The origin and legacy of the Etruscans through a 2000-year archeogenomic time transect:
The origin, development, and legacy of the enigmatic Etruscan civilization from the central region of the Italian peninsula known as Etruria have been debated for centuries. Here we report a genomic time transect of 82 individuals spanning almost two millennia (800 BCE to 1000 CE) across Etruria and southern Italy. During the Iron Age, we detect a component of Indo-European–associated steppe ancestry and the lack of recent Anatolian-related admixture among the putative non–Indo-European–speaking Etruscans. Despite comprising diverse individuals of central European, northern African, and Near Eastern ancestry, the local gene pool is largely maintained across the first millennium BCE. This drastically changes during the Roman Imperial period where we report an abrupt population-wide shift to ~50% admixture with eastern Mediterranean ancestry. Last, we identify northern European components appearing in central Italy during the Early Middle Ages, which thus formed the genetic landscape of present-day Italian populations.
The inference is basically what I concluded and stated this spring in my piece on ancient Rome:
By 1000 BC, on the eve of the Iron Age, a mixed ancestry derived from both Anatolian farmers and Indo-Europeans from the steppe was present all across continental Italy. Only on isolated islands such as Sardinia was the situation different, where Anatolian farmer societies continued unperturbed by Indo-European migrants, giving rise to the Nuragic civilization, with its massive fortifications. All the length of the Italian mainland, the ancestors of the Etruscans, Sabines, and Romans shared roughly the same genetic heritage, despite their ethno-linguistic differences.
The big novel find is that there is nontrivial Germanic ancestral shift after the fall of Rome. Some commenters have suggested this is actually just from rural Italians, who flooded back into the towns. This could probably be tested.
This statement of Richard Rocca is astonishing:
“I had mentioned this a while back, but Polada Culture / Early Bronze Age samples from the Adige Valley, which is the valley known for the great majority of Iron Age Rhaetian inscriptions, are surprisingly all G2a. The data quality is low (SNP testing) and some of the samples have second hand dating. If further testing proves this out, it is mind blowing to think that Otzi may have spoken a “Tyrsenian” language!”
I have been saying from so long that Etruscans had a component of the agriculturalists migrated from the Aegean Sea through the Balkans until central Europe leaving Rhaetians in the Alpine region when they migrated southwards. I said that they spoke Etruscan against the great linguist Schrijver who thought to Hattic.
These “Levantinists-Kurganists-Levantinists” continue to attribute R1b to Yamnaya and not that its origin is probably “Alpine”, that is, ultimately from the “Italian Refuge” of the Villabruna. I have some doubt now about my R-L23-Z2110-FGC24408 (but not about other R-Z2110 subclades) as come directly from there with Etruscans, but not about the other haplogrouo I have been examining in these years (R-L51-PF7589). My hg. may have come also from Sintashta with the charioteers, being in Italy only from 3100 years ago, even though probably documented in Latin noble families, but other subclades of R-Z2110 expanded from northern Europe and not from Yamnaya.
This paper seems to not support the idea of local population resurgence. However, Stanford (Antonio et. al 2019) and the Pre-print from the Reich Lab on Roman Danubian (Olalde et al. 2021) do support the idea. This one just takes the position the Eastern Mediterranean immigrants from Imperial era had a big impact, without interruption. Frankly, I think the demographic change after the fall of the empire, and local population resurgence ideas are too compelling, to ignore. While I respect the Max Planck Institute, I have to disagree with their modeling of modern Italians. But more importantly, does the Reich Lab, also disagree with them? I want to see the upcoming paper on Magna Graecia in Campania to see what this ancestry was like, and what implications it had for Italy, genetically. What about the inclusion of these people into a united-Italy under the Romans, and the implications it had genetically? I am not sure the reason for them to include Venosa samples from the middle ages, but it would have been a lot more helpful if they compared them to IA samples from the region.
A honest man:
Blogger Rob said…
Recent studies have shown that Beaker people had nuclear families. So we cant just explain away with absentee fathers. In fact, their male children in Iberia were buried with high quality gifts, meaning there was a strong link & connection.
But the truth of the matter is Bell beaker culture itself had little impact in Italy. it doesn’t make sense to speak of Bell Beaker when talking about LBA-IA early Etruscans.
September 25, 2021 at 12:07 AM
In fact I demonstrated that Sicilian Bell Beaker of Buffa 2000 BC was a “Flemish” (i.e. from the zone where Flemishes are living now). He is close to my daughter-in-law, 100% Flemish, practically German Franks, and has nothing at the autosomal level to do with me and neither with my Sicilian wife, with Norman mt, K1c1f3, and the Y R-L21 of the family, but practically close to me, i.e. largely “Roman”, in the autosome.
Said what I mainly think on the Open Thread but want to really reiterate, we really need samples from North Italy in this timeframe, to really begin to nail down how much of the Late Antiquity shift is not just a remergence from small villages but likely due to a drawing down of the empire’s northern population back into the centre and south.
Isabel Alves presentation gave some indication of what Late Antiquity Northern France from 300-500 CE was like, in the form of two samples – https://imgur.com/a/vtB6LG7 . Very roughly looks slightly less steppe than modern or medieval North and West French period, but matches Roman Britain samples better than Roman Iberia. There could be some rotation of populations across Europe rather than simply a migration of Germanic types.
For fun, here’s the current published entire European set* from 50 CE – 500 CE that are on Eurogenes G25, plotted on the Vahaduo West Eurasia PCA: https://imgur.com/a/Fwegidc
Interesting there was one Imperial Era (3CE) Etruria previously published, ETR001, and like this paper’s samples, this sample seems definitely shifted away from where the Etruscans would have been, but does sit overlapping more European end of the main cluster in the Imperial Rome set, and Europe shifted relative to the mean of the Roman samples from the Imperial period of 5CE. (There are of course the other Roman Imperial samples from Northern Spain and and the Roman soldier from Germany who are more like present day Spain / present day North Italians in some respects).
This plotting of ordinary Etrurians from the Imperial Era at a remove from the mean of the city of Rome does raise the question of whether Rome saw much more population turnover than was typical in North and Central Italy. (As much as this paper *does* validate that the shift happened during Roman Empire in North and Central Italy). I don’t think would be surprising to us at all, that Rome maintained a more Near Eastern profile than was typical, due to the dynamics of migration to an urban centre and the high mortality rates of an urban centre.
I might add the Etruscan Imperial Era from this paper when David gets round to it.
*Minus a HUN Elite, who I thought was too East Asian to meaningfully go on a West Eurasia plot. Also btw, for sample info UKR_Chernyakiv here are Ostrogoths as I understand it. Likewise the NW European like person from Slovakia, DA119, is supposedly a 20-25 year old chieftan who is described by the paper Damgaard 2018 as “Additionally, the inventory can be compared with other European elite Germanic chamber graves from the Late Roman Period”.
Wild stuff. So unlike the rest of peninsular Italy, the sons of the steppe who became the Etruscans adopted the language and customs of their EEF wives.
I quoted in our Italian blogs (and in private letters) what is written about this paper and the more and more shy critics to the agenda beyond these papers, not only of the Reichs lab (i.e. the Harvardians) but also Leipzig. I thank Razib Khan, whom we know who he is, and also Jovialis. Perhaps someone knows that I don’t like who uses nicknames, and pretty much always I understood who he is, from when I discovered Ken Nordtvedt in an anonymous skipper. We were probably on Rootsweb, 2007, or DNAforums 2008, of course banned from both later.
For understanding who is the person hidden in a nickname above all we have to understand the nickname.
Jovialis seems to know both Italian and Latin, it is close to Italian “gioviale”, but it presupposes Latin “*iovilus”, i.e. dedicated to Iuppiter (“godfather”) Iovis, which is at the origin of Iulius [*Iovilus], i.e. not only the gens Iulia but Caius Iulius Caesar par excellence. Perhaps a knowledgeable Italian-American who doesn’t dislike his origin, and not from Southern Italy in spite of that “godfather” (or also Sicilian Americans are changing their politics). But it is possible that this is the senhal of a known person, who probably doesn’t disagree with Gioiello.
@Matt
I’m guessing the extra northern input is from a variety of western and central european sources over time, plausibly including “internal” ones as you suggested, but here’s a general look just at the modern central Italians with the ancient sets, using aggregates from this basic set I made here: https://pastebin.com/GiQVDZ4t
Image: https://postimg.cc/MnQ0pLkg
The difference between the top two is I excluded the more steppe-rich CSN008, CSN009 and the Prenestini individual from the topmost model. Even in that one, the distance still looks pretty good for a general model but it’s interesting how much it drops proportionally for some moderns, it looks like those few somewhat more steppe-rich individuals do make a bit of a difference even if the %s remain basically the same. Strictly speaking, you don’t seem to need any more than some 30-something Near Eastern added to the Etruscan-Italic set model the moderns, even if the specifics are likely more complex. The “ItalicCEU” is arguably “Celtic” as you mentioned but just to keep to the paper’s labels there. Also plausible that some of that “Baltic” is, among other things, due to small admixture from Germanic-related sources since Scandinavia was a bit intermediate in “Baltic-like” ancestry between more western and eastern Europe.
The dynamics behind the Etruscan and Basque cases remain interesting…
Also a little thing I checked so far, I’m also guessing that they might have gotten a better model of Tuscan as C.Italy_Early.Medieval (main text) if they excluded that one J1 relatively outlying individual ETR014. The rest that are included in the preliminary G25 still are somewhat more southern than the modern Tuscan sample on average though one basically overlaps it.
https://postimg.cc/nCDfYq6B
@Forgetful, interesting model and proportions! I am quite surprised that the Armenian:Levant is so unstable up and down the peninsula, as doesn’t seem intuitive to me with the north-south cline, unless there’s a model where the north gets a different Near Eastern source to south, which seems a bit unexpected.
I’m not sure about the idea of using the modern Armenians in the model, though it’s not a huge problem so long as we understand that the sources are ancient for the Italic+Balkan+Italic_CEU+Levant set but modern for Armenian. The modern Armenian source is quite compressed towards the Levant compared to ancient samples we have so far. The modern Armenian set is closer to the Levant than to Armenia_LBA (https://imgur.com/a/051yeyU) circa 1000-900BCE, while not much shift in Levant. Understandable given sparse sampling in Turkey and Armenia post 1000 BCE.
Slight modification to your model, splitting the Balkan groups into Mycenean and other, and subbing in ancient Armenians for modern, and in some models removing North Balkan and/or adding DEU_Early_MA – https://imgur.com/a/74Mly8c.
Kind of interesting that in some of these models the Levant imprint marginally peaks in Lazio, though the total non-Italic related imprint peaks in Calabria (https://imgur.com/a/74Mly8c), excluding a Northeast outlier, boosted by Mycenaean related stream. Greek settlement in the South and still some legacy of Rome’s richer Levant influx? Also checked if this was Sardinian related (as an outside possibility), but doesn’t seem so on the face of it – https://imgur.com/a/oeRFT7O
Not very strong evidence, but possibly does accord with the idea that Levant related settlement was stronger in Rome and so still persists there more today.
Also some more IA regions on the PCA plots I made above: https://imgur.com/a/Ma8P6zx
So have the new Etruscan etc samples been updated on G25 PCA?
I’m probably one of the “commenter” you are referring to who ipotized a “local resurgence” in the late antiquity – early middle age.
The fact is, the extimated number of longobard invaders during the VI century is around 100.000 – 300.000 individuals, while the population of the italian peninsula was, at least, probably around 6-7 million people: no way they could have contributed to a 20% of the current population, even in the northern regions!
Looking at some of Vahaduo’s PCA, there definitely has to be some phenomena of re-emergence (or analogous) of Italic ancestry or an influx of SW European ancestry into Italy: https://imgur.com/a/Fw162oI
The cline formed by the averages of the present-day Italian regions, north-to-south, doesn’t overlap with a set of two-way models with DEU_Early_MA and ITA_Rome_Imperial at all.
Besides the Aosta Valley which resembles Provence (and some regions that I’ve excluded from the NE, who are North Balkan like), the cline of Italian provinces stretches up from Calabria to Veneto, and is enriched in “West Mediterranean” (EEF) ancestry; its not between ITA_Rome_Imperial and DEU_Early_MA, which would make for populations that are too rich in steppe ancestry and too low in EEF.
That shows up if you model them using distal sources too, they appear too Anatolia_N-rich. It’s a similar case in the Balkans where populations give a sense of being too Anatolia_N-rich in general to be simply Imperial average (which interestingly, from looking at it back when, also seems to include a couple of individuals from Italy that are IA Balkan-like in makeup – low/no Levant, modest CHG and steppe, little/no WHG – rather than Italic-NE mixes) – early Slavic mixes.
Theoretically, you need a more specific part of the wider set, it looks like, likely because the very varied set doesn’t give us a hypothetical average of the actually contributing population in the first place as has been mentioned.
Somehow completely missed a couple of recent comments above, they didn’t show up for me for some reason until now. Must be something on my part but weird.
Pretty much, as you’re basically suggesting, with modern Armenian I was trying to use a northern Near Eastern/CHG-Iran_N-rich source since we lack more detailed sampling from the area to cover the other part of the Near Eastern cline. Could use those more CHG/Pontic-like individuals from the Imperial set too, maybe, or even the Azeri Lowlands Neolithic/Chalcolithic ones. The non-Lchashen (since that one looks like it might have some low-level extra eastern stuff) ARM_MLBA-set is almost modern northern Caucasian like though its ancestry seems to come from solely earlier sources as we might expect from the chronology, while modern northern Caucasians give me an impression of requiring some partial Scythian/IA steppe-like admixture too, i.e. more complex interactions with the steppe through time. I didn’t want to use those specifically because they might represent transitional populations in the area and their very high steppe ancestry might be problematic while using the other sources, but it doesn’t seem to change things too much in your image.
I haven’t seen them updated on the main G25 datasheet yet. Only a preliminary one including just those that I saw either on Eurogenes or Anthrogenica.
@Forgetful, not anything on your part – sometimes I put too many links in my comments and they get hit by the Razib’s comments spam filter until he waves them through.
@Matt
Completely forgot about the rest. This is the full preliminary datasheet of all the individuals that could be included in G25:
https://pastebin.com/56gSCits
I don’t remember whom to give credit to for it, other than Eurogenes of course, so I can just repost it without specific attribution.
Probably quite a few interesting things to notice there but take a look at the odd MAS001 if you’d like. I’m not sure their Etruscan-Armenian_MLBA makes complete sense in the way it comes out in G25 so far.
Originally Posted by etrusco
Going by uniparentals, the continuity in Y line (and mtdna also?) in Italy seems stronger (from the iron age) than the autosomal one? Just asking
Originally answered by Richard Rocca
It’s really tough to say. Even if we take U152, it was all over by the end of the Bell Beaker period – France, Hungary, Germany, Poland etc. So, we will need a lot more refinement to see which U152 was brought in by Celts and Germanics to know for sure.
Addition: And I may add, U152 samples also exist from Tumulus Culture, Hallstatt, Tollense Battlefield samples etc. Simply too common everywhere to just be an Italic marker, so more subclade samples are required.
I make you note that 4000 years ago Celts and Germanics existed but not Italics or Etruscans!
It is very strange that the sample MAS001
“The C.Italy_Etruscan_MAS001 individual represents a single exception in our dataset showing a shift in PCA space toward Near Eastern populations ~200 BCE (Fig. 4A).
Instead, C.Italy_Etruscan_MAS001 can be modeled as a mixture between the C.Italy_Etruscan cluster and populations from the Caucasus, such as Bronze Age Armenians (Fig. 4B), indicating the sporadic presence of Iranian-related ancestry in Etruria at least by the second century BCE”
MAS001 mt T2h2 Y G-CTS796
finds the only two known samples so far in Italy and separated more than 11000 years ago. Of course we’d need deeper data, but this could be the proof that hg. G was older in Italy (I said that it was linked with the obsidian traffic in the Mediterranean Isles) than in the Caucasus.
https://www.yfull.com/tree/G-CTS796/
Neither the mt may be ascribed to Caucasus:
T2h2 A8041G * G12771A * T16172C formed 9500 ybp, TMRCA 3700 ybp
T2h2*
id:JQ702170.1
7 samples in GenBank out of 17.
About JQ702170 we don’t know the origin being a clone of the samples tested from FTDNA and donated to Behar for his reassessment and of course not tested in Israel as he said, but we have other 6 samples: JX153755 Denmark, JX153998 Denmark, KF161750 Denmark, KF162051 Denmark, KF163012 Denmark, and what is in Denmark was Southward before.
Of course there is also this sample, HQ610991, for the pleasure of Razib Khan:
Santhini,E., Vijaya Padma,V., Govindharaj,P., Sarveshwaran,V., Mohana Krishnan,L., Vijaya Kumar,K., Lalji,S. and Thangaraj,K.
TITLE Direct Submission
JOURNAL Submitted (14-NOV-2010) Biotechnology, Bharathiar University, Maruthamalai Main Road, Coimbatore, Tamil Nadu 641046, India
I’ll study all the haplotypes.
P.S., but with only 2 SNPs tested (HQ610991: A6120G, A8041G) I think that it isn’t T2h2 but probably these 2 very rare SNPs come in heteroplasmy from hg R*.
Anyway JQ702170 is from Slovakia with 3 private mutations: A546G A13907G T14502C, thanks to the old Ian Logan. Now we have to compare this Etruscan sample.
From my readings a figure of 200,000 to 300,000 is extremely high, given that the original figure was 60,000 Langobards and 20,000 Saxons, and, to the best of my recollection, the narrative was that the Saxons returned,
Regardless, we have ancient dna from a Germanic group invading Italy, and the men were all U-106. Even if we add I1 that would total under 10% of the yDna in Toscana.
As to U-152 in the modern area of Germany, I would speculate that the U-152 was there before the migration of the Germanic tribes, so I would be surprised if they carried a lot of U-152. They seem not to have been all that inclusive.
So, unless and until we find Germanic invaders into Italy who carried U-152, I would think the additional “northern” admixture most probably would consist of Gauls, and, perhaps, northern Italians, although for the latter we would need more samples for confirmation.
In terms of the modeling, I would think it would be interesting to try to model Early Medieval Tuscans as a combination of Etruscan first millennium BCE and Empuries, or the upcoming Lazaridis Roman_Greek sample, or Aegean IA. When the Campania paper comes out, there will, of course, be samples from actual Greek migrants to Magna Graecia, whom it isn’t implausible to think might have moved up the peninsula.
Quote Originally Posted by R. Rocca
“After seeing the Italic and Etruscan Iron Age samples, I would bet money that the Veneti didn’t “arrive” from anywhere but were also an autochthonous group. In fact, the Atestina/Este Culture is one of the few Iron Age cultures that has a crystal clear successor from the Iron Age to the Roman period”.
Ahahahahah, also Richard Rocca is becoming an autochthonist!
Of course the Novelist repeat the Harvardian novel: an Englishman cannot break the link between the Crown and the Finance:
“In my opinion, the Venetian Indo-Europeans who immigrated from the northeast in the Bronze Age (originally Eastern Central Europe) were like the other Italian peoples and later developed their own culture in the early Iron Age as you mentioned the Este culture, which in turn was influenced by the Villanova, Urnfield culture and parallel and almost timely to the Hallstatt phase. From a genetic point of view, you have absorbed a lot of the autochthonous population of the Po Valley over time and with the expansion into Dalamatia you were in contact with Illyrian and Liburnians (origin disputed whether Venetic or Illyrian or …?), Later a strong one came cultural influence from the Celts, but the Venetians would quickly be assimilated into the Roman Empire and their culture increasingly resembled Roman society. In the case of the Etruscans, we can similarly draw a parallel that genetically had a steppe input and, in turn, a genetic proportion of the autochthonous population of the Po Valley and with the big difference that they have a pre-Indo-European language (Etruscan) and that belongs to the Tyrsenian language probably does not come from the Aegean region but probably from the southern Alpine region and the Lemnian language was brought into the Aegean territory by the seamen and thus one can exclude the hypothesis that the Etruscans and Venetians originated from Anatolia, which was previously suggested”.
MAS001 is given now G-CTS5990/Z1903, it overwhelmingly Italian too at least between 8300 and 4700 ybp.