Genetic study The Genetic Legacy of the Roman Imperial Rule in northern Italy

[Vitruvius]

Otherwise the alternative is that there was (also) resurgence of an Iron Age genetic profile from more rural areas after the fall of Rome, this decreases both the northern European and the Imperial Rome contribution. If there is any truth in all the assumptions, it becomes really really difficult to get accurate results.

If we're speaking of a resurgence in iron age italic ancestry it seems apparent that was almost certainly that of northern Italic rather than IA Latin/Etruscan based off the current cline. If Latin/Etruscan IA rural resurgence was the cause for this shift, then modern Italians would have a much greater overlap with higher WHG Spaniard-like profiles than they do today.

Modern Italians have now lower WHG because they have also changed compared to the Early Iron Age population.

We don't have nearly enough data to determine this. We can say with more confidence they've changed compared to Latin/Etruscan populations, but as for Magna Graecian and N. Italic - the jury is still out.

If what Posth 2021 claims is true, the idea that an Imperial Rome profile was anywhere, then there needs to be a significant northern European DNA contribution to have a genetic profile of modern central Italians.

This idea doesn't add up from what I can see. WHG ancestry is too low in all Italians for this to be feasible I think. Also, EEF ancestry remains too high - even in the north.

 

[Pax Augusta]

If we're speaking of a resurgence in iron age italic ancestry it seems apparent that was almost certainly that of northern Italic rather than IA Latin/Etruscan based off the current cline. If Latin/Etruscan IA rural resurgence was the cause for this shift, then modern Italians would have a much greater overlap with higher WHG Spaniard-like profiles than they do today.

Yes, of course, the chances are greater for the north but the north is the largest genetic cluster in Italy, it doesn't really make sense to generalize about the whole north. In fact a Spaniard-like profile still exists in northern Italy and is found in some Alpine valleys not mixed with German and Slavic language minorities but it is not at all clear how common this genetic profile is.

We don't have nearly enough data to determine this. We can say with more confidence they've changed compared to Latin/Etruscan populations, but as for Magna Graecian and N. Italic - the jury is still out.

I remind you that much of southern Italy was Italic before it became partially Greek and that the Samnites of Campania had a similar profile to the Latins and Etruscans was said by Max Planck's geneticist Alissa Mittnik in a public meeting, not by a forum user. Of course, we do not know, as I said, whether it will be confirmed by a published study. Moreover, the few northern Italian samples like Broion are in line with Etruscans and Latins. But let's wait for the Proto-Illyrians who repopulated northern Italy.

It is really very naive to think that the Latins and Etruscans represented a genetic anomaly as you do, and that only they disappeared and that only central Italy changed and the rest of Italy remained unaltered because so far geneticists have focused only on them.

This idea doesn't add up from what I can see. WHG ancestry is too low in all Italians for this to be feasible I think. Also, EEF and Caucasian ancestry remains too high - even in the north.

That is not what the jury ruled.

 

[Palermo Trapani]

It is a very complicated calculation, for a number of reasons. Not least because the genetic profile of Imperial Rome already contains local Iron Age DNA as well as that of alleged newcomers from the eastern Mediterranean. If you use the model supported by geneticists, Local Iron Age + Imperial Rome + Langobards, cosidering that Imperial Rome already contains local Iron Age DNA, nearly 30-40 percent very roughly, look at all the work Jovialis has done over time, it turns out that the Iron Age DNA in modern samples is something between 30 and 40 percent, if instead you just use the two-way model (Local Iron Age + Imperial Rome), the local Iron Age inheritance in the modern populations exceeds 60 percent.

In no way should these models below be taken literally (the distances are off for many tagets). Also because the Greek presence in Italy could be a game changer. For convenience I am using the G25 right now but all the datasheet nomeclature has been changed and I find it very confusing now.

Taking a step backwards, we refer to Etruscans and Latins as the same thing because they are the only published ones, along with the Daunians from Apulia and Sicanians from Sicily. From what geneticists have anticipated so far, this Etruscan-Latin like genetic profile was also shared by other peoples of Preroman Italy, the Samnites of Campania (this was said by Max Planck's geneticist Alissa Mittnik), and the Picenes of the Adriatic coast, two Osco-Umbrian-speaking peoples. Studies have not yet been published and we do not know if it will be confirmed. It is very possible that geneticists will argue that Iron Age peoples of northern Italy also had a similar profile (and this is certainly true for some areas of the Po Valley since they were inhabited by Etruscans). And here comes another issue, that of North Balkan influences from the Late Bronze Age with fewer WHGs carrying other assumptions.

Modern central Italians form two clusters (Raveane 2019), one consisting of individuals from Umbria (samples from Perugia), Marche (samples from Ancona), and Latium (I ignore where the samples came from), and the other from Tuscany (samples mainly from central and southern Tuscany, unclear TSI) going a little further north towards the Emilia samples. Your question can be answered if it is known for sure whether it is true that the genetic profile of Imperial Rome was everywhere as shown so far (for Tuscany if I remember correctly 4 samples from the southernmost areas, bordering Umbria and Latium), a mix of local Iron Age individuals and Imperial-era migrants from outside Italy, but this genetic profile is quite close to the cluster of modern Italians of southern Italy, and this may lead to other alternative hypotheses, and that therefore it was actually (also) movements from southern Italy that helped spread this profile of Imperial Rome. But at this point it is also necessary to answer the question of what was the genetic profile of southern Italy in the Iron Age, which was inhabited by Italic, non-Italic, and Greek populations from 800 BC. If what Posth 2021 claims is true, the idea that an Imperial Rome profile was anywhere, then there needs to be a significant northern European DNA contribution to have a genetic profile of modern central Italians. Otherwise the alternative is that there was (also) resurgence of an Iron Age genetic profile from more rural areas after the fall of Rome, this decreases both the northern European and the Imperial Rome contribution. If there is any truth in all the assumptions, it becomes really really difficult to get accurate results.




The ancient Veneti spoke a language that shows some affinity to the Latin-Faliscan group, and the Raeti spoke a language related to Etruscan, and the area from Bologna to Verucchio has been inhabited by Etruscans since the earliest Iron Age. Draw your own conclusions. Then there is also northwestern Italy (Golasecca, Ligurians), which is a less clear issue even archaeologically.

Modern Italians have now lower WHG because they have also changed compared to the Early Iron Age population. The question of how much Iron Age DNA is left in modern central Italians is a question that applies to all Italians. Not surprisingly, this study on northern Italy is being released. Those on southern Italy will also be published later. Although it is not a given that geneticists will give all the most accurate answers.

Pax Augusta: Where did you get those sample averages to run those models. Thanks

 

[Pax Augusta]

Pax Augusta: Where did you get those sample averages to run those models. Thanks

From the G25, last datasheet, but it is not very good.

 

[Palermo Trapani]

From the G25, last datasheet, but it is not very good.

Pax Augusta: Ok, thanks. Did you use just those 3 Population averages as source populations to run those models?

 

[Vallicanus]

The debate between Pax Augusta and Vitruvius is fascinating and illuminating.

I need to add some data concerning Italian physical anthropology as researched by Livi and others.

The nasal index on the living subject, reported by Mori, shows a relatively high NI but this was largely due to a SHORT nose because the nasal breadth seemed to be narrow ( 33-34mm in Tuscany and the old province of Rome (present-day Lazio minus Rieti, Gaeta and the extreme south of Lazio today).

A short nose would exclude much immigration from areas like (Roman Age) Anatolia or Syria, alleged by supporters of Roman Italy overrun by Near Eastern slaves and migrants.

 

[Vallicanus]

Italian NI...Piemonte, Lombardia and Marche..all 67.3

Liguria 65.4
Veneto 69.7
Emilia 68.7
Tuscany 69.1
Umbria 70.4
Lazio 69.7
Abruzzi 69.8
Campania 69.7
Puglia 69.5
Basilicata 68.9
Calabria 68.0
Sicily 70.0
Sardinia 68.8

 

[Francesco]

1. This is very difficult to answer because it depends entirely on what the makeup was of the Aegean profiles that caused the shift. We need much better sampling of Magna Graecia to determine this. The more Anatolian these magna graecian profiles were, the higher of a percentage of latin/etruscan makeup would've survived. If Magna Graecians were simply modern Sicilian and Dodecanese like and less so Anatolian then I would estimate very little survived as a modern Sicilian profile seems to have become standard to C. Italy in the years 31bc to 500ad.
[...]

I wouldn't be so sure about which profile was more common in Central Italy in the imperial era: if we exclude Rome, in which many samples we have come from foreigners necropolis like Isola Sacra, we still have few samples from that time, during which there were also different funerary practices that could distort the results. This new study with more than 100 samples from northern Italy, on the other hand, seems a bit more promising.

 

[Pax Augusta]

Pax Augusta: Ok, thanks. Did you use just those 3 Population averages as source populations to run those models?

Yep, Palermo Trapani. I didn't have time and I did it quickly. I repeat, it should not be taken literally. The Italian cline is still that but G25 is not as accurate, for several reasons. First because of a problem created by academic studies (northeastern Italian average is very similar to Sappada, which is a German language minority originally and was used as a sample for Veneto in the studies, many areas of the Po Valley are under-sampled, Liguria is based on one sample, Piedmont is most likely Val Borbera, Swiss Italian is rather strange, etc. etc.), and then because of other problems created by those who run G25.

 

[Pax Augusta]

A short nose would exclude much immigration from areas like (Roman Age) Anatolia or Syria, alleged by supporters of Roman Italy overrun by Near Eastern slaves and migrants.

To be clear, I am not supporting anything at all; I remain more interested in ancient history than in the ethnogenesis of modern Italians. But I can't stand aprioristic theses, and the first to work with aprioristic theses were the geneticists of the past. For years I found myself rather lonely in forums arguing that the Etruscans were indigenous, and simply because that was what the specialists argued, but few read the scholarly texts, and the few who do read cherrypick only what they are interested in proving. I find it rather ironic that we have the opposite problem today.

 

[Vitruvius]

Yes, of course, the chances are greater for the north but the north is the largest genetic cluster in Italy, it doesn't really make sense to generalize about the whole north. In fact a Spaniard-like profile still exists in northern Italy and is found in some Alpine valleys not mixed with German and Slavic language minorities but it is not at all clear how common this genetic profile is.

Northern Italian overlap today with Spain is very low compared to IA Latin overlap. I think you are referring to exceptions here more than anything. Yes there is still some, but it's not prevalent like we saw with the IA Etruscans.

I remind you that much of southern Italy was Italic before it became partially Greek and that the Samnites of Campania had a similar profile to the Latins and Etruscans was said by Max Planck's geneticist Alissa Mittnik in a public meeting, not by a forum user. Of course, we do not know, as I said, whether it will be confirmed by a published study. Moreover, the few northern Italian samples like Broion are in line with Etruscans and Latins. But let's wait for the Proto-Illyrians who repopulated northern Italy.

I'm well aware of the IA southern Italics and Mittnik's study. I've been specifically referring to Mittnik's samples if you've been following along. What we see is that those southern Italics of Pontecagnano become Aegean shifted prior to Roman occupation. I'm not sure what the relevance of this is as so far I've referred to continuity with N. Italics and Magna Graecians and not southern or Central IA italics.

On the topic of Northern Italics and Broion, there is definitely overlap with the modern N. Italian cluster. I've posted this several times before and this is not simply my conjecture but what the actual study PCA shows. As a reminder BA Broion is only represented by the darker squares here. Are you going to tell me that the average does not look anything like where N. Italy is labelled on this PCA to you? Because that's not at all what I see.

It is really very naive to think that the Latins and Etruscans represented a genetic anomaly as you do, and that only they disappeared and that only central Italy changed and the rest of Italy remained unaltered because so far geneticists have focused only on them.

I don't understand what the propensity here is for putting words in my mouth or making this claim. I have never called or considered the IA Latins or Etruscans a genetic anomaly. I think they are very typical of central Italians of their day and also those of the bronze age. What's more is that it's clear this aegean shift did not just happen to central Italy but also to southern Italy. You, yourself just finished telling me about how southern Italics were Etruscan or Latin IA like in Mittnik's yet unpublished paper.

That is not what the jury ruled.

Doesn't change the validity of my point. EEF-like ancestry is too high for this scenario. If you want to go by the poorly contrived conjectures like that of Moots or other authors, modern Italians are supposedly combinations of German and Levantine migrants. Is that your stance?

 

[Pax Augusta]

If you want to go by the poorly contrived conjectures like that of Moots or other authors, modern Italians are supposedly combinations of German and Levantine migrants. Is that your stance?

No, of course not.

I'm only answering that for now, because I can't spend my whole days on a forum.

 

[Pax Augusta]

On the topic of Northern Italics and Broion, there is definitely overlap with the modern N. Italian cluster. I've posted this several times before and this is not simply my conjecture but what the actual study PCA shows. As a reminder BA Broion is only represented by the darker squares here. Are you going to tell me that the average does not look anything like where N. Italy is labelled on this PCA to you? Because that's not at all what I see.

View attachment 14473

What you call "N. Italian cluster" is ITA_N + ITA_C, you can read it with difficulty in the PCA, most likely a cluster of samples from northern Italy plus TSI (Tuscan). Certainly it is an interesting PCA, compared to the Bronze Age for ITA_N + ITA_C there is a shift toward the Balkans and not toward Anatolia IA or the Levant but the sample that join the "N. Italian cluster" (ITA_N + ITA_C) is a low-coverage sample from "Grotta Regina Margherita" (Collepardo, Frosinone, southern Lazio), whose symbol is in the shape of a rhombus, and not from the "Grottina dei Covoloni del Broion" (Colli Berici, VIcenza, Veneto), which is represented by a square.

There are 11 samples from the Broion published in the study, although the starting samples were many more. 4 are from the Chalcolithic (CA), 2 from the Early Bronze Age (EBA) and 5 from the Bronze Age (BA). 2 out of the 5 Bronze Age samples, one of which is low coverage, end up between Romania and Moldova, while 3, and none of these are low coverage, are located further west of the Ita_N + Ita_C cluster, below the Spaniards, more or less in the position where the Etruscans and Latins are also found. Two EBA Broion samples also have this position. While those from the Chalcolithic are all within the Neolithic cluster.

 

[Vitruvius]

What you call "N. Italian cluster" is ITA_N + ITA_C, you can read it with difficulty in the PCA, most likely a cluster of samples from northern Italy plus TSI (Tuscan).

Yes, I agree. The modern "C Italian" location on the PCA appears highly northerly shifted as modern C. Italians typically bridge North and south as far as their positioning on a PCA, but we don't see that here. I too would assume this population is somewhere from Northern Tuscany for the fact that they plot so far north. Modern N. Italian is also visible in a typical location near Spaniards and practically overlapping the C. Italians which is peculiar. The authors, in my opinion should've used a more expansive sample for modern C. Italians.

Certainly it is an interesting PCA, compared to the Bronze Age for ITA_N + ITA_C there is a shift toward the Balkans and not toward Anatolia IA or the Levant but the sample that join the "N. Italian cluster" (ITA_N + ITA_C) is a low-coverage sample from "Grotta Regina Margherita" (Collepardo, Frosinone, southern Lazio), whose symbol is in the shape of a rhombus, and not from the "Grottina dei Covoloni del Broion" (Colli Berici, VIcenza, Veneto), which is represented by a square.

Yes, there's definitely no shift towards anywhere southwards for both N. and C. Italy in this Bronze age PCA, I agree. Both seem to show a significant eastern pull towards northern Balkan influence which is why I've postulated for Protoitalic to derive from the Danubian basin. Northern Italy shows a stronger pull in this regard compared to central, as we have two N. Italic samples that show significantly more Yamnaya/steppe input comparatively to all C. Italians of this time frame. When we ignore the earlier Chalcolithic and EBA phases and focus strictly on the later bronze age samples N. Italy averages (albeit inhomogeneously) where modern N. Italians and EMA Bardonechia samples are. I have doubts of this being any coincidence.

There are 11 samples from the Broion published in the study, although the starting samples were many more. 4 are from the Chalcolithic (CA), 2 from the Early Bronze Age (EBA) and 5 from the Bronze Age (BA). 2 out of the 5 Bronze Age samples, one of which is low coverage, end up between Romania and Moldova, while 3, and none of these are low coverage, are located further west of the Ita_N + Ita_C cluster, below the Spaniards, more or less in the position where the Etruscans and Latins are also found. Two EBA Broion samples also have this position. While those from the Chalcolithic are all within the Neolithic cluster.

I view the Chalcolithic and EBA as a transitory period in which Italy either hadn't absorbed, or was still in the process of absorbing Steppe related ancestry. I probably should've specified that I was only looking at and focusing on the later Bronze age samples. I don't think the positioning the low coverage sample is inaccurate, based off the higher coverage sample that plots just next to it. Ultimately we'll need more samples but I think the homogenization of the later BA cluster in Broion, including the two high steppe samples, is what ultimately creates a modern like profile for northern Italians. This is what I would expect to see emerge in the early iron age.

 

[Vitruvius]

No, of course not.

I'm only answering that for now, because I can't spend my whole days on a forum.

Glad to hear it. We're on the same page, then.

 

[Vitruvius]

I wouldn't be so sure about which profile was more common in Central Italy in the imperial era: if we exclude Rome, in which many samples we have come from foreigners necropolis like Isola Sacra, we still have few samples from that time, during which there were also different funerary practices that could distort the results. This new study with more than 100 samples from northern Italy, on the other hand, seems a bit more promising.

I do think there is a survivorship bias for this period and I agree with you in this regard. I have no doubts about that. Foreign representation during this era is too high for reason that Roman tradition was incineration with burials being seen as exotic and specific to middle eastern cultures at the time. These foreigners are mainly represented by the "tail" which extends south of the Dodecanese in my opinion and includes levantines, proper anatolians and a couple north africans. These types were transitory and did not survive or leave any permanence in the region as late antiquity shows, but Sicilian-like profiles definitely seemed to represent the typical central Italian after the widespread conversion to Christianity and the normalization of burial practices from the years 300-500AD. In addition, we also have a profile of one Imperial Pompeian who never received funerary rites and died in the Vesuvius eruption. This individual plotted roughly over top of modern Calabrese, similar to many of the contemporary Roman samples.

For this reason I'm of the mindset that a typical Imperial Roman of both S. and C. Italy was Southern Italian like as a result of this Aegean shift. We see it broadly in Etruria as well. The shift was less dramatic than most make it out to be (as a result of survivorship bias), but I do think it happened. If there are pockets of modern Italians that are representative of the IA Latin and Etruscan profile averages I'm unaware of them, but N. Italians plot very close for the reasons I mentioned prior. Again I don't see this as coincidence. There are clearly very close genetic ties between bronze age N. and C. Italy.

That being said if we get a new study which clearly demonstrates the IA Latin profile in other parts of Imperial era Italy, I'd be happy to reassess my position. My hypothesis so far is simply that the IA N. Italic profile will be much more significant in its contributions to modern Italians when compared to IA C. Italic.

 

[Francesco]

I do think there is a survivorship bias for this period and I agree with you in this regard. I have no doubts about that. Foreign representation during this era is too high for reason that Roman tradition was incineration with burials being seen as exotic and specific to middle eastern cultures at the time. These foreigners are mainly represented by the "tail" which extends south of the Dodecanese in my opinion and includes levantines, proper anatolians and a couple north africans. These types were transitory and did not survive or leave any permanence in the region as late antiquity shows, but Sicilian-like profiles definitely seemed to represent the typical central Italian after the widespread conversion to Christianity and the normalization of burial practices from the years 300-500AD. In addition, we also have a profile of one Imperial Pompeian who never received funerary rites and died in the Vesuvius eruption. This individual plotted roughly over top of modern Calabrese, similar to many of the contemporary Roman samples.

Inhumation was also more common in the oscan context and maybe the C6 cluster from Antonio et al., the one wich is more numerous and resemble modern southern italian, is mainly composed of individuals of oscan background whose anecestors assimilated (a lot of) aegean ancestry in the previous centiruies.

We know for sure that Rome was heavily influenced both demografically and even linguistically by oscans (for the italian speakers, this text could be interesting: https://www.studietruschi.org/wp-content/uploads/2021/06/SE01_18.pdf) and if the latter had assimilated a lot af greek ancestry, a C6 cluster in the city of Rome makes total sense (I think I'll open onother thread on the C6 cluster, since I don't want to go off topic).

The Campanian aristocracy was the first "foreign" element to be coopted in the Senate (right after the samnitic wars if i recall correctly). Presumably, these aristocrats also bought with them their clientes, so a rapid expansion of this element in Rome - and later maybe even in its colonies - is not so out of the question.

 

[Vitruvius]

Inhumation was also more common in the oscan context and maybe the C6 cluster from Antonio et al., the one wich is more numerous and resemble modern southern italian, is mainly composed of individuals of oscan background whose anecestors assimilated (a lot of) aegean ancestry in the previous centiruies.

We know for sure that Rome was heavily influenced both demografically and even linguistically by oscans (for the italian speakers, this text could be interesting: https://www.studietruschi.org/wp-content/uploads/2021/06/SE01_18.pdf) and if the latter had assimilated a lot af greek ancestry, a C6 cluster in the city of Rome makes total sense (I think I'll open onother thread on the C6 cluster, since I don't want to go off topic).

The Campanian aristocracy was the first "foreign" element to be coopted in the Senate (right after the samnitic wars if i recall correctly). Presumably, these aristocrats also bought with them their clientes, so a rapid expansion of this element in Rome - and later maybe even in its colonies - is not so out of the question.

I wasn't aware that inhumation was more prevalent in Oscan speakers. This is interesting. Do you have any resources on Oscan funerary rites during the imperial era? I'd love to check them out. I agree with you on the source of this ancestry deriving from the descendants of Oscan speaking Italics who first conquered and assimilated Magna Graecians. I'll take a look at your link and read through this too because it looks interesting.

Another important note is how massively the citizen body of Rome expands between the year 115BC and 28BC as per the census. We see a 10 fold expansion of Roman citizens from 390k to 4 million within this 87 year time frame. Several important events occurred in these years, including the granting of citizenship to nearly the full Italian peninsula, as well as Caesar introducing convenient local municipal citizen registration in contrast to prior laws which required all new citizens to travel to Rome to register. Even so, the fact that the citizen body expanded so dramatically leads me to believe that central Italy was the most underpopulated region of Italy during this time frame and that both the north and the south of the peninsula were significantly more populous. We have individual Magna Graecian city states numbering 300k hundreds of years prior to this era, which is very telling as a comparison. When we look at it from this perspective, it would not take much immigration from south to central to induce the significant genetic turnover we speak of. This isn't even taking into account regional differences in fertility rates which very well could have also played a roll.

Tenney Frank did an excellent write up explaining the importance of the municipal law changes in detail and provides the census figures here: https://homepages.uc.edu/~martinj/Latin/Roman_Population/Frank - Roman Census Statistics.pdf

 

[Pax Augusta]

I wasn't aware that inhumation was more prevalent in Oscan speakers.

Inhumation was the the funerary rite typical of all the Osco-Umbrian-speaking peoples.

 

[torzio]

male cremation.......baltic amber urns

https://student-journals.ucl.ac.uk/pia/article/278/galley/407/view/