View Full Version : Do Yemenis and possibly other Arabians have Hunter-Gatherer ancestry?

07-07-2019, 07:22 PM
I notice that Yemenis predominantly and some Arabians carry MT haplogroups (L3I, L3X and L6 to name a few) that seem to come from palaeolithic easter african hunter-gatherers seeing as the haplogroups are restricted to East Africa and Arabia. However, the subclades and the TMRCA seem to indicate not a recent admixture say from the Arab slave trade but perhaps something much older.

There may have been a land bridge between the two continents during the LGM and with fluctuating weather conditions between 20,000-10,000 BC in the Area.


Could this have caused some hunter-gatherers to leave and cross over into the Arabian peninsula?

Arabians arrived from the North and Migrated South, during that time could they have encountered some hunter-gatherers and mixed with them?

07-08-2019, 06:43 AM
This would be fascinating - there is some earlier preceding groups, some which had very distinctive cultural assemblages that have been suggested as offshoots of far earlier preceding cultures from as early as 100,000 years ago, especially towards Oman.

This should tie in somewhat with a discussion that we had in another thread, where @NetNomad argued Somali mtdna with a seemingly more localized Horn African origin not coming from a variety of sources including Habesha and Oromo admixture.

Concerning these Horn African-looking L lineages in Arabians - the fact that Soqotris and other Yemenis sharing mtdna L clades with Horn Africans with deep divergences and a paleolithic Horn/NE African association would point in this direction. But I think we need more sampling, as the same mtdna lineages in patriachal regions with high absorbtion of exogenous maternal lineages - as Somalis are a product of - result in a high chance of not only a high diversity of mtdna subclades, but radically different origins for even the most closely related seeming branchings. This, and the very deep divergence and regional concentration between the various concerned L lineages here - and in general - supports this latter view.

Essentially my arguement is - don't expect Somali L3x carriers to share a common L3x matriarch who was assimilated into an early proto-Samaal camel herding community, but to come from a variety of sources (Omotic, a variety of unknown and unrelated pre-Afroasiatic paleolithic HG, Borana, etc), and with different time depths amongst Somalis. Those South Cushitic nomads had a fairly high frequency of L3x, and were mostly L3x1a. They also had L3h, but it's L3h1 - not the one most Somalis carry, which is L3h2. And while the South Cushitic and the Nilotic PIA samples had L3h1a, the latter had a wholly different subclade of L3h1a - L3h1a1, where the South Cushitic group was mostly L3h1a2, with a singleton L3h1a1. L3h1a alone is 40-55,000 years old, so this should not be shocking. L3h1a1 is 35 - 50,000 years old. These are extreme time depths when it comes to population genetics, and the amount of dead end, trace contributing, and ghost populations is massive. It makes answering anything definitively futile, as the same occurs for basically most L lineages known, whose most recent branchings usually date as far back as 7,000 years ago.

To further illustrate how divergent and how by sharing a not reasonably delimited lettered lineage can obscure how unrelated two samples can be - How difficult is it to trace L3a subclades to a single region when L3a's various immediate branchings occur anywhere from 60 - 30,000 years ago. And even if we only focused on finding the sources for Somali L3a, Somali L3a carriers have almost every other branch of this divergent lineage, and the resulting Somali L3a tmrca is the L3a matriarch itself, and we arrive at the same exact problem as before. This same thing is occuring in regards to L3f, a lineage estimated to be as anywhere from 45 - 75,000 years old, as some Somali samples are L3f1, others L3f2, and share a common tmrca anywhere from 25,000 - 60,000 years. If a population as homogenous as Somalis are like this - what do you expect for other populations throughout the African landmass? especially those in the Horn and Northeast Africa - which has the highest diversity of mtdna anywhere % wise.

While there is no doubt many of these tmrcas while decrease, there is a trend of many mrca's actually increasing. There is also no doubt that just as with the A00, D00, and D0 situation, many existing critical subclades and unknown mtdna lineages are yet to be found and mapped.

If we are to answer your question more definitively - I would look to whatever it is the Mahra carrying that differentiates them from Levantine Semitic groups. They were noted as also having more Neanderthal ancestry than other Arabians, and have very low ancestral East African-like ancestry, of the same kind and % as unmixed Levantine groups like Samaritans and Lebanese Christians.

Some we can probably say that at least for whatever pre-Semitic population the ancestors of the Mehri mixed with - they probably had higher Neanderthal, lesser ancestral East African-like ancestry (essentially noise-level), and lower to no Iran Chl ancestry. And looking at the ydna, may have been rich in some basal J lineages. I'd surmise that the West Semitic branching that constituted the common ancestors of the Mehri and other MSA groups and Habeshas mixed with this population, and the rather divergent and unexpected J1 lineages characteristic of these groups, such as J1-P56, and more basal J1-P58 lineages such as J1-L93. They maybe had high mtdna N*, but not of the same kind as Cushitic/Erythreans and more aboriginal NE African N lineages.

07-08-2019, 09:03 AM
More sampling should get us better dates and clarify the branching.

A quick aside -
The idea that some Yemeni's(and others) are admixed goes back at least as far as physical anthropologists like Coon and Eickstedt. The label they gave some coastal Yemeni's in their racial typing system was "Arabian Veddoid's" or something similar.
I'm not suggesting that there are a bunch of articles online trying to resurrect or vindicate racial typology using this example but it's good to be aware of the dragons on the map.