New conference on Bronze Age mobility in Europe

[Angela]

This is what I meant, Berun. Bear in mind that the Sardinian samples used in almost every genetic analysis are the HGDP ones from the isolated mountain plateau, not the western coastal cities with a bit of "other" ancestry. Some early PCAs were able to pick it up, even using only two dimensions. You can see that they are shifted left or west in the direction of WHG. They also overlap with the northern farmers, who were about 20% WHG.

You can also see it in ADMIXTURE when they drill down into the different areas of Sardinia. This is from Chiang e al using whole genomes, which is going to be more accurate. As you get away from that highland plateau, the "Yamnaya" increases.

Top half are the most isolated communities. In the coastal communities you're going to have the migration from the mainland, i.e. U-152. It would have been nice if they'd used more references, but this is at least more accurate than what we had before. Doubtless, some of the more minority ancestry, i.e. Punic, extra "Iran Neo" is going to get dumped into the closest of the three reference samples, or split up amongst them, but I doubt there was much of it.

Ed. Sorry, this is from Chiang et al Supplementary material.

Sorry again. I forget that not everyone is familiar with the different areas of Sardinia. You can locate the last six towns on the map, the ones with increased Yamnaya and decreased WHG. They're all the areas of the plain or the coast. The others are in or near the Ogliastra and the Barbagia.

 

[Pygmalion]

I doubt "urban" immigrations like those of Phoenicians and their Libyan associates left much or even any non-negligible impact in most places, including Sardinia, except where they really colonized the lands extensively. Cities were demographic sinks in antiquity, they usually tended to decrease and not increase their genetic impact over time, and that must have been especially true in the case of port cities, even more exposed to epidemics. And slaves, especially slaves in urban areas, must have left even less genetic contribution on the long term.

There were two main waves of levantine settlement in Sardinia. One which occurred in the 8th century bc when the Phoenicians from Tyre needed to set some permanent emporiums in the island since they began visiting it with intense frequency, and this first wave is usually considered to be a peaceful commercial migration according to current archaeologists. Archaeologically these emporiums yielded not only phoenician pottery but also large amounts of nuragic pottery and hybrid nuragic-phoenician pottery since their very foundation, which confirms mixed marriage and the presence of nuragic sardinian communities within those centers. And in most cases these cities were built inside local nuragic settlements such as that recently discovered at Nora (which as the name suggests was pre-phoenician) or that of Sulky. During this time large quantities of nuragic and hybrid nuragic-levantine pottery appears in phoenician settlements in North Africa, Sicily and Iberia, and nuragic pottery and artifact continued to be exported and sometime locally made in Tyrrhenian Italy and beyond. This fruitful collaboration between the Nuragics and the Phoenicians continued in the 7th century bc when some new settlements with mixed nuragic-phoenician planimetry and pottery were constructed, such as the recently excavated one at Nuraghe Sirai, which was founded over a previous nuragic settlement: https://www.academia.edu/31629324/L_età_del_Ferro_del_Nuraghe_Sirai_Layers_1_2016_

The second wave which occurred in the late 6th century bc would be that of the Carthaginians, which is considered to be a violent one by archaeologists, as both the destruction layers in the sardinian cities and the written sources suggest. During this time sardinian exports decreases drastically and the nuragic culture ceased to exist as a separate material culture. However a recent study on a 3rd century bc malaria ridden corpse from the punic necropolis of Caralis has shown that both his Y DNA (I2a1a1) and MTDNA were pre-phoenician/nuragic, and that thus at least by that time locals were part of the aristocracy of some of those punic or punic-conquered cities.

 

[Ygorcs]

True, but that isn't always the case - with urban immigration during the Roman period, mass (mainly Syrian and Jewish) immigration into Southern Italy left a huge presence that is clearly seen today in both modern phenotype and autosomal genetic admixture. I would have thought Sicily would have a lot of Greek and Phoenician ancestry too, but I agree that the impact of Mediterranean colonists tends to be overstated.

I am quite sure that impact is widely overestimated by some people. Sicilians and especially South Italians should have much more Levant_Neo and Iran_Neo and correspondingly less EEF and even less steppe ancestry than they do have. They should also have closer affinities to modern East Mediterranean people, and they do not, they cluster together with other Europeans and in most PCAs they are only slightly shifted to Near Eastern and North African populations, IIRC they are actually in some of the PCAs less close to West Asian samples than the Greek (mainly islanders). Besides, some of the main actual presence of apparently non-European origin includes E-M81, which probably has more to do with Berber and Arabized North African expansions in the Middle Ages, where some people did migrate, but as elite and also often as settlers in newly gained lands.

But I doubt most of that contribution came from Near Eastern slaves and even from urban poor migrants from the Near East. Any population growth was a huge feat in the ancient world, and most poor people in especially disease-prone areas like cities barely managed to leave any offspring in the long term. I have a real-life example from comparably much more developed times: Brazil from the mid 16th century to very recent years. Well, then, Brazil received some 5 million African slaves from 1530 to 1850, and it received roughly 6.5 million Eurasian, most of them after 1860, so with much less time to grow their numbers. But if you look at Brazil's African autosomal impact it is just around 20-25%. In Y-DNA lineages it is less than 10%, actually more like 5-7% only. Now project a similar situation back into Antiquity, only much worse than the modern and contemporary era. Of course not all Syrians and Jews migrated as slaves or very poor workers, but most of them were not exactly middle class far less elite.

 

[Cato]

What timeframe is ancient? Do you have a link to that abstract? I presume it isn't Reich as that said only Iran ancestry was new.

i've found the full abstract

The sequencing of ancient DNA (aDNA) has provided new understanding into human movement and demography for many regions around the globe. For mainland Europe, ancient DNA studies have revealed a dynamic history, with major inferred population influxes due to Neolithic and Bronze Age expansions. The population of the Mediterranean island of Sardinia has been notable in these studies–typically aDNA samples of the early Neolithic on mainland Europe cluster with modern Sardinian samples. The standing model is that Sardinia had a high influx of Neolithic ancestry followed by relative isolation from the mainland and subsequent Bronze Age expansions. To gain further insight, we analyze genome-wide capture data (~1.2 millions SNPs) of 26 ancient Sardinians spanning the Neolithic, Copper Age, and Bronze Age, including individuals from Sardinia's Nuragic culture. Merging this novel data with 998 previously studied aDNA samples from across Europe and throughout the last ten millennia, we are able to place the ancient Sardinian samples into the broader context of the peopling of Europe. We confirm that ancient Sardinian samples show a strong affinity to early Neolithic samples and a near complete absence of the “Steppe” ancestry associated with Bronze Age expansions on the mainland. Interestingly, we also detect elevated affinities with pre-Neolithic peoples of Europe. Moreover, we studied genetic change through time within Sardinia. To this end, we analyzed whole-genome sequence data from approximately 1,500 modern Sardinian individuals, densely sampled across much of the island. Using our ancient samples enables us to detect significant signs of recent admixture, in particular with a strong influence from the Mediterranean region. We also find that populations from the more isolated mountainous provinces of Sardinia are less admixed and have experienced high levels of genetic drift. Overall, our analysis allows us to shed new light on the intriguing history of the peopling of Sardinia"

So western and north-eastern Sardinians have significant post nuragic admixture, interesting....i guess that steppe level in nuragic was 0-4% like today Ogliastran plus extra WHG and the usual EEF

Utilizzando Tapatalk

 

[Jovialis]

I find this kind of analgsis pretty cpnfusing, not even mentioning this weird label "Neolithic" (Neolithic what? Anatolia? Europe? Levant?). Also confusing is the blending of Neolithic with Mesolithic and even Paleolithic admixtures. I think only a much more chronologically and geographically proximate set of proxy populations could be at least a bit clarifying. Iran-Neolithic was not the same as Iran-Mesolithic, BA steppe was not the same as EHG. I somehow guess that the pretty inflated WHG would be less prevalent using Chalcolithic populations of Europe and West Asia. Some ancestry, especially if it came from Anatolia, Caucasua and the Levant, may be assigned to the closest hunter-gatherer West Eurasian population, and the least admixed seems to be WHG. Maybe I am just rambling, but what I wanted to say is just that this model does not seem to be very useful for me. Also did you check if it has a really good fit? Not all such analyses are anywhere close to the truth. The algorithms just try to make their best considering the comparative sources it was given.

I took a look at the spread sheet for this calculator, and the modeling makes even less sense to me, when it analyzes actual ancient samples. I don't think this calculator is useful either.

 

[berun]

I doubt "urban" immigrations like those of Phoenicians and their Libyan associates left much or even any non-negligible impact in most places, including Sardinia, except where they really colonized the lands extensively. Cities were demographic sinks in antiquity, they usually tended to decrease and not increase their genetic impact over time, and that must have been especially true in the case of port cities, even more exposed to epidemics. And slaves, especially slaves in urban areas, must have left even less genetic contribution on the long term.

If you check up its history when Romans get the island the locals uprise against the new rulers and land holders carry Lybian names.

 

[Cato]

At this point i think that Romans/Italics had a bigger genetic impact in Sardinia than Phoenicians/Punics considering the Y DNA and mtDNA and autosomal data...

Western Sardinians have more Yamnaya than original Ogliastra (and less WHG)..Yamnaya can only come from Europe.

Utilizzando Tapatalk

 

[Angela]

At this point i think that Romans/Italics had a bigger genetic impact in Sardinia than Phoenicians/Punics considering the Y DNA and mtDNA and autosomal data...

Western Sardinians have more Yamnaya than original Ogliastra (and less WHG)..Yamnaya can only come from Europe.

Utilizzando Tapatalk

I agree. "Punic" y dna in Sardinia is very much a minority and the autosomal likewise. When you don't have mass migrations, it's difficult to change the genetic pattern.

As for the people of the Ogliastra and Barbagia, they remind me of those Andaman islanders who just killed that missionary. Strangers who dared to enter that isolated plateau just disappeared. Even the Romans with all their might left it strictly alone.

 

[Cato]

I was checking the 3 components graph by Chiang (latest version)...the data for Ogliastra are curious, for example Ilbono and Lanusei have 7% and 5% steppe while Gairo, a near village, 0%. I've read other studies and apparently people from that area rarely mix with each other, very high endogamy

Utilizzando Tapatalk

 

[Saetrus]

I was checking the 3 components graph by Chiang (latest version)...the data for Ogliastra are curious, for example Ilbono and Lanusei have 7% and 5% steppe while Gairo, a near village, 0%. I've read other studies and apparently people from that area rarely mix with each other, very high endogamy

Sardinians only show "Steppe" when you're forcing Sardinians to pick from LBK EN, Loschbour and Yamnaya as if they were the only components that existed, in unsupervised mode ADMIXTURE shows no Steppe in Sardinians only the Iran-related ancestry.


We could alternatively set your grandmother as a component and force all European populations to be modeled as partially your grandmother and Italians would be a high % your grandmother while Estonians only a tiny % but that doesn't mean all Europeans actually descend from your grandmother.

 

[Ailchu]

theoretically every form of migration, how small it may be, changes the genetic pattern of a population. it can be a little amount of people every few years and they can change the gene-pool a lot just takes a bit longer. but i wonder what kind of role drift plays here. if admixtures are small isn't it easier for them to get lost over time?

it's kind of strange that Loschbour decreases with increasing yamna but the EEF stays the same or inreases. this would mean that the imigrants were higher in EEF as the sardinians but had only yamna instead of additional whg.

 

[Angela]

Sardinians only show "Steppe" when you're forcing Sardinians to pick from LBK EN, Loschbour and Yamnaya as if they were the only components that existed, in unsupervised mode ADMIXTURE shows no Steppe in Sardinians only the Iran-related ancestry.


We could alternatively set your grandmother as a component and force all European populations to be modeled as partially your grandmother and Italians would be a high % your grandmother while Estonians only a tiny % but that doesn't mean all Europeans actually descend from your grandmother.

Whose unsupervised admixture runs? I really wish when posting you would cite the source of your information. It's one thing if it's from an academic paper. It's another if it's some unknown amateur. Also, what populations were used, what was the "goodness" of fit.

Then, it depends on the samples we're discussing. If you look upthread you'll see that in the big cities steppe can reach 10%. In some places in the Ogliastra and Barbagia it's zero or two-three percent. In addition, unless "your" runs were done on whole genomes, as Chiang's were, "yours" are not going to be as accurate.

Geographical distributions of modern yDna can be misleading in terms of time of arrival, direction of flow etc. but you can't totally discard it either. There's about 20% of R1b in Sardinia, the majority of it U-152. It's just outright silly to say there's no steppe at all in Sardinians.

 

[Saetrus]

Whose unsupervised admixture runs?

The very study same Chiang study.

Zero of the eastern European purple component in all Sardinian samples.

 

[Angela]

The very study same Chiang study.

Zero of the eastern European purple component in all Sardinian samples.

Saetrus, think about it. Do you think Chiang would contradict themselves this way?

Please read the key for the graphic and the text of the paper. The blue is "European" according to the key. What does that mean? They don't say, and I'm not precisely sure. The green, according to them, is found in not only the Caucasus, but the Middle East, North Africa, and "Southern Europe". I doubt that represents Iran Neo, unless you think that southern Iberia is 25% Iran Neo? They don't explain the significance of the purple, but while it may be eastern European that doesn't equate to steppe, unless you think the 40% and up of Neolithic ancestry in eastern Europe doesn't count? You can't give your own name to components.

Plus, right after the graph they continue to say that there is some steppe ancestry on Sardinia.

Is there a lot? No, which means the U-152 was almost certainly another male mediated migration. In addition, it obviously drifted to its current numbers.

The author has made the 2018 paper on the Sardinians available.
https://www.nature.com/articles/s41...NsTWDsHwCVhA9SIV60FGYsxedL6SaA0LXgE7RJ0fP5A==

 

[Cato]

I was checking the 3 components graph by Chiang (latest version)...the data for Ogliastra are curious, for example Ilbono and Lanusei have 7% and 5% steppe while Gairo, a near village, 0%. I've read other studies and apparently people from that area rarely mix with each other, very high endogamy

Utilizzando Tapatalk

the study

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0004654

Utilizzando Tapatalk

 

[Cato]

it's kind of strange that Loschbour decreases with increasing yamna but the EEF stays the same or inreases. this would mean that the imigrants were higher in EEF as the sardinians but had only yamna instead of additional whg.

Central Italians? Tuscans are EEF+Yamnaya in the Haak 2015 paper as the Spanish while Bergamo have some extra WHG

Utilizzando Tapatalk

 

[Ailchu]

Central Italians? Tuscans are EEF+Yamnaya in the Haak 2015 paper as the Spanish while Bergamo have some extra WHG

Utilizzando Tapatalk

but noone has as high EEF values as the sardinians in that study. so an increase of yamna should decrease EEF. but maybe it's just drift, some of the steppe could have been lost over time in favor of EEF, and sample bias that make it look like EEF did not change. or sardinians were just not homogenous even before new imigration.

 

[Pygmalion]

If you check up its history when Romans get the island the locals uprise against the new rulers and land holders carry Lybian names.

Ogliastrini carry italian names today and they're certainly not the same as mainland Italians. Of course the people living in coastal towns like Hamsicora / Hampsichora would have absorbed punic culture after almost 300 years of punic domination, Hamsichora's son Iostus doesn't even have a punic name though. And even the name Amsichora itself being berber is debetable, scholars have compared it to a berber women who appers in one of Plautus' comedies (180 bc) and is called Ampsigurra, I could make the same argument for Amsicora and Anassagora having the same root and that's a greek name. None of the sources mention him having berber ancestry, he's always defined dux sardorum.

 

[A. Papadimitriou]

If you check up its history when Romans get the island the locals uprise against the new rulers and land holders carry Lybian names.

Can you post the source for that claim?

 

[Pygmalion]

Can you post the source for that claim?

He's referring to Hamsicora https://en.wikipedia.org/wiki/Hampsicora

His name has been compared to Amsigurra, a berber woman in one of Plauto's commedies, and to a lake in North Africa. But usually the same scholars who make this comparison tend to emphasize Sardinians' North african ancestry excessively and unscientifically, going as far as saying that Sardinia was populated since prehistory from North Africa, or even taking Cicero's slanderous speech as a historical source, because in that speech he says that Sardinians were born out of the union of Phoenicians and Africans. Eitherway anyone who's familar with academical literature about ancient Sardinian knows that it's a common trend, especially among italian and sardinian scholars, to claim that Sardinians have North african origins.