Steppe Ancestry in western Eurasia and the spread of the Germanic Languages

[Capsian20]

Any news about data this study

[Anglesqueville]

No, no news about the data. I'm angry that we may have to wait for final publication. I am very impatient to be able to get my hands on the mass of new North Norwegian specimens. For my part, don't think that I'm inactive. If I don't publish anything, it's because I don't find anything very convincing on the only question that concerns me, that of the formation of the Eastern Scandinavian cluster. For days I have been trying to converge traditional methods (qpAdm, PCA, MDS) and IBD methods (I tried with non-phased data and their choice of software, and I tried with phased data and refinedIBD), but I remain in the fog.

[Capsian20]

Any news about data this study

[Anglesqueville]

For weeks I have been trying to reproduce the IBDs statistics from McColl et al. for the "early Scandinavian" and "east Scandinavian" clusters, taking as representatives those of its individuals which were published after imputation by Allentoft. Since I'm using imputed data (via GLIMPSE) by pros, I think I can have confidence in this data. Furthermore, I use the same IBDs research software as McColl (ibdseq, from Browning and Browning https://faculty.washington.edu/browning/ibdseq.html). Since McColl doesn't say anything about configuring this software, I first used the default settings, then tried others. It turns out that my results converge very poorly with McColl's conclusions regarding the constitution of this famous East-Scand cluster, in particular I do not see a clear precedence of Latvian_HG over the Scandinavian HGs emerging. I imagine that if I operated like the Eurogenious people I would loudly proclaim that I "debunked" McColl. I consider it more reasonable to think that something is wrong with my procedure. A little tired of all this I came back to the good old qpAdm, and looked if a group of HGs in addition to the left list including Latvian_LN and Poland_GAC would allow me to:
1) obtain reasonable adjustments
2) clearly separate the Mc Coll clusters.

I tried almost everything until finally, without really believing it, I tried sf11, the I1 from Stora Förvar (Gotland). Here are the results. I only indicate the best models, that is to say that when the only sources are Latvian_LN+Poland_GAC, it means that the complete model was infeasible, or that the coefficient of sf11 was too small to be significant (precisely lower than the standard error with a p-value for the alternative model much higher than 0.05).

Cluster early Scandia


Sweden_Neolithic_BAC.imputed_allentoft.ber1
Latvia_LN
poland_gac.imputed_Allentoft

best coefficients:    0.602    0.398
totmean:      0.602    0.398
boot mean:    0.602    0.398
      std. errors:    0.046    0.046

fixed pat  wt  dof    chisq      tail prob
          00  0    11    8.456        0.671999    0.602    0.398
   
Sweden_Neolithic_BAC.imputed_allentoft.ber2
Latvia_LN
poland_gac.imputed_Allentoft

best coefficients:    0.716    0.284
totmean:      0.716    0.284
boot mean:    0.717    0.283
      std. errors:    0.046    0.046

fixed pat  wt  dof    chisq      tail prob
          00  0    11    7.639        0.745223    0.716    0.284
 
 
Sweden_Neolithic_BAC.imputed_allentoft.RISE94
Latvia_LN
poland_gac.imputed_Allentoft

best coefficients:    0.650    0.350
totmean:      0.650    0.350
boot mean:    0.651    0.349
      std. errors:    0.047    0.047

  fixed pat  wt  dof    chisq      tail prob
          00  0    11    15.630        0.155451    0.650    0.350
 
Sweden_Neolithic.ial.NEO51
Latvia_LN
poland_gac.imputed_Allentoft

best coefficients:    0.582    0.418
totmean:      0.582    0.418
boot mean:    0.582    0.418
      std. errors:    0.042    0.042
     
fixed pat  wt  dof    chisq      tail prob
          00  0    11    14.244        0.219768    0.582    0.418
   
   
Cluster East Scandia

Sweden-LN-NEO261
Latvia_LN
poland_gac.imputed_Allentoft
sweden_mesolithic.ial.sf11

best coefficients:    0.406    0.433    0.161
totmean:      0.406    0.433    0.161
boot mean:    0.408    0.433    0.159
      std. errors:    0.076    0.053    0.092

fixed pat  wt  dof    chisq      tail prob
          000  0    10    6.039        0.812018    0.406    0.433    0.161
          001  1    11    9.421        0.583058    0.519    0.481    0.000
 
 
Sweden_N_Falköping.imputed_Allentoft_NEO220
Latvia_LN
poland_gac.imputed_Allentoft
sweden_mesolithic.ial.sf11

best coefficients:    0.482    0.444    0.073
totmean:      0.482    0.444    0.073
boot mean:    0.485    0.444    0.071
      std. errors:    0.079    0.052    0.095

fixed pat  wt  dof    chisq      tail prob
          000  0    10    13.035        0.221747    0.482    0.444    0.073
          001  1    11    13.763        0.246402    0.535    0.465    0.000


Sweden_N_Falköping.imputed_Allentoft_NEO225
Latvia_LN
poland_gac.imputed_Allentoft
sweden_mesolithic.ial.sf11

best coefficients:    0.416    0.382    0.202
totmean:      0.416    0.382    0.202
boot mean:    0.418    0.382    0.200
      std. errors:    0.076    0.052    0.096


  fixed pat  wt  dof    chisq      tail prob
          000  0    10    11.642        0.309704    0.416    0.382    0.202
          001  1    11    16.755        0.115321    0.557    0.443    0.000
   
   
Sweden_N_Falköping.imputed_Allentoft_NEO227
Latvia_LN
poland_gac.imputed_Allentoft
sweden_mesolithic.ial.sf11

best coefficients:    0.324    0.468    0.208
totmean:      0.324    0.468    0.208
boot mean:    0.325    0.468    0.208
      std. errors:    0.071    0.049    0.088


fixed pat  wt  dof    chisq      tail prob
          000  0    10    12.441        0.256644    0.324    0.468    0.208
          001  1    11    18.862      0.0636133    0.467    0.533    0.000
     
     
Sweden_N_Falköping.imputed_Allentoft_NEO224
Latvia_LN
poland_gac.imputed_Allentoft
sweden_mesolithic.ial.sf11

best coefficients:    0.405    0.445    0.150
totmean:      0.405    0.445    0.150
boot mean:    0.407    0.445    0.148
      std. errors:    0.074    0.051    0.089

fixed pat  wt  dof    chisq      tail prob
          000  0    10    5.074        0.886175    0.405    0.445    0.150
          001  1    11    8.186        0.696546    0.515    0.485    0.000
   
   
Sweden_LateNeolithic.ial.mc.oll010
Latvia_LN
poland_gac.imputed_Allentoft
sweden_mesolithic.ial.sf11


best coefficients:    0.364    0.491    0.145
totmean:      0.364    0.491    0.145
boot mean:    0.365    0.491    0.144
      std. errors:    0.073    0.049    0.085
     
fixed pat  wt  dof    chisq      tail prob
          000  0    10    7.594        0.668444    0.364    0.491    0.145
          001  1    11    10.906        0.451164    0.468    0.532    0.000
   
   
Sweden_LateNeolithic.ial.mc.oll009
Latvia_LN
poland_gac.imputed_Allentoft
sweden_mesolithic.ial.sf11

best coefficients:    0.316    0.376    0.308
totmean:      0.316    0.376    0.308
boot mean:    0.317    0.376    0.307
      std. errors:    0.072    0.052    0.091


fixed pat  wt  dof    chisq      tail prob
          000  0    10    3.696        0.960036    0.316    0.376    0.308
          001  1    11    15.878        0.145726    0.528    0.472    0.000
 
  (for this last case the p-value of the model without sf11 is 0.00048!)

The case of NEO220 is uncertain. But that of oll009 is mind-blowing. I don't know what will happen to the hypothesis of trans-Baltic migration resulting in this East_Scandinavian cluster. But as for the existence of this cluster, I would be surprised if the publication of McColl's exclusive data (I am thinking in particular of the Norwegians) calls it into question. Because this existence really seems incontestable to me.

[Angantyr]

I tried almost everything until finally, without really believing it, I tried sf11, the I1 from Stora Förvar (Gotland). Here are the results. I only indicate the best models, that is to say that when the only sources are Latvian_LN+Poland_GAC, it means that the complete model was infeasible, or that the coefficient of sf11 was too small to be significant (precisely lower than the standard error with a p-value for the alternative model much higher than 0.05).

Did you also try the PWC Neolithic HGs?

[Anglesqueville]

I tried almost everything until finally, without really believing it, I tried sf11, the I1 from Stora Förvar (Gotland). Here are the results. I only indicate the best models, that is to say that when the only sources are Latvian_LN+Poland_GAC, it means that the complete model was infeasible, or that the coefficient of sf11 was too small to be significant (precisely lower than the standard error with a p-value for the alternative model much higher than 0.05).

Did you also try the PWC Neolithic HGs?

Of course, I did. The results are among the worst.

[Angantyr]

I tried almost everything until finally, without really believing it, I tried sf11, the I1 from Stora Förvar (Gotland). Here are the results. I only indicate the best models, that is to say that when the only sources are Latvian_LN+Poland_GAC, it means that the complete model was infeasible, or that the coefficient of sf11 was too small to be significant (precisely lower than the standard error with a p-value for the alternative model much higher than 0.05).

Did you also try the PWC Neolithic HGs?

Of course, I did. The results are among the worst.

So worse than the other (i.e. non-SF11) SHGs? That is interesting, could it be their TRB ancestry that is wrong? Or some very strong drift?
By the way, does this go for both the Gotland PWCs and the assumed PWC individuals from Denmark, NEO33 and NEO898?

SF11 does stand out among SHGs in PCAs and such, and I have assumed that this is because he is very low coverage and also contaminated (see for instance Günther et al. 2018). Is there a risk that the imputation goes bananas on him for these reasons?

[Anglesqueville]

Did you also try the PWC Neolithic HGs?

Of course, I did. The results are among the worst.

So worse than the other (i.e. non-SF11) SHGs? That is interesting, could it be their TRB ancestry that is wrong? Or some very strong drift?
By the way, does this go for both the Gotland PWCs and the assumed PWC individuals from Denmark, NEO33 and NEO898?

SF11 does stand out among SHGs in PCAs and such, and I have assumed that this is because he is very low coverage and also contaminated (see for instance Günther et al. 2018). Is there a risk that the imputation goes bananas on him for these reasons?

I checked only Gotlands PWCs. About sf11 and imputed bananas, honestly, I don't know.

[La Tene]

Interesting I try to grasp what the consequences would be for the NW Block.

Initial these area contained the nodus of the BB of NW Europe and spread all over the place. So let's assume they spoke Italo-Celtic.

Do you have any idea when the BB network collapsed and the Italo-Celtic dialect began to diversify?

How reasonable is it to assume that out of NW block the North Sea Celtic of Schrijver emerged in IA?

maybe you could outline the bronze and iron age in the ‘eastern north sea’ zone for us as you seem knowledgable. It started with Rhenish type beaker (briefly) spreading as far as Denmark. Then you had barbed wire? Then Elp? Then Harpstedt?? Think i’m missing a lot of detail

Yes and you can add Sögel-Wohlde to the list as well as Urkeltentum used by mid 20th century German archeologist for what now is East North Sea Cluster   So among al those labels and frames where to begin?

In essence I see the eclips (group 1) as the NW block (aka Sögel Wohlde aka Elp etc etc etc etc):



And than I immediately correct you:
"NW Block imo might have done its own thing largely because it wasn’t a powerful core of a major network but instead lay on the periphery of the Atlantic, Nordic and C European networks who's powerhouses were elsewhere."

It was not a backward alley, it was the interface, so where Atlantic, Nordic and C European networks met!!!

I already showed that in these area there is from BB unto migration ages in some parts a big continuity in population, in settlements. I mentioned Eelde (just beneath the city of Groningen) in which in the deepest layer of a settlement a very rich BB grave is found and in which the settlement was just left about 400 AD, with the incoming Danish IA Germanics.

In genetic sense this area contains the core of the Bell Beakers, the ones who spread all over the place, and especially to the Isles (ggggg father alanarchae).


They were a mix of Single Grave and TRB West. See the paper which mentions high HG in the ENS cluster. Probably due to the high Ertebølle amount in TRB West (See Karsten Wentink's dissertation 2006, mentioned about 100 times   Imo ENS cluster represents the genetic continuity of these BB unto migration times!

In linguistic sense this NW Block is most probably following your own description:
"Italo-Celtic does have an extraordinary correlation with the populations that retained their bell beaker geveratic right until the light of history shone. My own view is that that whole  zone was likely covered in a myriad of Celto-Italic, para Celtic and para Italic dialects and it was relatively easy for them at elite level to also use a lingua Franca unified ‘high register’ dialect across vast areas as the importance of networking and power of the elites gradually increased through the bronze age."

And of course due to their position we can expect diverse cultural influences, from Nordic Bronze Age (that tipped probably Drenthe), from the Unetice/ Tumulus culture of Central Europe.

As archeologist as VandKilde state Sögel Wohlde (NW block) initial influenced the Valsømagle area (South Scandic). The fuze of both cultures made up in cultural sense the NBA. This NBA on his turn usurped the NE part of the NW block (unto Lüneberger heath etc). Maybe also in language development.

I guess that Schrijver (2017) has a point that the western part of the NW Block- for example the Frisii- staid in a much more Celtic groove. Until 400 AD with the influx of Danish IA ancestry:"the final take over of the whole NW Block by the Nordic Bronze Age heirs."

Except for the South of the Netherlands, see this only as indicative please:

Have any ENS samples been located outside of the Netherlands? More restricted cultures like Hilversum and Lower Rhine urnfield could perhaps be more of a safe bet for the core homeland of the ENS group. Unless there are more North French samples, and given that the "Belgae" supposedly start from the Seine if I'm not mistaken, I think there is a huge enough area between the Seine the and Meuse/Scheldt for a "gaulish colonised" Belgae zone, not necessarily including the ENS group core territories, (perhaps over running "Briton like" peoples of post MMDN- Deverel-Rimbury traditions of north coastal france). But if we got more ENS like samples from Belgica that could change the picture a bit, but they will probably be on an increasing gradient from south to the north.

East of the netherlands. Maybe the Germanic experts can chime in, but where exactly do the Rhine-Wesser Germanics of the IA come from? I gather originally by MIA-LIA they were located around the Wesser/Aller river basins, later parts of the Rhine, I dont know about the coastal areas. Are they local descendants of the Elp or Ems-Wesser cultures that got germanized/admixture, or direct intruders from the Elbe-Suebi zone, or a mix? Maybe different RW germanics have different ancestry from different pre germanic? Urnfield/EIA groups, such as lower Hesse mixed group?

[Ambiorix]

In the time since the release of the preprint, I have been carefully (re-)examining the supplemental material for the paper with a focus on the following: a) the Eastern North Sea (ENS) IBD cluster and its spread in ancient samples; my own haplogroup (R-DF19) and its correlation with not only the spread of R-S263, but with the spread of ENS more broadly. For my purposes, a large portion of what I am posting will contain information I shared with the R DF19 group at FamilyTreeDNA, but elaborated on with a narrower focus on the ENS cluster, how we might identify its spread with associated patrilineages, and some interesting migration vectors that can be gleaned from some of the ENS-enriched samples in the McColl et al. (2024) dataset (including samples that have so far not been discussed). All of this is of course my own interpretation of the data - thanks to @La Tene for the reminder!  

As a reminder, the authors clarify the characteristic features of the Eastern North Sea IBD cluster Supplementary Note S6.7.2; this group is notable not only because of differentiation from other (largely) Bell Beaker-derived groups through the inability to primarily be defined by BB ancestry, but also due to its exceptional continuity in relevant areas of the Netherlands even well into the Roman period – however, McColl et al. (2024)  emphasize the marginal contribution of this ancestry source to the future Frisians:

“S6.7.2 The Netherlands The Bell Beaker sub-cluster 0_2_1_2 WEuIsMl located primarily from the Eastern North Sea (ENS) region (present day the Netherlands) is unique in its high NWHG ancestry, low European Farmer, and inability to be modelled primarily as Bell Beaker ancestry, like most others from Bell Beaker sub-clusters (Figure S6.3.6).

When using early representatives of this cluster as a source, we see a large degree of genetic continuity from 3700 - 1700 BP. From Valkenburg (ZH) however, there are a number of individuals that do not fit the profile. The Roman cemetery Valkenburg Marktveld, located south of the auxiliary fort, was used between 50 – 300 CE for the entire military community that consisted of men, women and children, who lived in the vicinity of the auxiliary fort. Over 650 individuals were recovered, 145 of which are inhumations (41 adults, 104 children and 1529 infants); an extraordinary number as cremation dominates the Roman burial record in the Netherlands. The individuals included in this study are possibly associated with different departments of the Roman army, so the presence of non-local individuals is not unexpected (DeCoster et al., in prep.). For the individuals that do not fit the local profile, most are Celtic (similar to contemporaneous people from British Isles or France and one is similar to people deriving their ancestry from the Bronze Age Eastern Mediterranean. A single individual is modelled with North East European ancestry.

A transition by at least 1612 BP is apparent, Frisian individuals are modelled primarily as Southern Scandinavian ancestries, but possessing small amounts of the local ENS ancestry” (Supplementary Notes 2-7, pp. 82-83).

The Bell Beaker groups that migrated to the British Isles and predominantly contributed to the autosomal makeup of ancient Britons and Armoricans  are important to mention considering the fact their ancestry comes from Dutch Bell Beaker groups, associated also with the arrival of L21 into the British Isles. By contrast, many of the Dutch Bell Beaker (and SGC) samples so far found have been DF19+ (plus U106 IIRC), which I take to suggest that DF19 and L21 inhabited adjacent areas of the Netherlands in prehistory, with a large chunk of the L21-enriched population traversing the Atlantic façade in prominent numbers and not only settling in the Isles, but also Armorica and the coastline of northern France more generally. This would mean that groups enriched with lineages such as DF19, in association with R-U106 (specifically R-S263) could fill in areas left behind by L21 carriers, thus they would (I assume) have a larger share of earlier Single Grave Culture ancestry while still being strongly related to those emigrating Dutch Bell Beaker groups that contributed to the genetic landscape of Atlantic Europe. Additional support for a non-superficial similarity (in terms of genetic makeup) between the respective populations is found in the fact that a DF19>Z27257 (Batavian?) aDNA sample (also my closest ancient Y-DNA relative) is described by Speidel et al. (2024),  "High-resolution genomic ancestry reveals mobility in early medieval Europe" (https://www.biorxiv.org/content/10.1101/...5.585102v1) as carrying ancestry “which is currently indistinguishable from Iron Age/Roman populations of Britain, to the exclusion of other groups”:

"In southern Britain, the ancestries of Iron Age and Roman individuals form a tight cluster in our MDS analysis (Figure 3a), shifted relative to available preceding Bronze Age individuals from Ireland and Orkney, and adjacent to, but distinct from, available individuals in Iron Age and Roman central Europe (Austria, France and Germany). However, two burials from a Roman military fortress site in Austria (Klosterneuburg) carry ancestry which is currently indistinguishable from Iron Age/Roman populations of Britain, to the exclusion of other groups (qpWave cladality p-value = 0.6). At least one of these skeletons (R10657 - Verf. Fn. 806/10), a 25-35-year old male buried beneath a horse, shows osteological traits possibly linked to intense equestrianism, hinting that these may be non-local Roman soldiers. While one option is that they had ancestry from Britain, an alternative is that currently unsampled populations from large areas of western continental Europe carried ancestries similar to what we see in Iron Age southern Britain" (pp. 6-7).

Furthermore, Gretzinger et al. (2022) modelled the Netherlands_IA population as approximately 40% CNE-derived, with approximately 35% of their ancestry being WBI (British-like) and ca. 10% being IBS (Iberian-like) (Supplementary Information, p. 76). Considering the fact that R10657 plots just like the other DF19+ samples in the ENS group, in addition to the repeated observation across several independent papers that there is some prominent British component in relevant prehistoric Dutch populations, this suggests the British-like drift associated with the ENS IBD cluster is not simply because of shared Bell Beaker ancestry, but because the respective groups are prominently descended from the Dutch Single Grave and Bell Beaker traditions and thus uniquely linked to one-another.

Although Gretzinger et al. (2022) is problematized by the fact it failed to model ENS as a distinct genetic cluster, the Bronze and Iron Age Dutch population branch off a node intermediate between British Bronze and Iron Age populations and the CNE cluster in Extended Data Fig. 5 (although Netherlands_LIA is closest to prehistoric British samples out of the cohort). Simultaneously, we find that among the samples from Hoogkarspel reanalyzed by McColl et al. (2024), there is a gradient of ENS to Bell Beaker-derived ancestry present within the cohort, with 12081 having the most BB-derived/British-like ancestry while fellow resident I12082 is by and large ENS-derived (perhaps relevant to this is the fact Hilversum exercised a certain influence on Hoogkarspel, which could have also transmitted Y-DNA further north). If we use the Netherlands_BA and (most of the) Iron Age samples so far known and compare them with modern populations using, e.g., PCA charts and Distances in G25 (preferential treatment of course being given to other methodologies, since there are many caveats with using these sorts of tools and I doubt my usage of them), they plot very north, generally being positioned somewhat toward Norwegian and Icelandic samples, although one plots close to Irish samples (I12907) while I12906 plots West German-like. This suggests that in prehistory in the Netherlands, there could have existed a cline of British-like (specifically Irish and Scottish BA/IA-like) to Norwegian-like ancestry (the latter being especially prevalent) that prevailed in cultures such as Elp, Hoogkarspel, and undoubtedly also Harpstedt-Nienburg, which was disrupted by the arrival of neighbouring Southern Scandinavian-derived populations into the Netherlands from the Roman Period onward. That there was a close relation with the north extending beyond casual trade is suggested by the sporadic appearance of trace amounts of Southern Scandinavian ancestry among the Hoogkarspel culture samples, in accordance with the fact a cattle exchange network existed between Hoogkarspel, Elp and Nordic Bronze Age groups (including the nearby Luneburg culture), which I have proposed in the past also reflects marriage networks among community leaders as agents for cultural diffusion.

For reference, here are previously published samples with high degrees of ENS ancestry in the McColl et al. (2024) dataset on the North Europe PCA, plus the aforementioned R10657:


RE: the relationship between ENS and the Hilversum and Lower Rhine Urnfield phenomena – it is important to keep in mind, first and foremost, that the core ENS samples seem to be derived from areas where groups such as Elp, Hoogkarspel, and Harpstedt-Nienburg prevailed and spread their influence, so I think the crux of the ENS IBD cluster rests in the expansion of population in the northeastern Netherlands in prehistory (and persisting up to the Migration Period). It is interesting in this regard to see two entirely different profiles cohabitating the same spaces at least in the Hoogkarspel culture area – perhaps suggesting a difference in autosomal profile between people coming from the south and the east? All I can do is infer based on what McColl et al. (2024) has so far provided us, but certainly there is a precedent for assuming some sort of key differences if going by completely different cultural orientations for the respective groups (the Hilversum folks being “Atlantic” but inheriting their “British” goods from contacts to the south). Important to keep in mind is also the fact that if I am not mistaken, the collapse of the Hilversum culture was thorough, with RSFO Urnfielders moving into key areas in Belgium and the Southern Netherlands and leaving the last vestiges of the Hilversum culture marginalised in a small area of the southern-central Netherlands before disappearing entirely, so it is hard (albeit not impossible) for me to imagine them being demographically viable candidates for the persistence and spread of ENS-related ancestry. In this regard, sample I26829 is particularly interesting because he has the least amount of ENS and South Scandinavian ancestry among the Wervershoof cohort (who derive the vast majority of their ancestry from ENS), is positive for R-PF7589, and is much more southern-shifted overall, which considering the more quintessentially “Nordic” characteristics of the funerary assemblages and the consistency of ENS (+ trace Southern Scandinavian) in this group is rather exceptional. From memory, Wervershoof was a lucrative trade hub during the Bronze Age, so technically he could have been a trader from anywhere – however, due to the fact Hilversum was in many regards just as much a southern-oriented culture as it was “Atlantic”, I cannot help but wonder whether he himself could have gotten there by way of Hilversum? ENS of course could have been a much more widespread profile, but these are my (amateur) thoughts so far!

Returning attention to Y-DNA, I have always been of the mind that Y-DNA diversification is a relational phenomenon, i.e., certain Y-DNA haplogroups expand at the expense of others or spread out in conjunction with other, unrelated Y-DNA haplogroups, indicating complex alliances that different clans or cultures forge with one-another which lead to intermixing. As I shared earlier, previous samples that contributed to the ENS cluster identified by McColl et al. (2024) have been found to carry R-U106 (specifically S263 and downstream) and R-DF19, both of which are especially important to the ENS cluster considering their unique distribution pattern in moderns, seeming to have hotspots of diversity in similar (if not identical) areas such as Belgium, the Netherlands, Germany, etc. Considering the strong correlation between S263 and DF19, it is exceptionally important that McColl et aal. (2024) identify the arrival of S263 in Scandinavia with south-to-north movements starting in the Iron Age with ENS peoples, in conjunction with P312:

““Instead, its [i.e., R-Z19] sister lineage, R1b1a1b1a1a1b (R1b-S263), is absent in Scandinavia before the Iron Age (Figure S8), where it spreads, likely through an Eastern North Sea source, and becomes dominant in South Scandinavia during the Iron Age, before spreading through Northern Europe. This pattern strongly matches the one seen using autosomes, that detect gene flow back into Scandinavia related to the spread of Germanic languages. Another potential signal of this migration is the increase in frequency of R1b-U106 sister lineage, R1b1a1b1a1a2 (R1b-P312), that has a more continental distribution. and is almost absent in Scandinavia before 2,000 BP” (Supplementary Notes 2-7, p. 62).

The notion that this population was rich not just in DF19 and Z304, but also in P312 and S263 more broadly, would suggest that ENS was an expansive population within which subclade diversity could grow exponentially without much interruption (probably bigger than is typically assumed), which in my estimation might also involve certain L21 and DF27 subclades (although I would be amiss to suggest any certainty with this suspicion of mine). To elaborate on these considerations, I argue that the most important vector for the spread of modern DF19 has been the Elp, Wessenstedt, and Harpstedt-Nienburg cultures (especially the latter, which forms the basis for the later Rhine-Weser Germanics), with the preceding Sogel-Wohlde phenomenon also being relevant. Decoupling of Y-DNA and autosomal DNA would have naturally happened over time (as we see with the entirely southern Scandinavian-derived R-DF19 > Z302 sample from Denmark and the almost entirely ENS-derived E1b sample from Valkenburg*), but in any case the modern distribution of much of DF19 and S263 might be a good litmus test for determining how Y-DNA lineages that were in the ENS area might be distributed today. 

In a follow-up post, I will talk about some of the new samples from Bucy-le-Long in the dataset!

*potentially of interest is also the fact another E1b sample from LBA Denmark, CGG106737, is approximately half ENS

[Rodoorn]

Have any ENS samples been located outside of the Netherlands? More restricted cultures like Hilversum and Lower Rhine urnfield could perhaps be more of a safe bet for the core homeland of the ENS group. Unless there are more North French samples, and given that the "Belgae" supposedly start from the Seine if I'm not mistaken, I think there is a huge enough area between the Seine the and Meuse/Scheldt for a "gaulish colonised" Belgae zone, not necessarily including the ENS group core territories, (perhaps over running "Briton like" peoples of post MMDN- Deverel-Rimbury traditions of north coastal france). But if we got more ENS like samples from Belgica that could change the picture a bit, but they will probably be on an increasing gradient from south to the north.

East of the netherlands. Maybe the Germanic experts can chime in, but where exactly do the Rhine-Wesser Germanics of the IA come from? I gather originally by MIA-LIA they were located around the Wesser/Aller river basins, later parts of the Rhine, I dont know about the coastal areas. Are they local descendants of the Elp or Ems-Wesser cultures that got germanized/admixture, or direct intruders from the Elbe-Suebi zone, or a mix? Maybe different RW germanics have different ancestry from different pre germanic? Urnfield/EIA groups, such as lower Hesse mixed group?

My impression is that ESN is NE Dutch/NW Germany Bell Beaker derived.

Here with a nodus in North Dutch Northern Drenthe:


This post Bell Beaker group could indeed be very well the core of Rhine-Weser Germani, als labeled by the Romans.

About 400 AD there was an influx from a bunch of Saxon pirates especially around the NE Dutch and NW Germany coastal area. I'm convinced this is related tot the  IA Danish Isles ancestry (see Mc Coll  et al 2024).

The Rhine-Weser Germani aka Franks went partly SW wards towards the South Dutch, Belgian, Northern France area.

[Pylsteen]

Some time ago I counted the number of newly born subclades of P312 lineages in the FTDNA time tree,
for DF19 there seem to be several growth bursts:
1. Initial Bell Beaker peak (ca. 2100-2000 BC), mostly Z302, followed by lower growth ca. 2000-1600 BC (BB)
2. Bronze Age peak (ca. 1600-1200 BC), mostly DF88 and its subclade Z17122, followed by moderately lower growth. (Elp?)
3. Iron Age peak (ca. 900-100BC), all lineages but mostly Z17122, followed by lower growth during the 1st century BC. (Harpstedt? La Tene?)
4. renewed growth during the first millennium AD, also especially Z17122 (in Germanic context)

[Rodoorn]

An add to my previous posting on Y-DNA level and autosomal level.

My assumption is that the Salian Franks and Saxons represent on Y-DNA, level as on autosomal level a somewhat differentiated populations (which even guess into phenotype, but that aside). The Franks are the Roman times Rhine-Weser Germani. The Saxon influx are imo a bunch of pirates rooted in Danish IA ancestry.

The Salian Franks had their heartland in Salland and the Veluvian area (Central North in Lao 2013 et al).

Ok on Y-DNA level the Central (-North) area has a relative high amount of R1b S1116 (BB related?), from Maarten Larmuseau 2019:



See the red area in the Central (North). See also the gap in Friesland, the oldi Frisii left the scene abut 300 AD.

This is repeated on autosomal level. Imo yellow are the Saxons (incoming 400 AD>) and pink are the Salian Franks, and orange the ESN and you can see about the same image as on Y-DNA level, Lao 2013:



It makes imo very clear that the Salian Franks (pink, R1b S116)) were the old Rhine-Weser Germanics and the Saxons (yellow, lack R1b S116) were the incoming Danish Isles IA (mc Coll 2024) ancestry).

So all the rattling about a Sweden/Nordic-Harpstedt Nienburg connection, on genetic as on archeological level nonsens, imo wishful thinking. If this was the case then the Saxon influx (= imo Danish IA ancestry) would NOT be so differentiated from the Salian Franks one! And this is for sure the case....

[Ambiorix]

I looked through Figure 6.3.1.1 to identify the distribution of the ENS IBD cluster in various individual samples – importantly there are extreme outliers with high amounts of ENS ancestry (e.g., I20817 from Patterson et al. (2021), who also carried R-U152 > Z36 > Z37) which start to appear in Gaul. In this context, it is notable that the new samples from Bucy-le-Long contain several outliers with higher degrees of ENS-related ancestry and minor Scandinavian-related ancestry, suggesting that there was a southward movement of various elements from the north (especially, e.g., CGG022431 who carries R-S263); this potentially suggests intermediary or mixed groups taking protracted voyages southward even by the Early to Middle Iron Age, thinking of e.g., CGG023677 from Parançot, Bourgogne, but with the La Tene period containing varying but lower degrees of ENS ancestry among the non-outliers at Bucy-le-Long (note that the Remi and Suessiones, in whose territories Bucy-le-Long is situated, would have mostly been Marne culture descendants and continued to carry their legacy despite the intrusions of groups from the north). Worth mentioning is also an outlier with prominent ENS ancestry and traces of South Scandinavian ancestry (CGG022433) carrying haplogroup R1a-Z280 – I must admit I was surprised to see a R1a+ sample among these samples, but it would mesh well with the Belgae confederation having varying degrees of ancestry from groups bearing Germanic culture and language or otherwise descending from groups further north (especially since going by the aforementioned figure, CGG022433 does not have any Baltic affinity). Extrapolating this information, the presence of these "northern" haplogroups and ancestries suggest that some of the "Germanic" Y-DNA we see in present-day northern France and Belgium does not come entirely from later migrants like the Franks and Saxons, but instead could have been in the general area for a very long time*.

*My personal thoughts RE: the Franks are that due to the complex history of persisting “late Hallstatt” groups that mixed with Harpstedt-Nienburg (at least according to Hawkes and Dunning (1930), The Belgae of Gaul and Britain), coupled with the appearance of other mixed Celto-Germanic communities (such as the Aduatuci, the various constituent tribes of the Przeworsk culture, and some Suebic groups), many (if not most) Frankish tribes would not have been much different genetically from tribes like the Menapii, Morini, Tungri, etc. that they encountered and mixed with, i.e., I suspect the Franks already carried significant amounts of Gaulish ancestry.

[Rodoorn]

I looked through Figure 6.3.1.1 to identify the distribution of the ENS IBD cluster in various individual samples – importantly there are extreme outliers with high amounts of ENS ancestry (e.g., I20817 from Patterson et al. (2021), who also carried R-U152 > Z36 > Z37) which start to appear in Gaul. In this context, it is notable that the new samples from Bucy-le-Long contain several outliers with higher degrees of ENS-related ancestry and minor Scandinavian-related ancestry, suggesting that there was a southward movement of various elements from the north (especially, e.g., CGG022431 who carries R-S263); this potentially suggests intermediary or mixed groups taking protracted voyages southward even by the Early to Middle Iron Age, thinking of e.g., CGG023677 from Parançot, Bourgogne, but with the La Tene period containing varying but lower degrees of ENS ancestry among the non-outliers at Bucy-le-Long (note that the Remi and Suessiones, in whose territories Bucy-le-Long is situated, would have mostly been Marne culture descendants and continued to carry their legacy despite the intrusions of groups from the north). Worth mentioning is also an outlier with prominent ENS ancestry and traces of South Scandinavian ancestry (CGG022433) carrying haplogroup R1a-Z280 – I must admit I was surprised to see a R1a+ sample among these samples, but it would mesh well with the Belgae confederation having varying degrees of ancestry from groups bearing Germanic culture and language or otherwise descending from groups further north (especially since going by the aforementioned figure, CGG022433 does not have any Baltic affinity). Extrapolating this information, the presence of these "northern" haplogroups and ancestries suggest that some of the "Germanic" Y-DNA we see in present-day northern France and Belgium does not come entirely from later migrants like the Franks and Saxons, but instead could have been in the general area for a very long time*.

*My personal thoughts RE: the Franks are that due to the complex history of persisting “late Hallstatt” groups that mixed with Harpstedt-Nienburg (at least according to Hawkes and Dunning (1930), The Belgae of Gaul and Britain), coupled with the appearance of other mixed Celto-Germanic communities (such as the Aduatuci, the various constituent tribes of the Przeworsk culture, and some Suebic groups), many (if not most) Frankish tribes would not have been much different genetically from tribes like the Menapii, Morini, Tungri, etc. that they encountered and mixed with, i.e., I suspect the Franks already carried significant amounts of Gaulish ancestry.

There is some archeological prove that in the nodus of the BB NE Dutch/NW Germany BB namely in Northern Drenthe, for example in Eelde-Groote Veen-that there is a magnificent continuity from about 2200 BC (oldest layers)  until 400 AD (the settlement disappears, incoming Saxons)!!

This is the ESN continuity.....

The incoming Saxons used a Germanic language that was most probably (very) differentiated from the old Frisii one.

https://www.groningerarchieven.nl/actuee...derzetting