search this blog

Tuesday, August 12, 2025

Tripolje, Dereivka, and the Sredni Stog phenomenon (guest post)


This is another guest post by an anonymous contributor. Again, I don't necessarily agree with the author, but is he wrong? Feel free to let me know in the comments below.

The definition of Sredni Stog phenomenon (SSP) varies and is often a loosely applied term to refer to pretty much any individual in the Dnieper and Don regions between the ‘Neolithic’ and Yamnaya periods.

In order to elucidate the SSP, some brief remarks on the preceding Mariupol phenomenon are warranted. Understanding the Mariupol horizon is fairly straightforward – its development was catalysed by an intrusion of groups from somewhere west of the Dnieper, ~ 5500BC. The ‘proto-Mariupol’ group were genomic and economic ‘hunter-gatherers’, lacking any discernible EEF admixture, and with Y-hg I-L702 uniparental ‘trace-dye’. The Mariupol phenomenon predominantly impacted the lower reaches of the Dnieper and Azov steppe (Lower Dnieper- Azov group, ‘LDA’), but extended toward the Don, Volga and even the Kuban steppe in an attenuated form. The elevated levels of “Ukr N’ in Golubaya Krinitsa and the Y-hg I2a-L702 individual at Berezhnovka attest to this movement. The Mariupol phenomenon is associated with the development of formal cemeteries, linking them with Late Neolithic mixed farmer/HG groups in the northeast Balkans. Individuals were buried in a ‘supine straight-legged’ inhumation, with grave goods such as boar-tusk pendants for select males and adorning shell beads for females. This might signal the emergence of gender-differentiation in burials and the rise of local leaders or ‘chiefs’.

Data sets treat the ‘Neolithic’ in Ukraine as a monolithic phenomenon, however it is important to note that Dereivka stands apart – it is 200km north of other “Ukraine_N” sites such as Volnienski, Igren and Vovnigi, beyond the ‘Dnipro bend’. Moreover, male individuals from Dereivka are overwhelmingly assigned to Y-hg R1b-V88 and the burial pose (N-S) at Dereivka deviates from the more common E-W orientation seen elsewhere. Quite a few of the published ‘Sredni Stog’ individuals are from Dereivka, and often earlier than 4500BC, and N. Kotova assigns them to the Dnieper-Donets culture. Moreover, the recently published middle Don individuals, such as those from Golubaya Krinitsa and Vasilyevsky Cordon -17, are also not Sredni Stog, but can be thought of as ‘partially Mariupolised hunter-gatherers’. In another example, the (undated) ‘Sredni Stog’ individual I27930 from Igren was assigned to Y-hg Q and he can be modelled as a 2-way mix of EHG & WHG. This individual is actually from the Mesolithic.

So what occurred during the Sredni Stog period? In contrast to the Mariupol phase, the population dynamics associated with SSP are complex: at least three external flows can be highlighted (i) the advance of Tripolje communities from the Carpathians to the Dnieper (ii) arrival of South-Caucasian/CHG agro-pastoralists in the north Caucasus, and (iii) arrival of ‘central Asian’ populations in Volga-Caspian region (represented by “TTK individual’); in addition to intra-steppe shifts and flows. Notwithstanding, the ‘ideological background’ of SSP is rooted in the Mariupol horizon. The stereotypical SSP burials features individuals buried on their back, but increasingly with legs up-flexed. And we see the beginnings of kurgan constructions, which vary from stone cairns to soil-thrown barrows. Most are buried in simple pits, however some have more complex ‘catacomb’ pits.

What happened in the Dniester-Dnieper-Don region during the SSP? We can begin by orientating ourselves with a PCA to observe two main clines developing. One cline develops between ‘Ukr_N and EEF and a second cline pulls toward Lower-Volga Caucasus groups. The first cline mostly comprises of ‘Farmers’ from Tipolje and ‘hunter-gatherers’ from Dereivka. The second cline consists of individuals from Dereivka and the lower Dnieper-Azov group pulling toward Lower Volga-Caucasus groups.


Admixture analysis with qpAdm reveals 3 groups within the 2 broad clines. The first group can be thought of as ‘core Sredni Stog’. These individuals are 2-way mixes of ‘Ukr_N’ and ‘Steppe Eneolithic’ (sometimes Progress works, sometimes Remontoye or Berezhnovka). They are both males and females. In our examples, the females are from Kopachiv Yar (4000 BC) and Dereivka (3500BC). The males come from Dereivka (4300BC), Moluykhiv Bugor (4000BC), Vynohrado (4000BC); they are all derived for Y-hg I2a-L703+. These results represent a blending of social networks between the LDA and various lower Volga-Caucasus groups, and the subsequent expansion by LDA further West. The terminus ante quem of 4300 BC matches the corrected dating of the Kuban steppe sites such as Progress & Vojnuchka.


Another subset comprises of individuals from Dereivka and Verteba cave who situated on an ‘Ukr-N’ < - - > EEF cline. Many of the earlier Dereivka individuals are almost 100% Ukr_N. Verteba cave Tripolje can be modelled as 80% EEF + 20% Ukr_N. One individual from Dereikva (I3719) falls outside the Dereivka <-> Tripolje clin, as he plots further ‘south’ with Balkan-LBK farmers. Consistently, he comes out as ~100% EEF with qpAdm. Dating to ~4700BC, he precedes the arrival of Tipolje groups to the region by hundreds of years. FtDNA have assigned him to I2-Z161- FTH81, which is distinctive to the LDA haplotypes and is phylogenetically linked to a Czech LBK individual.


A third group consists of individuals with more complex 3-way ancestries, consisting of EEF, Ukr-N and Steppe_En and/or Maikop. These come from late Dereivka and late Tripolje groups, in archaeological literature often termed as ‘Soldanesti’, ‘Zhivotilovksa-Volchansk’, Cernavoda (Kartal). Once again, the males from Soldanesti and Cernavoda derive from LDA-related Y-hg I2a-L703 in some shape or form.


Conclusions:

1) Firstly, we note that the Dereivka group was subject to early EEF influence, as soon as eastern LBK groups reached Ukraine after 5000BC. However, their main interaction occurred with the younger, Tripolje group, which expanded toward the Dnieper after ~ 4300 BC.

2) In the LDA group we observe patrilineal continuity. These clans created expansive social-networks. They initially mixed with groups from the lower Volga-Caucasus area. Some then moved west, and ‘took over’ the Tripolje region and acquired high levels of EEF.

3) As a third conclusion, we can reject the commonly held notion that Tripolje was ‘conquered by Yamnaya pastoralists’. Our analysis instead highlights that their core structure fragmented as they became intertwined with powerful networks to their west (Trpolje) and east (Sredni Stog). The ‘take-over’ was due to the expansion of LDA/ SS groups. Mixed groups emerged such as Cernavoda and Soldanesti, which retained Tripolje ancestry and some cultural traditions. By the time Yamnaya groups reached the Dniester forest steppe, Tripolje had been long gone.



See also...

‘Proto-Yamnaya’ Eneolithic individuals from Kuban steppe c. 3700 BC ? (guest post)

207 comments:

«Oldest   ‹Older   201 – 207 of 207
Norfern-Ostrobothnian said...

Well adding Iran N would break the model because there's probably bidirectional geneflow. What we'd need to find the true Iran N source of ANE would be DNA from cultures such as Tutkaulian or other Central Asian Epipaleolithic cultures.

Norfern-Ostrobothnian said...

@Shomu
I don't think Iran N has IBM related ancestry, it and AASI seem to act as proxies for something unique to it and CHG (possibly Caucasus UP too but not necessarily). CHG and Iran N are also not identical, so you'd have to add geneflow specific to those groups too.

Rob said...

@ Norfern
''Well adding Iran N would break the model because there's probably bidirectional geneflow. What we'd need to find the true Iran N source of ANE would be DNA from cultures such as Tutkaulian or other Central Asian Epipaleolithic cultures.''

Sure, but why would it break the model even if there was bidirectional gene flow ? If the qpAdm set up obeys chronology, it can cope with such complexities. That's why we can sensibly perform proximal modelling in post-3000 bc Europeans who have overlapping strands of ancestry, for example.

Rob said...

@ ArmandoR1b
I had a deeper look at the Bourgoise et al study. If we look at the supplement, every CW sub-group has steppe ancestry, and variable about of 'Late Neolithic' (GAC/ TRB) ancestry. This obviuosly means that the key link to CW burial is steppe ancestry.
However, they conclude that there is no correlation between steppe ancesty and CW burials because of their model assumption - if CW burials spread due to steppe migrations, CW burials in the East should be older than the West. Their analysis shows that Fatyanovo burials and Polish-NW Ukrainian CW dont seem to be particularly 'old', and are the same age as CW burials in Germany, for ex. The former is not surprising, because we know its a later CW offshot, the latter is a bit surprising and must mean something.
However, their data shows two major regions of early burials (look at their radiocarbon curves for CW Fig 1) - Czechia and the East Baltic. So although I disagree with the final conclusion, the great benefit of this study is their analysis of chronology which has improved our understanding.

I wont go into Bell Beaker for now, that is far more complex

Norfern-Ostrobothnian said...

@Rob
Iran N ANE is more related to the ANE in TTK than to any available ANE source, so that causes the model to fail. If we had an ANE sample that was more related to the ANE in TTK than the ANE in Iran N is then the model would likely not fail as bad.

Norfern-Ostrobothnian said...

Also consider that my ANE source here is MA1, which is not at all as proximal to the ANE in Iran N as the ANE in TTK would be.

Dospaises said...

@Rob "the great benefit of this study is their analysis of chronology which has improved our understanding." Yes, which should also cause future investigations of the details which is good.

«Oldest ‹Older   201 – 207 of 207   Newer› Newest»