Population structure in Italy using ancient and modern samples

Thanks, yeah, that all makes sense.

One of the other things I learned in this book on Carthage is more detail on Greece following the Bronze Age collapse: worse than the Dark Ages in Central and Western Europe after the fall of Rome: severe depopulation, little to no trade or manufacturing, and they even forgot how to write.

I wonder if there was a little continued migration from the east? Or, the sample from Spain might have been an islander. We really need more Greek samples from later periods both on the mainland and on the islands. Particularly interesting would be samples from the period of Magna Graecia in the large Greek city states which sent settlers to Sicily and Southern Italy.

Have you done any analysis of people from Crete? Are there available samples from Rhodes or the Dodecanese?

Yes, in my opinion the Empuries Greek sample does look like an Aegean islander, it seems to have even less steppe ancestry and to be much more Minoan-shited, but also less Iranian/Caucasian-shifted than the Mycenaean. I wonder how IA Cretans were like autosomally because of that...

Unfortunately I can only work with the samples used in the Global25 datasheets. They're so poor, too broad, as far as the Greek population is concerned. They just subdivide them into "Greek" (I assume mainland Greece, but there's quite a bit of structure even there), "Greek_Trabzon", "Greek_Central_Anatolia" and "Greek_Crete". There are also Cypriots. It's a pity they don't have anything more region-specific. The "Greek" average population sample looks particularly problematic, in each and every model I have made they look like they have a huge chunk of "northern" ancestry not found in Mycenaean or Empuries Greek samples. I presume many if not most of the samples used for that "Greek cluster" are northern Greeks.

Modern Cretans, even more than modern Greeks (I reckon mainlanders, as I said), have some really strange results using the same old reference populations (the former preference to Yamnaya and Catacomb becomes a preference for CWC, mainly CWC_Poland and CWC_Baltic). I assume it doesn't have only to do with subsequent autosomal changes (like "northern" influences, mainly Slavic ones), but also with additional millennia of genetic drift making the ancient fits much less perfect, so the algorithms are much more likely to choose other "unlikely" aDNA samples to explain the modern genetic structure, especially since virtually all BA steppe samples (whether they are CWC, BB, Yamnaya, Catacomb, Sintashta etc.) are very similar to each other. Anyway, they seem to be much more Levant and Caucasus/Iran-shifted than the Mycenaean Greek samples. So Aegean islanders are the source of part of the Iran and Levant affinities of Italians? I think that's likely. As for steppe ancestry, it's clear that both mainland Greeks (mainly north in this datasheet? I believe so) and Cretans received a significant input from more steppe-rich populations since the BA.

Here are some new calculations I have just done using only the reference steppe-related populations that appeared in the ancient Mycenaean and Empuries samples:

[1] "distance%=1.1921 / distance=0.011921"

Greek_Crete

Kura-Araxes_Kalavan 31.30
Levant_N 17.10
Tisza_LN 14.90
Balaton_Lasinja_CA 12.50
Yamnaya_Ukraine 6.75
Yamnaya_Bulgaria 6.10
Tepe_Hissar_ChL 4.75
Yamnaya_Samara 3.30
Balkans_N 1.70
Minoan_Lasithi 1.60

[1] "distance%=1.8034 / distance=0.018034"

Greek

Anatolia_EBA_Isparta 34.15
Tisza_LN 24.95
Yamnaya_Bulgaria 14.70
Yamnaya_Ukraine 10.10
Comb_Ceramic_Estonia 5.20
Catacomb 5.05
Greece_N 3.75
Levant_N 2.10

[1] "distance%=1.0167 / distance=0.010167"

Cypriot

Levant_N 25.05
Armenia_ChL 16.90
Kura-Araxes_Kalavan 15.80
Hajji_Firuz_ChL 14.80
Tisza_LN 7.05
Tepe_Hissar_ChL 5.65
Balkans_N 4.95
Starcevo_N 4.45
Catacomb 4.10
Anatolia_EBA_Isparta 1.25



[1] "distance%=1.612 / distance=0.01612"

Mycenaean

Greece_N 36.60
Balkans_N 22.95
Kura-Araxes_Kalavan 17.85
Yamnaya_Ukraine 8.80
Anatolia_EBA_Ovaoren 6.35
Maykop 3.55
Vucedol_no_steppe 2.55
Morocco_EN 1.35

[1] "distance%=2.018 / distance=0.02018"

Minoan_Lasithi

Greece_N 38.15
Anatolia_EBA_Isparta 21.40
Vucedol_no_steppe 12.25
Barcin_ChL 10.70
Balkans_N 9.15
Anatolia_EBA_Ovaoren 5.15
Hajji_Firuz_ChL 3.20
 
Italy is quite unlike other places in Europe in that there isn't that much going on south of the Po. Some coastal settlements sure, but inland there's just the mountain shepherds whose material culture didn't change much throughout the metal ages. No complex cultural layers etc. indicative of migrations.

I'm not an expert on soils and stuff, but I think the reason for this might be that southern Italy isn't exactly prime agricultural estate (again, Sicily excepted). I believe evidence suggests that the Greeks didn't venture much beyond the coasts. They probably saw no real reason to do so, and mobile highland tribes are of course notoriously difficult to deal with.

So basically you think the Neolithic Central Italians and South Italians were already pretty much like the modern inhabitants (significant Levantine, Iranian, steppe ancestry and everything else), since the inland people's material culture didn't change much throighout the Metal Ages? That's basically what I can take from your perspective on the impossibility of successive migrations to Italy contributing to the cumulative non-EEF admixtures in those regions. I find that scenario very unlikely. Also, I believe you might be underestimating the positive effects of a dynamic, increasingly cosmopolitan coastal culture in Italy perhaps leading to their descendants to gradually become dominant also in numbers (and therefore in genetic impact).
 
Yes, in my opinion the Empuries Greek sample does look like an Aegean islander, it seems to have even less steppe ancestry and to be much more Minoan-shited, but also less Iranian/Caucasian-shifted than the Mycenaean. I wonder how IA Cretans were like autosomally because of that...

Unfortunately I can only work with the samples used in the Global25 datasheets. They're so poor, too broad, as far as the Greek population is concerned. They just subdivide them into "Greek" (I assume mainland Greece, but there's quite a bit of structure even there), "Greek_Trabzon", "Greek_Central_Anatolia" and "Greek_Crete". There are also Cypriots. It's a pity they don't have anything more region-specific. The "Greek" average population sample looks particularly problematic, in each and every model I have made they look like they have a huge chunk of "northern" ancestry not found in Mycenaean or Empuries Greek samples. I presume many if not most of the samples used for that "Greek cluster" are northern Greeks.

Modern Cretans, even more than modern Greeks (I reckon mainlanders, as I said), have some really strange results using the same old reference populations (the former preference to Yamnaya and Catacomb becomes a preference for CWC, mainly CWC_Poland and CWC_Baltic). I assume it doesn't have only to do with subsequent autosomal changes (like "northern" influences, mainly Slavic ones), but also with additional millennia of genetic drift making the ancient fits much less perfect, so the algorithms are much more likely to choose other "unlikely" aDNA samples to explain the modern genetic structure, especially since virtually all BA steppe samples (whether they are CWC, BB, Yamnaya, Catacomb, Sintashta etc.) are very similar to each other. Anyway, they seem to be much more Levant and Caucasus/Iran-shifted than the Mycenaean Greek samples. So Aegean islanders are the source of part of the Iran and Levant affinities of Italians? I think that's likely. As for steppe ancestry, it's clear that both mainland Greeks (mainly north in this datasheet? I believe so) and Cretans received a significant input from more steppe-rich populations since the BA.

Here are some new calculations I have just done using only the reference steppe-related populations that appeared in the ancient Mycenaean and Empuries samples:

[1] "distance%=1.1921 / distance=0.011921"

Greek_Crete

Kura-Araxes_Kalavan 31.30
Levant_N 17.10
Tisza_LN 14.90
Balaton_Lasinja_CA 12.50
Yamnaya_Ukraine 6.75
Yamnaya_Bulgaria 6.10
Tepe_Hissar_ChL 4.75
Yamnaya_Samara 3.30
Balkans_N 1.70
Minoan_Lasithi 1.60

[1] "distance%=1.8034 / distance=0.018034"

Greek

Anatolia_EBA_Isparta 34.15
Tisza_LN 24.95
Yamnaya_Bulgaria 14.70
Yamnaya_Ukraine 10.10
Comb_Ceramic_Estonia 5.20
Catacomb 5.05
Greece_N 3.75
Levant_N 2.10

[1] "distance%=1.0167 / distance=0.010167"

Cypriot

Levant_N 25.05
Armenia_ChL 16.90
Kura-Araxes_Kalavan 15.80
Hajji_Firuz_ChL 14.80
Tisza_LN 7.05
Tepe_Hissar_ChL 5.65
Balkans_N 4.95
Starcevo_N 4.45
Catacomb 4.10
Anatolia_EBA_Isparta 1.25



[1] "distance%=1.612 / distance=0.01612"

Mycenaean

Greece_N 36.60
Balkans_N 22.95
Kura-Araxes_Kalavan 17.85
Yamnaya_Ukraine 8.80
Anatolia_EBA_Ovaoren 6.35
Maykop 3.55
Vucedol_no_steppe 2.55
Morocco_EN 1.35

[1] "distance%=2.018 / distance=0.02018"

Minoan_Lasithi

Greece_N 38.15
Anatolia_EBA_Isparta 21.40
Vucedol_no_steppe 12.25
Barcin_ChL 10.70
Balkans_N 9.15
Anatolia_EBA_Ovaoren 5.15
Hajji_Firuz_ChL 3.20

Very interesting. Can't wait to get more ancient Greek samples.

The old "academic" Greek sample is not terribly helpful for these purposes. It's from Thessaloniki (ancient Salonica) in Greek Macedonia. Samples from the Peloponnese obviously are in the possession of the authors of the paper on the genetics of that area, but either he won't release them or no one cares to include them.

1624108_orig.png


The small stretch on the eastern side of the Italian peninsula, ie the Adriatic side, is the area where Maciamo found the highest levels of J2 in Italy.

map-of-ancient-greek-world.jpg
 
The old "academic" Greek sample is not terribly helpful for these purposes. It's from Thessaloniki (ancient Salonica) in Greek Macedonia. Samples from the Peloponnese obviously are in the possession of the authors of the paper on the genetics of that area, but either he won't release them or no one cares to include them.

Ah, now that makes A LOT of sense. I was having trouble reconciling with the idea that mainland Greeks look so different from South Italians and Cretan Greeks in these Global25 average population samples (compare below). They look way too northern, as if having a really substantial post-Mycenaean genetic flow. East Sicilians have a lower distance from the Mycenaeans than those (Thessaloniki) Greeks, though that does not necessarily imply that's entirely because they descend from Mycenaeans more than those Greeks (it could be just a more similar admixture composition). On ther other hand, Cretan look too "eastern". What could explain so much extra Kura-Araxes-like and Levant_BA-like ancestry even already including Mycenaean + Minoan_Lasithi or Mycenaean + Minoan_Lasithi + Empuries? I really don't know.

It's a real pity such a historically important area as Greece doesn't have more aDNA samples and a much more regionalized distribution of samples in these datasheets.

[1] "distance%=1.3689 / distance=0.013689"
[1] "distance%=1.0389 / distance=0.010389"[1] "distance%=1.3687 / distance=0.013687"
Cypriot Greek_Crete Greek



Levant_BA_South 32.3Mycenaean 30.9Mycenaean 31.80
Kura-Araxes_ARM_Kalavan 27.6Kura-Araxes_ARM_Kalavan 24.0Latvia_BA 29.00
Minoan_Lasithi 19.8Levant_BA_South 17.4Minoan_Lasithi 23.25
Mycenaean 15.2Minoan_Lasithi 13.8Kura-Araxes_ARM_Kalavan 12.30
Latvia_BA 5.0Latvia_BA 13.8Levant_BA_South 3.65

Including the Empuries sample...

[1] "distance%=1.1886 / distance=0.011886"
[1] "distance%=0.9303 / distance=0.009303"[1] "distance%=1.1675 / distance=0.011675"

Cypriot Greek_Crete Greek



Levant_BA_South 32.05Mycenaean 23.95Iberia_Northeast_Empuries2 31.95
Iberia_Northeast_Empuries2 29.65Kura-Araxes_ARM_Kalavan 20.70Latvia_BA 26.95
Kura-Araxes_ARM_Kalavan 25.45Iberia_Northeast_Empuries2 20.65Mycenaean 20.50
Mycenaean 6.90Levant_BA_South 16.65Anatolia_Isparta_EBA 12.65
Minoan_Lasithi 3.20Latvia_BA 12.40Kura-Araxes_ARM_Kalavan 5.85
Latvia_BA 2.75Anatolia_Isparta_EBA 5.65Levant_BA_South 2.10
 
Ah, now that makes A LOT of sense. I was having trouble reconciling with the idea that mainland Greeks look so different from South Italians and Cretan Greeks in these Global25 average population samples (compare below). They look way too northern, as if having a really substantial post-Mycenaean genetic flow. East Sicilians have a lower distance from the Mycenaeans than those (Thessaloniki) Greeks, though that does not necessarily imply that's entirely because they descend from Mycenaeans more than those Greeks (it could be just a more similar admixture composition).

It's a real pity such a historically important area as Greece doesn't have more aDNA samples and a much more regionalized distribution of samples in these datasheets.

There are three samples being widely used already. One is all from Thessaloniki, the other from Crete and there is a third sample which is from a data bank and cannot be traced to any particular region afaik, so it could be from different regions. Now, if see the PCA below, you will notice that, among the Greek samples, the Cretans are the closest to Myceneans although they do not plot with them (The grey dots that do, are Sicilian and Maltese). Myceneans and Minoans are between modern Greeks and modern Middle Easterners. Unless the first samples are not representative of the Cretan and Thessaloniki populations, I would not expect modern continental Greeks to end up more Middle Eastern than Cretans. And since Sicilians are closer to Myceneans than Cretans are, the chance that continental Greeks will be between Sicilians and Myceneans is slim.

nature23310-f1.jpg



In fact, if you look at the leaked image posted above, assuming it is correct, it seems they already have continental Greek samples from several regions and most of the Greeks that are visible are quite far from Myceneans. Only some are kind of close to the northwesternmost Mycenean sample (and maybe some more Greek marks are covered under the yellow group), but some other modern samples (Sicilians??) and some Romans are at a similar distance, if not closer. There could be several explanations for this. This is the problem with autosomal DNA, it is impossible to trace ancestry to through time without a complete record of genetic profiles from all times and populations.

It will be interesting to discover how Greeks from different regions differ autosomally from each other.

 
Ah, now that makes A LOT of sense. I was having trouble reconciling with the idea that mainland Greeks look so different from South Italians and Cretan Greeks in these Global25 average population samples (compare below). They look way too northern, as if having a really substantial post-Mycenaean genetic flow. East Sicilians have a lower distance from the Mycenaeans than those (Thessaloniki) Greeks, though that does not necessarily imply that's entirely because they descend from Mycenaeans more than those Greeks (it could be just a more similar admixture composition). On ther other hand, Cretan look too "eastern". What could explain so much extra Kura-Araxes-like and Levant_BA-like ancestry even already including Mycenaean + Minoan_Lasithi or Mycenaean + Minoan_Lasithi + Empuries? I really don't know.

It's a real pity such a historically important area as Greece doesn't have more aDNA samples and a much more regionalized distribution of samples in these datasheets.
Here is a study that includes modern Peloponnesian samples from every region of the Peloponnese. Many in fact cluster with Sicilians, and seem to be in between Sicilians and Cretan/Dodecanese Greeks, just like geography suggests, https://www.nature.com/articles/ejhg201718. This is not an ancient sampled study, but it does give you an idea of how close modern southern Italians are to modern southern Greeks, which is obviously suggestive of the ancient respective populations as well.
 
Here is a study that includes modern Peloponnesian samples from every region of the Peloponnese. Many in fact cluster with Sicilians, and seem to be in between Sicilians and Cretan/Dodecanese Greeks, just like geography suggests, https://www.nature.com/articles/ejhg201718. This is not an ancient sampled study, but it does give you an idea of how close modern southern Italians are to modern southern Greeks, which is obviously suggestive of the ancient respective populations as well.

Nothing unexpected from a comparison that leaves out all the rest of the Balkans. Of course Greeks are similar to Sicilians on a European level.

In any case, we should wait for the Italy paper since they seem to have all the answers, if the leaks are true.
 
So basically you think the Neolithic Central Italians and South Italians were already pretty much like the modern inhabitants (significant Levantine, Iranian, steppe ancestry and everything else), since the inland people's material culture didn't change much throighout the Metal Ages? That's basically what I can take from your perspective on the impossibility of successive migrations to Italy contributing to the cumulative non-EEF admixtures in those regions. I find that scenario very unlikely. Also, I believe you might be underestimating the positive effects of a dynamic, increasingly cosmopolitan coastal culture in Italy perhaps leading to their descendants to gradually become dominant also in numbers (and therefore in genetic impact).

No, I think the southerners either came with the Apennine or the descendent Fossa culture.

I think you're overestimating the importance of cities. Even Ravenna, Rome and Syrakus did not change the natural north-south cline in Italy. Ancient cities were frequently depleted.
 
The the early Apennine culture also roughly coincides with the appearance of J2a in Pannonia and J2b in Dalmatia and Sardinia. Not sure what happened there honestly, but archaeologically there are strong similiarities between the Balkans, the Carpathians and Italy in the MBA.
 
Last edited:
Here is a study that includes modern Peloponnesian samples from every region of the Peloponnese. Many in fact cluster with Sicilians, and seem to be in between Sicilians and Cretan/Dodecanese Greeks, just like geography suggests, https://www.nature.com/articles/ejhg201718. This is not an ancient sampled study, but it does give you an idea of how close modern southern Italians are to modern southern Greeks, which is obviously suggestive of the ancient respective populations as well.

Indeed. Wish he'd included mainland Southern Italy, but the differences aren't big. They probably also overlap with the Peloponnese.



stxHjn9.png


Also interesting in terms of our discussions:
qxPjWMq.png


Does Eurogenes never include these samples in his "calculators"? Why not?

It seems the new paper is from Stanford. I sure hope Spencer Wells is not involved. He's been wrong almost as often as Eurogenes, and knows almost as little about ancient history as does the latter.
 
I love, just absolutely love, how neither Eurogenes nor anybody at anthrogenica has made the slightest reference to the fact that they were absolutely and completely wrong about the autosomal signature of the Etruscans.

At least "Agamemnon" has had the decency to make himself scarce. The rest have no honor whatsoever.

Yet, they continue to make dogmatic pronouncements without even having the papers and samples in front of them. They know exactly what happened and why.

I wonder if it has occurred to these geniuses that on top of everything else, it's going to be very difficult to figure out if the group found at Ostia, to speculate wildly, were "locals", or resident merchants from elsewhere. The only recent paper which has addressed this issue is the Langobard one. I certainly hope that they did isotope analysis on these finds.

Don't bother to accuse me of not wanting my husband to be partly descended from Jews/Semites. His best friends, my best friend, people closer to us than family in some ways, are Jews. We'd be proud to be related to them.

This is about "evidence"; finding it, piecing it together without biases. Think about the people who lead this "charge": a documented anti-semite, nordicist/slavicist who said Southern Italians should be kicked out of Europe, and a pro-Arab agitator and another anti-semite who hates his Sicilian father. Oh, and they know next to nothing about ancient history. Could they have been any more wrong about the Etruscans? Why was I right? Because I know the archaeology and the history and I didn't let my personal preferences influence my thinking. I've spent my life doing that and I wasn't about to become a different person because it has to do with Italian ethnogenesis. Anyone who has read my posts over the years will, in fact, not be surprised to hear that if I'm to be completely honest I'm a bit disappointed that Etruscans have apparently turned out to be this "northern". I'd actually have preferred them to be more "Cretan" like. I didn't let that preference influence my thinking, however. I just wish others were as honest.
 
Please correct me if I'm wrong, but aren't the Etruscan more or less the indigenous peoples of the central Italian peninsula? Are they related to the old Neolithic peoples of Anatolia and so on?
 
No, I think the southerners either came with the Apennine or the descendent Fossa culture.

I think you're overestimating the importance of cities. Even Ravenna, Rome and Syrakus did not change the natural north-south cline in Italy. Ancient cities were frequently depleted.

Oh I see. I asked that because you said in your previous comment that there had hardly been any major changes in the Metal Ages as a whole. The Apennine culture is a very intriguing one for me, indeed.

As for my reference to "dynamic, increasingly cosmopolitan coastal culture in Italy", I'm not referring necessarily to cities. I'm referring to the increasingly productive (economic progress ultimately derived mainly from primary sector activities: agriculture, animal husbandry, timber extraction, mining etc.) rural zones. Cities only ever appeared as a consequence of significant and earlier rural progress.
 
My comment was very civil. He thinks that Italics were the natives from the neolithic era. That's simply not possible and certainly that's not what the PCA shows.

Greeks in the PCA have a different symbol from the rest of southern Europe (Greek Macedonia, Greek Thessaly...). Starting from the Greeks, almost everything else is reconstructed.


The original PCA

HM671qo.png


Interesting! So Etruscans were close to modern Iberians, North Italians and Tuscans.
 
stxHjn9.png


Also interesting in terms of our discussions:
qxPjWMq.png

It's interesting how much Cretans are shifted toward modern Anatolians and Levantines in comparison to the other Greek samples. The same pattern is seen using ancient (BA) Anatolian and Levantine samples. That cannot be attributed to Minoan ancestry, because even including the Minoan_Lasithi sample the Cretans still require extra Levantine and especially BA Anatolian admixture. Was there something in the post-Minoan history of Crete that I don't know?
 
Please correct me if I'm wrong, but aren't the Etruscan more or less the indigenous peoples of the central Italian peninsula? Are they related to the old Neolithic peoples of Anatolia and so on?


According to their genetic position the Etruscans seem to be a mixture of Neolithic European + Bronze age Bell Beaker and probably also with steppe ancestry. More or less as they are today the North Italians or Tuscans. Maybe Etruscans just a little more to the west than the modern population.
 
Anyone who has read my posts over the years will, in fact, not be surprised to hear that if I'm to be completely honest I'm a bit disappointed that Etruscans have apparently turned out to be this "northern". I'd actually have preferred them to be more "Cretan" like. I didn't let that preference influence my thinking, however. I just wish others were as honest.

Based on how many samples can we draw a conclusion on the origin of the Etruscans? Till now two unpublished studies were leaked, pre announced. These have limited amount of Iron Age samples, some leaks write about Neolithic ancestries of Iran and Levant plus Steppe ancestry, and some are talking about the ancestry of Early Farmers plus Bell Beaker ancestry. Without knowing the historical and archaeological background for each sample, and having so few samples, and not knowing the locations and the exact time of the individuals, and not knowing anything about their Y-DNA, how can we draw conclusions so easily? Which of the leaked ancestries are equal to the Proto Etruscans, and which to the Proto Latins?
 
According to their genetic position the Etruscans seem to be a mixture of Neolithic European + Bronze age Bell Beaker and probably also with steppe ancestry. More or less as they are today the North Italians or Tuscans. Maybe Etruscans just a little more to the west than the modern population.

Which one of the three belongs to the Proto Etruscans and the Proto Etruscan language?
 
Based on how many samples can we draw a conclusion on the origin of the Etruscans?

There have already been studies on Etruscan samples that had supported what the leak claims.
Genetics is not saying anything new, scholars have been thinking for years that the Etruscans were native.
 

This thread has been viewed 326385 times.

Back
Top