Ancient I1 samples list and discussion

[JonikW]

Going to acknowledge my laziness up front guys, haven't done much reading.  Looking at the ftdna Discover tree, the 2 oldest pre-I1 samples look like they come from Spain.  Do you agree with the migration path their globetrekker feature offers?  It has I-Z2699 forming in Spain around 14-15kya after coming through Doggerland and then fanning back northwards as it warmed up.  It makes the most sense to me to see it as a branch that became western/southwestern early on and then spread SW --> N.  Is this controversial or is the western origin more or less settled?

I think nothing is truly settled at this stage and I personally wouldn't want to place a bet on which refugium was involved for pre-I1 SNPs. That I-Z2699 route may well be correct but I don't think this tool is the best way to look at whatever evidence there is. If you'll indulge me I'll say why I think Globetrekker is a useless resource when it comes to I1 itself at least. Here's a screenshot of my own story to illustrate my points:



First, it puts I1 itself in northern France, which was obviously absurd based on both ancient and modern distribution even before the early samples in Allentoft and McColl. (I haven't wasted my time examining the Globetrekker methodology but it seems to be based on modern distribution and so takes France as a kind of central point between Scandinavia, Germany etc and Britain.)

It then puts my Z140 SNP in western Germany close to the Dutch border. This early NBA marker actually belongs somewhere in Sweden in my view (I suspected southern Sweden before the samples in McColl and I'm sticking to that for now).

Things get even worse for me with S12289. I think of this as my “favourite” SNP because it's late BA (about 800 BCE), there is a good data set, and the ancient and modern distribution offer a lot of clues about its origin point.

I've been saying for years that the evidence has long suggested that S12289 was largely focused on the Anglian area by the IA, hence its modern distribution including most significantly its heavy weighting in some areas of England and elsewhere in the Isles today courtesy of the Migration Period. (It pains me to say that Globetrekker mistakenly puts the origin of S12289 in England because of that modern weighting, completely oblivious of known history including the Anglo-Saxon settlements, and the known ancient samples. Even a cursory look at the overall sample set would show any amateur historian the true story.)

My theory gained a boost when Gretzinger’s Anglo-Saxon paper gave us the earliest samples found in Britain, in the Anglian settlement at West Heslerton (x2) and at Dover Buckland. (In the ancient record we also have a stray c10th century CE find from Tarquinia in Italy, a VK sample from Ribe in Jutland, a late Medieval sample from Lübeck and – thanks to McColl – the earliest example yet found, from the Danish island of Bornholm in the Roman IA.)

I'm sticking to the area on this map I've posted before (see below) for the origin point of S12289. The marker may have formed in the east of that boundary while its greatest legacy has been courtesy of the western zone. I may be wrong, but I'm not likely to be anywhere near as wrong as Globetrekker. If I was FTDNA I would mothball this tool until they've worked out how to make it smarter. I'm a massive fan of the other tools they now provide and this lets them down.

[Strider99]

Last year around this time of the year, an article written by Razib Khan about the Nordic Bronze Age and the expansion of I1 was discussed in the predecessor to this thread back on Anthrogenica. I believe it was initially posted by JonikW, but I have lost the link to the article. I see that Razib has now written a new article about the same topic, which appears to be based on the final print of Allentoft 2024. From what I recall, none of us in the old ancient I1 thread were subscribed to Razib, and I still haven't subscribed, so I am unable to read the full article. 

Nonetheless, the "previewed" version of the article features an image that rather neatly illustrates the phylogenetic structure of M253, which I think may deserve a mention in this thread: 



I'd be interested to hear if anyone here is subscribed and if so, what your thoughts about his conclusions are. Last time, the consensus among regulars of this thread seemed to be that the path he outlined for I1 in his first article (which was an expansion from Poland during the BA) is highly unlikely based on ancient DNA and modern subclade distribution. JMcB then posted his own suggestion with an area circled on a map of the modern distribution that seemed far more likely at the time (and holds up even better in light of more recent samples). Razib seems to have changed his mind now, and is instead arguing for a dispersal from Eastern Scandinavia. 

I completely agree with that JonikW delineates above in regards to Globetrekker, I might add. I believe most people familiar with both the modern phylogeny and the ancient DNA would agree (and that's not just true for I1, but for many other haplogroups given a clearly inaccurate path by that tool).

[Riverman]

Last year around this time of the year, an article written by Razib Khan about the Nordic Bronze Age and the expansion of I1 was discussed in the predecessor to this thread back on Anthrogenica. I believe it was initially posted by JonikW, but I have lost the link to the article. I see that Razib has now written a new article about the same topic, which appears to be based on the final print of Allentoft 2024. From what I recall, none of us in the old ancient I1 thread were subscribed to Razib, and I still haven't subscribed, so I am unable to read the full article. 

Nonetheless, the "previewed" version of the article features an image that rather neatly illustrates the phylogenetic structure of M253, which I think may deserve a mention in this thread: 



I'd be interested to hear if anyone here is subscribed and if so, what your thoughts about his conclusions are. Last time, the consensus among regulars of this thread seemed to be that the path he outlined for I1 in his first article (which was an expansion from Poland during the BA) is highly unlikely based on ancient DNA and modern subclade distribution. JMcB then posted his own suggestion with an area circled on a map of the modern distribution that seemed far more likely at the time (and holds up even better in light of more recent samples). Razib seems to have changed his mind now, and is instead arguing for a dispersal from Eastern Scandinavia. 

I completely agree with that JonikW delineates above in regards to Globetrekker, I might add. I believe most people familiar with both the modern phylogeny and the ancient DNA would agree (and that's not just true for I1, but for many other haplogroups given a clearly inaccurate path by that tool).

The situation for I-M253 is very similar to J-L283 and E-V13, because what these three haplogroups have in common is that their former base and roots being largely annihilated, even if they were once widespread like E-L618 for E-V13, during the steppe expansion. We deal in all three cases with a single individual or possibly small clan which barely survived the steppe invasion and then, only a couple of generations later, started to expand from within an Indoeuropeanised context.
The best comparison to come up with is that of an orphan or adopted child, which kind of lost its parents and starts a completely new life, new identiy, with a new family, only to rise to prominence within that new family and in the end even replacing the old relatives of the parents.

It is extremely difficult to pin the deeper origin down, because the earlier history of the branch plays no role for its later success. Because in such cases it doesn't matter where the ancestor was picked up as much, as where the branch survivors rose to prominence.

As for SNP tracker and Globetrekker etc., most of these tools have kind of predictable biases, based on two factors: Known mass migrations and testing bias.

Here again the same patter can be observed for E-V13 and I-M253:
1) There are way more samples from countries like England and Ireland than Germany, France and anything East or Southward. The big plus for I-M253 is that Sweden is well-tested too and this balances things out a bit. But the Western pull is still an issue.
2) Known mass migrations which might have moved whole clusters of branches away from the former homeland. Clearly I-M253 branches moved South in large numbers, therefore if calculating a pattern from modern samples, we have to expect a pull South.

SNP Tracker is sometimes a bit closer to reality than Globetrekker still, at least for some branches. This is especially true for I-M253, which Globetrekker puts into North Western France.
And here comes the most interesting part: Despite the likely pull both towards West and South, I-M253 has a relatively Eastern positioning:




Keep in mind that this is, based on modern sampling, still too Western!

You can immediately see the distribution (Slavic replacement!) and testing bias by looking at the map for descendants. The pull towards the British Isles is just ridiculous and we all know that data from the Roman Iron Age would have looked very, very different, with very little on the British Isles, and lots of dots from Poland and Ukraine for example:




Therefore the modern distribution and testing bias surely creates, still, even on SNP Tracker, a Western pull.

Therefore going East and a bit South would bring you towards Brandenburg or Poland.

[MrI1]

@Running4Borders

I made a video regarding this topic titled Pre-I1 Migration Routes in the link below which addresses this question.

https://youtu.be/HCbmaM-T0MI?si=zLn0DeEKQNOZo1t8

[JMcB]

Last year around this time of the year, an article written by Razib Khan about the Nordic Bronze Age and the expansion of I1 was discussed in the predecessor to this thread back on Anthrogenica. I believe it was initially posted by JonikW, but I have lost the link to the article. I see that Razib has now written a new article about the same topic, which appears to be based on the final print of Allentoft 2024. From what I recall, none of us in the old ancient I1 thread were subscribed to Razib, and I still haven't subscribed, so I am unable to read the full article. 

Nonetheless, the "previewed" version of the article features an image that rather neatly illustrates the phylogenetic structure of M253, which I think may deserve a mention in this thread: 



I'd be interested to hear if anyone here is subscribed and if so, what your thoughts about his conclusions are. Last time, the consensus among regulars of this thread seemed to be that the path he outlined for I1 in his first article (which was an expansion from Poland during the BA) is highly unlikely based on ancient DNA and modern subclade distribution. JMcB then posted his own suggestion with an area circled on a map of the modern distribution that seemed far more likely at the time (and holds up even better in light of more recent samples). Razib seems to have changed his mind now, and is instead arguing for a dispersal from Eastern Scandinavia. 

I completely agree with that JonikW delineates above in regards to Globetrekker, I might add. I believe most people familiar with both the modern phylogeny and the ancient DNA would agree (and that's not just true for I1, but for many other haplogroups given a clearly inaccurate path by that tool).

Unfortunately, I never subscribed, so I can’t get any further than you did. I used to read his blog but he hardly ever posts there anymore. So I deleted it.

His new map is an improvement over the earlier one, though.


The two Indo-European revolutions and Germania’s
https://www.razibkhan.com/p/the-two-indo...evolutions

[MrI1]

The new preprint has a wealth of information that I’m probably going to cover over multiple videos. My first video covers the East Scandinavian Cluster associated with Haplogroup I1.

https://youtu.be/Vm3mHH_lg38?si=0wF-KKMbVwmvHc3p

[Strider99]

Radko brought attention to a very interesting study that was published yesterday over in this thread. 

Unveiling Hunnic legacy: Decoding elite presence in Poland through a unique child’s burial with modified cranium

Highlights
• Skull deformation insights – rare Hun practices discovered in Southern Poland.
• Genetic diversity in double grave demonstrates European and Asian ancestry.
• Radiocarbon dates indicate the early presence of Huns north of the Carpathians.
• Mortuary practices reveal the different social positions of Hunnic and local children.
• Children buried in grave with a dog, cat, and crow serving as their companion animals.

Abstract
This article presents a double burial from Czulice indicating elements of the Hunnic culture. Individual I, aged 7–9, and Individual II, aged 8–9 with a skull deformation, were both genetically identified as boys. Individual II, who exhibited genetic affinity to present day Asian populations, was equipped with gold and silver items. In contrast, Individual I displayed European ancestry. The application of strontium isotope analysis shed light on the origins of the individuals. Individual I was non-local, while Individual II was identified as a local, but also falling within the range commonly associated with the Pannonian Plain. Stable isotope analysis suggested a diet consisting of inland resources. Through radiocarbon dating, this burial was determined to date back to the years 395–418 CE, making it the earliest grave of its kind discovered in Poland. The analyses have provided new insights into the nature of the relationship between the Huns and the local inhabitants.

Czu001 appears to be M253+. His autosomal profile is quite interesting:

Distance to: poland_huns:czu001
0.03510637 Norwegian
0.03654482 Danish
0.03667875 Swedish
0.03676891 Icelandic
0.04263559 Irish
0.04297734 Dutch
0.04406862 Scottish
0.04423722 Orcadian
0.04541657 Shetlandic
0.04555471 Welsh
0.04583222 English
0.04721248 English_Cornwall
0.04794601 German_Hamburg
0.04986822 French_Brittany
0.05237936 Afrikaner
0.05280204 German
0.05782225 German_Erlangen
0.05922289 French_Pas-de-Calais
0.06022684 German_East
0.06073925 BelgianA
0.06116635 Czech
0.06136925 BelgianB
0.06358391 Polish_Silesian
0.06409859 Slovakian
0.06434639 Austrian

poland_huns:czu001,0.121791,0.133034,0.075801,0.070737,0.042469,0.017012,0.00188,0.011307,0.006749,-0.002369,-0.005196,0.004946,-0.013379,-0.013762,0.021444,0.025722,0.004433,-0.003294,-0.002891,0.008754,0.013102,0.000618,-0.002095,0.019882,-0.001197

Thanks also to @Orentil for offering some interesting speculation as to who this boy might have been and how he ended up in such a burial context. Truly a fascinating sample! Here is what Orentil wrote, and I think it sounds very plausible: 

A Gepid or Herulian boy? (if it is true that the Visigoths and East Goths already showed major shifts at that time). I assume an IBD analysis would rather group him to EastScand then SouthScand.

PS: Maybe a hostage boy of a Germanic elite family at a Hunnic court (see posts below) in Southern Poland, dying together with his Hunnic friend, very interesting find

On a related note, some new I1 samples that I believe have yet to be discussed in this thread are the recent Gepid samples from this study. The Gepid period samples are quite I1-rich, and some of them look surprisingly northern-shifted autosomally (in particular samples RKO002 and RKF275). The former carried I1-F2642 and the latter I1-Z138. 

To end this post with some good news: I was pleased to hear that there will likely be new samples from the IA site of Gamla Skogsby on Öland in the not too distant future. The place is pretty close to Torslunda, which is interesting. Some of the remains at that site should be very old. Thanks to Jerrark over at skandinavisktarkeologiforum for sharing the news.

Edit: I took a look at the downloadable Supplementary Data 3 which has a few interesting tidbits. "To further investigate the apparent connection between our test samples and ancient individuals we performed D-statistics (Patterson et al., 2012) of a form (Yoruba, czu001: GermanyEarlyMed.SG, X) and (Yoruba, czu002: HungaryHunEliteIA, X). Despite low numbers
of overlapping SNPs, we find that each of the two individuals is consistently closer to their respective chosen reference sample (i.e. GermanyEarlyMed.SG for czu001 or HungaryHunEliteIA for czu002) than any other ancient group. Moreover, most of the significant estimates (|Z| ≥ 2) mostly imply distance in opposite geographic directions supporting inference regrading divergent
ancestry of the two individuals (Data S4). Thus, it seems czu001 has northern/central European ancestry, very similar to previously described contemporary individuals from early medieval
Bavaria with normal skull shape (Veeramah et al., 2018). While czu002 with signs of skull modification, despite presenting only >6,000 overlapping SNPs in f-statistics analyses,
consistently displays closer relation to Asian populations as well as individuals previously identified as Huns (Gnecchi-Ruscone et al., 2021)."

I hope that we'll get to see czu001 get a more elaborate Y-DNA assignment at FTDNA in the future. The study assigns him only as "I1;I1∼" due to low coverage.

[Strider99]

Courtesy of member Radko: czu001 has now received a deeper assignment by FTDNA. L22>P109>FGC21611.