Evidence for dynastic succession among early Celtic elites

[rafc]

Using FTDNA frequencies for groups like V13 will lead to underestimation, as many V13s on FTDNA are only tested up to M35, and so not counted in the V13 frequency for a country. For M35 FTDNA gives a frequency of 6% in France, which is reliable and is an upper limit for V13. I think it's reasonable that V13 will be ~4% and other E-M35 ~2%. Given that E-M35 apart from V13 is mostly found outside of Europe, it's also likely nearly all arrived in France during the Roman era.

[GHurier]

Using FTDNA frequencies for groups like V13 will lead to underestimation, as many V13s on FTDNA are only tested up to M35, and so not counted in the V13 frequency for a country. For M35 FTDNA gives a frequency of 6% in France, which is reliable and is an upper limit for V13. I think it's reasonable that V13 will be ~4% and other E-M35 ~2%. Given that E-M35 apart from V13 is mostly found outside of Europe, it's also likely nearly all arrived in France during the Roman era.

From YFULL (counting only fully classified samples) :
12 E-V13
31 E-M35
Thus, closer to a 1/3 ratio than a 2/3.

I would say that ~(2 +/- 1)% of E-V13 seems realistic.

Considering the under-sampling of France for YFULL, I won't claim anything strong based on so few samples about clades cross-correaltion.
But, for E-M35 we clearly see the Roman-era signal. The E-M35 diversity injection in France is dominated by movements during the Roman-era.
For E-V13, there is not enough sample to get significant signal. But there is no Roman-era coupling involving French samples alone (All of the decoupling is achieved during IA). But if we add German samples (69 samples when adding French and German ones), then the Roman signal becomes clear, while some ~800 BCE decoupling also appear significant.
Therefore, most likely some E-V13 reach Germany at least with IA-diffusion. 

Emerging from likely cremating populations ... E-V13 ancient DNA sampling might also present a low-bias due to funerary-customs/Y-DNA correlations.
Also, most of actual presence of E-V13 in France seems to be either in South-Western part of France (not yet sampled during IA) or near the German border.

Anyway, the specific case of France would require more modern sampling to have a clear idea from a statistical analysis of the Y-DNA tree.

[Riverman]

Using FTDNA frequencies for groups like V13 will lead to underestimation, as many V13s on FTDNA are only tested up to M35, and so not counted in the V13 frequency for a country. For M35 FTDNA gives a frequency of 6% in France, which is reliable and is an upper limit for V13. I think it's reasonable that V13 will be ~4% and other E-M35 ~2%. Given that E-M35 apart from V13 is mostly found outside of Europe, it's also likely nearly all arrived in France during the Roman era.

From YFULL (counting only fully classified samples) :
12 E-V13
31 E-M35
Thus, closer to a 1/3 ratio than a 2/3.

I would say that ~(2 +/- 1)% of E-V13 seems realistic.

Considering the under-sampling of France for YFULL, I won't claim anything strong based on so few samples about clades cross-correaltion.
But, for E-M35 we clearly see the Roman-era signal. The E-M35 diversity injection in France is dominated by movements during the Roman-era.
For E-V13, there is not enough sample to get significant signal. But there is no Roman-era coupling involving French samples alone (All of the decoupling is achieved during IA). But if we add German samples (69 samples when adding French and German ones), then the Roman signal becomes clear, while some ~800 BCE decoupling also appear significant.
Therefore, most likely some E-V13 reach Germany at least with IA-diffusion. 

Emerging from likely cremating populations ... E-V13 ancient DNA sampling might also present a low-bias due to funerary-customs/Y-DNA correlations.
Also, most of actual presence of E-V13 in France seems to be either in South-Western part of France (not yet sampled during IA) or near the German border.

Anyway, the specific case of France would require more modern sampling to have a clear idea from a statistical analysis of the Y-DNA tree.

French E-V13 testers are almost as rare on YFull as German ones. FTDNA is more representative.
But it really looks like much of the French E-V13 is associated with Germanics and more Northern and Eastern Celts beside the Roman association.


That's another fun fact, that E-V13 is lower in the classical Gallo-Roman than in the Germanic-Celtic sphere.

A lot of the French E-V13 seem to be German or descend from Germanics. The hotspots are in Alsace-Lorraine. In Iberia too there is an association with Gothic-Suebian-Sarmatian-Frankish settlement as well, beside older layers.
The question is were those picked it up.

[corrigendum]

Using FTDNA frequencies for groups like V13 will lead to underestimation, as many V13s on FTDNA are only tested up to M35, and so not counted in the V13 frequency for a country. For M35 FTDNA gives a frequency of 6% in France, which is reliable and is an upper limit for V13. I think it's reasonable that V13 will be ~4% and other E-M35 ~2%. Given that E-M35 apart from V13 is mostly found outside of Europe, it's also likely nearly all arrived in France during the Roman era.

French E-V13 testers are almost as rare on YFull as German ones. FTDNA is more representative.
But it really looks like much of the French E-V13 is associated with Germanics and more Northern and Eastern Celts beside the Roman association.

I don't consider FTDNA to be as representative as YFULL because of low testing resolution and the fact that FTDNA's database is heavily skewed by family projects. In the case of E-V13 in France, this doesn't really matter because the difference of E-V13 % between the two projects is small: FTDNA 2%, YFULL 2.7%.

Such percentages are in line with the fact that E-V13 was very minor in the pre-Roman era as it hasn't still been found in France or areas near France in the IA, but starts to increase during and after the Roman era throughout western Europe.

As such, the idea that at least a modest part of E-V13 in France comes from Germanic tribes or northern/eastern Celts is untenable based on the archaeogenetic record. 

The majority of E-V13 very likely dates to the Roman era and a part of western European E-V13 likely comes from post-Roman migrations from south to north. Roman era E-V13 in western Europe should have been part of the local population for ~500 years by the end of the late antiquity, but even between 400-700 CE we get such E-V13 profiles near the Germany-France border (Alt-Inden, Westfalen) and the France-Italy border (Collegno, Piemonte):

Code:

Distance to: Germany_Late_Antiquity-Early_Medieval_Alt-Inden_009
0.02812736 Italian_Calabria
0.02935658 Greek_Dodecanese
0.03000592 Greek_Dodecanese_Rhodes
0.03054806 Greek_Kos
0.03175097 Italian_Campania
0.03208173 Greek_Crete_Lasithi
0.03298907 Sicilian_East
0.03326448 Italian_Jew
0.03343843 Romaniote_Jew

Target: Germany_Late_Antiquity-Early_Medieval_Alt-Inden_009
Distance: 2.7323% / 0.02732252
56.8 TUR_Barcin_N
13.6 Yamnaya_RUS_Samara
10.8 IRN_Ganj_Dareh_N
9.2 GEO_CHG
9.0 Israel_Natufian
0.6 Han

Code:

Distance to:    Italy_North_EarlyMedieval_Langobards_3:CL38
0.02699630    Greek_Kos
0.02791665    Greek_Dodecanese_Rhodes
0.02806734    Greek_Crete
0.02848391    Greek_Crete_Lasithi
0.02904599    Italian_Calabria
0.02908877    Greek_Dodecanese
0.03073418    Ashkenazi_Germany
0.03076571    Greek_Cyclades_Amorgos
0.03084933    Italian_Campania
0.03091443    Greek_Crete_Heraklion
0.03134585    Romaniote_Jew

Target: Italy_North_EarlyMedieval_Langobards_3:CL38
Distance: 2.8039% / 0.02803927
50.2    TUR_Barcin_N
16.6    Yamnaya_RUS_Samara
13.0    GEO_CHG
11.8    Israel_Natufian
8.2    IRN_Ganj_Dareh_N
0.2    Han

[GHurier]

French E-V13 testers are almost as rare on YFull as German ones. FTDNA is more representative.

We completely agree about FTDNA data being way better for France/Germany sector.
Sadly, when I compute clade-correlation, I need, in an easily usable shape :
1. The total Y-tree (clades and TMRCAs)
2. All samples and locations
If I can get that by downloading ~20 webpage for YFULL ... I still have to figure out how to retrieve such information from FTDNA in a reasonable way.

Thus, up to now I 'm playing with the YFULL samples ... even if indeed, using FTDNA tree would be way better for France.

[Riverman]

French E-V13 testers are almost as rare on YFull as German ones. FTDNA is more representative.

We completely agree about FTDNA data being way better for France/Germany sector.
Sadly, when I compute clade-correlation, I need, in an easily usable shape :
1. The total Y-tree (clades and TMRCAs)
2. All samples and locations
If I can get that by downloading ~20 webpage for YFULL ... I still have to figure out how to retrieve such information from FTDNA in a reasonable way.

Thus, up to now I 'm playing with the YFULL samples ... even if indeed, using FTDNA tree would be way better for France.

You should ask how the creator of this site did it, because he seems to use FTDNA data:
http://scaledinnovation.com/gg/gg.html?nm=tools

[Riverman]

I don't consider FTDNA to be as representative as YFULL because of low testing resolution and the fact that FTDNA's database is heavily skewed by family projects. In the case of E-V13 in France, this doesn't really matter because the difference of E-V13 % between the two projects is small: FTDNA 2%, YFULL 2.7%.

On FTDNA all the E-V13 which didn't test at a higher resolution than E-M35 will end up there. The true frequency in many countries, even more if subtracting Ashkenazi Jewish testers, is way higher than the raw numbers suggest.

FTDNA is clearly way more representative for Germans than YFull, there can be no doubt about that. Absolutely not. Whole major branches are missing, especially for E-V13, at YFull, many of which have high, ancient diversity and many testers at FTDNA. This means a lot, because Germans are severely underrepresented on both sites, its just worse on YFull.

What this suggests is that we deal with many E-V13 branches which had early, major founding and diversification events in Germany or close to Germany. Like additional hotspots are clearly in Hungary, Czechia, Slovakia, Poland - those are all interconnected, much closer than they are to the Balkans South of the Danube, which is quite telling, especially since the Southern Balkan/Albanians in particular are so well-tested.

Here some numbers to digest from FTDNA:
E-M35: Germany (7), France (5)
E-V13: German (3), France (2)

Especially for Germany you can be sure that not 4 percent are "other E-M35" than E-V13. That's just not real. At the same time we can see that countries like Belorussia have far too much E-M35 because of the overrepresented Aschkenazi vs. local testers. This plays into Germany and France as well, but obviously less so than in Belorussia, Ukraine and Romania for example.

Therefore the frequencies are not so much better at FTDNA for some countries, mainly because of the level of testing and the relatively high percentage of Ashkenazi testers, but the representation within the E-V13 (or other) branches is way better, especially Germans and French.

[jdbreazeale]

French E-V13 testers are almost as rare on YFull as German ones. FTDNA is more representative.

We completely agree about FTDNA data being way better for France/Germany sector.
Sadly, when I compute clade-correlation, I need, in an easily usable shape :
1. The total Y-tree (clades and TMRCAs)
2. All samples and locations
If I can get that by downloading ~20 webpage for YFULL ... I still have to figure out how to retrieve such information from FTDNA in a reasonable way.

Thus, up to now I 'm playing with the YFULL samples ... even if indeed, using FTDNA tree would be way better for France.

You should ask how the creator of this site did it, because he seems to use FTDNA data:
http://scaledinnovation.com/gg/gg.html?nm=tools

That's Robin Spencer <[email protected]>.  He's a very active co-admin of the FTDNA "England GB Groups EIJ" project.

[Hodo Scariti]

All the other males are G2a too?

[La Tene]

I don't know much about LAN001 but:
I17607 800-550 BCE Louny, Stradonice, Czech Republic P312>DF19>DF88>FGC11833>S4281  Czech_IA_Hallstatt C or D Hallstatt

Interestingly, all those Central European IA groups have their own northern outlier - Hallstatt and la Tène alike.

The cores of those groups are essentially Gaulish-like, but :
Czech IA La Tène I120509 plots "north of the Arctic circle" on PCAs. France Grand Est COL11 is surrounded by Swedes.
Czech Hallstatt Bylany DA112 is Dutch-like, while DA111 lands half way between Brittany and the Basque country.
In that respect, the Slovakia IA Vekerzug group (not Gaulish-like, for that matter) is a caricatural case, with people plotting among Lithuanians, and others plotting in Lazio.

I wonder if this is due to the fact that those groups were still "in the making", sort of, awaiting genetic homogenization, or whether those outliers were traders or mercenaries who died a long way from where they belonged.

More selfishly, I am very much interested in the Gaulish-like looks of those German, Czech, Austrian, even Slovenian IA samples. Two years ago, Claire Elise Fischer's Gaulish paper concluded that they could detect no genetic discontinuity in Iron Age France. I pointed out on a French forum that given the genetic similarity in Western and Central Europe at the time, migrations might pass undetected. C.E. Fischer was kind enough to answer and explain that, although several scenarios could be considered, they had prudently opted for "cultural diffusion" rather than demic movement on the grounds that archaeologists preferred that option. Problem is, French archaeologists seem to have a(n) (ideological) problem with whatever concerns migrations of IE-speaking groups or sub-groups. It reeks of nazi propaganda to their eyes and is currently politically incorrect.

Still, for the sake of my own private identity quest, I can't help wondering when I hit on maps like this one :

Im not skilled with the genetic calculations, but with regard to proto celts, isn't it easy by now to analyse/compare BA france and BA central europe? I get that movements in the Hallstatt/LA Tene era could make the central Europeans "Gaulish like" (or vice versa), but my question is, were the central europeans "gaulish like" in the BA? Because is if not, and they are "Gaulish like" in the IA, then wouldn't that imply west-to-east direction of gene flow? After all, its seems almost everybody here has abandoned the "Celtic from the East" theory. So ergo, how the hell could central europeans in the BA be "Gaulish" or "Celtic" like?

I had been in under the impression that the Tumulus and Hallstatt cultures were intrusive "central european" cultures that imposed themselves on the native Gaulish populations in the centre-eastern france-rhineland.

while DA111 lands half way between Brittany and the Basque country

Could this be some kind of west french profile?

[Andour]

while DA111 lands half way between Brittany and the Basque country

Could this be some kind of west french profile?

I posted a Vahaduo PCA some way up this thread (https://genarchivist.com/showthread.php?...6#pid12746) with the new Hallstatt samples plus the geographically close Lech Valley Bell Beaker and Early Bronze Age samples - plus a few Gaulish Iron Age groups. If you look at it, you'll notice a few things :
- The Lech Valley Bell Beaker samples are extremely heterogeneous, as if the fusion between incoming and local populations was incomplete yet.
- The Lech Valley Early Bronze Age samples are entirely contained within the area occupied by the new SW German Hallstatt Celtic samples. Which means these Hallstatt people were similar to their Bronze Age predecessors : no movement from the West. No increase in WHG.
- France IA Grand Est overlaps with both Lech BB and Lech EBA groups. Although there's a slight pull to the West, those pops were quite similar.
- Both France IA South-East and IA South-West (Occitanie) land outside the Hallstatt sphere, and are much more Basque-like.

Now if you add modern Auvergne samples to this PCA, you'll find it lands right in the middle of the West German Hallsatt Celtic group. Logically, given the geographic position of the Auvergne, if there hadn't been any movement from the East at some time in the Hallstatt period, the Auvergne samples should end up about half-way between the SE and SW French_IA samples. Such is not the case. There is no pull to the south-east (which might be explained by Roman veterans settling here), no pull to the north (Frankish or Burgundian). There's a pull to the east...

Czech, Hungarian, and Slovenian La Tène samples are very much Gaulish-like. I am prepared to concede that this may be due to movements from the West. But these recent German Hallstatt samples take us one step further back in time, before the La Tène eastward move. So I'll reiterate my views here that when archeologists dig into it deeper and cooperate more closely with geneticists, they may well have to give up the "no discontinuity" rant.

I was reading a fascinating paper recently on the early Iron Age in France. Interestingly, it developed like this : In the 10th and 9th centuries BC...
- plain settlements are abandoned.
- hillforts are erected everywhere.
- notable changes in ceramics appear.
- new weapons are also found.
- exchange networks are re-oriented.
--- Conclusion of the author : we do not observe any cultural discontinuity... !!

I don't know if DA111 moved to Bohemia from the West. I honestly think he was just a local who for some reaon happened to inherit more than his fair share of WHG. There might have been a residual pocket of outliers around (a bit like those Hassleben Gauls in the midst of Germans). But I am increasingly convinced that Early Hallstatt witnessed a massive move from SW Germany into Central East France, which the subsequent "conquests" (Roman or Germanic) failed to obliterate.

[La Tene]

while DA111 lands half way between Brittany and the Basque country

Could this be some kind of west french profile?

I posted a Vahaduo PCA some way up this thread (https://genarchivist.com/showthread.php?...6#pid12746) with the new Hallstatt samples plus the geographically close Lech Valley Bell Beaker and Early Bronze Age samples - plus a few Gaulish Iron Age groups. If you look at it, you'll notice a few things :
- The Lech Valley Bell Beaker samples are extremely heterogeneous, as if the fusion between incoming and local populations was incomplete yet.
- The Lech Valley Early Bronze Age samples are entirely contained within the area occupied by the new SW German Hallstatt Celtic samples. Which means these Hallstatt people were similar to their Bronze Age predecessors : no movement from the West. No increase in WHG.
- France IA Grand Est overlaps with both Lech BB and Lech EBA groups. Although there's a slight pull to the West, those pops were quite similar.
- Both France IA South-East and IA South-West (Occitanie) land outside the Hallstatt sphere, and are much more Basque-like.

Now if you add modern Auvergne samples to this PCA, you'll find it lands right in the middle of the West German Hallsatt Celtic group. Logically, given the geographic position of the Auvergne, if there hadn't been any movement from the East at some time in the Hallstatt period, the Auvergne samples should end up about half-way between the SE and SW French_IA samples. Such is not the case. There is no pull to the south-east (which might be explained by Roman veterans settling here), no pull to the north (Frankish or Burgundian). There's a pull to the east...

Czech, Hungarian, and Slovenian La Tène samples are very much Gaulish-like. I am prepared to concede that this may be due to movements from the West. But these recent German Hallstatt samples take us one step further back in time, before the La Tène eastward move. So I'll reiterate my views here that when archeologists dig into it deeper and cooperate more closely with geneticists, they may well have to give up the "no discontinuity" rant.

I was reading a fascinating paper recently on the early Iron Age in France. Interestingly, it developed like this : In the 10th and 9th centuries BC...
- plain settlements are abandoned.
- hillforts are erected everywhere.
- notable changes in ceramics appear.
- new weapons are also found.
- exchange networks are re-oriented.
--- Conclusion of the author : we do not observe any cultural discontinuity... !!

I don't know if DA111 moved to Bohemia from the West. I honestly think he was just a local who for some reaon happened to inherit more than his fair share of WHG. There might have been a residual pocket of outliers around (a bit like those Hassleben Gauls in the midst of Germans). But I am increasingly convinced that Early Hallstatt witnessed a massive move from SW Germany into Central East France, which the subsequent "conquests" (Roman or Germanic) failed to obliterate.

Do the non outlier samples from Bucy le long and Parançot also overlap with Baden Wurttemberg Hallstatt samples on the PCA?

[Andour]

I can't retrieve the G25 coords for Parançot (not sure I ever had them). Would be glad to have them, by the way.

Below is a PCA with the German Hallstatt samples, the IA samples from Hauts de France (I mean, those I have in my lists), Alsatian samples from IA1 and IA2.
The Hauts-de-France sample in green is the Bucy-le-long sample (BFM265). He lies at the fringe of the German samples. We should bear in mind that the Hallstattian zone of influence in France never really concerned the Hauts-de-France, which seems to have had its own specific history.
The sample labelled Gaul is an Alsatian sample dated around 500 BC - still rather distant from the IA2 Alsatian samples.
The samples labelled France_GrandEst_IA are dated from 675 BC to 565 BC. Look where they stand.
Noteworthy is the shift to the east of the Alsatian samples between IA1 and IA2.

[La Tene]

I can't retrieve the G25 coords for Parançot (not sure I ever had them). Would be glad to have them, by the way.

Below is a PCA with the German Hallstatt samples, the IA samples from Hauts de France (I mean, those I have in my lists), Alsatian samples from IA1 and IA2.
The Hauts-de-France sample in green is the Bucy-le-long sample (BFM265). He lies at the fringe of the German samples. We should bear in mind that the Hallstattian zone of influence  in France never really concerned the Hauts-de-France, which seems to have had its own specific history.
The sample labelled Gaul is an Alsatian sample dated around 500 BC - still rather distant from the IA2 Alsatian samples.
The samples labelled France_GrandEst_IA are dated from 675 BC to 565 BC. Look where they stand.
Noteworthy is the shift to the east of the Alsatian samples between IA1 and IA2.

Oh I think I made a mistake, your PCA does not include samples from sites listed in the supplementary  Archaeological Supplementary Material for Western Eurasia (S1B) part of the paper  Steppe Ancestry in western Eurasia and the spread of the Germanic Languages because its a pre print right? So the new samples could not be analysed yet, right? So the Hauts de France, France Grand Est IA, Alsace IA1, IA2 samples are from the previous publications about genetic history of France?

Samples from sites like Durnberg, Halltatt in Austria
Camp du Château,  Champ Peupin,  Moidons and Parançot,  Leuglay, tumulus des Montagnottes,  Maisey-le-Duc,  Nod-sur-Seine,  Pothières,  Sainte Colombe-sur-Seine,  Tumulus de La Forêt de Châtillon,  Vix mont Lassois in Gaul
Saarland, Rubenheim

[poilus]

I can't retrieve the G25 coords for Parançot (not sure I ever had them). Would be glad to have them, by the way.
Below is a PCA with the German Hallstatt samples, the IA samples from Hauts de France (I mean, those I have in my lists), Alsatian samples from IA1 and IA2.
The Hauts-de-France sample in green is the Bucy-le-long sample (BFM265). He lies at the fringe of the German samples. We should bear in mind that the Hallstattian zone of influence  in France never really concerned the Hauts-de-France, which seems to have had its own specific history.
The sample labelled Gaul is an Alsatian sample dated around 500 BC - still rather distant from the IA2 Alsatian samples.
The samples labelled France_GrandEst_IA are dated from 675 BC to 565 BC. Look where they stand.
Noteworthy is the shift to the east of the Alsatian samples between IA1 and IA2.

Oh I think I made a mistake, your PCA does not include samples from sites listed in the supplementary  Archaeological Supplementary Material for Western Eurasia (S1B) part of the paper  Steppe Ancestry in western Eurasia and the spread of the Germanic Languages because its a pre print right? So the new samples could not be analysed yet, right? So the Hauts de France, France Grand Est IA, Alsace IA1, IA2 samples are from the previous publications about genetic history of France?
Samples from sites like Durnberg, Halltatt in Austria
Camp du Château,  Champ Peupin,  Moidons and Parançot,  Leuglay, tumulus des Montagnottes,  Maisey-le-Duc,  Nod-sur-Seine,  Pothières,  Sainte Colombe-sur-Seine,  Tumulus de La Forêt de Châtillon,  Vix mont Lassois in Gaul
Saarland, Rubenheim

I do not know if the PCA includes all the IA Alsatian samples available, but 40 AI1 genomes from Duttlmenheim should be published, the sonner the better. 
 [img][/img]
 
o