Sunday, November 19, 2023

Genomic insights into the Seima-Turbino Phenomenon

Hey there friends,

Some of you may have been wondering where my posts have been all summer. I hope so at least because it has been like nearly six months since I last posted. Basically, I decided it was best for me to have a posting hiatus as I wanted to dedicate my time to a project I had been working on. While I first continued working on some of my research topics for future blog entries while also working on that project, I felt that I wasn’t being productive and movig at a snail’s pace on both fronts, so I decided to pause all that and strictly focus on said project. I'm super pumpd about it, I have been working on it for a year or so. I won’t reveal what the project will be yet, but let’s just sy I’m putting the Scythia back in MusaeumScythia, big time.


So what is up with this post then? Nearly two months ago there were two very interesting preprints released and I cannot pass by an opportunity to rant about the topic they covered. I had been planning to write something about this topic for a long time, waiting for the right time. So I am briefly breaking my hiatus to present you another MuseumScythia classic, reviewing the insights of these two preprints as they relate to the Seima-Turbino phenomenon.

Bronze Age Northern Eurasian Genetics in the Context of Development of Metallurgy and Siberian Ancestry

Ainash Childebayeva, Fabian Fricke, Adam Benjamin Rohrlach, Lei Huang, Stephan Schiffels, Outi Vesakoski, Lena Semerau, Franziska Aron, Vyacheslav Moiseyev, Valery Khartanovich, Igor Kovtun, Johannes Krause, Sergey Kuzminykh, Wolfgang Haak



The Eurasian Bronze Age (BA) has been described as a period of substantial human migrations, the emergence of pastoralism, horse domestication, and development of metallurgy. This study focuses on individuals associated with BA metallurgical production, specifically the Seima-Turbino (ST) phenomenon (∼2,200-1,900 BCE) associated with elaborate metal objects found across Northern Eurasia. The genetic profiles of nine ST-associated individuals vary widely ranging between ancestries maximized in individuals from the Eastern Siberian Late Neolithic/BA, and those of the Western Steppe Middle Late BA. The genetic heterogeneity observed is consistent with the current understanding of the ST metallurgical network as a transcultural phenomenon. The new data also shed light on the temporal and spatial range of an ancient Siberian genetic ancestry component, which is shared across many Uralic-speaking populations, and which we explore further via demographic modeling using additional genome-wide (2 individuals) and whole genome data (5 individuals, including a ∼30x genome) from northwestern Russia

Postglacial genomes from foragers across Northern Eurasia reveal prehistoric mobility associated with the spread of the Uralic and Yeniseian languages

Tian Chen Zeng, Leonid A. Vyazov, Alexander Kim, Pavel Flegontov, Kendra Sirak, Robert Maier, Iosif Lazaridis, Ali Akbari, Michael Frachetti, Aleksey A. Tishkin, Natalia E. Ryabogina, Sergey A. Agapov, Danila S. Agapov, Anatoliy N. Alekseev, Gennady G. Boeskorov, Andrey A. Chizhevsky, Anatoly P. Derevianko, Viktor M. Dyakonov, Dmitry N. Enshin, Alexey V. Fribus, Yaroslav V. Frolov, Sergey P. Grushin, Alexander A. Khokhlov, Egor P. Kitov, Pavel Kosintsev, Igor V. Kovtun, Kirill Yu. Kiryushin, Yurii F. Kiryushin, Nikolai P. Makarov, Viktor V. Morozov, Egor N. Nikolaev, Marina P. Rykun, Tatyana M. Savenkova, Marina V. Shchelchkova, Svetlana N. Skochina, Vladimir Shirokov, Olga S. Sherstobitova, Sergey M. Slepchenko, Konstantin N. Solodnikov, Elena N. Solovyova, Aleksandr D. Stepanov, Aleksei A. Timoshchenko, Aleksandr S. Vdovin, Anton V. Vybornov, Elena V. Balanovska, Stanislav Dryomov,  

Garrett Hellenthal, Kenneth Kidd, Johannes Krause, Elena Starikovskaya, Rem Sukernik, Tatiana Tatarinova, Mark G. Thomas, Maxat Zhabagin, Kim Callan, Olivia Cheronet,  Daniel Fernandes, Denise Keating, Matthew Ferry, Candilio Francesca, Lora Iliev, Kadir Toykan Ozdogan, Kirsten Mandl, Matthew Mah, Adam Micco, Megan Michel, Inigo Olalde, Fatma Zalzala, Swapan Mallick, Nadin Rohland, Ron Pinhasi, Vagheesh Narasimhan, David Reich



The North Eurasian forest and forest-steppe zones have sustained millennia of sociocultural connections among northern peoples. We present genome-wide ancient DNA data for 181 individuals from this region spanning the Mesolithic, Neolithic and Bronze Age. We find that Early to Mid-Holocene hunter-gatherer populations from across the southern forest and forest-steppes of Northern Eurasia can be characterized by a continuous gradient of ancestry that remained stable for millennia, ranging from fully West Eurasian in the Baltic region to fully East Asian in the Transbaikal region. In contrast, cotemporaneous groups in far Northeast Siberia were genetically distinct, retaining high levels of continuity from a population that was the primary source of ancestry for Native Americans. By the mid-Holocene, admixture between this early Northeastern Siberian population and groups from Inland East Asia and the Amur River Basin produced two distinctive populations in eastern Siberia that played an important role in the genetic formation of later people. Ancestry from the first population, Cis-Baikal Late Neolithic-Bronze Age (Cisbaikal_LNBA), is found substantially only among Yeniseian-speaking groups and those known to have admixed with them. Ancestry from the second, Yakutian Late Neolithic-Bronze Age (Yakutia_LNBA), is strongly associated with present-day Uralic speakers. We show how Yakutia_LNBA ancestry spread from an east Siberian origin ~4.5kya, along with subclades of Y-chromosome haplogroup N occurring at high frequencies among present-day Uralic speakers, into Western and Central Siberia in communities associated with Seima-Turbino metallurgy: a suite of advanced bronze casting techniques that spread explosively across an enormous region of Northern Eurasia ~4.0kya. However, the ancestry of the 16 Seima-Turbino-period individuals--the first reported from sites with this metallurgy--was otherwise extraordinarily diverse, with partial descent from Indo-Iranian-speaking pastoralists and multiple hunter-gatherer populations from widely separated regions of Eurasia. Our results provide support for theories suggesting that early Uralic speakers at the beginning of their westward dispersal where involved in the expansion of Seima-Turbino metallurgical traditions, and suggests that both cultural transmission and migration were important in the spread of Seima-Turbino material culture.

If you haven’t already, please read these two articles first before reading further. They cover several topics, both looking into the Uralic languages and the Seima-Turbino phenomenon.  I will be personally focusing on the second part in this entry. The rest of the topics covered in these two articles are interesting as well, but are beyond my scope.

When it comes to blending genetics, linguistics and archaeology I think the second article from Harvard has outdone pretty much every other ancient DNA article there is. That said, it is far from flawless and I am not without critique. There is a lot to get into, and hopefully the authors will take some of the critiques I have and process them.

Seima-Turbino theories

The Seima-Turbino (transcultural) phenomenon is an event in prehistoric Northeastern Europe and Siberia which took place around 2200-1900 BC. Archaeological evidence showcases a sudden distribution of bronze artefacts made with lost-wax casting methods, a technology entirely unknown to the region, which prior to only housed stone age hunter-gatherers or pastoralists with more simpler methods of metallurgy. In addition do we see a collection of archaeological sites clearly related to this spread, the majority of these sites containing cenotaphs.

After 1900 BC the “phenomenon” ceased but the style of metalworking continued to radiate through Eurasia, with metal objects based on this style eventually reaching from Fennoscandia to Southeast Asia[1]. The metallurgical influences of this stylistic form not only reached far, but also had a significant impact on the metallurgical traditions in the western and eastern Eurasian steppes, northern China, and the whole north eurasian belt.

The range and distribution of the Seima-Turbino sites and objects can be seen in the first and second map below:

But perhaps this map makes things a bit clearer:

Over the years many different theories have come about, trying to explain this almost inexplicable event in north eurasian prehistory. Some of these are quite sound while others are far from it. I don’t have time to go over all of these, but the main points of contention generally seem to be around the direction of diffusion and the populations involved.

A common theory, particularly when first proposed, was that the Seima-Turbino phenomenon developed in Eastern Europe and then spread to Siberia. After all, the first discoveries come from Europe as do most Seima-Turbino sites, such as Seima and Turbino which give name to this phenomenon. 

Originating in northeastern Europe, the Seima-Turbino metallurgicak style then spread eastwards. For the proponents of the European homeland of Proto-Uralic this wave seemed like a good explanation for the spread of Proto-Uralic from Eastern Europe towards Siberia. It explains an origin in Eastern Europe, a fitting material distribution in terms of where Uralic languages were spoken historically and a geographic proximity towards Indo-Iranians, explaining the hundreds of loanwords coming from this language group into Uralic.

Even in this day and age you will find articles arguing for a European homeland, such as Asko Parpola’s article from last year arguing for an European origin of the Seima-Turbino phenomenon [2]. I won’t comment on the linguistic homeland debacle but Parpola pulls off a serious case of “bad archaeology” to give archaeological credence to the European origin of the Seima-Turbino phenomenon.

As you might have guessed from my tone, I’m more in favour of a homeland in Siberia. Ultimately this fits the best with the C14 dates and the typology of the objects that point towards an eastern origin of this phenomenon [3]. 

Evgeny Chernykh and Sergey Kuzminykh are amongst the archaeologists which argue that the origin of the Seima-Turbino phenomenon lies further to the east, and have done it the best out of all of them I think. According to them the origin of the Seima-Turbino phenomenon is found in the region where the Altai and Tian Shan intersect, bridging Central Asia and Siberia. 

Beyond this they also seem to argue that at the core of this phenomenon in their view was a group from the Altai/Dzungaria region that were metalworking pastoralist peoples. They were the source of this new metalworking tradition and spread across Siberia into Europe [4]. The authors dubbed them 'The forerunners of Genghis Khan", referring to their expansion from Mongolia towards the west.

It seems indisputable that armed riders were the nucleus of the “Seima-Turbino” nomadic cultures. Both their numerous weapons and the micro-sculptures with which they adorned their princely daggers and knives prove this. It is hard to say if their families and herds accompanied them on their long migration, as was often the case in invasions by later nomadic warriors, but it is doubtful that these groups were numerous. Most likely, the comparatively small amount of metal from the complexes across theirhuge arching territory, which marked their rapid move to the west, also indicates the comparatively small number of warriors who came such great distances from the east. Generally their aggression was directed toward the west. The routes taken by these

warriors—following the sun—was most likely determined by courses of the numerous rivers of the Siberian plain, among which they chose the largest for navigation. Initially focused within the Ob and, perhaps more significantly, within the Irtysh Basins, their routes lead them west. After crossing the Lower Ural River, the Seima-Turbino Transcultural Phenomenon spread into the vast Volga-Kama Basin, perhaps via the Chusovaya Basin, and then out to the Oka River. They were traveling not only in the summer but probably also during the winter. Certainly, it is clear from their material culture (see fig. 15.1 and earlier discussion of this remarkable artifact) that they were familiar with skis, and the warriors clearly adapted their small but robust horses for travel across deep snow. Did they also use boats? It seems likely, or at least possible, given their association with rivers, though they left no direct evidence for us to consider. Studying the map detailing their distribution (fig. 15.1) gives rise to another mysterious question. Why did these native peoples of the mountain-steppe of Dzungaria and Mongolia suddenly set off into the forests of the Ural, Volga, and Kama Basins? What pushed them to change the natural “habitat,” which was presumably so familiar to them? Only one thing seems certain—the decision was forced upon them.

I also have to give Anthony’s description of the Seima-Turbino phenomenon in his book the horse, wheel and language credit because I find it quite sober and accurate as well:

Beyond the western Altai/middle Irtysh core area the Seima-Turbino horizon was not a culture. It did not have a standard ceramic type, settlement type, or even a standard mortuary rite. Rather, Seima-Turbino metal-working techniques were adopted by emerging elites across the southern Siberian forest-steppe zone, perhaps in reaction to and competing with the Sintashta and Petrovka elites in the northern steppes. A The Eurasian Steppes 44J series oforiginal and distinctive new metal types quickly diffused through the forest-steppe zone from the east to the west, appearing in late Abashevo and Chirkovskaya cemeteries west of the Urals almost at the same time that they first appeared east of the Urals, beginning about 1900 BCE. The rapidity and reach of this phenomenon in the forest zone is surprising. The new metal styles probably spread more by emulation than by migration, along with fast-moving political changes in the structure of power.

My own views

As I mentioned earlier I personally side with the eastern hypothesis, and more or less agree with the placement of the nucleus of this development of metalworking in the Altai regions of Mongolia and Dzungaria as proposed by Chernykh and Kuzminykh.

One reason is the quite convincing argument based on faunal depictions. Animals such as ibexes,wild horses, tigers or snow leopards, all native to the Tian Shan and Altai mountains and not present in more western regions of SIberia or Europe are regularly depicted in Seima-Turbino artefacts.

The primary reason which convinces me is that Seima-Turbino technology ultimately is based on lost-wax casting methods. Lost-wax casting also was not independently developed, as you see a sudden appearance and spread of this form of casting rather than a gradual development over the course of centuries.

In the Eastern European steppes lost-wax casting (or any form of casting) was not practised, but it was in southern central asia since the 4th millennium bc [5]. This technological impulse had to spread into northern populations slightly before 2200 BC, which eliminates Abashevo/Sintashta populations as a mediator of this impulse. It slightly predates the BMAC as well, meaning that Anthony’s scenario as described here probably doesn’t cut it either:

“Lost-wax and hollow-mold casting methods probably were learned from the BMAC civilization, the only reasonably nearby source (perhaps through a skilled captive?).”

So what can? Some archaeologists such as Lonnett and Frachetti suggest that Afanasievo people migrated into southern Central Asia along the Inner Asian mountain corridor, but David Anthony disputes the suggested evidence as being a sign of Afanasievo contacts, and suggests it represents contact with the indigenous populations of Central Asia, and I agree with him. 

These contacts must have arose somehow because later on in the succeeding Chemurchek culture do we see genomic evidence that these contacts occurred, as several of the Chemurchek samples from Xinjiang and Mongolia carry southern Central Asian ancestry.

Unfortunately the archaeology of the Chemurchek culture is rather poor. We only have burials to go off for the most part and not a great deal has been uncovered. But this limited information has already given us important clues as Chemurchek burials dated to the third millennium B.C show casting moulds [6], which were not seen in these regions prior. These are some casting moulds for what seem to be daggers and adzes from Chemurchek-1:

Unlike Chernykh and Kuzminykh I do not believe there was movement of pastoral tribesmen from the Altai across a vast swathe of Siberian and European landmasses, leaving behind a trail of monument and sanctuary sites.It is an illustrious story but I think these two archaeologists almost portray it as if this is the only solution to this archeological mystery, and I disagree with that.

I think that for the most part a lot of the spread of these artefacts are simply trade and local emulation as Anthony described it. You have Seima-Turbino goods as far flung as Borodino on the Moldovan/Ukrainian border. Those artefacts did not get there because some Western Siberian warrior buried his goods on the western black sea coast obviously. 

It does get a lot more interesting when we are talking about the main Seima-Turbino sites though. What exactly happened here is still a mystery. What we generally see is that around these Seima-Turbino sites you have a continuation of material culture, before and after the cemeteries you more or less have the same peoples (during the bronze age this shifts of course). It is likely indicative that if these sites were built by eastern migrants, these groups likely wouldn’t have been large as they did not leave an archaeological imprint. They either left, fizzled out, got assimilated into the existing peoples, or were killed off. Can’t say really.

One thing seems certain to me though, whatever spread in Europe brought a Western Siberian flavour with them. The anthropomorphic statues and masks we see at Galych or Pyshma show a clear Western Siberian/Central Asian cultural heritage, with parallels found in various ancient Siberian and Central Asian and modern Siberian, Central Asian and even South Asian peoples.

We know that Indo-Iranians were interacting with these Siberian populations through archaeology and through a heavy dose of linguistic influence in Uralic languages, followed by a massive spread of Indo-Iranian populations across southern Siberia so I expect them to be represented quite decently in this phenomenon as well.

Another theory that never has convinced me is the one where the spread of this network is imaged to be the result of Uralic tribes expanding across northern Eurasia, leaving behind a sudden but vast archaeological trail at fairly northern latitudes, as proposed by some scholars and increasingly more popular on the internet. 

To clarify, this does not mean that I think the Uralic ethnogenesis and dispersion is entirely unrelated to this prehistoric event, but I do not believe that the distribution of Seima-Turbino sites is an archaeological representative of the spread of Uralic peoples across Siberia and into Europe, nor do I believe that they were the core population of the Seima-Turbino phenomenon. I do believe that the existing trade networks established during this period between Europe and Siberia could have been a driver for the migration route that Uralic people took, while also being a driver for Indo-Iranian migrations to the east.

So to summarise, my position for the last few years has consistently been:

  • The initial impulse of Seima-Turbino came from the Altai region by way of technological influence from southern central asia

  • The spread of this metallurgy tradition was likely not accompanied by a cohort of pastoral Altai tribesmen migrating into Siberia and spreading westwards, although the presence wouldn’t be impossible and would explain some linguistic theories

  • This material influence spread amongst the proximate Siberian populations, some still hunter-fishers whereas some were becoming pastoral, where we see the first Seima-Turbino epicentres develop.

  • Indo-Iranians were about to go on a massive expansion through Siberia and surely would have gotten wind of this and likely would be trading and competing with the tribes in Siberia, meaning pioneering groups definitely went into this direction.

  • All this development would have triggered hunter-gatherer populations from the east to migrate westwards to also participate in these networks. The Siberian ancestors of Uralic would be one of such eastern Siberian peoples, with this time period being their first interaction with Indo-Iranian speaking peoples (leaving an early layer of Proto-Indo-Iranian loanwords), likely deep in Siberia.

  • The appearance of the Seima-Turbino monument sites in Europe are not the result of Uralic tribes having a massive expansion from Siberia. Whoever built them likely left no lasting results in the region

Finally got that out of the way. Now that we have ancient DNA from Seima-Turbino sites and contemporary sites in Siberia, we can to a degree figure out how these hypotheses match up with the ancient DNA record.

DNA evidence

The combined data of these two articles provided us with over a dozen samples from two Seima-Turbino sites, Rostovka and Satyga-16. Childebayeva’s article contained six samples from Rostovka, and Zeng’s article contained 5 samples from Rostovka and two from Satyga-16. Since the samples were sequenced separately by different laboratories we actually have some overlap in published samples. 

The samples from other sites such as Tatarka Hill or Chernoozero-1 do not come from Seima-Turbino sites of cemeteries, despite the Zeng’s team kind of making it appear like they are sometimes, thus I won’t be discussing these samples here.

Before going into the sites and samples, I quickly want to point something out about the models used and how they should be interpreted. In Childebayeva's article, the Rostovka samples were modelled with three sources: Steppe_MLBA, WSHG and Yakutia_LNBA. Neolithic West-Siberians from the Transurals, Indo-Iranians and eastern Siberians beyond lake Baikal. This is a pretty good combination which got decent results as most samples got passing qpadm values.

But this should not be taken too literally in all cases. The genetics of northern Central Asia and Siberia were intertwined, with various populations genetically contributing to one another. This means that “WSHG” alone is not going to capture all the genomic variety, and this will naturally shift some ancestry to steppe_MLBA and Yakutia_LNBA populations.

It is important to keep in mind that Steppe_MLBA and Yakutia_LNBA populations were very new to this area and couldn’t have been there longer than a few generations, thus small frequencies of either streams are unlikely to be genuine ancestry from these populations.

The modelling of the samples in Zeng’s article is a bit more complex so this becomes less of an issue. They also use a combination of admixture and qpadm which I think was quite nice. The only issue is that some models seem a bit messy in terms of population choice in my opinion.


The Rostovka cemetery is one of the main Seima-Turbino sites, considered to be somewhat of the “centre” of the Seima-Turbino phenomenon by proponents who adhere to an east-west direction for the spread.  Rostovka is known for some of the more interesting finds, such as the blades decorated with what seemingly seems to be a man on skis being pulled by horses.

The Rostovka cemetery is also a special site because it contains a significant number of burials. A cemetery with burials, not that strange right? Yes but it just so happens to be that most Seima-Turbino sites in Europe lack burials but contain cenotaphs, burial monuments without remains. Most of the graves at Rostovka seemed to have been reopened with the remains purposely damaged, and I wonder if this is connected to the cenotaphs somehow.

Given the various theories regarding the origin of the Seima-Turbino, the Rostovka cemetery is a key site to investigate with the help of genome sequencing. If the spread of the Seima-Turbino phenomenon was accompanied by a movement of pastoralists from the direction of the Altai and Dzungaria, this should be reflected in Rostovka as it is a pretty eastern cemetery and one of the oldest ones. The people buried here cannot be a few generations removed from this original exodus event thus the DNA imprint should still be visible somewhat even after generations of mixing.

Similarly if the Seima-Turbino phenomenon over time came to include different peoples but was significantly driven by the expansion of Uralic populations, we should see this uralic core represented as well, either in terms of the demographics at Rostovka or in noteworthy graves at this site.

A brief overview of the genomic ancestry of the samples can be seen in the various infographic from the two articles below:

What I will do now is group these various samples into three clusters, go over their data and burial information and provide more context about what we got from Rostovka, one sample at the time.

Steppe_MLBA cluster

This one is simple and just constitutes the batch of samples who have steppe_mlba populations as their core genetic component. These were Proto-Indo-Iranian speaking peoples that were radiating out of their Eastern European homeland into Siberia and Central Asia, about to go on a massive expansion.

Krasnoyarsk_BA cluster

This one is also pretty straightforward. Remember Kra001/Krasnoyarsk_BA, our homie from Neftoprovod which caused a bit of upheaval because he seemed to be the closest related sample to the direct paternal ancestors of European Uralic speakers we got, found deep in Siberia? This cluster is named after him and refers to his genetic component. The samples with this component at its core will be under this cluster.

Western Siberian cluster

I couldn’t think of a better name because I think it is a bit ill-fitting since there are various opinions on what constitutes Western Siberia. Just draw a very big circle around Rostovka that still is small enough to not include Europe or regions beyond lake Baikal, and then just imagine all the populations that lived there. Technically this would also include parts of eastern Siberia in this definition, and parts of Central Asia in everyone's definition.

The reason for this is because in Siberia and Central Asia there was continuous gene flow between populations from all these areas, so they cannot be detached like the following components. I also have a feeling that the Seima-Turbino phenomenon would have triggered demographic expanses of certain populations due to changes in subsistence economies as well as more long-distance mobility, shifting the genetic cline which had formed during the neolithic and eneolithic.

Steppe_MLBA cluster


M/F: Male

Y-DNA: R-M417 (R1a)

Mtdna: R1a

Date: 4150-3800 BP (stratigraphic context)

Burial description:

Grave 8, in which RO003 was buried, is particularly rich. It contains two lanceheads of the types KD-10 and KD-14, a socket ax of type K-20, a dagger blade of the type NK-6, several small flint arrowheads, two rectangular flint blades, and a bone handle.

Autosomal profile:

ROT003 was modelled as fully Sintashta-like.

I find it quite interesting that this individual had a quite rich burial at Rostovka despite being of Indo-Iranian origin and probably a migrant to the region. This is quite significant as it shows that Indo-Iranians not only had a presence, but achieved high status within this network.



Y-DNA: R-M417 (R1a)

MTDNA: U5a1+@16192

Date: 2137-1919 BC

Burial description:

“The inventory of grave 33, in which ROT016 was buried, contains a lancehead of type KD-24, two gold rings, several small flint arrowheads, two stone objects, which are usually described as arrow-smoothers, and a big number of perforated bone battens. These battens are interpreted as the remains of a piece of body-armor.”

Autosomal ancestry:

ROT016 was modelled as having 51.1% Sintashta_MLBA Ancestry, 33% Yakutia_Lena_4780_2490 ancestry and 15.9% WSHG ancestry.

Interestingly this was one of the burials at Rostovka which displayed remnants of bone armour in his grave. Considering the lancehead, I wonder if this means that this depiction of a Rostovka warrior  was inspired by ROT016:


In that case I guess it is a bit of a plot twist that this individual was half Indo-Iranian with R1a-M417 (likely Z93). There is more going on however because in Zeng’s article another sample was attributed to burial 33, I32816, and this sample was completely different from a genetic perspective.


M/F: Male

Y-DNA: - (not given)

Mtdna: R1a

Date: 1926-1754 BC

Burial description

Burial 7 (individual ID I33398) was located at a depth of 40 cm from the modern surface. It was

observed as a rectangular patch of gray earth filled with charcoal and burnt soil. The dimensions of the burial were measured to be 70x220 cm, with the west-northwest–south-southeast orientation. During the excavation, a strip of charcoal was uncovered at the level of the initial filling, running from the northwest to the southeast. Additionally, pieces of charcoal were found in separate clusters,approximately 20 cm above the bottom of the grave, suggesting the presence of remnants of a burnt wooden ceiling or lining. Among the findings within the burial, on the first layer, a roughly worked stone artifact, a fragment of a deer antler possibly used as a handle, and an ellipsoid-shaped bone object with holes at the ends were discovered. The skeletal remains of an adult male, estimated to be around 40 years old, were found in a bent position on the left side with the feet pointing to the west. However, the preservation of the skeleton is poor, with only the left and right femur and tibia intact. Disorganized ribs and a humerus were located beneath the left thigh, while the lower jaw was found under the right ulna. The cluster of finds near burial 7 included an asymmetric quartzite knife, several fragments of bone armor, a fragment of a stone object, and a small stone.

Autosomal ancestry

I33398. We find that this individual can be modeled as an unadmixed representative of a Steppe_MLBA-related source (100% Russia_Srubnaya, p=0.053), in keeping with his behavior in PCA and ADMIXTURE (Main Text Fig. 1 center, 4B; Extended Data Fig. 4, 5). 

At first I assumed this was a duplicate of ROT003, but I3398 and ROT003 were attributed to different graves, burial 7 and 8 respectively. Maybe this is an error and these are the same individual after all, since they are both unmixed steppe_mlba samples with the same mtdna clade.


M/F: Female

Y-DNA: -

Mtdna: U5b1b2


Burial description:

Burial 34 (individual ID I32898) was situated at a depth of 65 cm from modern surface. Prior to its discovery, an indistinct spot was identified at the lower horizon of the chernozem. Within this spot, a flake and two stone arrowheads with straight bases were found. The dimensions of the grave were 90x280 cm, and it was oriented from southeast to east-northwest. Inside the grave, the well-preserved remains of a man, aged between 25 and 30 years, were uncovered. The individual was lying on his back in an extended position with his head facing south-southeast. Adjacent to the right elbow of the deceased, a bronze celt was discovered, pointing towards the eastern wall. The celt's sleeve contained remnants of a wooden handle. Beneath the celt, a substantial bronze spear with a forked rod and a hook below a feather was found. Golden earrings in the form of rings with open edges were situated near the temples of the skull. On the pelvic bones, there was a lamellar bronze knife with traces of a wooden scabbard. Adjacent to the pelvic bones on the left side, a stone arrowhead with a straight base was positioned, while two similar points were located near the left knee joint. A bone puncture was observed near the left hand. On the left side of the skull, a bronze point-awl with a bone handle and the remains of a bronze artifact in the form of an awl inserted at an angle into the bone handle were discovered. In the southern corner of the grave, to the left of the skull, a large fragment of ceramics was recovered from the filling material of the grave, near the level of the frontal part of the skull.

Autosomal profile:

In the first set of qpAdm, a set of 3-way admixture models pass (p>0.15), with sources drawn from the FSHG cline contributing ~10%, Yakutia_LNBA ancestry sources contributing ~20%, and Steppe_MLBA ancestry contributing ~70% of ancestry. A similar set of models pass in the second set of qpAdms, with ~10% ancestry from the FSHG cline, East Asian ancestry from Russia_Tatarka_BA (~20%), with the remainder coming from a Steppe_MLBA-related source (~70%; p>0.14).]

It is odd how the burial description mentions a single burial of a male, but this sample belonged to a female sample. ROT017 in the first article is also said to have come from burial 34, and perhaps this description would be of his rather than this female sample? In the archaeological notes of I32898 there aren’t any mentions of several individuals buried in grave 34.

Beyond these samples there are also several samples in the Western Siberian cluster which carrysubstantial Steppe_MLBA cluster. These are:

  • ROT004

  • I32369

  • ROT011

Krasnoyarsk_BA cluster


M/F: Male

Y-DNA: N-L392

Mtdna: G2a1

Date:1938-1700 calBC

ROT002 XY 211,602 - N1a1a1a1a L392 G2a1 1938-1700 calBC (± 2σ)* 

Burial description:
The individual ROT002 was buried together with a dagger blade of the type NK-4, several nonspecific flint-tools, a grindstone, two ceramic vessels and a bone-arrowhead, similar to arrowheads, which were found among the ST-materials of the Kaminskaya Cave on the western site of the Urals2.

Autosomal profile:

ROT002 was modeled as 87.4 - 92.3% Yakutia_Lena_4780_2490 ancestry, with the remainder either being steppe_MLBA, EHG or WSHG depending on the model.

ROT002 is the same individual as I32545.


Burial description:

Burial 5 (individual ID I32545) was situated at a depth of 30 cm from the modern surface. It appeared as a rectangular patch of gray earth filled with charcoal and burnt soil, with dimensions measuring 65x150 cm. The orientation of the burial was east-northeast–west-southwest. The skeleton of a presumed male, approximately 25 years old, was found lying on his back with the head facing east. The preserved bones include the ulna and humerus of the arms, the femur and lower leg of the left leg, as well as the pelvic bones, although they were displaced. The skull, chest bones, spine, and phalanges were in a very poor state of preservation. Within the grave, several artifacts were discovered, including a bronze socketed spearhead, four jasper flakes, and stone arrowheads. The spearhead tip was likely not placed within the grave but left after most of the filling took place, and it was found stuck into the wall of the grave.

Autosomal profile:

I32545. We find that this individual can be modeled with near-unadmixed ancestry from Yakutia_LNBA (~97% Yakutia_LNBA) at the p>0.01 level; this is improved with the addition of Russia_Tatarka_BA in the second qpAdm setup, even with Yakutia_LNBA moved to the references (~97% Russia_Tatarka_BA, p=0.06). All passing models for this individual include a Yakutia_LNBArelated source.

In Zeng’s article they also mention this about the sample:

ROT002, the individual with the highest proportion of north Siberian ancestry, was assigned to 399 the N1a1a1a1a (N-L392) haplogroup. This Y-hg has also been found in two published BOO individuals . The lineage N-L392 is one of the most common in present-day Uralic populations, and was estimated to have split from other N-lineages around 4,995 years ago.This finding further highlights the importance of Y-hg N-L392 as being linked to the dissemination of proto-Uralic, but potentially also involving the ST network. 

This suggests that this individual was not just related to Uralic people in general, but seems a close relative of the Siberian populations which migrated into Europe and became ancestral to European Uralic speakers.

Childebayeva’s article makes it a point of note that the burial of ROT002/I32545 is rather typical and a bit sparse:

The rather sparse burial of ROT002 does not stand out 465 typologically in contrast to most other burials of the cemetery, despite the genetic attribution to the Eastern Siberian LNBA .

In addition to ROT002/I32545 there also were these samples that were modelled with a considerable Kra001 component, but ultimately are mixed. As I mentioned earlier, it is hard to quantify if this ancestry actually came from those components or if it represents another stream of east asian ancestry. I32898 seems to be modelled pretty securely with about 20% Krasnoyarsk_BA type ancestry.

Western Siberian cluster


M/F: Male

Y-DNA: Q-M346 (Q1b)

Mtdna: H1

Date: 2202-1983 BC

Burial description:

ROT004 was buried together only with a little asymmetrical fint-blade and a piece of bone, which could be a fragment of a body-armor, like in grave 33. 

Autosomal profile:

ROT004 was modeled as 34% Sintashta, 56.3% WSHG and 9.7% Yakutia_Lena_4780-2490.

Based on the autosomal ancestry, haplogroups and burial goods, I think this might be a duplicate of sample I32369. I am not sure though because it isn’t clear from the archaeological description.



Y-DNA: R-M73 (R1b)

Mtdna: A10

Date: 4150-3800 BP (stratigraphic context)

Burial description:

The grave of ROT006 contains only an asymmetrical flint-blade and a very little metal awl.

Autosomal profile:

ROT006 was modelled as 88.3% WSHG, 0% Sintashta_MLBA and 11.7% Yakutia_Lena_4780-2490.


M/F: Male

Y-DNA: C-L1373

Mtdna: C4

Date: 2054-1174 calBC (± 2σ)* 

Burial description:
The grave goods, which were found together with individual ROT011 in grave 23 consist of a dagger blade of type NK-6, a rectangular flint blade, a piece of copper ore and a nonspecific polished stone tool.

Autosomal profile:

ROT011 was modelled as 39.5% Sintashta, 39.4% WSHG and 21.1% Yakutia_Lena_4780_2490.

So a few interesting things about ROT011. First off, the sample has nearly 40% sintashta ancestry yet neither his Y-chromosome or Mtdna clades come from Sintashta-related peoples. C-L1373 is a clade of East Asian origins and likely comes from the Western Siberian side, but the mtdna C4 is ultimately of ANE origin and was present in both West and East Siberian populations.

What seems most probable to me is that one parent was around 75% steppe_MLBA,  and the other parent had no steppe_MLBA ancestry. The first parent would probably be its mother and the second one his father. The lower C14 bounds for this sample are also really low at 1174 BC, nearly 1000 years after the Rostovka cemetery.



Y-DNA: -


Date: 4150-3800 BP (stratigraphic context)

Burial description:

“The grave of ROT013 contains no finds at all”

Autosomal ancestry:

In the first model, ROT013 was modelled as 11.1% Sintashta_MLBA, 79.2% WSHG and 9.7% Yakutia_Lena_4780_2490 . In the second model with 7% Yakutia_Lena_4780_2490 and 93% WSHG ancestry.


M/F: Male

Y-DNA: C-F9992


Date: 2133-1919 BC

Burial description:

“The grave of ROT013 contains no finds at all”

Autosomal ancestry:

In the first model, ROT015 was modelled as 5.5% Sintashta_MLBA Ancestry, 62.8% WSHG ancestry and 31.7% Yakutia_Lena_4780_2490 ancestry. In the second model ROT015 was modelled as 30.2% Yakutia_Lena_4780_2490 and 69.8% WSHG ancestry.

Another interesting point was this:

Based on these analyses, we identified a pair of second-degree relatives (ROT011 and ROT015), both of whom are males carrying the Y-haplogroup C2a, and could either represent a grandson/grandparent, a nephew/uncle pair or paternal half-siblings, consistent with overlapping radiocarbon dates for both individuals (Table 1).

Given the close genomic and paternal suggestion with the lack of significant steppe_mlba input, this supports my suggestion for the genetic profile of ROT011 earlier. Someone with a profile like his mixed with someone with ~75% steppe_mlba ancestry forming a profile resembling ROT11, which could explain the lack of steppe_mlba haplogroups despite the substantial contribution.


M/F: Male
Y-DNA: Q-L53
MtDNA: H101
Age; 3150-3800 BP (stratigraphic context)

Burial description:
ROT017 was buried in grave 34 with an inventory of a dagger blade of type NK-4, a socket ax of type 20, a lancehead of type KD-10, two gold rings, a big ceramic fragment, several small flint blades, and a little bone awl.

Autosomal ancestry:

ROT017 was modeled as 51.2% Yakutia_Lena_4780_2490, 17.5% Sintashta_mlba and 31.4% WSHG.

So interestingly, the autosomal profile of this individual seemed to contain around 50% Eastern Siberian ancestry. I initially wanted to place this sample in the Krasnoyarsk_BA cluster but neither the paternal or maternal haplogroups seem directly connected to the Kra001-related populations, making it hard to say from which Siberian populations this eastern ancestry comes from. On the paternal side the sample had Q-L53, which could be an indication it could be partially derived from ANE/East Asian mixed populations. 

Another point I wonder about is that perhaps the burial description given for sample 132898 is actually of ROT017 since they come from the same burial and ROT017’s gender matches the archeological description of a male.



Y-DNA: Q-L940 (Q1b)

Mtdna: H1b

Date: 2300-1700 BC (stratigraphic context)

Burial description:

On top of the second skeleton in the southwestern portion of the grave, a third skeleton, individual 3m (individual ID I32369) was positioned, likely that of a male individual. The skull of this individual was placed above the left half of the second skeleton's pelvis, with the back of the head facing upwards. Adjacent to the skull, the handle of a bone knife was preserved. The pelvic bones (two of them) from this third skeleton were discovered in the eastern half of the grave.

Autosomal profile:

I32369. This individual can be modeled as a 3-way admixture with major fractions of ancestry from a Steppe_MLBA-related source (~30-44%) and FSHG ancestry (~51-61%) and minor contributions from an East-Asian related source, in keeping with his behavior in PCA and ADMIXTURE (Main Text Fig. 1 center, 4B; Extended Data Fig. 4, 5). 

This profile is possibly the result of a predominantly indo-iranian mother and a Western Siberian father. Q-L940 is an early branch of Q which has subclades in both Europe and Siberia. H1b though is a clear west asian mtdna line and comes from his european side, making it likely this is a western siberian Q1b clade.

As I mentioned earlier in regards to ROT004, between these two samples there is a match in paternal haplogroups, maternal haplogroups and autosomal profile, both largely descending from a west-Siberian population with 30-44% contribution from a Steppe_MLBA related person. If these turn out to be two different people, then they must be siblings because this is too close of a match.


M/F: Male
Y-DNA: C-Y11990
MtDNA: C1e
Age: 2300-1700 BC (stratigraphic context)

Burial description:

Burial 33 (individual ID I32816) Burial 33 (individual ID I32816) was located at a depth of 25 cm from modern surface. It was identified by a patch of dark coaly earth with an indefinite shape. The dimensions of the grave were measured as 25x200 cm. At a distance of 30 cm from the southeastern wall, a forked spear was discovered, with its tip inserted into the ground. Within the southeastern half of the grave, a significant accumulation of bone armor plates was found. These plates were in a disordered pile, and among them, a portion of a woman's skull, estimated to be aged 20-25 years, was also discovered. The lower layer of plates rested at the bottom of the grave, extending in some cases towards the bottom or the wall. In the northwestern half of the grave, tibia bones and a tibia were found. Based on their position, it can be inferred that the skeleton was lying on its back in an extended position, with the head facing southeast. Among the armor, a golden ring was found closer to the south-eastern wall, while a second golden ring of larger diameter was found closer to the center of the grave. Two abrasives were located in the southern corner of the grave. Within the armor, a fragment of an unfinished bone puncture was also discovered. In the center of the grave, a stone arrowhead with a straight base was present. The filling material of the grave yielded additional artifacts, including the second half of a bone puncture, another stone arrowhead with a straight base, and a bone knife handle with a hole for the blade. The handle was adorned with a delicate pink geometric ornament. The collection of findings within grave 33 also included a flint flake, fragments of ceramics, a stone flake, two more ceramic fragments, fragments of animal bones, another stone flake, additional ceramic fragments, an accumulation of ceramic fragments, and further fragments of animal bones.

Autosomal ancestry

I32816. A 3-way model passes for I32816 in the first set of qpAdm (~54% Altai_N + ~30% EHG + ~15% Yakutia_LNBA; p=0.09); this model is congruent with the behavior of this individual in 233 ADMIXTURE (Fig. 4B). A similar set of models pass in the proximal qpAdms, with multiple proximal FSHG sources equally sufficing for all of this individual’s FSHG ancestry. However, as ADMIXTURE also shows that this individual possesses significant Cisbaikal_LNBA ancestry, we repeated the proximal qpAdm with Cisbaikal_LNBA added to the references, rotated into the sources. We find a wide range of passing 3-way and 4-way models for I32816, but now all passing models involve Cisbaikal_LNBA ancestry rotated into the sources at ~15-20%.

Now an interesting thing here is that in the second article, the sample from grave 33 is completely different from ROT016 in the first one. It is unclear from the burial descriptions provided in both articles if burial 33 had several males buried in it, or just one. It might be worthwhile to mention that I32816 was only period dated, between 2300-1700 BC, whereas ROT016 has radiocarbon dating.

Additionally, a material counterpart to the genetic link between some individuals from ST necropolises and populations of Northeast Siberia can be found in suits of armor made of bone plates, which have been found from the Glazkovo and especially the Ymyyakhtakh cultures. One set was buried with a male of the Yakutia_LNBA population mentioned in this study: N4a1.SG from the Kyordyughen site 83; others hail from the Upper Yenisei region where Kra001.SG and the Tatarka individuals were situated 83, and three were buried in Rostovka, one with an admixed male reported in our study (I32816 from Grave 33 61) that bore both Yakutia_LNBA and Cisbaikal_LNBA ancestries.

I found the bolded part in this section a bit shoehorned in. It seems like the authors want to make a genomic connection to the bone armours, of northeast Siberian origin, and the genetic profile of this individual. But this individual was mostly of western siberian origin, and without double checking the model itself it is possible this individual is entirely of local western siberian eneolithic stock, given how it simultaneously needs more western and more eastern streams of ancestry relative to altai_N.  One thing that is odd is that in the section about genetics they never discuss the sample having both ancestries, just that they modelled it with it both in qpadm and that Cisbaikal_LNBA shows up in ADMIXTURE. 

In any case it seems highly unlikely to me that this individual was wearing bone armour because of an East-Siberian great-grandparent or distant ancestry from such a population, and I don’t really see why this was brought up. If the individual  was I32545 I would have understood it as it was a recent migrant from northeastern origins.


M/F: Male
Y-DNA: C-Y11990
MtDNA: C1e
Age: 2300-1700 BC (stratigraphic context)

Burial description:

Individual 4 (individual ID I32899) exhibited a specific arrangement of remains. Above the leg bones of the second buried individual, situated in the western half of the grave, lay the skeletal remains of a fourth individual, likely a woman aged 40-45 years. The preserved elements from this individual included the skull with the face upward, the left tibia, the right half of the pelvis, the right ulna, and the left clavicle. Following the removal of all the skeletons, a patch of mixed earth was discovered at the bottom of the grave in its eastern half. This patch was deepened by 15 cm from the bottom of the grave. Within the deepest section of this patch, three bronze objects were found. The first was a forked tip of a stake, above which was another similar tip with a spike. On top of these tips lay a bronze one-eared celt adorned with a ladder ornament and hatched triangles, measuring 14 cm in length. Notably, all these bronze items were positioned with their blades facing west. The accumulation of finds associated with burial 8 included the following items: eight stone flakes, a bronze socketed spearhead featuring a long open socket deeply embedded, with its point facing downward into the mainland, a fragment of a spearhead crafted from shale, a chert plate resembling a knife, a blank for a knife made of diorite, and a pot.

Autosomal profile:

I32899. No models pass for this individual in the first set of qpAdms (highest p=0.007), but when populations of the Eneoltihic Trans-Ural steppe are included, a model with 77% Russia_Ural_Kulmetovo_C_6.0-6.1kya + 23% EHG (p = 0.25) passes. 


M/F: Male
Y-DNA: C-F1699
Age: 2300-1700 BC (stratigraphic context)

Archaeological description:

Burial 24 (individual ID I25555) was situated at a depth of 15 cm from modern surface and appeared as a dark patch of mixed earth, measuring 70x200 cm. Pieces of burnt soil were present within the grave's fill. The remains of the deceased were discovered in the grave, albeit in a poorly preserved state. The western corner contained fragments of the thighs, pelvis, tibia, and right hand. In the eastern half of the grave, the lower jaw of a woman, approximately 40 years old, was found. The inventory associated with burial 24 comprised several items. These included a stone arrowhead with a straight base, a fragment of a clay mold used for casting a spearhead, a bronze chisel, and a bronze plate knife. Furthermore, the accumulation of finds within the grave consisted of a flake of clay slate, a bronze forked tip of a stake featuring a spike at the base of the feather. The sleeve of the stake was equipped with an eyelet and four convex rollers. Additionally, a badly corroded bronze two-eared celt was present, featuring a round sleeve and a slightly convex blade without ornamentation. Moreover, a bushed bronze

spearhead, cast in a double-leaf mold and possessing two through holes in the bushing, was also part of the assemblage. All these artifacts were found together, in close contact with each other, with their points facing downward and stuck into the ground. In proximity to these items, a fragment of an arrowhead made of clay shale and a fragment of rock crystal were also discovered.

Autosomal profile:

I25555. 4 3-way models pass for I25555 in the first set of qpAdm, involving a substantial fraction of ancestry from the FSHG cline (either ~45%Tyumen_HG or ~55% EHG), Steppe_MLBA ancestry (~15- 35%) and either Mongolia_N or Yakutia_LNBA ancestry (~22-28%; p>0.10). A wide range of 4-way models pass in the proximal qpAdm, with either 2 FSHG sources or 2 East Asian sources; Yakutia_LNBA-related sources, Mongolia_N_North, and China_YR_MN can individually or in combination account for the East Asian ancestry of this individual. However, the range of models that pass contrasts with results of ADMIXTURE (Extended Data Figure 10, Fig. 4B), where this individual appears to possess ancestry from the FSHG cline alongside a complex mix of East Asian ancestries, which may be due to this individual’s low coverage (0.04x). As the main East Asian source for this individual in ADMIXTURE is Cisbaikal_LNBA, we repeated the proximal qpAdm for all ST individuals with Cisbaikal_LNBA added to the references, rotated into the sources. We find a wide range of passing models for I25555, but once again with either Cisbaikal_LNBA, Mongolia_N_North or Yakutia_LNBA sufficing for this individual’s East Asian ancestry. We suggest once again that this may be due to the poor coverage. All models containing Cisbaikal_LNBA in the sources and references can be found in SI Data S7, Table 3B.


M/F: Female
Y-DNA: -
MtDNA: R1b1
Age: 2035-1900 cal BC

Burial description:

Burial 28 (individual ID I25558) was situated at a depth of 60 cm from modern surface and appeared as a patch of dark mixed soil, measuring 80x180 cm. Its orientation was from east to west. Within the grave, the remains of a man aged 30-35 years were discovered. The preserved bones included the leg bones, which were found in an orderly manner in the western half of the grave, as well as fragments of the skull, ribs, and the left ulna. It appears that the deceased individual was laid with their head facing east. Notably, the position of the bones in the right foot suggests that the foot was turned towards the left heel during the person's lifetime. Curiously, one of the bones ended up on top of all the other bones, although no evidence of disturbance in the grave was observed at this location.

Archaeological description:

I25558. We find that this individual can be modeled with near-complete descent from Tyumen_N_HG (~93%); when populations from the Eneolithic Trans-Ural steppe are added, this individual can be modeled with complete descent from Russia_CombPitWare_Gladunino_C_4.7kya (~100%, p=0.25), in keeping with her behavior in PCA and ADMIXTURE (Main Text Fig. 1 center, 4B; Extended Data Fig. 4, 5). 

Nothing too special going on with this sample, but interestingly this is another example where the gender of the individual does not match up with the assumed gender in the archaeological supplementary.

This wraps up the samples from Rostovka. At Rostovka we see quite a heterogenous bunch of individuals in the same cultural context. There are 11 samples where the dominant component was derived from populations who lived between the Ural mountains and lake Baikal. It is important to keep in mind that this does not mean that these samples were all locals. The local populations should sit somewhere on the cline between WSHG and ESHG populations, yet some samples are richer in WSHG ancestry than others, and other samples have substantial eastern ancestry.

Four samples come from people whose primary ancestral component was derived from steppe_mlba components. In addition, several of the primarily Western Siberian samples seem to have significant steppe_MLBA contributions, both through maternal and paternal geneflow.

We also find an individual which was highly related to the Krasnoyarsk_BA population, entirely derived from it, with a Y-chromosome haplogroup closely related to those of European Uralic speakers if not directly ancestral to them. Some other samples may have minor genetic contributions from these populations also.

The fact that you have unmixed Europeans and Eastern Siberians buried at the same site, in the same cultural context is quite amazing honestly. It is a testament to how strong and wide-ranging these networks were, connecting people from far away.


Satyga-16 is one of the more northern Seima-Turbino sites, situated  on the eastern side of the Urals mountains. The site was described as a Seima-Turbino monument founded by Seima-Turbino people going towards the northwest, traversing on the river routes.  You can see its location on this map below at #6:

Here is a layout of the burials at Satyga-16.

Honestly I kind of wished we got more samples from Satyga-16 because it seems like quite a large burial ground and it would have given us more of a clue regarding the populations which lived here. There is a unique factor going on at Satyga-16 because all the individuals found are ultimately migrants or derived from migrants, unlike what we see at Rostovka.


M/F: Unknown
Y-DNA: -
MtDNA: -
Date: 2288-2058 BC

Burial description:

Burial 17 (individual ID I32551) suffered significant damage due to disturbance by an uprooting, making it difficult to reconstruct its shape and parameters. It is believed that two graves were originally present at this location, oriented at an angle to each other. This is supported by the preserved cruciform outlines of the shared depression, as well as the presence of one skull within the pit and the discovery of another individual's skull not far from its eastern wall. Presumably, a burial (17a) was initially constructed in this spot, aligned along the west-southwest axis, containing the remains of a 25-30-year-old man. Later, another pit (17b) was dug above it, oriented along the NW-SSE line, approximately 0.5 m wide, in which a woman of about 35 years old was buried. The sloping walls of this grave descended steeply to the bottom, reaching a depth of 0.2 m into the subsoil. In the redeposited layer of the uprooting pit, a bronze scraper, a bronze object resembling a chisel with remnants of a bone handle, a stone arrowhead, and a tool on a plate were found. Above the graves, an accumulation of jar vessel fragments was recorded, while fragments of a third vessel were uncovered to the south of the burials (Trufanov, 2011). Burial 17 is dated to 2288-2058 calBCE (3770±25 BP, PSUAMS-12647).

Autosomal profile:

I32551_d. This individual can be modeled as a product of a two-way admixture (~30% Yakutia_LNBA + ~70% Srubnaya_LNBA) in the first set of qpAdms at the p>0.01 level (p=0.04). A similar model passes when FSHG populations of the Eneolithic Trans-Ural steppe are added, with Russia_Tatarka_BA as the Yakutia_LNBA-related source (p=0.07). All passing models for this individual include a Yakutia_LNBA-related source. 

This sample is interesting as well because according to the models the sample had about 70% steppe_mlba ancestry and 30% Krasnoyarsk_BA ancestry, with seemingly no FSHG contribution. Unfortunately we do not really have other points of information due to insufficient coverage. No where is it mentioned if this sample was taken from the male or the female individual in this burial. In any case, this profile does make you wonder about the story of how these people arrived and were buried at this northern latitude.


M/F: Male
Y-DNA: R-Z2122
MtDNA: Z1a1a
Age: 2571-2348 BC

Burial description:

Burial 6 (individual ID I32552). The grave pit is subrectangular in plan with rounded corners, 220x46 cm in size, oriented along the west-east line. The eastern wall is cut by a small rounded pit, 35 cm in diameter. The walls of the grave pit are even, gently descending to the bottom, noted at a depth of 10- 18 cm in the mainland. The eastern half of the pit is 5-10 cm deeper than the western one. The filling is brown sandy loam, covering a thin (3-6 cm) layer of pale yellow sandy loam in the bottom part, saturated with fine coals and remains of wood decay. Two carbonaceous strips (134x12x2-10 cm) located parallel to each other were unearthed 10-20 cm above the bottom along the long axis of the grave. Twelve fragments of pottery were found above the burial. At the bottom of the pit near the western wall, a skull and teeth of an adult male 30-40 years old were found. Next to the skull lay a bronze dagger in a birch bark sheath. Small fragments of ceramics were found in different places of the grave (Trufanov, 2011).

Autosomal profile:

I32552. In the first set of qpAdm, this individual can be modeled as three-way admixtures between ~47% Yakutia_LNBA + ~5-20% Srubnaya_LNBA, with the other two sources drawn from the FSHG cline (p>0.44). These models are in keeping with the behavior of this individual in ADMIXTURE (Fig. 4B, Extended Data Figure 10). In the second set of qpAdms, with populations of the Eneolithic/Early Bronze Age Trans-Ural Steppe as sources, 3-way admixture models pass (p>0.10), with ~40% each of Russia_Tatarka_BA-related and FSHG-related ancestry, and ~15-20% ancestry from the Steppe_MLBA-related source. All passing models for this individual include a Yakutia_LNBA-related source.

This sample at Satyga-16 is very interesting to say the least. Half of his ancestry seems from Kra001-related populations, with the other half being a mixture of Western Siberian and Stepppe_MLBA ancestries.The y-chromosome haplogroup of the individual was R1a-Z122, coming from the steppe_MLBA side. The maternal haplgroup is also interesting as it is Z1a1a, which interestingly does seem to have somewhat of a distribution related to Uralic languages. Based on the ancestry proportions and uniparental markers you can surmise that what we have here is someone who had an Steppe_MLBA/Western Siberian mixed father and an Uralic mother.

I32552 was dated to 2571-2348 BC, which seems a bit too early by my book as the Satya016 site is supposed to be contemporary to Rostovka. The ancestry at such an early date at this location would not make sense either. It is possible that this sample is offset a bit due to reservoir effects, which could occur if the individual had a diet rich in fish. Perhaps he was born and raised at these latitudes?

Just as Rostovka, Satyga-16 just shows how wild the interconnectivity of the Seima-Turbino period was. Here you have in the northern part of the Transurals a male which was a descendant of three populations that a few generations ago had not even met, with his paternal origins coming from the most southern of the three. The other individual had the vast majority of his ancestry come from the same southern Steppe_MLBA populations

Other points I want to mention

Before getting to my review of the Seima-Turbino in light of these new findings there are some other things I came across in these two articles that I have to address. Some things just kinda bugged me and I feel like pointing it out.

Samus culture

One thing I was a bit surprised to come across was that the label for the Tatarka Hill samples contained “Samus”, likely referring to the bronze age Samus culture of Western Siberia. When I decided to dig in to the supplementaries I found this:

“The Yakutia_LNBA and Russia_Tatarka_BA populations fall close to each other and also close to present-day Nganasans. Uralic populations form a cline stretching from present-day West Eurasians to the Yakutia_LNBA and Samusakaya populations. “

And this:

“Discussion. We find that the qpAdm results are broadly congruent with those of unsupervised analyses (PCA and ADMIXTURE) and discuss them in context. First, we find in all cases that ST individuals can derive all their Yakutia_LNBA ancestry from a Samus’-related population, even with Yakutia_LNBA itself moved into the set of references. This implies that Yakutia_LNBA ancestry in the ST individuals could be mediated by a source closely related to the population in the Samus’ culture, without direct admixture from populations with such ancestry from further further afield in East Siberia. Second, we find that FSHG sources from the Trans-Ural region suffice to account for the FSHG ancestry among many ST individuals, with at least one model passing for most individuals that has FSHG ancestry deriving near-completely from the local FSHG sources we used in the proximal qpAdm setup. Lastly, we find connections further afield to the far west (WHG ancestry) and the far east (nonYakutia_LNBA East Asian ancestry).”

It seems the authors have figured that the Tatarka hill samples are representative of the Samus culture, or come from Samus culture-related people. From what I can tell the attribution of the Tatarka hill site to the Samus culture seems to be derived from the Tatarka Hill site being mentioned as a Samus-Khizorovo type site in Kuzminykh’s 2011 article, which the authors referred to in their supplementaries.

Samus-Khizorovo refers to a type of axe heads which were derived from the Seima-Turbino tradition. We see these in the Samus culture sites but also at sites of the Koptyaki culture such as Shaitanskoye II from which these objects come from:

Samus culture and Samus-Khirizovo type should thus not be equated because one refers to a specific material culture and another to a particular metallurgical tradition found in several material cultures.

The cemetery at the Tatarka Hill contains graves of various period, the latest phase dates from the Middle Ages. The most part of the burials dates back to the Late Bronze Age. The funeral rite and material culture of these burials reveal similarities with the monuments of the forest-steppe “andronoid” cultures of Western Siberia (such as Elovka 2, Sopka-2, Chernoozerye-1). Among the metal artifacts listed are single-edged and, less commonly, double-edged knives, “horned” bracelets, rings covered with gold foil, various plate decorations, etc. In burials 14 and 15, two false-eared celts of type K–54 of the Samus-Kizhirovo tradition were found

Another group of burials is characterized by connections with the cultures of the Baikal region, as indicated by stone burial structures and jade rings (Kuz’minykh, 2011). A group of burials in shallow grave pits, oriented in the northwest to southeast direction, contained individuals laid in supine positions with their heads to the southeast. The grave pits were partially covered by stone slabs. The burials lack grave goods, and their chronology could be estimated only based on radiocarbon datings. We sequenced four individuals from this group of burials and dated two of them. The radiocarbon dates (2288-2058 calBCE (3770±25 BP, PSUAMS-9048) and 2021–1884 calBCE (3585 BP±20, PSUAMS7545)) reveal the attribution of these burials to the Seima-Turbino period and distinguish them from the total mass of the Late Bronze Age Andronovo-like burials. The freshwater reservoir offset of these dates might have explained the incongruity of the C14 data and the archaeological context, but the isotopic data reveals the highest values of δ13C among all our analyzed samples from Siberia and a relatively low value of δ13N. The δ13C/δ13N ratio better supports a foraging-hunting diet than the intensive fish consumption of the studied individuals. Based on this observation, we assume a specific Early-Middle Bronze Age phase at the Tatarka-Hill cemetery.

Furthermore as you can read in their supplementaries it is the “Andronoid '' burial group at the Tatarka hill which actually contain these Samus-Kizhirovo objects. The earlier burials from which the Tatarka hill genomes come from lack these Samus-Kizhirovo artefacts, in fact they don’t really have any grave goods at all. Their burial position differs from Samus as well, since the Tatarka hill burials are supine with heads oriented towards the south, whereas Samus culture positioned their burials in crouched positions with the body towards the left and the head towards the northeast, similar to the Okunev burial rites. The Tatarka Hill samples have very similar burial rites to the individuals from Neftoprovod however.

The authors also refer to these Andronoid burials as Andronovo-like, but “Andronoid '' refers to a collection of material cultures in bronze age Siberia that have varying degrees of Andronovo culture influence. Some Andronoid cultures like Irmen and Karasuk are derived from Andronovo people with material culture that diverged from Andronovo, but other Andronoid cultures may only have some Andronovo ancestry or simply just material influence from them. 

Skull  and reconstruction of an Irmen culture female

One of the Andronoid sites mentioned in particular was Chernoozerye-1. A sample in this study (I6787) came from this site and was modelled as a mix of WSHG, Yakutia_LNBA and WHG, without steppe_MLBA ancestry.

So there are several things going on here. The Tatarka hill site is not classified as belonging to the classic Samus culture. Then you have the presence of Samus-Kizhirovo celts at the Tatarka hill site, found amongst the Andronoid type burials. It is these burials that give this site a Samus-Kizhirovo connection whereas the burials from which the genomes come from have different cultural contexts and predate these by centuries. Thus there is no basis to give these a Samus-khirzovo affiliation. Simply put, the Tatarka Hill samples are not from the Samus culture, and should not be labelled as such.

Hunter-gatherer or pastoralists?

Another point I saw was that they assigned hunter-gatherer status to cultures that are typically described as pastoralists in archaeological literature. See here:

“This may suggest that individuals from Rostovka with unadmixed FSHG ancestry may derive from the metallurgical foragers of the nearby Odinovo and Krotovo cultures—the cultures best known to have engaged in the systematic casting of ST Artifacts”

While it isn’t incorrect to describe Odino as hunter-gatherers, this does not paint the full picture. The odino culture starts as as being full-on hunter-gatherers in the third millennium B.C, but the remains of Odino culture individuals from the late third millennium B.C already show a dietary shift caused by the addition of domestic livestock to their diet [7]. In the Krotovo culture of the second millennium B.C we regularly find the remains of domesticated sheep, cattle and to a lesser extent horses. While fishing, hunting and collecting wild millet would make up a significant portion of their diet [8], “Metallurgic foragers' would not be the correct nomenclature for their lifestyles. 

Krotov culture burial from Ust-Tartas-1

So this in itself would not be a much an issue, but the eason why I brought it up is because this matches up with the hunter-gatherer focused angle” of the authors, for a lack of better description. The article focuses a lot on hunter-gatherer groups and their interconnectivity across this northern eurasian belt, for example the words hunter and gatherer appear over 30 times each whereas pastoral(ism/ist) on the other hand is mentioned five times. This is a bit peculiar when the Seima-Turbino phenomenon is described as having metalworking pastoralist populations at it’s core.

I have a feeling this is in part because of the forager nature of Proto-Uralic populations, which seems to be the main focus of their article, but within the context of the seima-turbino phenomenon I find this to be a misunderstanding or misrepresentation of the actual archaeology. And this is a perfect segue to another point which has been bugging me a bit:


Personally I am a bit dissapointed in how Uralic-centric both articles seem, especially in the context of the Seima-Turbino phenomenon. Obviously I understand that the spread of this language family is an enigma which needs some real explaining, but given how the article features the Seima-Turbino phenomenon heavily I feel it does injustice to the main populations involved. There is very little talk about the Siberian populations at the helm of the Seima-Turbino phenomenon, which I find odd because these ultimately are the main populations at play here as shown from the genomes. 

The early presence of Indo-Iranians in Siberia, unmixed and mixed at two Seima-Turbino sites is barely discussed. The findings of significant Indo-Iranian participation seems like a big deal  to me, because it is the prequel to the later massive expansions Indo-Iranians had through Siberia and Central Asia, topics the Harvard lab have extensively covered in the past. In Zeng’s article, Indo-Iranian was mentioned about 5 times and Uralic about 60. Yeniseian , another language whose distribution was discussed, was mentioned about 20 times for example. 

Zeng’s article in my opinion also goes as far to significantly downplay the role that pastoral populations had within the Seima-Turbino phenomenon,when they precisely had such an important role and are described as the main drivers in basically all academic works covering the phenomenon.

All in all, I have a feeling both of the teams were working with the goal to unravel the illustrious origins of Uralic people and were aware/informed of the theory connecting the Seima-Turbino phenomenon to the dispersal of Uralic languages. The data of the genomes from Seima-Turbino and contemporary sites seem to only have purpose as a backbone to talk about the dispersal of the Siberian component of Uralic speakers and the Uralic language itself.

Zeng et al. goes even a bit further because the presentation of the data makes it somewhat seem like the Tatarka Hill samples were Seima-Turbino or Seima-Turbino related sites. I know this because I know several people who precisely got that impression at first.

When keeping this in mind, take a look at this quote:

Extreme diversity should not obscure the fact that the ST phenomenon was the context within which Yakutia_LNBA ancestry first dispersed westwards—almost to the Urals—for the first time, as suggested by the presence of such ancestry at all four ST sites we sampled (Tatarka, Rostovka, Satyga-16, and Chernoozerye-1). The presence of unadmixed Yakutia_LNBA ancestry at Tatarka Hill, to the west of the Kra001 individual from two centuries earlier, shows that the Yakutia_LNBA ancestry that penetrated onto the forest-steppe of this region, far Southwest of the Northeast Siberian forest zone, may have persisted to later contribute to the Yakutia_LNBA ancestry in ST necropolises. Intriguingly, Kra001 is just one out of a group of burials at Nefteprovod-2 dated from ∼4.2 to ∼3.9kya that show strikingly similar burial rites as the individuals at Tatarka Hill (SI Section II.F.iii.b-c). This confluence of cultural and genetic similarities suggests that a coherent and culturally distinctive population mediated the intrusion of Yakutia_LNBA ancestry westwards into the Krasnoyarsk-Kansk forest-steppes before 4.2kya, which then persisted in the region and later genetically impacted ST necropolises.”

Here they flat out refer to Tatarka Hill as a Seima-Turbino site and the whole paragraph seems to serve the purpose to “centralise” the Seima-Turbino phenomenon around this Uralic-related component, but what we see in ancient DNA data in my opinion does not warrant this at all.

In terms of the Krasnoyarsk_BA component within the context of actual Seima-Turbino sites (Rostovka andd Satyga-16) We have one unmixed sample at Rostovka, and some samples that may derive a minority of their ancestry from this population. At Satyga-16 we have two samples with this component, but one has it as a minority component and the sample with significant amounts acquired it through maternal geneflow. Basically, the genetic evidence for these populations having a significant demographic presence, being the driving force or the elites of the network is simply not there.

If the Seima-Turbino sites in Siberia paint the above picture, then the main European Seima-Turbino sites which are generally argued to have been left by this Siberian bundle + Abashevo newcomers from Eastern Europe are unlikely to show a much larger participation of this population either I would say. It is clear to me that the idea of the Seima-Turbino phenomenon being the archaeological imprint of the Uralic migration into Europe needs to be discarded.

The entry of Uralic speakers based on the ancient and modern genomes we have was most likely done by a population that upon reaching the Ural mountains would have been largely if not entirely Kra001/Yakutia_LNBA type populations. This can be seen in the lack of Western Siiberian ancestry in European Uralic speakers except for a small figure in the most eastern Permic speakers, and some of these European Uralic speakers have very significant amounts of eastern Siberian ancestry.

In my opinion this would also suggest that this population had a subsistence economy similar to those populations, being hunter-gatherers (with the capability of making bronze weaponry). Perhaps there was a growing sense of familiarity with other subsistence economies developing but they would largely be hunter-fishers. This has demographic implications as hunter-gatherer communities, even sedentary fishing based ones, have very small population sizes.

Perhaps this explains the modern Uralic cline then? A small population of hunter-gatherers entered Europe, and during their expansion mixed with their demographically larger neighbouring populations thus significantly altering profile, with these later profiles having a larger demographic spread after adopting new subsistence economies.

Yes kind of, but it is a bit more complex.

If we look at the BOO samples we can take a few things away;

  • The BOO population was undoubtedly derived from the same Siberian populations European Uralic speakers (Magyars excluded) are derived from, carrying the same paternal line N-L1026.

  • The profile of the BOO people began forming around the start of the second millennium B.C, indicating that around this time they start mixing with Eastern European Hunter-gatherer or West-Siberian hunter-gatherer populations around the Urals

  • The BOO people either took part in the same mixing process that other European Uralic speakers did but split off when this shifted from EHG to Steppe_MLBA populations, or the BOO people split from the European Uralic people prior to any of these mixing events, having a separate mixture event with EHG/WSHG populations

  • The BOO people were hunter-gatherers, and arrived on the Kola Peninsula halfway in the second millennium BC

We also have NEO538 from iron age Russia, one of the samples from Allentoft’s upcoming article. This sample is very interesting as it dates to 800 BC, which is close to being contemporary to the earliest Estonian Iron Age samples, likely being Baltic Pre-Proto-Finnic populations. This sample was about 1000 kilometres to the east of Estonia, yet its autosomal profiles couldn’t be more different from the Est_IA populations. Here is a slightly convoluted Gobal25 model for both samples to highlight their genetic differences:

Target: Estonia_IA.SG:s19_0LS10_1_noUDG.SG

Distance: 3.3036% / 0.03303590

52.6 Estonia_BA.SG

28.4 Russia_Srubnaya

7.6 Latvia_MN_Comb_Ware.SG

5.4 Lithuania_EMN_Narva

4.8 Russia_Krasnoyarsk_BA.SG

1.2 Sweden_LNBA

0.0 Russia_Tyumen_HG


Target: NEO538:NEO538

Distance: 3.8797% / 0.03879739

42.6 Russia_Krasnoyarsk_BA.SG

27.2 Russia_Srubnaya

18.8 Latvia_MN_Comb_Ware.SG

9.6 Estonia_BA.SG

1.8 Lithuania_EMN_Narva

0.0 Russia_Tyumen_HG

0.0 Sweden_LNBA

There is nearly a 10-fold difference in terms of Krasnoyarsk_BA-related ancestry. Note that this sample also has no Western Siberian ancestry, meaning this kra001-rich profile did not arise from recent migrations from Siberia into Europe. This is about as much if not slightly more Kra001 ancestry than seen at Bolshoy Oleni Ostrov. There is not much information to be gathered from the burial context but given the time and place it is unlikely for this individual to have come from a hunter-gatherer population. 

To me this sample shows that the Uralic cline in the iron age was not a simple matter of east versus west, other factors such as northern vs southern latitudes and perhaps more importantly, the climates along these latitudes seem to play a role in the degree of Kra001-related ancestry in populations. It also shows that some iron age Uralic populations still carried 40% Kra001-like ancestry, vastly more than any of the regionally proximate Uralic populations today.

I have no idea which linguistic group of Uralic this sample would have belonged to, but its location would indicate a relation to Pre-Proto-Finnic or Pre-Proto-Saamic populations. Or perhaps this individual came from further east and would have been an early speaker of Mordvinic or Permic that migrated westwards.

So to get back to the Seim-Turbino question, how does this relate?

In my eyes it seems indicative that the core European Uralic population after entering Europe initially lived in areas which were mainly inhabited by the remnant EHG populations and mixed with them first. This can be seen in the high amounts of EHG in NEO538 and the BOO samples. This is probably where the first differentiation between Uralic speakers in Europe develops as some would be more EHG-mixed than others. 

This is then followed by centuries of interaction with Indo-Iranian speakers west of the Urals via the Srubnaya complex, a period where the lifestyles of these populations gradually shifted to pastoralist-oriented lifestyles. Somewhere in the later bronze age there must have been an expansion of Baltic_BA-like people to the east, bringing slash-and-burn agricultural practices with them. Depending on where the Uralic speakers lived they acquired more steppe_mlba or baltic_BA related ancestry. It is likely that circa 1400-1000 BC there is a significant expansion of such populations towards the south and west, which I think is related to the climatic shifts mentioned earlier.

If we look at the Seima-Turbino sites in Europe though, we see a distribution that follows different lines. Aside from the Kaninskaya cave all the sites in Europe are either close to or within territory inhabited by the Abashevo culture. Here are the maps again:

In this scenario, the migrants upon entering Europe circa 2000-1800 BC would have immediately interacted with these pastoralist populations and intermixing with them would have lead to a rapid mass loss of Kra001 ancestry for Abashevo/Srubnaya ancestry due to their respective demographic differences. To create the profile seen with NEO538 or BOO, these populations would then have to move north and mix with Eastern European hunter-gatherers to a very significant degree, which in light of the demographics of Abashevo/Srubnaya versus late EHG populations just seems untenable.

If the authors were hoping to unravel the story of the Uralic expansion, rather than focusing on Seima-Turbino so heavily the authors should have rather focused on the post Seima-Turbino material cultures of the 2nd millennium bc in both Europe and Siberia. 

The second millennium B.C is a period of change in northern Eurasia, with major events such as the Indo-Iranian expansion occurring. The later bronze age also saw climatic developments which lead to demographic declines in many regions. This is part of a general trend in Eurasia in the later bronze age, leading to events such as the bronze age collapse. In the eurasian steppes, the demographic decline more or less left the steppes right open for the taking by the first true steppe nomads of the iron age, who did not suffer these effects in their more eastern homeland.

My best bet is that this climatic scenario plays a significant part in the Uralic expansion as well. The scant genomic evidence seems to point towards a later demographic expanse of populations of Kra001-EHG mixed ancestry, who then significantly mixed with proto-Iranian speaking Srubnaya people, followed by a mixing event with possible Para-Balto-Slavic speaking peoples. All of this points to later bronze age affairs, rather than one of a major expansion at the start of the 2nd millennium BC. But we will find out eventually as I imagine we will get samples in the future that shine a light on this matter.

Seima turbino revisited

The genetics of the peoples at the Seima-Turbino cemeteries are simply put, all over the place. They come from various different populations with different subsistence economies, some from nearby and others from far away. Many samples were mixed offspring descending from genetically distinct populations that had only recently begun interacting with one another. Despite the data only coming from two Seima-Turbino sites, with just a few samples we can piece a lot together about the Seima-Turbino transcultural phenomenon.

Altai-Dzungarian core

The first aspect that should be addressed is the theory of this pastoralist core from Altai-Tian Shan migrating from those regions across Siberia and Eastern Europe. As mentioned before I mostly agree with the hypothesis of Chernykh and Kuzminykh, but this is where my opinions differ.

I haven’t exactly seen anything in these DNA samples that points to the scenario proposed by them. I will say that with the combination of several ancestry streams involved, measuring smaller amounts of these components might be difficult, so it might be that some of these samples with partial steppe_MLBA ancestry could actually have ancestry from this region instead. 

Chemurchek stelae featuring an archer

When autosomal ancestry isn’t clear enough, haplogroups should do the trick. In the current batch of samples and data we have very robust evidence for Sintashta populations with their particular Y-dna markers, but not for the populations such as Chemurchek, Okunevo, or other populations from this area. Keep in mind that with the publishing of the genomes we could potentially uncover a closer link than we can currently estimate from the presented data.

However, a lack of such ancestry in these samples does not rule out an involvement or presence of such peoples in the Seima-Turbino phenomenon. Maybe they were buried at different sites? Maybe they were the cenotaph people?

Until further evidence shakes this up, the current ancient DNA record is not supportive of the movement of a Seima-Turbino pastoralist core from the Altai region spreading across Western Siberia. The samples at Rostovka seem to derive from various locations in Siberia or northern Central Asia without any clear connections to the Afanasievo/Chemurchek or Okunev-related cultures. Perhaps describing the Seima-Turbino phenomenon as "bronze age forerunners of Genghis Khan" was not too accurate after all.

Indo-Iranians and the Seima-Turbino phenomenon

At Rostovka we see quite a decent representation of Indo-Iranian peoples. Obviously the first sample to be pointed out was ROT003, an unmixed Indo-Iranian buried in a wealthy burial at Rostovka. In addition, one of the heavier equipped warriors was of paternal Indo-Iranian descent, and there were several samples with majority or full Indo-Iranian ancestry.

Abashevo charioteers by Yakov Petrov

If the trajectory between Rostovka and Satyga-16 was the result of a small military group expanding from A to B, the expanding military group should in theory have been derived from the three groups at Rostovka; Indo-Iranians, Western Siberians and Early Uralic peoples. At Satyga-16 the result of this “trajectory” seems to be an individual with 70% steppe_MLBA ancestry, and another which seems to have paternal Indo-Iranian descent and maternal Uralic descent. Given the latitude of this site I find the Indo-Iranian representation to be quite significant, as both individuals from this site had Indo-Iranian ancestry. 


To me this suggests that  beyond having a presence within the Seima-Turbino network, Indo-Iranians had prominent positions within the Seima-Turbino. Indo-Iranians with prominent burials, being significantly present in at least two Seima-Turbino sites, with several individuals within these contexts being of paternal Indo-Iranian descent.

Quite of the mixed samples seem to have had female mediated Indo-Iranian ancestry as well. These samples were primarily of Western Siberian origin and are perhaps indicative of the the political relations between these two groups in the Seima-Turbino phenomenon. Indo-Iranian men and their paternal descendants were part of these groups and had favourable positions within the network, but they also engaged in female exogamy with the native Siberian population as several of these seem to have mtdna clades accompanying this geneflow. This process was seemingly continued in the bronze age, as female individuals of probable Andronovo origin have been found in Krotovo culture contexts.

If you read between the fingers when the the Seima-Turbino phenomenon is covered in academic literature, there quite often is  description of a strong dichotomy between Seima-Turbino and Abashevo-Sintashta populations, or narrative of adversarial relationships between these two. To a degree there is some justification behind this as they are two distinct material groups and you see a lack of adoption of Seima-Turbino weaponry in the earlier phases of the Steppe_MLBA associated material cultures. But genetics does not show a “clean break” between the two, and shows that Indo-Iranians were significantly involved in the Seima-Turbino phenomenon, and not just in the west.

The early presence of Indo-Iranian ancestry deep in Siberia is a perfect explanation for the later massive expansions of Indo-Iranians circa 2000 BC. During this time period early pioneering groups and warriors were already travelling deep in Siberia, which led to an increasing familiarity with this new territory in their homeland. Soon afterr do we see these pastoral populations expand massively through Southern Siberia, ending up as far east as the Minusinsk Basin in what seems to be about a century or less even.

And hey I guess Parpola wasn’t all wrong either, because Abashevo-derived people do seem to play a significant role in the ST-phenomenon, moving in a east-west trajectory from Eastern Europe deep into Siberia.

Seima-Turbino and the Uralic expansion

Similarly the Uralic expansion westwards can also be tied into the Seima-Turbino phenomenon as you might have picked up from both articles. At Tatarka Hill do we see a direct migration from eastern Siberian populations that end up in a fairly southwestern Siberian location without admixing in the process, likely an indication of rapid population movement. 

It is in these regions that through river networks they could have had trade connections with the more southern pastoral populations without being geographically adjacent to them. Similar to how the Cossacks expanded through Siberia in a matter of decades through a series of interconnected rivers and tributaries, in my opinion this likely enabled the Uralic populations to have a swift and rapid expansion circa 2000 BC as well, but towards the west. 

The divergence of Uralic languages have been argued to have a fast breakup, with samoyedic splitting fairly early and the other branches splitting not much later. This is strengthened by how most Uralic speaking populations in Europe seem to lack WSHG-related ancestry while sometimes having significant Kra001-related ancestry. As discussed before, to me this suggests that the Uralic population that spread into Europe did so with minor mixing with Western Siberians and Indo-Iranians in Siberia, likely the result of rapid movements across sparsely populated regions at more northern latitudes.

At Satyga-16, the evidence for Kra001-related ancestry comes in precisely such a mixed manner, whereas at Bolshoy Oleni Ostrov we are looking at a population that is mostly a EHG/Kra001 mixture, meaning that their Siberian ancestors arrived with the vast majority of their ancestry still being like late neolithic Yakutian populations. 

The Uralic presence/participation in the Seima-Turbino phenomenon seems quite limited. Rostovka shows a cohort of Western Siberian populations and Indo-Iranians, with one yakutia_LNBA male amongst them. Some other samples may have smaller amounts of this component but it could also be from other peoples at such quantities rather than the result of direct admixture from a recently-migrated Yakutia_LNA population.

Satyga-16, perhaps also part of this trend of higher rates of Kra001 ancestry at more northern latitudes in the current ancient DNA record (Satyga-16, shows a more considerable presence of this component in comparison to Rostovka. However at Satyga-16 the Yakutia_LNBA component was carried by one individual with 25-30% of this component, the remainder being 75-80% steppe_mlba ancestry, and one individual which had 50% of this component. Given that this individual had Y-DNA R1a-Z2122 andm tdna w1, the situation is likely the result of a Indo-Iranian-West Siberian mixed father and a Yakutia_LNBA mother.

If the samples at Satyga-16 are representative of the groups and dynamics of the Seima-Turbino builders in Europe, then the Seima-Turbino wave into Europe is unlikely to be the representative of the Uralic expansion into Europe, or an archaeological representative of their entry in Europe. That is a big if however since we are talking about a site with two genomes, but the data at Rostovka certainly adds to that. It remains to be seen what other Seima-Turbino sites with human remains will reveal, and it likely will forever remain a mystery who built the various cenotaphs in Europe.

Kaninskaya cave.

OneSeima-Turbino site that might be interesting to mention is the one at Kaninskaya grave, which you may have seen mentioned several times in the articles and in my post. This cave was used as a religious site throughout various time periods, with the oldest evidence dating back to the Seima-Tubino period. This makes the Kaninskaya cave the northernmost Seima-Turbino site, situated on the European side of the Urals.

As you can see in this section about the Kaninskaya cave, the site is noted forth aving arrowheads and fishhooks with strong connections to contemporary populations in eastern siberia [9]:

Rostovka is also known for its abundant bone inventory. The most interesting finds are plate armor (graves 3, 6, 33). These shells are very similar to the defensive armor from the graves of the Glazkov culture on Ust-Ilga on the Lena Judah. Perevoznoy on the Yenisei, which A.P. Okladnikov (1955, pp. 233, 234, 248, 250, 252, fig. 118) at one time considered to be examples of the oldest bone armor in Northern Eurasia. Here it is appropriate to recall the bone implements from the Kaninskaya cave on Pechora, where bone arrowheads and composite fishhooks were found, which also have obvious analogies in Eastern Siberia (Burov, 1983, p. 43, fig. 7).” 

I find the connections the northern Kaninskaya cave finds have with eastern Siberian traditions to be quite interesting. The goods in burial 5 from Rostovka, the Yakutia-LNBA one, were linked to these finds in this cave as well, showing similarities in the style of arrowheads.

Kaninskaya cave’s northern location at that time period was a sparsely populated area only inhabited by remnant EHG foragers. It could be that what we see here is an archaeological reflection of an Kra001-related eastern Siberian population entering Europe, bound to start interacting with local hunter-gatherer populations. Does that sound familiar? Unfortunately these connections for now are purely archaeological but as they say; where there is a smoke, there is fire.

One thing I got from these two articles that have me a little concerned is that we have two different archaeological supplementary notes, and they do not fully match up. We see samples with different genetic profiles being attributed to the same burials as well as possible duplicates attributed to different burials. It might be that some burial descriptions got mixed up. This in itself is not an issue but we do not often get DNA articles where two different teams sample from the same site, and in this case we do and it doesn’t fully cut out. Does this mean I have to question the archaeological description given to samples in other studies? I would like to see the authors address the mismatches in DNA samples or archaeological descriptions.

All in all, the two articles have provided us with very interesting insight into the Seima-Turbino phenomenon and it will be very interesting when these samples are published. Childebayeva's article had a more typical approach and their team wrote a decent article that will add to our existing knowledge, but won't exactly break the presses. Their conclusion in regards to the Seima-Turbino phenomenon and the Uralic languages was quite sober as well:

With the new data from ROT, we are able to assess a recent proposal which suggested that Uralic languages could have been used within the ST network leading to the initial spread of Uralic languages across the Eurasian forest steppe. After performing various tests of genetic structure of the ST-associated individuals, we report genetic profiles on an ancestry cline that generally mirrors the genetic distribution of modern-day Uralic-speaking populations of the northernmost forest-tundra (taiga and tundra) ecological cline. Our findings show that the ST-associated individuals from Rostovka likely did not originate from a single location but rather represent people from a wide geographical area. Seima-Turbino was a latitudinal phenomenon on the same east-west axis where also the hypothetical homelands of the ancestral Uralic subgroups were positioned. Thus, our genetic results are temporally and geographically consistent with the proposal that Uralic languages could have spread within the ST network, but are neither a clear nor a direct proof. Further ancient human DNA data from northern Eurasia will help elucidate the details of the wider spread of ancient Siberian ancestry and its association with proto-Uralic speaking groups.

Zeng et al. in my opinion was far more daring, beyond it also being of larger scope of course. The upside of this is that in terms of combining ancient DNA with other fields such as archaeology and linguistics you get quite a fantastic article, and one that in my opinion has outdone most articles when it comes to this. I barely even discussed it in this write-up but I really liked that they connected Yeniseian languages to Cisbaikal_LNBA ype populations - a suggestion I have been making for many years now. 

The downside is that they might have gotten a little too excited in their attempts to explain it all, were a bit too hasty to make DNA-to-archaeology-to-linguistic connections which ultimately lead to some mistakes in their pre-preprint in addition to perhaps a misdirected scope in terms of sampling. Most of these issues can be solved through rewriting but some issues simply would need more sampling from different populations.


  1. White, Joyce & Hamilton, Elizabeth. (2009). The Transmission of Early Bronze Technology to Thailand: New Perspectives. Journal of World Prehistory. 22. 357-397. 10.1007/s10963-009-9029-z. 

  2. Parpola, Asko. (2022). Location of the Uralic proto-language in the Kama River Valley and the Uralic speakers' Expansion east and west with the 'Sejma-Turbino transcultural phenomenon’ 2200-1900 BC. Археология Евразийских степей. 258-277. 10.24852/2587-6112.2022.2.258.277. 

  3. Marchenko, Z., Svyatko, S., Molodin, V., Grishin, A., & Rykun, M. (2017). Radiocarbon Chronology of Complexes With Seima-Turbino Type Objects (Bronze Age) in Southwestern Siberia. Radiocarbon, 59(5), 1381-1397. doi:10.1017/RDC.2017.24

  4. Nomadic Cultures in the Mega-Structure of the Eurasian World Еvgenij N. Chernykh

  5. Salvatori, Sandro & Vidale, Massimo & Guida, Giuseppe & Masioli, Elisa. (2009). Ilgynly-depe (Turkmenistan) and the 4th Millennium BC Metallurgy of Central Asia. Paléorient. 35. 47-67. 10.2307/41496961.

  6. Wei, Peter & Jia, Peter & Betts, Alison. (2010). A re-analysis of the Qiemu'erqieke (Shamirshak) cemeteries, Xinjiang, China. Journal of Indo-European Studies. 38. 

  7. Марченко Жанна Валерьевна, Панов Всеволод Сергеевич, Гришин Артем Евгеньевич, and Зубова Алиса Владимировна. "Реконструкция и динамика структуры питания одиновского населения Барабинской лесостепи на протяжении III тыс. До Н. Э. : археологические и изотопные данные" Вестник археологии, антропологии и этнографии, no. 3 (34), 2016, pp. 164-178.


  9. Памятники сейминско-турбинского типа в Евразии: Эпоха бронзы лесной полосы СССР