The Jews of Yemen were, until the airlifts of 1949 and 1950, one of the most isolated Jewish communities in the world, and one of the oldest: Jewish presence in south Arabia is attested by the third century AD, and in the fourth to sixth centuries the kingdom of Himyar, which ruled most of Yemen, adopted a form of monotheism that its own inscriptions and later Christian and Muslim sources describe as Jewish. That history poses a genetic question that no other community poses in quite the same way: is Yemenite Jewish ancestry a Levantine community that settled in Arabia, an Arabian population that adopted Judaism, or both? This post is about what the genome-wide and ancient data can say. It is part of the series introduced in Jewish ancestry and ancient DNA, and as everywhere in it: ancestry is not identity, and nothing here is a statement about who is Jewish.
What the living genomes show#
In the genome-wide studies of Jewish populations, Yemenite Jews cluster with the other Middle Eastern Jewish communities and with Levantine populations rather than standing apart, but they are also, of all the large Jewish groups, the one that sits closest to the populations of the Arabian Peninsula, and the one with a detectable affinity to them that the Iraqi, Iranian and Sephardic communities do not share (Behar et al. 2010). The Y-chromosome and mitochondrial pictures agree: the community's paternal lineages are dominated by the J1 and J2 branches common across the Levant and Arabia, and its maternal lineages include both Near Eastern and specifically Arabian branches (Non et al. 2011). The community also shows strong endogamy, with the elevated runs of homozygosity that mark a small, closed population over many centuries.
The plain reading is that both histories are true. A Levantine-derived community settled in Yemen, and over the following centuries it absorbed local Arabian ancestry, whether through conversion in the Himyarite period, through marriage, or both. Genetics cannot separate those mechanisms; it can only measure the result.
The two ancient references#
What ancient DNA adds is a pair of reference populations that were unavailable a decade ago.
The Bronze Age Levant. The Canaanite genomes of Megiddo, Hazor and Sidon describe a population made of the local Levantine Neolithic lineage and a substantial Iranian- or Caucasus-related layer (Haber et al. 2017; Agranat-Tamir et al. 2020). This is the Canaanite source in our catalog, and it is the reference for the Levantine half of the question.
The pre-Islamic Arabian Peninsula. Population-genetic work on Arabia shows a population built on a deep Natufian-related Levantine lineage, with substantial Iranian-related ancestry that arrived through Bronze Age contact across the Gulf and an African contribution that accumulated along the Red Sea (Almarri et al. 2021; Lazaridis et al. 2022). Arabians also retain the smallest Neanderthal contribution of any Eurasian population, a mark of their early separation. This is the Arabian Peninsula source, and it is the reference for the Arabian half.
The two references overlap: both descend from the Natufian lineage, and both carry Iranian-related ancestry. What distinguishes them is the African-related component and the balance of the rest. That overlap is the whole difficulty of modelling a Yemenite genome, and it is why the standard error on each weight matters more here than the weight itself.
What a qpAdm model does with a Yemenite genome#
qpAdm writes your genome as a mixture of ancient source populations, tested against a set of outgroups, and returns a weight, a standard error and a Z-score for every source and a p-value for the model that can reject it. The method is described in Understanding qpAdm and the numbers in the reading guide.
Hunter-Gatherer and Neolithic Farmer era. The Natufian source carries a larger share than in any other Jewish community, beside the Anatolian and Iranian farmer sources; the Western Steppe Herder and Western Hunter-Gatherer sources are at or near zero; and a Sub-Saharan African source may earn a small place, as it does for most Arabian genomes. That last source is the one to watch: its Z-score tells you whether the African-related share is real in your file or noise.
Classical Antiquity era. The model is Canaanite against the Arabian Peninsula source, and the search is spent almost entirely on the question of how the model divides a genome between them. Because the two overlap, a low-coverage file may return a model in which both sources pass the p-value but one of them has a standard error above the bar, and that model is not published. A higher-coverage file separates them more cleanly. The report shows the standard errors, so you can see how confidently the split was made; you can check your file's coverage for free in the file check before ordering. The analyst also tests whether the Eastern Mediterranean source fits the Levantine side better than Canaanite, and whether an Anatolian source earns a place; for most Yemenite genomes it does not.
What the model does not say#
- It does not decide the Himyarite question. A large Arabian weight is consistent with conversion, with marriage, and with a Levantine founding population that was itself partly Arabian. The sources are ancient populations; the model measures ancestry, not events.
- It does not say that any individual from Megiddo or from a Bronze Age Arabian burial was your ancestor.
- It does not measure Jewishness. A Yemeni Muslim genome and a Yemenite Jewish genome may return similar models, and that is a statement about shared ancestry, not about either community.
Related reading#
- Jewish ancestry and ancient DNA: the series introduction
- Mizrahi Jewish DNA, the communities Yemenite genomes are most often compared with
- The Canaanite source population
- Maternal haplogroups explained, for the mitochondrial side of the Yemenite question
- Why qpAdm for Jewish genomes
References#
- Agranat-Tamir, L. et al. (2020). The genomic history of the Bronze Age Southern Levant. Cell, 181(5), 1146–1157.
- Almarri, M. A. et al. (2021). The genomic history of the Middle East. Cell, 184(18), 4612–4625.
- Behar, D. M. et al. (2010). The genome-wide structure of the Jewish people. Nature, 466, 238–242.
- Haber, M. et al. (2017). Continuity and admixture in the last five millennia of Levantine history from ancient Canaanite and present-day Lebanese genome sequences. American Journal of Human Genetics, 101(2), 274–282.
- Lazaridis, I. et al. (2022). The genetic history of the Southern Arc: a bridge between West Asia and Europe. Science, 377(6609), eabm4247.
- Non, A. L. et al. (2011). Mitochondrial DNA reveals distinct evolutionary histories for Jewish populations in Yemen and Ethiopia. American Journal of Physical Anthropology, 144(1), 1–10.
