Unraveling Roman Mobility: Archaeogenomic Insights from Anatolia to the Italian Peninsula

[lockdownboredom]

https://open.metu.edu.tr/bitstream/handle/11511/113454/MscThesis_Elifnaz_Eker_Final_NEW.pdf

 

[Riverman]

They confirm once more the Anatolian-Levantine shift in the Roman era. But in this paper they also propose a limited Southern European influx into Anatolia, making the exchange a bit more bidirectional, though the proportions are vastly different if comparing admixture into Italia (and Greece for that matter) vs. Southern Europeans-Balkan into Anatolia.

 

[Pax Augusta]

It is not strictly a scientific paper but rather a master’s thesis from a Turkish university, not even a doctoral dissertation. The thesis reiterates the conclusions of previously published studies (Antonio, Lazaridis, Moots, Posth...) and includes 95 newly sequenced genomes from various regions of Anatolia, covering the period from the 6th century BCE to the 10th century CE. The historical section contains numerous significant errors. In the future, this thesis could be revised to become a peer-reviewed article.

I read quickly and it doesn't seem to me that this work offers significant new insights, it merely repeats established findings with several errors. It aligns with the conclusions of prior studies: central Italy during the Iron Age was populated by locals; there was a gradual increase in migrants from the eastern Mediterranean; during the Roman Empire, these migrants caused a demographic shift; and the arrival of migrants from central and northern Europe in central Italy after the fall of the Roman Empire contributed to the formation of modern Italians. However, since all the new samples come from Anatolia, it cannot be said that it further confirms the Anatolian-Levantine shift in Roman times. Such confirmation would have required samples from the European regions of the Roman Empire rather than Anatolia. As Lazaridis has already argued, the findings are consistent with the hypothesis that Anatolia in the Second Iron Age was a significant source of eastern migrants during the Roman Imperial period. Regarding Italy, Iron Age samples from many areas of northern and southern Italy are still lacking. Geneticists continue to focus on central Italy, as the most comprehensive studies to date have concentrated on Rome and its surrounding areas. Some specific studies on Southern Italy, for example, would be very useful at this point.

Interesting that a Turkish university on the one hand advocates a migrationist thesis for Europe and on the other a genetic continuity for Turkey, a bit in contradiction to the recent study of modern Turks where samples from the Balkans to the areas closest to the Levant of Turkey were used, trying to emphasize the idea that Turks range from Europe to West Asia.

 

[Vallicanus]

One significant historical error is that the Lombards (Langobards) conquered Rome. They often besieged the city but never captured it.

 

[Jovialis]

It is not strictly a scientific paper but rather a master’s thesis from a Turkish university, not even a doctoral dissertation. The thesis reiterates the conclusions of previously published studies (Antonio, Lazaridis, Moots, Posth...) and includes 95 newly sequenced genomes from various regions of Anatolia, covering the period from the 6th century BCE to the 10th century CE. The historical section contains numerous significant errors. In the future, this thesis could be revised to become a peer-reviewed article.

I read quickly and it doesn't seem to me that this work offers significant new insights, it merely repeats established findings with several errors. It aligns with the conclusions of prior studies: central Italy during the Iron Age was populated by locals; there was a gradual increase in migrants from the eastern Mediterranean; during the Roman Empire, these migrants caused a demographic shift; and the arrival of migrants from central and northern Europe in central Italy after the fall of the Roman Empire contributed to the formation of modern Italians. However, since all the new samples come from Anatolia, it cannot be said that it further confirms the Anatolian-Levantine shift in Roman times. Such confirmation would have required samples from the European regions of the Roman Empire rather than Anatolia. As Lazaridis has already argued, the findings are consistent with the hypothesis that Anatolia in the Second Iron Age was a significant source of eastern migrants during the Roman Imperial period. Regarding Italy, Iron Age samples from many areas of northern and southern Italy are still lacking. Geneticists continue to focus on central Italy, as the most comprehensive studies to date have concentrated on Rome and its surrounding areas. Some specific studies on Southern Italy, for example, would be very useful at this point.

Interesting that a Turkish university on the one hand advocates a migrationist thesis for Europe and on the other a genetic continuity for Turkey, a bit in contradiction to the recent study of modern Turks where samples from the Balkans to the areas closest to the Levant of Turkey were used, trying to emphasize the idea that Turks range from Europe to West Asia.

I shared this post on X, great insights.

 

[Jovialis]

It seems that many researchers, both academic and hobbyist, continue to ignore the existence of Magna Graecia and overlook the fact that ChromosomePainter can discern haplotype-ancestry clusters, despite citing a study that explicitly uses it.

 

[kingdavid]

nice
but where are the uniparental markers ?

 

[MOESAN]

It is not strictly a scientific paper but rather a master’s thesis from a Turkish university, not even a doctoral dissertation. The thesis reiterates the conclusions of previously published studies (Antonio, Lazaridis, Moots, Posth...) and includes 95 newly sequenced genomes from various regions of Anatolia, covering the period from the 6th century BCE to the 10th century CE. The historical section contains numerous significant errors. In the future, this thesis could be revised to become a peer-reviewed article.

I read quickly and it doesn't seem to me that this work offers significant new insights, it merely repeats established findings with several errors. It aligns with the conclusions of prior studies: central Italy during the Iron Age was populated by locals; there was a gradual increase in migrants from the eastern Mediterranean; during the Roman Empire, these migrants caused a demographic shift; and the arrival of migrants from central and northern Europe in central Italy after the fall of the Roman Empire contributed to the formation of modern Italians. However, since all the new samples come from Anatolia, it cannot be said that it further confirms the Anatolian-Levantine shift in Roman times. Such confirmation would have required samples from the European regions of the Roman Empire rather than Anatolia. As Lazaridis has already argued, the findings are consistent with the hypothesis that Anatolia in the Second Iron Age was a significant source of eastern migrants during the Roman Imperial period. Regarding Italy, Iron Age samples from many areas of northern and southern Italy are still lacking. Geneticists continue to focus on central Italy, as the most comprehensive studies to date have concentrated on Rome and its surrounding areas. Some specific studies on Southern Italy, for example, would be very useful at this point.

Interesting that a Turkish university on the one hand advocates a migrationist thesis for Europe and on the other a genetic continuity for Turkey, a bit in contradiction to the recent study of modern Turks where samples from the Balkans to the areas closest to the Levant of Turkey were used, trying to emphasize the idea that Turks range from Europe to West Asia.

Are they not considering only ancient Anatolians pop's?Sure modern Turks are another thing. Complicated history.

 

[Pax Augusta]

Are they not considering only ancient Anatolians pop's?Sure modern Turks are another thing. Complicated history.

The time span considered ranges from the 6th century BCE to the 10th century CE, although at the beginning of the paragraph on continuity in Anatolia, there is an ambiguous statement: “The analysis showed us that Anatolian genomes consistently overlapped with modern South Asian populations across all four time periods,” suggesting continuity up to the present. This statement is ambiguous because it relies on the assumption that there are no significant genetic differences among various South Asian populations. However, this does not truly prove continuity. The focus then returns to the time span from the 6th century BCE to the 10th century CE, just before the linguistic ancestors of the Turks arrived in Anatolia. As you also noted, the genetic structure of the Turks is far more complex, and the idea of continuity from the 6th century BCE is very weak. This is especially true since Turkey, in addition to hosting even today numerous ethnic groups, became the seat of the Ottoman Empire from the 14th century onward, an empire no less extensive than that of ancient Rome.

Another issue with this MSc thesis is the apparent absence of ancient samples from Constantinople. In contrast, the vast majority of samples from the Italian Peninsula come from Rome and its surroundings, all already published, with the majority drawn from the studies by Antonio (2019, 2024) and Moots (2022). Only one sample was used from Posth 2021 (for technical issue?), while Aneli 2022 contained no new samples as I remember. Essentially, the entire thesis heavily relies on the conclusions of Antonio and Moots.

This tendency among geneticists to mutually support each other’s work is well known; it was already evident in the 1990s studies, which relied solely on mtDNA from modern samples and produced studies that were, in reality, quite misleading. It is undeniable that the Roman Empire was cosmopolitan, with many foreigners relocating to Rome in pursuit of opportunity. However, I would like to see geneticists provide more concrete evidence and fewer studies built on speculation or aligned with their preconceived notions. And then if geneticists would study a little more ancient history, archaeology and anthropology, it wouldn't be bad at all. But I know, one cannot demand too much.

 

[HereToLearn]

I have a general observation that applies to all studies, not just this one: while I'm not a statistician, I understand that small sample sizes can lead to unreliable conclusions. Specifically, limited samples are more prone to overrepresenting extreme cases, which can distort the true incidence rates found in the general population.

It is, for example, apparent to me that a sample size around 10 (or even 20?) should not even be put out there for a serious debate. Am I wrong?

 

[MOESAN]

I have a general observation that applies to all studies, not just this one: while I'm not a statistician, I understand that small sample sizes can lead to unreliable conclusions. Specifically, limited samples are more prone to overrepresenting extreme cases, which can distort the true incidence rates found in the general population.

It is, for example, apparent to me that a sample size around 10 (or even 20?) should not even be put out there for a serious debate. Am I wrong?

a small s(say 5 to 10)

I have a general observation that applies to all studies, not just this one: while I'm not a statistician, I understand that small sample sizes can lead to unreliable conclusions. Specifically, limited samples are more prone to overrepresenting extreme cases, which can distort the true incidence rates found in the general population.

It is, for example, apparent to me that a sample size around 10 (or even 20?) should not even be put out there for a serious debate. Am I wrong?

A small sample (say: 5 to 10) in the same place at the same period of time, showing homogeneity without familial links is reliable at the allover autosomes level. Otherwise, I agree, it isn't of great value.

 

[Pax Augusta]

I have a general observation that applies to all studies, not just this one: while I'm not a statistician, I understand that small sample sizes can lead to unreliable conclusions. Specifically, limited samples are more prone to overrepresenting extreme cases, which can distort the true incidence rates found in the general population.

It is, for example, apparent to me that a sample size around 10 (or even 20?) should not even be put out there for a serious debate. Am I wrong?

Good question. In the case for example of this table that comes from this master's thesis, the 22 samples from southern Europe (600 BCE-1CE) are nothing but the usual ones already published by Antonio 2019 and Moots 2022, i.e., the sum of the 10 samples from Latium (1 from Veio, 3 from a coastal cliff from Civitavecchia, 6 from different necropolis from Latium vetus), and 1 from Abruzzo plus an additional 11 samples from Moots 2023 from Tarquinia, obviously also Latium. All of these samples together are from at least three different ethnic groups (Etruscans, Latins, and the one from Abruzzo may be a Proto-Picene), plus the possible non-locals, of which at least a couple are condidered Celts and another couple Levantine.

These 22 individuals come from at least 7 or more different localities in Latium, and they have very different dates, the oldest ones (the Protovillanovan from Abruzzo and the Villanovan-Etruscan from Veio) being dated around 900-800 B.C., while for example the two fully Levantine samples from a necropolis in Tarquinia are dated between 150 and 50 B.C., between 700 and 800 years after the oldest ones. Are 22 samples therefore sufficient? Actually, no. But they are sufficient for certain narratives. There are three additional cases: two individuals from a Latin necropolis show a mixed genetic signal from the eastern Mediterranean but still cluster with modern Italians, likely in the central-southern cluster. One individual from a Civitavecchia necropolis appears to have partial North African ancestry, possibly Punic or Sardinian-Punic. In total, 15 out of 22 individuals seem to be entirely local, while 7 are non-local. However, of these 7 non-locals, 2 are European, and the number could be considered 4 if we include those who plot with modern Italian populations. Moots repeatedly claims that 40% of the individuals are non-local, originating from other parts of the Mediterranean. In reality, the proportion of non-locals is lower. The presence of non-locals is undeniable, but the stated percentage does not align with the published data.

 

[Jovialis]

I put appropriate capitalization for the title of the thread, I know it was just a copy and paste from the paper. But I disliked the fact that it looked like virtual screaming.

 

[Vitruvius]

The idea of strong genetic continuity in Anatolia from 600BC-1000AD is something pretty well evidenced and this can be further extended to the Chalcolithic era. I don't see that part as controversial. All bets are off after 1000AD and the migration of the Turks of course.

This study has really very little to do with Italy from my perspective. While Anatolian migrants may have made up a minor part of the contribution to the genetic shift between IA and imperial Central Italy - the vast majority of it was more likely from Magna Graecian Greek Italiotes who had IA/Classical era origins in Greece proper. Those that did come from Anatolia would've been Hellenized, regardless, and the genetic differences were rather small between the two . By the Roman era we now know there was no genetic distinguishable autosomal difference between Peloponesian Greeks of Tenea and Greco-Anatolians, meaning that the minority of individuals who plot like anatolians around Rome in the Imperial era may not actually themselves be direct anatolian transplants.

 

[arxaiogenetiki.blogspot]

The idea of strong genetic continuity in Anatolia from 600BC-1000AD is something pretty well evidenced and this can be further extended to the Chalcolithic era. I don't see that part as controversial. All bets are off after 1000AD and the migration of the Turks of course.

We've had proto-/early Turkic samples for a while and we can look for that ancestry in modern Turks. It ranges from 0 to 5-10% depending on the area. If you assume a C. Asian proxy then just double it to 0-20%. It's safe to assume modern Turks are largely (>80%) IA Anatolians, although I haven't looked into the possibility of Armenian/Caucasus influence, which would offer a similar result to what already exists. (And I'm aware there are sides very invested into it being true, and others sides very invested in wanting it not to be true, so there's no point going there unless a paper addresses it explicitly)

ed.* with some Balkan/Greek influence in the West although this could already exist there since the IA or Hellenistic times.

 

[Vitruvius]

We've had proto-/early Turkic samples for a while and we can look for that ancestry in modern Turks. It ranges from 0 to 5-10% depending on the area. If you assume a C. Asian proxy then just double it to 0-20%. It's safe to assume modern Turks are largely (>80%) IA Anatolians, although I haven't looked into the possibility of Armenian/Caucasus influence, which would offer a similar result to what already exists. (And I'm aware there are sides very invested into it being true, and others sides very invested in wanting it not to be true, so there's no point going there unless a paper addresses it explicitly)

ed.* with some Balkan/Greek influence in the West although this could already exist there since the IA or Hellenistic times.

Influence from the Armenian highlands diluted the neolithic Anatolian base genome heavily during the Chalcolithic era. If we use the four neolithic Armenian/Azeri genomes we have access to as a source for this profile the ancestry displacement of Barcin Neolithic ancestry in Anatolia is something like 60% and we begin to see this type of caucasian ancestry creep into Greece and Sicily even by the end of the neolithic (in very small proportions). Anatolia stabalizes after the Chalcolithic and retains its Caucasian heavy structure up until the arrival of the Turks in ~1000AD, but during the full extent of the bronze age, iron age and antiquity it perpetually displaces the Greek genetic structure as time marches on. This influence is felt strongest in the islands and peloponese. Northern Greece remains to be seen - there will certainly be influence there but it will probably be lesser. Similarly the Greeks which arrived in Italy during the Magna Graecian era also displaced the local Oscan/IA latin like structure in favor of their Caucasian heavy aegean derived ancestry. The ladder is of course retained and normalized in Southern italy to this day, albeit under the Italian ethnological context instead of that of Greek.

 

[arxaiogenetiki.blogspot]

Influence from the Armenian highlands diluted the neolithic Anatolian base genome heavily during the Chalcolithic era. If we use the four neolithic Armenian/Azeri genomes we have access to as a source for this profile the ancestry displacement of Barcin Neolithic ancestry in Anatolia is something like 60% and we begin to see this type of caucasian ancestry creep into Greece and Sicily even by the end of the neolithic (in very small proportions). Anatolia stabalizes after the Chalcolithic and retains its Caucasian heavy structure up until the arrival of the Turks in ~1000AD, but during the full extent of the bronze age, iron age and antiquity it perpetually displaces the Greek genetic structure as time marches on. This influence is felt strongest in the islands and peloponese. Northern Greece remains to be seen - there will certainly be influence there but it will probably be lesser. Similarly the Greeks which arrived in Italy during the Magna Graecian era also displaced the local Oscan/IA latin like structure in favor of their Caucasian heavy aegean derived ancestry. The ladder is of course retained and normalized in Southern italy to this day, albeit under the Italian ethnological context instead of that of Greek.

Yep, that's what I meant by "Chalcolithic". Late Neolithic Greece is synchronous with this, and in Greece it was actually a pretty decent amount. In Minoans it even reached roughly half(!) of their ancestry.

IIrc Anatolia retains a stable average of CHG/Iran but has increased fluctuations during the Iron Age (see Lazaridis et al 2022). Basically a Mycenaean influence resulting in slightly lower CHG/Iran in some parts. Then the Hellenistic/Late Antiquity era is similar, same average but with outliers. I think this mostly corresponds with geography.

Btw wasn't Bronze Age Sicily already somewhat similar to Mycenaeans? But I might remember wrong.

 

[Vitruvius]

Yep, that's what I meant by "Chalcolithic". Late Neolithic Greece is synchronous with this, and in Greece it was actually a pretty decent amount. In Minoans it even reached roughly half(!) of their ancestry.

IIrc Anatolia retains a stable average of CHG/Iran but has increased fluctuations during the Iron Age (see Lazaridis et al 2022). Basically a Mycenaean influence resulting in slightly lower CHG/Iran in some parts. Then the Hellenistic/Late Antiquity era is similar, same average but with outliers. I think this mostly corresponds with geography.

Btw wasn't Bronze Age Sicily already somewhat similar to Mycenaeans? But I might remember wrong.

The Minoans certainly had influence from BA anatolia, yes, but it was much less than what is seen in LBA Crete under the Mycenaean context.

BA Sicily also had some minor incoming BA anatolian-like influence but this was more likely from Greece itself than direct Anatolian population movements. You can think of Sicily as downstream from Greece which itself was downstream from Anatolia, which during the Chalcolithic was downstream from the Armenian highlands. Ancestry seems to expand from east to west in this fashion.

 

[arxaiogenetiki.blogspot]

The Minoans certainly had influence from BA anatolia, yes, but it was much less than what is seen in LBA Crete under the Mycenaean context.

BA Sicily also had some minor incoming BA anatolian-like influence but this was more likely from Greece itself than direct Anatolian population movements. You can think of Sicily as downstream from Greece which itself was downstream from Anatolia, which during the Chalcolithic was downstream from the Armenian highlands. Ancestry seems to expand from east to west in this fashion.

~50% was less than in LBA Mycenaean Crete? Or do you mean something else? Mycenaeans themselves in general have an Eastern pull which is additional to the one Minoans had (otherwise their CHG/Iran would be lower than it is due to dilution) but it's definitely lower than 50% iirc.

 

[Vitruvius]

~50% was less than in LBA Mycenaean Crete? Or do you mean something else? Mycenaeans themselves in general have an Eastern pull which is additional to the one Minoans had (otherwise their CHG/Iran would be lower than it is due to dilution) but it's definitely lower than 50% iirc.

This is what I'm getting at. This "Eastern" pull was just further progressive Anatolian genetic influence. There was also a simultaneous increase of EHG ancestry which corresponds to influence from the northern or north east Balkans, which should not be ignored as well. The differential input of both of these ancestries will likely create a north/south cline once we have more northern Greek samples to compare against our southern Greeks for these same time periods.