search this blog

Wednesday, June 3, 2015

Mitochondrial DNA from medieval Bulgaria

Human Biology has a new preprint on ancient mitochondrial DNA (mtDNA) from three medieval Bulgar sites. It includes data for 13 samples belonging to 10 different haplogroups: H, H1, H5, H13, HV1, J, J1, T, T2 and U3.

Abstract: Ancient (proto-) Bulgarians have long been thought to as a Turkic population. However, evidence found in the past three decades show that this is not the case. Until now, this evidence does not include ancient mitochondrial DNA (mtDNA) analysis. In order to fill this void, we have collected human remains from the VIII-X century AD located in three necropolises in Bulgaria: Nojarevo (Silistra region) and Monastery of Mostich (Shumen region), both in Northeast Bulgaria and Tuhovishte (Satovcha region) in Southwest Bulgaria. The phylogenetic analysis of 13 ancient DNA samples (extracted from teeth) identified 12 independent haplotypes, which we further classified into mtDNA haplogroups found in present-day European and Western Eurasian populations. Our results suggest a Western Eurasian matrilineal origin for proto-Bulgarians as well as a genetic similarity between proto- and modern Bulgarians. Our future work will provide additional data which will further clarify proto-Bulgarian origins; thereby adding new clues to current understanding of European genetic evolution.

Nesheva, D V.; Karachanak-Yankova, S; Lari, M; Yordanov, Y; Galabov, A; Caramelli, David; and Toncheva, Draga, Mitochondrial DNA Suggests a Western Eurasian origin for Ancient (Proto-) Bulgarians (2015). Human Biology Open Access Pre-Prints. Paper 69.


Chad Rohlfsen said...

I'm seeing the same thing too. I'm not picking much up in the way of East Asian ancestry in Bulgarians and hardly any in Hungarians. Bulgars and Magyars must not have left much of a legacy, or were extremely West Eurasian like.

Mike Thomas said...

A welcome study, but it makes the false assumption that "Bulgars" (to use propper terminology instead of 'proto-Bulgarians') are somehow 'Asian" or 'eastern'. Clearly, they were west Eurasian, and this fact does not disprove that they weren't Indo-Europeans (most likely Oghuric speakers). Even if Altiac type languages first originated in Mongolia, or what have you, there is no reason to assume that 6th century CE Turkic tribes should be genetically oriental.

It seems there is a lot of political pressure in Bulgaria these days to disprove the Turkic hypothesis, and now theyre making unsubstantiated leaps in logic

Krefter said...

Only the HVR1 region was tested. Based on HV1 the haplogroups are...

U3-no extras. xU3a, U3c.
J-no extras. xJ1b, J1d, J2b, J2a1a. Most likely J1c.
J1b1a1-no extras.
T2-16274. xT2b, T2c1, T2e, T2f, T2a1b.
T-no extras. xT1, T3, T2b, T2c1, T2e, T2f, T2a1b. Most likely T2.
H13a2c-no extras.
H5-no extras.
R-16092-16142-16266-16278. H or H1t1a1 seems most likely.
R-no extras=2. Most likely H.
R-16111-16173-16183-16278. Most likely H.

Here's, Ancient Euro mtDNA.

All those haplogroups first appear in the Neolithic(except for rare exceptions).

U3 is known but rare in the Neolithic, and nowhere else. It's very popular in West Asia. J-no extras(prob. J1c) is the dominant form of J in Neolithic, and rare in the Bronze age Steppe. J1b1a1 is a popular form of J in Bronze age Steppe and rare in the Neolithic. It is a Euro-specific form of J1b.

All the medival Bulgarian T is probably T2. All forms of T2 existed in Bronze age Steppe and Neolithic. Although T2 is more typical in for the Neolithic. H5 existed in Bronze age Steppe and Neolithic.

H13 can't be defined with HV1, but subclades of H13a have been found in Bronze age Steppe and Central Europe with HV1 but haven't in Neolithic Europe.

HV1(all HV so far is HV-16311) hasn't been found in Ancient Europe and is popular in West Asia today.

andrew said...

I guess I missed the memo about ancient Bulgars being Turkic. If I'd gotten it maybe I would have been more surprised.

Krefter said...


I've looked at mtDNA J in Basque, BeloRussians, Polish, Russians, and Norwegians. Almost 100% is: J1c, J1b1a1, J2b, and J2a. Most of their J(more so for Basque, as expected if J1c is more typical of EEF) is J1c.

I've also looked at mtDNA J in Palestine, Jordan, Syria, Saudi Araba, and Yemen. The majority of their J is, J1b(xJ1b1a1) and J1d, both of which are very rare in Europe. The remaining of their J is mostly "J*"(probably J1c), J1b1a1, J2a1a, J2a2b, and J2b, and all are also well-represented in Europe.

Looking at Pre-Historic European mtDNA J(I know I've said this before). All but two of Neolithic Js are J1c(aka "J*"), and those two are typical Euro(rare in West Asia) J2b1a. J is rare in Yamnaya and Catacomb, and most are J2b and J1b1a1. In German Corded Ware, Unetice, and Bell Beaker most are J1c, but some are J2b and J1b1a1.

Mike Thomas said...

Looking at 7th and 8th century burial evidence, Bulgar graves consist of a good 30-40% (!)

So Bulgars must have made a significant impact on moderrn Bulgarians. The simple conclusion is that Bulgars- whatever their supposed language- were nothing other than a west eurasian group. Which makes sense because theyd always lived around the Pontic region, and not China or Mongolia as outdated literature suggested.

These results are not surprising, and are similar to those from ancient Magyars and pechenegs. Many different linguistic groups lived on the Pontic steppe for centuries, and millennia. Not surprising given the region constantly served as a vector for language spread (but not origins).

Hakan Tokaj said...

This is not very compelling. mtDNA only, from sites that are, at the earliest, centuries after the foundation of the first Bulgarian empire -- on similar basis you would "prove" a Mesoamerican origin for "Ancient (Proto-) Mexican Mestizos".

All Turkic peoples analyzed to date, even in the Caucasus and Anatolia or in the Volga plain or forest zone west of the Urals, have an East Eurasian component of ancestry, including the Chuvash, the only extant Oghuric population (so, too, do their Finnic and Ugric neighbors). No, they probably would not have been predominantly East Eurasian by the time of their arrival in SE Europe, but "originally" all indications are that they would have been much more so. Where male-driven expansion processes predominate we cannot hope to see left symmetric impressions in patrilineage and maternal line; for illustration in studies of Volga-Ural Tatars, we see the mtDNA pool is substantially Western Eurasian (84% here:; for other Tatar populations, it seems even more).

"These results are not surprising, and are similar to those from ancient Magyars and pechenegs."

The case of the Magyars gives us additional reason for caution. Considerations:

(1) the exact context of the remains is important, as ancestry could have remained variable by status into this time period:

(2) replacement of a quite differentiated superstrate population is conceivable within historical times. we have not yet any genome-wide ancient Magyar data, but some proxy evidence -- here the primary European lactase persistence allele, at something upwards of 70% in modern Hungary -- suggests they were not interchangeable with modern Hungarians (in line with the latter having relatively little ancestry from the elites who introduced the Hungarian language).

"In ancient Hungarians, the T-13910 allele was present only in 11% of the population, and exclusively in commoners of European mitochondrial haplogroups who may have been of pre-Hungarian indigenous ancestry." (

Anonymous said...

This are not Proto-Bulgars/Steppe Bulgarians and they spent already enough time in Bulgaria to mix heavily with the pre-Bulgarian population. I don't expect steppe Bulgarians to be predominantly east Asian but they certainly not lacked east Asian admixture.

Romulus said...

There is that mtdna H13 again, Yamnaya, Bell Beaker, Minoan, and now Medival Bulgarian. It sure gets around.

Mike Thomas said...

Hakan tokaj

" No, they probably would not have been predominantly East Eurasian by the time of their arrival in SE Europe, but "originally" all indications are that they would have been much more so."

Perhaps; but the pre-proto-Bulgars had already been in Europe (ie pontic steppe) for two centuries before arriving South of danube. In fact we don't exactly know when Oghuric groups first arrived in EE- with the Huns seems probable.

But I agree full genome and Y sequencing is desirable but not always possible

Krefter said...

"Post and Discuss New/Old mtDNA Data"

Anything to do with Modern mtDNA. I'm almost done with studies with data from West Asia and Pakistan, and from Russia and Poland. Everything non-West Eurasian is open-game. There's a lot to discover there.

Alexandros said...

If I remember well, 2 samples from Iron age Bulgaria (Thracians) came out to be U3b and HV, which indicates that the direct maternal ancestors of these Medieval Bulgarians most likely resided in modern day Bulgaria from the Iron age, rather than being a 'newer' addition in the area from the steppe or the Volga region (i.e. Bulgars).

By the way, an off-topic question, what is the reason for not using those Iron age Thracians ( in the several calculators including ancient DNA samples (West Asia K8, Teal K9, etc.)? Is it a matter of too few SNPs, high contamination, something else?

Krefter said...


All we can say is, Iron age and Medieval Bulgarian mtDNA follow the pattern we would expect if there was maternal continuum. The sample set isn't large enough and there is't enough coverage to say there's strong evidence of continuum, IMO.

Mike Thomas said...


"2 samples from Iron age Bulgaria (Thracians) came out to be U3b and HV, which indicates that the direct maternal ancestors of these Medieval Bulgarians most likely resided in modern day Bulgaria from the Iron age, rather than being a 'newer' addition in the area from the steppe or the Volga region (i.e. Bulgars)."

non sequitur, for reasons which should be obvious

andrew said...


H13 indeed does has a very broad distribution, but is not as strongly West Asian specific as your list of examples would suggest. It is found at significant levels in the Levant and Arabia as well.

ur coron said...

@Mike Thomas

Dear Mike Thomas,

At the link i have read a discussion between Prof. Dr Angel S Galabov(author of Proto Bulgarians study) and the associated research site.

If you read the arguments Prof. Dr Angel S Galabov is using, you can see that you were right with the assuming that they have political goals and make unsubstantiated leaps in logic.

Dnagames said...

Great one!

Ж.Войников said...

The study of Desislava Nesheva and co-authors (2015/2016) provides interesting and puzzling results from studied samples. The following mt-DNA hpllogs: H, H1an2, H1r1, H1t1a1, H1a2, H2a2a1, H5, H7f, H13a2c1, H14b1, HV1, J1b1a1, T2, T2g2a (old T3) and U3, U4a2b, U4c1 . They also meet in modern times. Bulgarians.
They have been published in her autoreference to the dissertation.

Ж.Войников said...

D. Nesheva also explores representatives of the Yamnal Culture, who lived in the ІІІ millennium BC, which he mistakenly describes as "Thracians". The results are: H6a1a, H7a1a, N1a1a1a, T2e2a, HV0, HV1a'b'c, J1c (2 samples), K1c1, H (2 samples), U5a1.).
For Thracians in the Balkans can talk only about 1000 BC. BC.