View Full Version : Maternal genetic continuity in northern Mesopotamia since the Neolithic

07-08-17, 15:27
See Reyhan Yaka et al:

"North Mesopotamia has witnessed dramatic political and social change since the Bronze Age, but the impact of these events on its demographic history is little understood. Here we study this question by analysing the recently excavated Late Iron Age settlement of Çemialo Sırtı in Batman, southeast Turkey. Archaeological and/or radiocarbon evidence indicate that the site was inhabited during two main periods: the first half of the 2nd millennium BCE and the first millennium BCE. Çemialo Sırtı reveals nomadic items of the Early Iron Age, as well as items associated with the Late Achaemenid and subsequent Hellenistic Periods. Mitochondrial DNA (mtDNA) haplotypes from 12 Çemialo Sırtı individuals reveal high genetic diversity in this population, conspicuously higher than early Holocene west Eurasian populations, which supports the notion of increasing population admixture in west Eurasia through the Holocene. Still, in its mtDNA composition, Çemialo Sırtı shows highest affinity to Neolithic north Syria and Neolithic Anatolia among ancient populations studied, and to modern-day southwest Asian populations. Population genetic simulations do not reject continuity between Neolithic and Iron Age, nor between Iron Age and present-day populations of the region. Despite the region's complex political history and indication for increased genetic diversity over time, we find no evidence for sharp shifts in north Mesopotamian maternal genetic composition within the last 10,000 years."

Just looking at Table 1 on page 17 of the supplement where they show the percentages between Cemialo Sirti in southeast Turkey and Northern Syria, I'm not seeing it, but maybe it's made clear upon close reading of the whole paper.

Also, these don't seem very resolved.

07-08-17, 15:46
It's nice to have some Mesopotamian mtDNA from the Neolithic and Iron Age. It's unfortunate though that these sequences are only based on the HVRI and HVRII regions. The table on page 17 merely lists the top haplogroups (H, J, K, T...) without any subclade. Unless they managed to get deeper subclades but didn't report them, I don't see how they could come to the conclusion that the maternal haplogroups have remained largely unchanged over the last 10,000 years. Besides the percentages are irrelevant as there are only 10 Neolithic and 12 Iron Age samples! You can't possibly get reliable frequencies from that. These are yet other scientists who jump to conclusions based on patchy data just so that their paper's title sound more interesting. That's unfortunately no way of conducting science.

In fact, when we look at their data, the first thing that is blindingly obvious is the lack of similarity between the Neolithic and the Iron Age sets of haplogroups. The Neolithic samples contain 40% of haplogroups K, which is completely absent from the Iron Age samples. The Neolithic also has L, N* and R0, which are equally absent from the Iron Age. In contrast, the Iron Age has a lot of H (41.7%, against 10% in the Neolithic), one M, two R (other than R0) and has the typically European haplogroup U2. The only sign of continuity is the presence of two HV in each dataset, but they could belong to different subclades with different origins, as there are 17 subclades of HV. Neither set of ancient samples has any haplogroup J or T, which now represent 38% of SE Anatolian lineages, nor U (except one U2), W or X, nor any East Asian mtDNA (A to G), which makes up 7.6% of modern SE Anatolian lineages (obvious Turkic origin as they are absent from modern Syria).

How could they confuse such a strong discontinuity for continuity? I am mystified.

07-08-17, 16:05
I agree with Maciamo. How could they come to those conclusions without deeper subclade analysis? By itself, H is meaningless and it's not even showing continuity on that super broad scale.

I feel like the amateur bloggers could beat this paper with one arm tied behind their backs....