allowing for a relatively easy transmigration across the Middle East with herds of goats.
Neolithic J1 goat herders were almost certainly not homogenous tribes consisting exclusively of J1 lineages, but in all likelihood a blend of J1 and T1 lineages. So much is evident from the presence of both J1 and T1 in north-east Africa, Yemen, Saudi Arabia, but also in the Fertile Crescent, the Caucasus and the mountainous parts of southern Europe. Maternal lineages also correlate. Wherever J1 and T1 are found in high frequency, mtDNA haplogroups HV, N1 and U3 are also present, as well as J, K and T to a lower extent (=>see Correlating the mtDNA haplogroups of the original Y-haplogroup J1 and T1 herders). It is unclear whether goats were domesticated by a tribe that already comprised both J1 and T1 lineages, or if the merger between the two groups happened during the Neolithic expansion, when two separate tribes would have bumped into each others, intermixed, and thereafter propagated together.
Bronze Age expansion of J1-P58
Kitchen et al. (2009) estimated through a Bayesian phylogenetic analysis that Semitic languages originated in the Levant around 3,750 BCE, during the Early Bronze Age. It evolved into three groups: East Semitic (an extinct branch that comprised Akkadian), Central Semitic (which gave rise to Aramaic, Ugaritic, Phoenician, Hebrew and Arabic), and South Semitic (South Arabian and Ethiopian).
J1-P58, the Central Semitic branch of J1, appears to have expanded from the southern Levant (Israel, Palestine, Jordan) across the Arabian peninsula during the Bronze Age, from approximately 3,500 to 2,500 BCE. Camels were domesticated in Somalia and southern Arabia c. 3,000 BCE, but did not become widely used in the southern Levant before approximately 1,100 BCE. Camels played an important role in the further diffusion of J1-P58 lineages, notably with the Bedouins in the desertic parts of the Middle East and North Africa. Bedouins now make up a substantial percentage of the population of Sudan (33%), Libya (15%), the United Arab Emirates (8%) and Saudi Arabia (5%).
The two most common Jewish subclades of J1 downstream of P58 are L816 and ZS227. The latter includes the Cohanim haplotype. Most of the other branches under P58 could be described as Arabic, although only P858 seems to be genuinely linked to the medieval Arabic expansion from Saudi Arabia.
Based on very limited data, the main Lebanese subclades of J1 appear to be J1-YSC234 and J1-YSC76. Both subclades have also been found in Sicily, Andalusia and Portugal, which suggests that they were already found among thePhoenicians. However, since the Arabs conquered the same regions as those colonised by the Phoenicians, it is too early to reach such a conclusion.
| Is J1-P58 the main Arabic paternal lineage ? |
|---|
Looking at the Y-DNA haplogroup frequencies by country it is easy to assume that J1 is a marker of Arabic ancestry because J1 reaches its maximum frequency in and around the Arabian peninsula. Things are more complicated, however. It is important to make a clear distinction between people who speak Arabic and those who are genetically Arabic. These are two completely different things. For comparison, people who speak languages descended from Latin (French, Spanish, Portuguese, Italian, Romanian) are not necessarily descended from the ancient Romans of the Latium. Even those who do may not have more than a tiny fraction of their genome which was inherited from actual Roman ancestors. This is why most Romance-language speakers today cannot be considered as genetically Roman. Most present-day Arabic speakers outside the Arabian Peninsula are likewise only very partially or not all Arabic genetically. In the northern half of the Middle East, most of the people who call themselves Arabs of today are in fact mainly descendants of other historic peoples, such as the Phoenicians, Assyrians, Babylonians, or even the Hurrians. Most of these peoples are predominantly J2, with many minority haplogroups (E1b1b, G, J1, L, Q, R1a, R1b, T). The confusion comes from the fact that the Arabic language, which appeared in little more than 1,500 years ago, is much more recent than the haplogroups J1 and J2, which are both over 15,000 years old. Even the J1-P58 subclade, associated with the historical Arabs, very clearly predates the Arabic language. The common ancestor of the J1-P58 men dates back to approximately 3500 to 4000 years ago, a time that corresponds to the development of the Central Semitic dialects. That is why that this haplogroup is common to both the Jews and the Arabs. Yet hardly anyone would argue that Jews are Arabs genetically. The two peoples are Semitic, but Jews are not Arabs, and vice versa. The true lineage of the historic Arab people (so mainly from Jordan and Saudi Arabia) is J1-YSC235 and its subclade,J1-P858. The other subclades of J1 cannot be considered to be the paternal descendants of first speakers of Arabic. These other J1 lineages were Arabicized alongside other haplogroups (J2, Q1b, etc.) during the Islamic expansion from the 7th century onward. It is true that these more genuinely Arabic J1-P858 lineages are found everywhere in the Arabic-speaking world today, but they only represent a small minority of lineages everywhere but inside the Arabian peninsula. More importantly, J1-P858 is not the only haplogroup that spread with the Arabic expansion linked to the diffusion of Islam. Nowadays only 40% of Saudis and 30% of Jordanians belong to J1. E1b1b-M34 is another important Arabic lineage, being found in 25% of Jordanians and 10% of Saudis. Like J1-P58, E-M34 it is also shared with their Semitic cousins, the Jews. Haplogroup E1b1b is considered the prime candidate for the origin and dispsersal of Afro-Asiatic languages across northern and eastern Africa and south-west Asia. The Semitic languages appear to have originated within a subclade of the M34 branch of E1b1b. One specific deeper subclade is surely associated with the development of Arabic language and with J1-P858, but it hasn't been identified yet. Note that E-M34 itself is many thousands of years old and is also found in non-Semitic countries, including Turkey, Greece, Italy, France and Spain. |
Geographic distribution
Distribution of haplogroup J1 in Europe, the Middle East & North Africa
Frequencies og haplogroup J1 in Europe and West Asia tend to vary considerably from one regional community to the next. The highest local percentages in Europe are found in Greece, Italy, France, Spain and Portugal and hardly ever exceed 5% of the population. However Italy, France and Spain also have areas where J1 appears completely absent. Even in northern Europe, where the nation-wide frequencies are below 0.5%, very localised pockets of J1 have been observed in Scotland, England, Belgium, Germany and Poland. Larger sample sizes are needed to get a clearer picture of the distribution of J1 in Europe.
Surpisingly, even in the Caucasus and in Anatolia, the region where this haplogroup is thought to have originated, there are wide discrepancies between regions. For example, the Kubachi and Dargins from Dagestan in the Northeast Caucasus have over 80% of J1 lineages, while in their Ingush neighbours, 200 km to the north, it barely reaches 3%. East Anatolia around Lake Van sees over 30% of J1, whereas south-west Anatolia has only 2%. Even within Kurdistan frequencies vary greatly. The small sample sizes for each region is surely to blame.
In Arabic countries, J1 climaxes among the Marsh Arabs of South Iraq (81%), the Sudanese Arabs (73%), the Yemeni (72%), the Bedouins (63%), the Qatari (58%), the Saudi (40%), the Omani (38%) and the Palestinian Arabs (38%). High percentages are also observed in the United Arab Emirates (35%), coastal Algeria (35%), Jordan (31%), Syria (30%), Tunisia (30%), Egypt (21%) and Lebanon (20%). Most of the Arabic J1 belongs to the J1c3 variety.
Surpisingly, even in the Caucasus and in Anatolia, the region where this haplogroup is thought to have originated, there are wide discrepancies between regions. For example, the Kubachi and Dargins from Dagestan in the Northeast Caucasus have over 80% of J1 lineages, while in their Ingush neighbours, 200 km to the north, it barely reaches 3%. East Anatolia around Lake Van sees over 30% of J1, whereas south-west Anatolia has only 2%. Even within Kurdistan frequencies vary greatly. The small sample sizes for each region is surely to blame.
In Arabic countries, J1 climaxes among the Marsh Arabs of South Iraq (81%), the Sudanese Arabs (73%), the Yemeni (72%), the Bedouins (63%), the Qatari (58%), the Saudi (40%), the Omani (38%) and the Palestinian Arabs (38%). High percentages are also observed in the United Arab Emirates (35%), coastal Algeria (35%), Jordan (31%), Syria (30%), Tunisia (30%), Egypt (21%) and Lebanon (20%). Most of the Arabic J1 belongs to the J1c3 variety.
Subclades
The above tree was created mostly with the data from the Haplogroup J Project at Family Tree DNA, with some additional deep subclades from Victar Josef Mas' phylogenetic tree.
J1 can be divided in two main groups: the very large J1-P58 subclade, and the other branches of J1.
J1-P58 (J1b2 on the ISOGG tree, formerly known as J1c3) is by far the most widespread subclade of J1. It is a typically Semitic haplogroup, making up most of the population of the Arabian peninsula, where it accounts for approximately 40% to 75% of male lineages. The dominant lineage in the Arabian peninsula is J1-L147.1. The numerous subclades downstream of L858 represent the tremendous expansion of J1 lineages linked to the propagation of Islam and the Arabic language from Saudi Arabia from the 7th century CE.
L147.1 is also the Cohen Modal Haplotype. Roughly half of all Cohanim belong to the L147.1 subclade. In the Hebrew Bible the common ancestor of all Cohens is identified as Aaron, the brother of Moses.
J1-P58 is thought to have expanded from eastern Anatolia to the Levant, Taurus and Zagros mountains and the Arabian peninsula at the end of the last Ice Age (12,000 years ago) with the seasonal migrations of pastoralists. Arabic speakers recolonised the Arabian peninsula in the Bronze Age from the north-west of the peninsula, close to modern Jordan. The rise of Islam in the 7th century CE played a major part in the re-expansion of J1 from Arabia throughout the Middle East, as well as to North Africa, and to a lower extent to Sicily and southern Spain.
Z2223, M365.1, L136 and Z1828 could all have been minor lineages accompanying the main Neolithic haplogroup G2a. That would explain their very wide geographic distribution and low frequency outside their region of origin.
Like haplogroup G, J1 might have been of the principal lineages to bring domesticated animals to Europe. Both G and J1 reach their maximal frequencies in the Caucasus, some ethnic groups being almost exclusively J1 (Kubachis, Kaitaks, Dargins, Avars), while others have extremely high levels of G (Shapsugs, North Ossetians). Most of the ethnic groups in the North Caucasus have between 20 and 40% of each haplogroup, which are by far their two dominant haplogroups. In the South Caucasus (Georgia, Armenia, Azerbaijan), haplogroup J2 comes into the admixture and is in fact slightly higher than either J1 or G. Most of the Caucasian J1 is at present J1*, meaning that at present no common SNP has been identified that could form a new subclade. Armenia stands out of the lot by having a substantial J1c3d minority (at least one third of all J1, i.e. roughly 4% of the population).
J1 can be divided in two main groups: the very large J1-P58 subclade, and the other branches of J1.
J1-P58 (J1b2 on the ISOGG tree, formerly known as J1c3) is by far the most widespread subclade of J1. It is a typically Semitic haplogroup, making up most of the population of the Arabian peninsula, where it accounts for approximately 40% to 75% of male lineages. The dominant lineage in the Arabian peninsula is J1-L147.1. The numerous subclades downstream of L858 represent the tremendous expansion of J1 lineages linked to the propagation of Islam and the Arabic language from Saudi Arabia from the 7th century CE.
L147.1 is also the Cohen Modal Haplotype. Roughly half of all Cohanim belong to the L147.1 subclade. In the Hebrew Bible the common ancestor of all Cohens is identified as Aaron, the brother of Moses.
J1-P58 is thought to have expanded from eastern Anatolia to the Levant, Taurus and Zagros mountains and the Arabian peninsula at the end of the last Ice Age (12,000 years ago) with the seasonal migrations of pastoralists. Arabic speakers recolonised the Arabian peninsula in the Bronze Age from the north-west of the peninsula, close to modern Jordan. The rise of Islam in the 7th century CE played a major part in the re-expansion of J1 from Arabia throughout the Middle East, as well as to North Africa, and to a lower extent to Sicily and southern Spain.
- J1-L147.1 is the main Arabic cluster as well as the Cohanim haplotype (YCAII=22-22) among Jews.
- J1-L444 is a small Arabic subclade defined by DYS531=12.
- Both L616 (aka L615) and L859 share an STR value for DYS485 around 14.
- J1-Z640 (aka Z641 or Z644) is mostly a European subclade, although it has been found also in Turkey and in the Arabian peninsula. Z640+ members typically have DYS561=14. Its subclade L174.1 can be identifed by the STR value DYS594=11.
- J1-L1253 corresponds to DYS557>18. It appears to be limited to Britain and Ireland.
- J1-L823 corresponds to DYS452=30.
- J1-L1279 corresponds to DYS497=16.
- J1-L817 is a major Jewish cluster, also defined by L818 and DYS392=13. Some members belong to its subclade L816.
- J1-L93 (aka L92.1) and J1-L386 are both defined by DYS446=13. L93 has a DYS641 value smaller or equal to 10. L386 can be identified by YCAIIa=17.
As of early 2013:
- J1-Z2223 matches the STR value DYS446=12. It has been found in Anatolia, Germany and in the British Isles.
- J1-M365.1 has been only found in very low frequencies in Western Europe (Western Iberia, French Pyrenees, Belgium and England) and northern Iran. It may have been a minor lineage of Neolithic farmers/stockbreeders.
- J1-L136 lineages negative for P56 or P58 are generally restricted to Europe. L136* is found in Eastern, Central and Mediterranean Europe. It also probably has a Neolithic origin.
- J1-P56 is a minor Arabic cluster found on the Red Sea coast of Saudi Arabia and Yemen. It can be identified by the STR value DYS641=11.
- J1-Z1828 is defined by the STR values DYS436=11 and DYS388<15. This is the second most common top level subclade after J1-P58. It is particularly frequent around the Taurus and Zagros mountains and in the Caucasus, but has also been found at low frequencies in western Turkey, Greece, South Italy, Central Europe, France, and the British Isles. The L1189 subclade seems to be mostly European, while the Z1842 subclade is chiefly restricted to the Caucasus, Zagros and Taurus.
- J1-M62 is an isolated lineage found in a Crimean Tatar of Uzbekistan (probably a private SNP).
Z2223, M365.1, L136 and Z1828 could all have been minor lineages accompanying the main Neolithic haplogroup G2a. That would explain their very wide geographic distribution and low frequency outside their region of origin.
Like haplogroup G, J1 might have been of the principal lineages to bring domesticated animals to Europe. Both G and J1 reach their maximal frequencies in the Caucasus, some ethnic groups being almost exclusively J1 (Kubachis, Kaitaks, Dargins, Avars), while others have extremely high levels of G (Shapsugs, North Ossetians). Most of the ethnic groups in the North Caucasus have between 20 and 40% of each haplogroup, which are by far their two dominant haplogroups. In the South Caucasus (Georgia, Armenia, Azerbaijan), haplogroup J2 comes into the admixture and is in fact slightly higher than either J1 or G. Most of the Caucasian J1 is at present J1*, meaning that at present no common SNP has been identified that could form a new subclade. Armenia stands out of the lot by having a substantial J1c3d minority (at least one third of all J1, i.e. roughly 4% of the population).
