V13: splitting the problem

[rafc]

Hi, as mentioned on the other thread, the "splitting the problem" thread from Anthrogenica is now only accessible through the Genoplot archives. It thought it could be useful to repost them here, the text has not been changed.

First part:

A few years ago someone posted a thread with the title "Why is E-V13 so confusing?", and I think it's a sentiment we can all share.
Even the authors of the Southern Arc paper were puzzled:

Its absence in Bronze Age southeastern Europe (n=107) is in remarkable contrast with its
ubiquity in the present day, leading us to hypothesize that either it did exist there prior to our
sampling but in a specific region from which we have no samples or it arose elsewhere and
migrated to southeastern Europe just prior to the earliest sampled individuals.

I posted in the past on how I think there are two quite different types of subclades in V13. First we have what I would call the ‘Eastern clades', they are very present in Western-Europe, but also in the Middle-East and Arabia. I call them 'Eastern', because they have less presence in the Western Balkans compared to other clades. In these 'Eastern clades' I would put a lot of basal clades under BY3880, but also
https://www.yfull.com/tree/E-FGC44169/
https://www.yfull.com/tree/E-Y16729/
https://www.yfull.com/tree/E-Y145455/

I put this in contrast with what I would call the 'Western clades', which are very present in the Western Balkans, for example:
https://www.yfull.com/tree/E-CTS9320/
https://www.yfull.com/tree/E-FGC11450/
https://www.yfull.com/tree/E-L241/.
They are comparably less present in Western-Europe, are virtually absent in the Middle-East and Arabia, but also have a large presence in Central and Eastern-Europe. To me this implies both groups must have had some spatial separation, and followed different trajectories in the Iron Age/Roman Era. The 'Eastern clades' must have been in a good position to be recruited by the Roman army, as their presence on the British isles proves. Maybe their presence in the Middle-East and Arabia has the same origin, or maybe there was a different mechanism at work.

I think different solutions are needed for both groups of clades, and the confusing nature of V13 is mainly due to this. I will try to propose a different solution for each. The Ancient DNA we got over the past few years gives a good explanation for the 'Eastern clades'. I put the samples from the Early Iron Age Psenicevo and Babadag cultures on this graph:



I also put some contemporary Steppe samples on the graph to illustrate the difference. I think everyone agrees that in the whole Bronze Age the genetic profile in the area of the later Babadag culture (around the Danube close to the Black sea) was very steppe-like, so the Babadag sample (MJ-12) looks like it must have its BA-roots somewhere else, it can hardly be derived from the MBA/LBA cultures of the region. This fits nicely with the fact that the Babadag settlements started “ex nihilo” around 1000BC. Babadag is comparable to Psenicevo, not just in the use of the typical stamped pottery, but also in the use of specific rituals with the dead. These rituals clearly point to one and the same population. The logic seems Psenicevo spread to Babadag, since Babadag was a newcomer in the region. The PCA above fits perfectly with this. The Babadag samples sit exactly where you would expect, if you had the Psenicevo-samples migrating northwards, and mixing with Steppe-like locals.

Now since we only have one Babadag sample, we could question how representative it is, but I also put the Late Iron Age samples from Glinoe with high ANF on the graph, and it's striking how comparable the profile is. Glinoe was close to the area of the Saharna-culture, which again used stamped-pottery and the same rituals with the dead as Psenicevo and Babadag. So it’s possible the Glinoe samples represent a local profile since the Early Iron age. I think this group of cultures is a good fit for the 'Eastern clades' of V13 for several reasons:

• It’s clear that they would have been mainly close to the Black sea, that puts them in a good positions to be recruited by Greeks and Macedonians, and in any case gave good access to the Eastern-Med through Greek colonies that were present as of the 8th century BC.
  
• Most of the area later became part of the Roman empire, and was the favorite recruiting ground for the army.
  
• The Psenicevo samples were nearly all V13, as was one of the two Late Iron Age samples from Glinoe (the Babadag is female)
  

I think this neatly explains the presence of the 'Eastern clades' in the Middle-East, Arabia and Western-Europe (especially the British Isles). So we have the first part of the solution: at the start of the Early Iron Age these 'Eastern clades' were probably mainly present in the Southeastern corner of present day Bulgaria, an area which I think matches good with the very ANF-profile of these clades.

However, this Psenicevo to Babadag phenomenon is not a good match for the 'Western clades'. Compared to the 'Eastern clades' they must have been more inland, to explain their lack of presence in the Middle-East and Arabia. We know from the early Roman era Ancient DNA that V13 was prominent in Viminacium (~33%) and it was also found in Nis. Now the question is where it came from. Based on leaks, it seemed for a long time that there was little V13 in Serbia in the Iron age. These leaks have now been debunked, and with that I believe a solution to the problem is possible.

Given the location of Viminacium and the regions where the 'Western clades' are prominent, I think the Basarabi phenomenon (8th-6th century BC) is the main vector for the spread of these clades. The heartland of Basarabi is in the region of Viminacium, and given that not long after the disappearance of Basarabi we have the Celtic invasions in the Balkans, I find it hard to see which other mechanism could have brought these clades to this region. The Basarabi phenomenon was also present in Romania, and seems like a good vector to bring these Western clades to Central and Eastern-Europe. The question is then how these 'Western clades' got to the Basarabi heartland.

In the LBA this was the region of Encrusted pottery. This is thought to have originated in Hungary and spread downstream along the Danube with migrations, taking the habit of urnfields with them. Ancient DNA studies point to them being very HG, and not having V13. Now since we split the problem, the 'Western clades' are not obliged to have the very ANF profile of course, but it still feels very unlikely the encrusted pottery could have been V13-rich. Basarabi is probably made up of elements (maybe culturally, maybe genetically) of the LBA Encrusted pottery, the LBA Belegis II culture of the Banat, and the Early Iron Age Insula Banalui culture. The Belegis II again seems unlikely as a source for V13, as we have some idea of what were the genetics there in the first half of the second millennium BC (not V13), and it’s a long way from where the 'Eastern clades' resided (I would expect some geographic continuity). Insula Banalui by contrast is another version of the Stamped ware phenomenon that spread in the Eastern Balkans, and was also spread in Northwest Bulgaria.



So my best guess would be that the 'Western clades' were originally in a position just west of the 'Eastern' ones, and then at the onset of the Iron Age moved North-westwards. Just like Babadag was a mix with locals, I would also expect a mix with locals here, creating a more HG shifted profile, towards Serbian BA samples (or even the Encrusted profile), and in terms of Y-DNA it seems logical also L283 and R1b would be present (and some other minor clades). With this placement there is a geographic continuity for the V13 clades, and I think it gives a reasonable explanation for the difference in behavior of the two groups of clades.

In terms of language, I think that early Iron Age group that spread out spoke the ancestor of Thracian and Dacian, with the 'Eastern clades' later forming the backbone of Thracian speakers, while the 'Western clades' were at the basis of Dacian speakers (if there was any difference between the two languages at all).

A sketch of how I see the spread of V13 in the Iron Age:

[rafc]

Part 2:

The first post was mainly about V13 in the Iron Age. There are two things I want to add: first on V13 in the Bronze age.

Above I argued that the bulk of V13 must have been in Southern Bulgaria at the onset of the Early Iron age, spreading North and North-West from there. The Psenicevo samples had a very specific autosomal profile, being very high in ANF. No other such profile has been found in the many Bronze age samples that have been tested around Europe and Western-Asia, so it's hard to see it coming from somewhere else. Together with the lack of V13 in Bronze Age test results, it seems very likely to me that there was no substantial change from LBA to EIA. V13 and its specific profile were already present in Southern Bulgaria, what changed is just that they started expanding in the EIA.

On top of that, the specific genetic profile and dominance of one Y-haplogroup might find a good explanation in what came before. The Middle Bronze Age period in Bulgaria (say 2100-1500BC) was a period of severe depopulation, this kind of bottleneck explains the genetic changes and the dominance of one group. V13 might have profited from the bottleneck by being in a certain location or having a certain lifestyle/subsistence strategy, but I doubt we will ever figure that out for sure.

With this period we also touch upon the subject of the roots of the Thracians, which according to some are later arrivals than Yamnaya. Catacomb culture and, more often, Multi-Cordoned ware (MCW, KMK or Babyno) have been suggested. Now the interesting thing is that we have one sample from MBA Bulgaria (I2163), and it plots close to contemporary Multi-Cordoned ware samples (and other Steppe like samples) from Ukraine and Moldovia.



I2163 was R1a which was also found in MCW, and is the only Y-DNA group besides V13 which was found in Kapitan Andreevo (Psenicevo). There are scholars who argue Thracian is close to Baltic languages, a link with something Corded-Ware derived might fit with that, but there are many theories on the affiliations of Thracian out there, and since I'm not a linguist I have no clue which ones make sense.

In any case, all the above seems to fit well with a scenario where a very ANF (more or less neolithic) population, dominated by V13, survived somewhere in the EBA. The V13 group then profited of the MBA bottleneck to expand in Bulgaria (maybe the Yamnaya admixed EBA population had moved on to the Western Balkans and Greece?) There they mixed with an incoming group of MCW, creating the basis for the LBA/EIA population. The question here is probably how likely it is such a very ANF population survived (but we know some very HG ones did!). Note that they didn't necessarily survive within Bulgaria, but it is certainly a possibility. Another issue is the low levels of R1a in the later Balkans compared to V13, this would mean the population bringing the Thracian language was much smaller than the local V13 population. Z2103 (associated with Yamnaya) is by contrast found more often. So I'm not sure the scenario is realistic from that point of view. (but we see comparable phenomenons in Mycaneaens and Phrygians)

Example of how the MBA cline could have looked:



Although this scenario for how the LBA/EIA V13 rich group came to be looks possible to me, I have to add that it's at odds with what the Southern Arc paper and the Reich lab have claimed: that all Paleo-Balkanic people/languages derive from one Yamnaya migration. So if we follow their view, the EIA V13 group is simply a descendant of the EBA populations of Bulgaria, and Thracian should derive (together with every Paleo-Balkanic language) from the common language spoken by the Yamnaya migrants. This implies the MBA sample above was a dead end. Given the depopulation in the MBA, it seems unlikely to me that the one sample we have would not be relevant, but nothing is excluded, it's only one sample after all. Maybe someone knowledgeable could check with formal stats whether the Kapitan Andreevo sites can work as a combination of very ANF samples from Late Chalcolithic/EBA (like I2519) and either Yamnaya or the MBA sample?

[rafc]

Part 3:

After digging in the Bronze age history of V13 I want to look at what happened after the Early Iron age to try and explain how the V13 distribution became the madness it is today. I thought one extra post would be enough, but now I see I might need one or two extra to get the Roman and post Roman part covered.

So to recap, I proposed that in the Early Iron age the ‘Eastern clades’ of V13 spread north along the Black Sea coast up to the Danube and Dniester. They spread a typical form of ‘Stamped ware’ and typical rituals with the dead. I proposed they correspond to what classical authors called Thracians.
I also proposed that the ‘Western clades’ of V13 lived in an area to the West/Northwest of the Eastern Clades. I propose they played a major role in the genesis of the Basarabi phenomenon (800-600BC), which likely happened in South-West Romania. I think looking at this recent map with the spread of Basarabi, most people can easily see a link with the peoples that the classical authors call Dacians, and I proposed the ‘Western Clades’ form the basis of these Dacians. This fits of course with the supposed relatedness of the Thracian and Dacian languages, although they are not well attested.



I also want to re-illustrate the PCA position of the EIA samples we have from Kapitan Andreevo, Diamandievo and the BGR_IA sample, all plotting close together, and the EIA MJ-12 sample and LIA Glinoe samples which also plot together. In my view they all represent the same profile but MJ-12 and Glinoe (both from near Moldovia) are admixed with a local Steppe-like population that was present there in the LBA (see my first post in this thread for a detailed reasoning why):



I will refer to both groups together as the “Thracian cline”. Now for the “Western clades” we don’t have direct samples, so it’s hard to say whether they would fall into the same cline, although I think much points to that, but I’ll get back to that. Apart from the samples on the graph above we have some other V13 from the Iron age, which can help understand what happened later, as well as some non-V13 which falls in the same cline. The latter category is much harder to use, since samples can easily plot in the same area without any relatedness, and especially that upper left part of the cline has samples from many other parts of the Balkans that just happen to have the same proportion ANF/Steppe and low HG, most likely without a real connection. Since the male samples from that Thracian cline are dominated by V13, this can be a useful clue to determine relatedness or not.

The bottom-right part of the cline is more unique, so there is a very good chance that I7233 (female, 850BC) from Northern-Macedonia had roots to the east. From Bulgaria itself we have I19500 (3d century BC, although only archeologically dated), who is also V13. Interestingly he plots somewhat to the northeast of the older samples, due to slightly increased CHG. It is only one sample, but it does seem that what we see here is the effect of East-Med ancestry spreading in the Hellenistic world. Since the 5th century already Thrace was involved with Greeks and Persians, and in the 4th century it had been conquered by the Macedonians who had founded some cities there. So it would seem there was a shift to more East Med going on, and even though we have no coastal samples yet, I would bet that in and near coastal cities the shift was much bigger (I19500 is from the interior). I also suspect that V13 would form a sizable part of the population of these coastal cities (together with Greek, Anatolian and Levantine groups). I don’t think a same shift is visible in the samples from Glinoë who dated from the same era, probably because they were much further from the Hellenistic world.



Much closer to the upper left corner we find I18832 (260BC) an outlier from a La Tène context on the Austrian-Hungarian border. The sample is V13, so while its placement allows many origins, I think origins in the Eastern-Balkans are most likely. While this sample looks like he could be born in the Eastern-Balkans, there are two other samples who decidedly do not: I16272 (Prague, 300BC) and I18527 (Gvor, Hungary, 250BC). In the Czech La Tène samples there is one V13 in 28 samples, in 16 Hungarian La Tène samples there is one local looking V13 and one Eastern Balkan like V13. Assuming that indeed in the Bronze Age V13 was confined to Southern Bulgaria, there should be a mechanism that brought V13 here, long enough before 300BC so they could become indistinguishable from local samples. The fact that the V13’s took on the local profile, and form a small minority shows this was a mechanism with limited impact.



We can identify this mechanism through some other samples. We can observe that two samples from Slovakia, on the Hungarian border (I12097 and I11722) plot very close to the bottom right part of the Thracian cluster. This profile seem unexplainable by anything else than a migration from the Eastern Balkans. The samples are dated 650-500BC and belonged to the so-called Vekerzug culture. This culture is seen by many as the result of Thracian and Scythian influence or even migration into the Pannonian basin. This seems to be confirmed by the genetics, and this would make it a first example of V13 hitchhiking along with Steppe-nomads from the Eastern Balkans into Pannonia, something I think repeated multiple times afterwards.



The total of Vekerzug samples plot all over the PCA, actually comparable to Avar age Pannonia. This is very likely illustrating the combination of locals (very WHG), “Thracian” newcomers and in-betweens that are admixed between the two. So I think the La Tène V13 samples in Gvor and Prague are well explained by that first wave of V13 arriving in the 7th to 6th century BC with Vekerzug, and being completely La Tène like by the third century BC. And what do we see? Indeed, the Vekerzug samples had a V13 (I14465), already admixed and shifted towards the local Central European population.

We also have a Croatian V13 sample from the LIA, who looks like a local of his region. He is from northern-Croatia, close enough to where the La Tène V13's were found to assume a comparable mechanism, although other explanations certainly exist.

[rafc]

Part 4 was mostly about the Danube Limes samples, but based on the pre-print. Now we have the actual samples, it should be reviewed so I skip it.

Part 5:

The next part will focus on the post-Roman samples from Pannonia. Above the sample I20802 (3d century AD) was not mentioned. Even though it dates from the Roman period, it comes from outside the Roman empire. It’s described as Sarmatian. In the first century AD the Iazyges (often called Sarmatians, which is a term that also covers other tribes) had invaded Eastern Pannonia, where they remained in power for a few centuries. I20802 would suggest that a large part of these Iazyges were actually Dacians coming along, having a lot of V13.

We also have graves that date from the late Sarmatian age (end 4th, early 5th century), and have typical Sarmatian customs. They have mostly the same genetics as the 3d century one, and cluster quite close together. Out of 8 male samples, one is V13. There is also one female sample that plots close to the upper left of the “Thracian cline”. None of the samples show CHG ancestry, making any Roman genetic impact in this period very unlikely. My interpretation is that most of the Sarmatian samples represent a mix of descendants of Dacian migrants of the first century AD, higher Steppe Sarmatians and higher WHG Central-Europeans (locals?). The one female sample shows there was still new influx from the east, the paper mentions that this cemetery shows connections with the Chernyakhov culture.



After the Sarmatian graves we have actually a gap of no less than 150 years until the early Avar samples. Some of the Sarmatian graves still overlap with the reign of the Huns (c. 433-454), but the following period, where former Hunnic allies took over Pannonia, the Goths in the west and Gepids in the east, is a blank slate in the east. After the large gap from the later Sarmatian/Hunnic period to the Early Avar age, the genetics in Easter Pannonia have change considerably. The PCA below shows samples from the 7th century, but not the East-Asian ones. Now only 3 samples plot where the tight cluster of Sarmatians did: a V13-CTS9320, an R1a, and a female. It seems like most of the previous population disappeared or was deeply changed. The CTS9320 can probably be seen as a holdover from the Sarmatian era population.

If it was population change, then it was not caused by the very Germanic Gepids, as only four female samples plot very Germanic. Many samples now plot more Eastern shifted, and especially remarkable are several samples, nearly all V13, which look like they come from the Thracian cline but with slight CHG shift. This is pretty much what we can expect of the local profile in Romanized areas of the Eastern to Central Balkans in late Antiquity. So why do we find it here and now? We can also observe that while there is much V13 in that CHG shifted Thracian cline, there is nearly no V13 in the samples west of the Thracian cline.



The only explanation I see is that the Avars must have forcibly brought in new people. Maybe from an area like Viminacium, maybe from further away in Thracia proper. The Avars did supposedly take 270,000 captives with them after they unsuccessfully besieged Constantinople in 617. Maybe the other profiles represent profiles from Illyria and other neighboring areas?
The Avar-era paper itself had nothing useful to say on this issue, but the paper on Avar origins did have this:

In the early Avar Empire, the archaeological record demonstrates that the population of the Carpathian Basin remained quite heterogeneous, at least culturally. The late-Roman element was reinforced by captives from the Balkan provinces and elsewhere settled as dependent laborers.

I think that is indeed what happened. If we look at the samples from the 8th century, it’s clear that now V13 plots mainly west of the Thracian cline, where before the bulk of samples was already, so they lost their original profile through admixture with other populations in Pannonia, who were likely also brought in by the Avars, but from other areas (Slavs, Germans, Illyrians, ...).



This would suggest the bulk of the Central European V13 arrived in early Medieval times, as transplants from the Roman Balkans, while a small part were holdovers from earlier waves.

[rafc]

Part 6:

I'll be continuing on following the implications of splitting the V13 problem in the Eastern and Western clades. In the last post I tried to show that the majority of the V13 in Avar age Pannonia came from Romanized areas in the Eastern and or Central Balkans, and were forcibly relocated to Pannonia by the Avars, mainly to the area between the Danube and the Tisza that had been more or less depopulated in the previous period, acting as a no-man's land between Lombards to the west and Gepids to the east.

If many of these V13 came from cities like Viminacium, where the Avars were certainly active, this provides a nice explanation of why V13 branches in the Western Balkans appear close to those in Central Europe. This is assuming that indeed a lot of V13 got assimilated in proto-Albanians and moved west with them.

The V13 in Italy is a bit more puzzling to me. The samples from Collegno and Venosa (assuming the Venosa one resembles others from the site) seem very East-Med, like you would expect from local samples in this era. They show up in areas that up to today have elevated levels of V13. The autosomals seems to suggest they were already in Italy for a while, or they came from an area with a very comparable profile. The (current) lack of V13 in Iron age Italians on the other hand, suggest V13 did not arrive before the LIA. For what happened in between there are multiple options. I know many people dislike it, but Thracian slaves on Latifundia in Sicily and Southern Italy still seems like an option to me. V13 could also have come from Greek Colonies on the Black sea shores, although it seem odd it would lead to such percentages. A final possibility is fleeing Balkanites in the late Antiquity, although that seems to less fit the autosomals. Maybe a combination of all makes the most sense. But while some of it fits well for Sicily and Southern Italy, I have some more trouble seeing it work in Liguria.

For V13 in Greece I think the same must more or less apply. A combination of Thracian slaves, Hellenistic migration, and fleeing Balkanites in Late antiquity, with later addition of migrations from the Balkans by Arvanites and Vlachs.

[rafc]

Part 7:

In the first post I used the MJ-12 sample to illustrate that Babadag looked like Psenicevo + some Steppe admixture. I would like to expand on that point a bit in this post.

First on MJ-12 (radiocarbon dated 992-830 BC), she comes from the site that I marked in yellow on the below image, in case people wonder if she really is Babadag:



Then I put all the MBA/LBA and Iron age samples from the Ukraine and Moldavia on this graph:



In fact, the group marked "MLBA_STEPPE" contains all the samples from Bulgaria over Romania, Moldavia, Ukraine, into Russia up to the Volga and the Ural (generally dubbed Srubnaya). So it's quite impressive how homogenous this group is. In terms of Y-DNA it's almost all R1a. Equally impressive is that in the Iron age no samples plot there anymore, that whole profile vanished. Instead we see 4 clusters, and a cline inbetween that most likely represents a mix of the clusters.

Clusters 3 and 4 have the same profile as the Siberian Saka and Tagar groups, and so represent Cimmerian/Scythian newcomers from the east, with obviously more East-Asian ancestry than the MLBA_STEPPE group. For Cluster 2 I proposed they were a mix between a more BGR_IA/Kapitan Andreevo profile and some Steppe. This was probably the MLBA_STEPPE group, as MJ-12 still seems very low in East-Asian Ancestry (the LIA samples with the same profile have a bit more), and the oldest Steppe input seems to have been the Saka group, which seems like a less good fit for this mix.

The more mysterious part is Cluster 1, which had only sample in Glinoë, but appeared more often deeper in Ukraine. Just like the Glinoe they were culturally Scythian, even if not genetically. In Cluster 2 we have a V13 of course, which fits with a Southern origin in Bulgaria, while Cluster 1 has a R1b-U152/L2, which clearly points to Central Europe. Now, it is well known that at the end of the LBA the so called Gáva-Holigrady culture expanded eastwards into Transylvania, Moldavia and North-West Ukraine (you might have read one or two posts on that in this forum). Coincidentally, LBA samples from Hungary form a great fit to create cluster 1 by mixing with the MLBA_STEPPE group. So just like with the Stamped ware South to North scenario, we see a match between genetics and archeological data.

So there probably was a first movement of people coming from the west and mixing with local Steppe groups to create cluster 1 (say circa 1200-1100BC) and then around 1000BC a second movement from people coming from the south up to Dniester where they mixed with local Steppe groups (Coslogeni) to create cluster 2. At around the same time the first new Nomadic groups would have come from the east to the pontic steppe also (probably first cluster 3 and then cluster 4). And the cline represents the mix between the four clusters. Even though they all apparently had a Scythian culture, many of them were not so genetically, in fact none of the LIA Glinoe samples looked very "Scythian" (while the EIA Glinoe samples did). I put the proposed genesis of IA groups in Moldovia & Ukraine on this graph:



So this might also be a good clue as to how the elusive Gava ancestry looked.

[rafc]

Part 8:

One of the the "Eastern Clades" I would like to highlight is the FGC44169 group (4100bp on Yfull, 4078BP on FTDNA), of which the majority is S7461 (4100/4048), and under that BY5022 (4000/3817). The basal FGC44169 branch BY6357 is found in Bulgaria, and the Glinoë V13 was FGC44169+ and is positive for one BY6357 SNP out of two, but negative for all below.

Directly under the biggest branch, BY5022, we have a branch with an Iranian (FT215369) a scientific Turkish sample (FGC76600), a Saudi branch (BY112334), and a branch with an Armenian and many Western Europeans and a Czech (BY5084), and finally, the big FGC57496 group. Also BY5022+ was aDNA from Viminacium (R6756) who had a profile identical to the Kapitan Andreevo samples.

The FGC57496 branch then splits in FT186096 (Armenian + Irishman), BY30352 (Azerbaijani and Greek), BY160588 (Bulgarian and Croatian), BY5145 (Bulgarian and Iraqi), BY5465 (Bulgarian, Lebanese, Egyptian and aDNA from Kapitan Andreevo and Avar age Hungary) and FGC44154.

FGC44154 splits in BY55940 (Arabian), Y132363 (Western European) and FGC44177.

FGC44177 splits in V1444 (Druze from Lebanon), BY5050 (Lebanese-Syrian and English and aDNA from Avar age Hungary) and FGC44175 (a diverse group with many Italians)

So in terms of aDNA we have samples from this branch in Glinoë, Kapitan Andreevo, Avar age Hungary and a sample from Viminacium that looks just like the Kapitan Andreevo one. In terms of modern distribution it's clearly very diverse with a large presence in the Middle-East and Arabia, but only a very limited presence in the Western Balkans. This branch fits very well with the ones that might have been spread over the Eastern Balkans with Stamped ware, and were also part of Hellenistic and Roman mobility. The Avar age samples can likely be attributed to population displacements by the Avars.

After this post, Pribislav replied:

Great post rafc. Just wanted to add there's also sample I10946 from Himera, Sicily (780-400 BC) at FGC44169>BY6357 level. Negative for downstream clades A9723 and FT27407.

And I added:

Thanks, I forgot about that one. Since he is also in BY6357 he is "close" to the Glinoë sample. It's also somewhat odd that I10946 is autosomally so close to I10950, and yet has a very different subgroup of V13 (albeit also a rare one). Their presence as mercenaries in Sicily fits well with the idea that the Eastern clades spread far around due to their access to the Greek world via the Greek ports on the coast there. But the autosomal profile of those Himera samples remains enigmatic. They must come from somewhere more inland, given the lower ANF and higher WHG. Probably a very large area might match these autosomals, from the Carpathians to the Adriatic, and north up to Slovakia. But obviously not very southern. Given that they are both V13, the BY6357 that seems more eastern, and that many other Himera mercenaries appear to have Black sea ties, I'm most inclined to place them in the eastern part of that area, something like Transylvania. We don't have samples from that region yet, but I think logically they would fall between Moldavian and Hungarian profiles, so there should be some overlap with the I10946 and I10950.

That fits with what I proposed higher, that V13 in the Pannonian basin was the result of V13 being dragged along by Cimmerian/Scythian movements to central Europe. I think the same might have brought V13 to Transylvania from the east which fits well with BY6357 showing up in Glinoë and I10946. The autosomal profle is then the result of mixing with a local population.

[Southpaw]

Apparently those E-V13 in Late Chalcolithic Bulgaria were newcomers from somewhere in between Northern Carpathian - Ukrainian Steppe. High mountain herders. Very likely early spinoffs i would say. Very problematic to put a finger all in all.

[Riverman]

Apparently those E-V13 in Late Chalcolithic Bulgaria were newcomers from somewhere in between Northern Carpathian - Ukrainian Steppe. High mountain herders. Very likely early spinoffs i would say. Very problematic to put a finger all in all.

A group between Petresti and Tripolye-Cucuteni (subgroup). They mixed GAC groups first and Western steppe herders second.
Possibly they split with some moving South with Cernavoda-Usatovo related groups, while others moved West to form Cotofeni.
That there was a mixture zone of TCC, GAC and Western steppe groups from the corded decorated ceramic groups is well known, and their centre of gravity became Transylvania with first Decea Muresului and then Cotofeni.
If splinters moved South to Bulgaria, it just would make sense. But their centre remained in the Carpathians.

The main issue is always the cremation gap. We only get samples from splinters, assimilated people and those which switched to inhumation, like with Babadag, Mezocsat and Basarabi.

Basarabi is a safe bet, less safe is how it came up in the first place, but Oltenia, an evolution from Vartop (Channelled Ware with Encrusted Pottery substrate) is my current favourite.

[rafc]

Part 5:
The next part will focus on the post-Roman samples from Pannonia. Above the sample I20802 (3d century AD) was not mentioned. Even though it dates from the Roman period, it comes from outside the Roman empire. It’s described as Sarmatian. In the first century AD the Iazyges (often called Sarmatians, which is a term that also covers other tribes) had invaded Eastern Pannonia, where they remained in power for a few centuries. I20802 would suggest that a large part of these Iazyges were actually Dacians coming along, having a lot of V13.

We also have graves that date from the late Sarmatian age (end 4th, early 5th century), and have typical Sarmatian customs. They have mostly the same genetics as the 3d century one, and cluster quite close together. Out of 8 male samples, one is V13. There is also one female sample that plots close to the upper left of the “Thracian cline”. None of the samples show CHG ancestry, making any Roman genetic impact in this period very unlikely. My interpretation is that most of the Sarmatian samples represent a mix of descendants of Dacian migrants of the first century AD, higher Steppe Sarmatians and higher WHG Central-Europeans (locals?). The one female sample shows there was still new influx from the east, the paper mentions that this cemetery shows connections with the Chernyakhov culture.

After the Sarmatian graves we have actually a gap of no less than 150 years until the early Avar samples. Some of the Sarmatian graves still overlap with the reign of the Huns (c. 433-454), but the following period, where former Hunnic allies took over Pannonia, the Goths in the west and Gepids in the east, is a blank slate in the east. After the large gap from the later Sarmatian/Hunnic period to the Early Avar age, the genetics in Easter Pannonia have change considerably. The PCA below shows samples from the 7th century, but not the East-Asian ones. Now only 3 samples plot where the tight cluster of Sarmatians did: a V13-CTS9320, an R1a, and a female. It seems like most of the previous population disappeared or was deeply changed. The CTS9320 can probably be seen as a holdover from the Sarmatian era population.

If it was population change, then it was not caused by the very Germanic Gepids, as only four female samples plot very Germanic. Many samples now plot more Eastern shifted, and especially remarkable are several samples, nearly all V13, which look like they come from the Thracian cline but with slight CHG shift. This is pretty much what we can expect of the local profile in Romanized areas of the Eastern to Central Balkans in late Antiquity. So why do we find it here and now? We can also observe that while there is much V13 in that CHG shifted Thracian cline, there is nearly no V13 in the samples west of the Thracian cline.

The only explanation I see is that the Avars must have forcibly brought in new people. Maybe from an area like Viminacium, maybe from further away in Thracia proper. The Avars did supposedly take 270,000 captives with them after they unsuccessfully besieged Constantinople in 617. Maybe the other profiles represent profiles from Illyria and other neighboring areas?
The Avar-era paper itself had nothing useful to say on this issue, but the paper on Avar origins did have this:

In the early Avar Empire, the archaeological record demonstrates that the population of the Carpathian Basin remained quite heterogeneous, at least culturally. The late-Roman element was reinforced by captives from the Balkan provinces and elsewhere settled as dependent laborers.

I think that is indeed what happened. If we look at the samples from the 8th century, it’s clear that now V13 plots mainly west of the Thracian cline, where before the bulk of samples was already, so they lost their original profile through admixture with other populations in Pannonia, who were likely also brought in by the Avars, but from other areas (Slavs, Germans, Illyrians, ...).

This would suggest the bulk of the Central European V13 arrived in early Medieval times, as transplants from the Roman Balkans, while a small part were holdovers from earlier waves.

I took a look at the new paper on Avars and it mostly fits with the above. First an erratum: the previously published late Sarmatians did not have a V13, not sure how that error got in there.
Then for the new paper: the Sarmatian era samples plot closely with the ones previously published. There is a V13 in this new batch and as before, the Sarmatian era samples look like a mix between Sarmatian (Iazyges) newcomers and local survivors from the Iron Age. The V13's could be part of these survivors, or have come along with the Sarmatians (maybe both).

The Gepid era samples confirm the population break in Eastern Hungary I mentioned above, with the Avar era population looking very different than the Sarmatian era one. And if we ignore the real Avar samples, V13 is the most widespread haplogroup. I think that what I wrote above on the forced population relocations by the Avars is still the main reason for this presence of V13 in Avar age Hungary. Something like Viminacium could work as a source, except that the Avar era samples have none of the very East-Med profiles. Maybe the Avar era population was mainly coming from the countryside, while the very East-Med profiles would have mainly been urban (and also seem quite typical for Viminacium). The presence of 'Western clades' like CTS9320 and L241 also points to the central-Balkans. This could suggest that the rural population of the Central Balkans was mainly V13 (and that the Thracian profile was extending here, the L241 from the new avar batch is L241 for example).

[Riverman]

I took a look at the new paper on Avars and it mostly fits with the above. First an erratum: the previously published late Sarmatians did not have a V13, not sure how that error got in there.
Then for the new paper: the Sarmatian era samples plot closely with the ones previously published. There is a V13 in this new batch and as before, the Sarmatian era samples look like a mix between Sarmatian (Iazyges) newcomers and local survivors from the Iron Age. The V13's could be part of these survivors, or have come along with the Sarmatians (maybe both).

The Gepid era samples confirm the population break in Eastern Hungary I mentioned above, with the Avar era population looking very different than the Sarmatian era one. And if we ignore the real Avar samples, V13 is the most widespread haplogroup. I think that what I wrote above on the forced population relocations by the Avars is still the main reason for this presence of V13 in Avar age Hungary. Something like Viminacium could work as a source, except that the Avar era samples have none of the very East-Med profiles. Maybe the Avar era population was mainly coming from the countryside, while the very East-Med profiles would have mainly been urban (and also seem quite typical for Viminacium). The presence of 'Western clades' like CTS9320 and L241 also points to the central-Balkans. This could suggest that the rural population of the Central Balkans was mainly V13 (and that the Thracian profile was extending here, the L241 from the new avar batch is L241 for example).

The local populations along the Danube in Serbia and Bulgaria in particular are highly likely to have been dominated by E-V13 in the Roman period. But if talking about the Tisza-Transtisza zone, actual Daco-Romans and tribal Dacians come defintiely into play, especially for the Sarmatians.

If talking about the Scythians and Sarmatians, just like if talking about the Celts and Germanics, we need to distinguish different sites and the era from the regional groupings. Like during the Scythian and Vekerzug period it is absolutely sure that the local population was concentrated in the cremating Vekerzug people, especially those of the Sanislau group. Some of the Scythian cemetaries in Transylvania are definitely not local and intrusive, therefore the chance for having lots of E-V13 is low, whereas others are more mixed and these should yield some E-V13 from Gáva, Mezocsat-Late Gáva and Basarabi people in the region.

Even the E-V13 samples from the Western/Northern main branches along the Danube are not sure where they come from, they might very well have been the result of resettled Daco-Romans and Dacians to a large degree. I wouldn't wonder if even in the same Danubian region with the same dominance of E-V13 we will encounter different branches and subclades depending on the time and context these burials are from.

If the Northern E-V13 branches would have been much further to the South, their early (Iron Age) Westward and Northward distribution would be less efficient presumably and it would have been much more likely that they would have bled into the South and West Balkans. Both is not proven yet, the old Balkan samples are pretty specific and distinct.

Therefore I highly doubt they were much South of the Danube to begin with, but rather concentrated in areas like Banat, Oltenia, Transylvania, Transcarpathia, Eastern Slovakia, that zone which basically equals Western Romania and adjacent regions.

If you look at the map on the left, you can see the elevant Early Iron Age cultures which likely all got E-V13:
- Basarabi is the central group, mostly along the Danube
- Mezocsat is the Late Gáva-mixed population with Cimmerian influences and elites added
- Thracian tribes, basically post-Psenichevo in the South East
- Babadag to the East

https://pbs.twimg.com/media/F19jfq2WYAQh...me=900x900
https://twitter.com/AlbHistory/status/16...90/photo/2


There would be additional groups to the North and East, but those are the central ones which are all highly likely to have been rich in E-V13, and the epicentre being, very clearly, Basarabi in that time frame. The Basarabi zone being also, even in Bulgaria, characterised by North Thracian/Dacian place name endings from the written historical sources.

Just a reminder, Gáva, Belegis II-Gáva, Psenichevo and Basarabi, all 4 layers colonised in succeeding waves Moldova:



https://www.researchgate.net/profile/Aur...rpatic.png

On the various Channelled-Stamped Ware expansions into Moldova, starting with Gáva-Holigrady, Belegis iI-Gáva/Chisinau-Corlateni, Psenichevo-Babadag, Basarabi:

Early Holihrady Hallstatt culture in the regions
from the eastern Carpathians and from the Western Podolia
(cca 1200 - cca 800 (?) a. Chr.), related to the culture with
G?va grooved pottery, originating in the Carpathian basin,
forming all together the so-called block
cultural G?va-Holihrady-Grăniceşti (Свєшнiков
1964, 40 et seq .; Смирнова 1969, 7-33; idem 1976,
18 pcs .; idem 1990; Крушельницька 1990; idem
1993, 56 et seq .; Malev 1981; Крушельницкая,
Малеев 1990, 123 et seq.).

Early Hallstatt culture Chisinau-Corlateni
(c. 1200 - cca 900 BC) from the region of
The forest-steppe of the Dniester-Siret River is connected
of the culture with grooved pottery type Belegi? II
from the Middle Danube basin (L?szl? 1994; Leviţki
1994а; Sava, Levitsky 1995, 157 et seq.).

Tămăoani-Holercani cultural group with ceramics
engraved and polished (ca. 1100 - ca. 1000 BC),
known in the area between Siret and Dniester, in the area
of the confluence of the steppe with the forest-steppe, represents the periphery
north and northeast of the cultural block of
at the Lower Danube Babadag I-Tămăoani-Holercani
(and the Balta group on the left bank of the Dniester) (H?nsel
1976, 122 et seq .; L?szl? 1986, 65 et seq .; Levitski 1994b,
219 et seq .; Vanchovov 1993, 28-39; Nicic 2008).

Early Hallstattian culture Cozia-Saharna with
incised and stamped pottery (approx. 1000 - approx
800 BC) from the forest-steppe region of interfl
the Dniester-Siret River represents the eastern limit of
of the Eastern Balkan cultural bloc
P?eničevo II-Babadag II (H?nsel 1976, 134
et al .; L?szl? 1989, 111 et seq .; Kashuba 2000, 255 et seq .;
Niculiţă, Zanoci, Arnăut 2008).

Middle Hallstatt culture Soldanesti (c. 800
- cca 700 a. Chr.) 1, widespread in the southern region
of the Middle Dniester basin, is a component part
of the Basarabi cultural complex, which
originates in the Middle Danube area (Мелюкова
1958, 64-76; 1979апушнян 1979; Гольцева, Кашуба
1995, 32-37; Kaşuba 2008a, 37-50)

https://www.researchgate.net/publication...0_p_49-102

Therefore any Eastern steppe groups could have picked up lineages from all these 4 main waves either from West or East of the Carpathian mountain range. E-V13 was already close to the steppe anyway, since Gáva presumably, but latest with the colonisations of Psenichevo and Basarabi.