Intriguing, but don't all somalis have minor arabian roots? And isn't the person fully somali genetically? With the high prevalence of T especially in some clans?
Sure, Somalis have some level of limited residue of Arabian-like ancestry. Haplogroups introduced to Somalis do not need to have any autosomal effects.
I hypothesize those most South Arabian elements come from mixing with Cushitic people in Eritrea before we arrived on the northern Somali coast. I hold the belief that non-Agaw, genetically Somali-like groups ( probably Lowland Cushitic) had an independent but synonymous Southern Arabian mixture when those ancient Yemenis arrived in the northern Horn (a separate Arabian group that deeper to northern Eritrea), our mixture representing a distinctly removed lifeway horizon from the highland Agaw. We basically mixed with those and received some of it. To what degree we mixed with those Cushitic lowlanders is hard to say.
I think this is somewhat of an evidence:
That Somali sample came to the Somali region with the E-V32 wave. Notice the Tigray samples that belong to the same tree. Tigres today are basically mixed of various Cushtic groups, and I have said before that much of their sub-structure is definitely East Cushitic. But if you notice the Y-DNA comes from Arabia. So it fits nicely with my hypothesis.
The other factor is that people say it fluctuates from negligible to 12% (I don't remember exactly, since it was years ago since I measured this on G25). Well, on the macro Eurasian scale, it is all even across the board with insignificant differences. This means it was all baked from the beginning during a minor population size founder effect and probably happened while E-V32 migrated south to Eritrea, mixed with a similar tribe that had some new Arabian in them, and then went further to the Somali coast.