Tracing back Phoenician & Arabic DNA in modern Spaniards using Haak 2015's admixtures
I have updated the genome-wide section of my Genetic history of the Iberian peninsula by analysing Haak 2015's admixtures (K=20).
Haak et al.'s autosomal data shows that the Basques and other North Spaniards differ from other Spaniards by the absence of Bedouin-like (purple), Caucaso-Gedrosian (greyish green), and East African (pink) admixtures. These three components are found among the Southwest Asians (Arabs) and North Africans. These undeniably represent the genetic contributions of the Arabs and Berbers from the Moorish period, but also probably to a considerable extent that of the ancient Phoenicians.
The Bedouin-like admixture is the dominant component and accounts for approximately 10% of the Central and South Spanish DNA. This admixture peaks in Saudi Arabia and Yemen and some of it could indicate medieval Arabic ancestry. Then comes the Caucaso-Gedrosian (5%), which is found mostly in the Middle East, but is absent from Morocco and most of Algeria. This admixture is found in Tunisia (8%) and Sardinia (3%) though, which strongly suggests that the Phoenicians brought it to the West Mediterranean. The East African admixture only makes up 1% of Spanish genomes, the same percentage as in Sicilians and North African Jews. Berbers and Egyptians have about 10% of this admixture. Neolithic farmers would have contributed most of the 50% of the orange admixture, which represents the Early European Farmer admixture taken from actual Neolithic samples. Some Neolithic admixture would have come from the Phoenicians and the Moors. Comparing the admixtures found in Lebanon, Sardinia and Tunisia, it seems that the ancient Phoenicians had about one third of Bedouin-like (purple), one third of Caucaso-Gedrosian (greyish green) and one third of Neolithic Farmer (orange).
Since the Caucaso-Gedrosian was probably brought to Spain mostly by Phoenicians (being nearly absent from Algerian, Mozabite and BedouinB), it can be inferred that the Phoenicians contributed approximately 12% of the DNA in an average South Spanish genome (4% for each of the three admixtures). The other 6% of the Bedouin-like admixture would be medieval Arabic in origin. Using the proportions of modern Saudi Arabs as a proxy, we can estimate that the Bedouin-like admixture made up 75% of medieval Arabs' genomes. That would give a total of about 8% of Arabic DNA in a South Spanish genome today.
What about the 80% left ?
The remaining 80% of the Central/South Spanish genome is half West European Mesolithic Hunter-Gatherer (WHG) and half Early European Farmer (EEF). The Haak 2015 paper shows that Early Neolithic farmers from northern Spain (Catalonia) had between 0 and 10% of Mesolithic HG admixture (the rest being only EEF). This had increased to nearly 20% in all samples by the Middle Neolithic (so 80% EEF and 20% WHG). The assimilation of the WHG population undoubtedly continued during the Late Neolithic and Chalcolithic, probably reaching the parity observed in the Basque people in the 2nd millennium BCE.
The complete absence of Caucaso-Gedrosian admixture (i.e. the "Armenian-like" component of Yamna samples) among Neolithic farmers as well as modern Basques and North Spaniards (including Catalans) means that by the time R1b spread around Iberia their carriers' genomes had become so diluted that they had lost virtually all their Yamna admixture.
Considering that the vast majority of Spanish R1b belong to the DF27 branch, and mostly the deep L176.2 subclade for the Catalans and the M153 subclade for the Basques, it is not unlikely that the Spanish R1b has its roots in Southwest France, and that the introgression to the Iberian peninsula only dates from the Iron Age to Middle Ages. R1b could have come in several waves, from the Hallstatt Celts (500 BCE) and ancient Vascons/Gascons to the Franks (800-1000 CE), and spread southward with the Reconquista.
Hallstatt Celts would have carried R1b-U152, though, and the fact that U152 only makes up between 1 and 5% of the Spanish male lineages today can either mean that:
A) the Celts didn't have a significant impact on the Iberian population (unlikely considering that over half of the peninsula was Celtic speaking before the Roman conquest)
B) Celtic paternal lineages were replaced by haplogroups of later invaders, be them Germanic (R1b-S21, I1, I2a2a, R1a) or of Gascon origin (R1b-DF27). Surprisingly it is essentially the Gascon/Basque lineages that predominate today, even though M153 is only about 1300 years old. This would suggest a very recent, and indeed medieval replacement of a large percentage of paternal lineages.
I have updated the genome-wide section of my Genetic history of the Iberian peninsula by analysing Haak 2015's admixtures (K=20).
Haak et al.'s autosomal data shows that the Basques and other North Spaniards differ from other Spaniards by the absence of Bedouin-like (purple), Caucaso-Gedrosian (greyish green), and East African (pink) admixtures. These three components are found among the Southwest Asians (Arabs) and North Africans. These undeniably represent the genetic contributions of the Arabs and Berbers from the Moorish period, but also probably to a considerable extent that of the ancient Phoenicians.
The Bedouin-like admixture is the dominant component and accounts for approximately 10% of the Central and South Spanish DNA. This admixture peaks in Saudi Arabia and Yemen and some of it could indicate medieval Arabic ancestry. Then comes the Caucaso-Gedrosian (5%), which is found mostly in the Middle East, but is absent from Morocco and most of Algeria. This admixture is found in Tunisia (8%) and Sardinia (3%) though, which strongly suggests that the Phoenicians brought it to the West Mediterranean. The East African admixture only makes up 1% of Spanish genomes, the same percentage as in Sicilians and North African Jews. Berbers and Egyptians have about 10% of this admixture. Neolithic farmers would have contributed most of the 50% of the orange admixture, which represents the Early European Farmer admixture taken from actual Neolithic samples. Some Neolithic admixture would have come from the Phoenicians and the Moors. Comparing the admixtures found in Lebanon, Sardinia and Tunisia, it seems that the ancient Phoenicians had about one third of Bedouin-like (purple), one third of Caucaso-Gedrosian (greyish green) and one third of Neolithic Farmer (orange).
Since the Caucaso-Gedrosian was probably brought to Spain mostly by Phoenicians (being nearly absent from Algerian, Mozabite and BedouinB), it can be inferred that the Phoenicians contributed approximately 12% of the DNA in an average South Spanish genome (4% for each of the three admixtures). The other 6% of the Bedouin-like admixture would be medieval Arabic in origin. Using the proportions of modern Saudi Arabs as a proxy, we can estimate that the Bedouin-like admixture made up 75% of medieval Arabs' genomes. That would give a total of about 8% of Arabic DNA in a South Spanish genome today.
What about the 80% left ?
The remaining 80% of the Central/South Spanish genome is half West European Mesolithic Hunter-Gatherer (WHG) and half Early European Farmer (EEF). The Haak 2015 paper shows that Early Neolithic farmers from northern Spain (Catalonia) had between 0 and 10% of Mesolithic HG admixture (the rest being only EEF). This had increased to nearly 20% in all samples by the Middle Neolithic (so 80% EEF and 20% WHG). The assimilation of the WHG population undoubtedly continued during the Late Neolithic and Chalcolithic, probably reaching the parity observed in the Basque people in the 2nd millennium BCE.
The complete absence of Caucaso-Gedrosian admixture (i.e. the "Armenian-like" component of Yamna samples) among Neolithic farmers as well as modern Basques and North Spaniards (including Catalans) means that by the time R1b spread around Iberia their carriers' genomes had become so diluted that they had lost virtually all their Yamna admixture.
Considering that the vast majority of Spanish R1b belong to the DF27 branch, and mostly the deep L176.2 subclade for the Catalans and the M153 subclade for the Basques, it is not unlikely that the Spanish R1b has its roots in Southwest France, and that the introgression to the Iberian peninsula only dates from the Iron Age to Middle Ages. R1b could have come in several waves, from the Hallstatt Celts (500 BCE) and ancient Vascons/Gascons to the Franks (800-1000 CE), and spread southward with the Reconquista.
Hallstatt Celts would have carried R1b-U152, though, and the fact that U152 only makes up between 1 and 5% of the Spanish male lineages today can either mean that:
A) the Celts didn't have a significant impact on the Iberian population (unlikely considering that over half of the peninsula was Celtic speaking before the Roman conquest)
B) Celtic paternal lineages were replaced by haplogroups of later invaders, be them Germanic (R1b-S21, I1, I2a2a, R1a) or of Gascon origin (R1b-DF27). Surprisingly it is essentially the Gascon/Basque lineages that predominate today, even though M153 is only about 1300 years old. This would suggest a very recent, and indeed medieval replacement of a large percentage of paternal lineages.
Last edited: