Genomic and dietary discontinuities during the Mesolithic and Neolithic in Sicily

Mesolithic and Neolithic in Sicily

Highlights

1.Genetic transition between Early Mesolithic and Late Mesolithic hunter-gatherers.
2.A near-complete genetic turnover during the Mesolithic-Neolithic transition.
3.Exchange of subsistence practices between hunter-gatherers and early farmers.


Summary

Sicily is a key region for understanding the agricultural transition in the Mediterranean, due to its central position. Here, we present genomic and stable isotopic data for 19 prehistoric Sicilians covering the Mesolithic to Bronze Age periods (10,700-4,100 yBP). We find that Early Mesolithic hunter-gatherers (HGs) from Sicily are a highly drifted lineage of the Early Holocene western European HGs, while Late Mesolithic HGs carry ∼20% ancestry related to northern and (south)eastern European HGs, indicating substantial gene flow. Early Neolithic farmers are genetically most similar to farmers from the Balkans and Greece, with only ∼7% ancestry from local Mesolithic HGs. The genetic discontinuities during the Mesolithic and Early Neolithic match changes in material culture and diet. Three outlying individuals dated to ∼8,000 yBP, however, suggest that hunter-gatherers interacted with incoming farmers at Grotta dell’Uzzo, resulting in a mixed economy and diet for a brief interlude at the Mesolithic-Neolithic transition.

Link:https://www.sciencedirect.com/science/article/pii/S2589004222005144

Anfänger said:

Mesolithic and Neolithic in Sicily

Highlights

1.Genetic transition between Early Mesolithic and Late Mesolithic hunter-gatherers.
2.A near-complete genetic turnover during the Mesolithic-Neolithic transition.
3.Exchange of subsistence practices between hunter-gatherers and early farmers.


Summary

Sicily is a key region for understanding the agricultural transition in the Mediterranean, due to its central position. Here, we present genomic and stable isotopic data for 19 prehistoric Sicilians covering the Mesolithic to Bronze Age periods (10,700-4,100 yBP). We find that Early Mesolithic hunter-gatherers (HGs) from Sicily are a highly drifted lineage of the Early Holocene western European HGs, while Late Mesolithic HGs carry ∼20% ancestry related to northern and (south)eastern European HGs, indicating substantial gene flow. Early Neolithic farmers are genetically most similar to farmers from the Balkans and Greece, with only ∼7% ancestry from local Mesolithic HGs. The genetic discontinuities during the Mesolithic and Early Neolithic match changes in material culture and diet. Three outlying individuals dated to ∼8,000 yBP, however, suggest that hunter-gatherers interacted with incoming farmers at Grotta dell’Uzzo, resulting in a mixed economy and diet for a brief interlude at the Mesolithic-Neolithic transition.

Link:https://www.sciencedirect.com/science/article/pii/S2589004222005144

Click to expand...

Thanks for posting this paper, I have been waiting for 2 years for the published version to come out, I remember discussing the pre-print version back in 2020. Hopefully, the Genomes are available. I am particularly interested in how the Neolithic Sicilian samples in this paper overlap with the Neolithic Lazio-Rome samples from Antonio et al 2019. as well as the Bell Beaker Samples from Sicily and Northern Italy and Otzi the Iceman.

Cheers.

"early Neolithic farmers are genetically most similar to farmers from the Balkans and Greece, with only ∼7% ancestry from local Mesolithic HGs. "

:bigsmile:

That's basically minoan-like, with a tiny bit less CHG.

---

---

---

---

---

---

---

---

---

---

:sun:

Based on my previous analysis, and the findings of this paper; the vast majority of Sicilian DNA was brought in the Neolithic. I think all those EBA-LBA samples had been mixed into the population at large.

Jovialis said:

Based on my previous analysis, and the findings of this paper; the vast majority of Sicilian DNA was brought in the Neolithic. I think all those EBA-LBA samples had been mixed into the population at large.

Click to expand...

I tend to agree. Those Neolithic samples from this paper seem to very similar to the models you ran on ancient Italian samples using Dodecad.

Palermo Trapani said:

Thanks for posting this paper, I have been waiting for 2 years for the published version to come out, I remember discussing the pre-print version back in 2020. Hopefully, the Genomes are available. I am particularly interested in how the Neolithic Sicilian samples in this paper overlap with the Neolithic Lazio-Rome samples from Antonio et al 2019. as well as the Bell Beaker Samples from Sicily and Northern Italy and Otzi the Iceman.

Cheers.

Click to expand...

You're welcome.

I think the genomic data will be available here: https://www.ebi.ac.uk/ena/browser/view/PRJEB50762?show=reads

There are 4 males for which the haplogroups could be determined, one being the R1b Bell Beaker derived individual from the Bronze Age. Anything about the other threes haplogroups? Its written its in the supplementary, data s4, but I couldn't find a link to the supps.

Riverman said:

There are 4 males for which the haplogroups could be determined, one being the R1b Bell Beaker derived individual from the Bronze Age. Anything about the other threes haplogroups? Its written its in the supplementary, data s4, but I couldn't find a link to the supps.

Click to expand...

The Pre_Print version had the Y DNA Haplogroups in a supplement that had a link to access it. I think I may have posted it 2 years ago in a thread about the pre-print version of the paper. But, I think the pre-print did not have an R1b listed among the 4 Males in the sample, so additional testing must have resulted in them re-classifying the Bronze Age sample to R1b (UZZ57) and giving it a new ID #. Alternatively, the UZZ57 sample may be a new one that was not in the pre-print that they were able to sequence and one of the older samples, perhaps in the review process was dropped due to low-coverage, failed to meet quality control standards, etc.

In fact, UZZ61 is in the pre-print and version for publication was Y C1a2, UZZ40 and UZZ81 are in both versions, carried Y I and I2a2, respectively. UZZ33 on the other hand which was in the pre-print version and was defined as Y haplogroup H was not included in the version that went through peer review and is to be published.

In the supplements of the pre-print version, they clearly state that archeological evidence has determined human presence at Grotto dell Uzzo (in Trapani province of Sicily). Perhaps they had this Bronze age sample and decided to include it in published version to provide a longer time-series link from the Late Mesolithic up to the Bronze Age since there are 19 other Bronze Age Sicilians published in the literature, the Sicilian_Bell Beaker plus the 18 Bronze Age Sicilians from the Fernandes et al 2020 paper.

Jovialis said:

"early Neolithic farmers are genetically most similar to farmers from the Balkans and Greece, with only ∼7% ancestry from local Mesolithic HGs. "

:bigsmile:

That's basically minoan-like, with a tiny bit less CHG.

Click to expand...

This might be the reason why Neolithic Sicily is Minoan-like:

Recently, it was shown that some Early Neolithic farmers from peninsular Italy might retain an additional ancestral component related to early farmers from Iran (Ganj Dareh) and/or HGs from the Caucasus (CHG) (Antonio et al.,2019). This raises the possibility that they descended from a different group compared to the known western Mediterranean and central European farmers (Antonio et al., 2019; Mathieson et al., 2018)

Click to expand...

I wonder if the Anatolian_N source is coming from Tepecik Ciftlik instead of Barcin, but is this enough to explain the excess of CHG/Iran_N in Sicily or are there other sources? I am sure your knowledge on this issue is more profound than mine.

Anfänger said:

This might be the reason why Neolithic Sicily is Minoan-like:



I wonder if the Anatolian_N source is coming from Tepecik Ciftlik instead of Barcin, but is this enough to explain the excess of CHG/Iran_N in Sicily or are there other sources? I am sure your knowledge on this issue is more profound than mine.

Click to expand...

Anafinger: While they raise the CHG/Iran_Neolithic signal documented in Antonio et al 2019 in the Neolithic samples from Lazio, they don't address it in this study with the 7 Neolithic samples from Grotto Dell Uzzo_Sicily (Trapani), best I can tell (unless it is talked about in the supplements). They document (p.15) that the Neolithic Sicilians in this paper share genetic affinity with EN groups in the Balkans, as well as Central Europe associated with the continental route of the Early European Farmer expansion rather than the so called Mediterranean route. Note that as the authors state, nearly all of the EEF ancestry groups, including those associated with Mediterranean route are a subset of the genomic diversity found among EEF from Barcin and the North Agean (pp.3-4).

Are there Cardial samples from Iberia, say, which have been or are available for testing to see if they retain that CHG/Iran Neo signal, or did that get washed out by the intrusion of more WHG ancestry.

I ask because long ago Dienekes found that CHG/Iran Neo signal in Otzi, but not Gok 2.

Also, is it present in the Central European farmer samples, or groups like later Cucuteni?

The two possible scenarios are 1) it was present in both streams of the Neolithic, but was washed out in areas where there was more WHG intrusion because it was a relatively small component, or 2) one stream carried it and one stream didn't.

Regardless of the answer, the amount of CHG/Iran Neo wouldn't have been large enough to remove the group from the larger cluster of European farmers.

I'm juggling a lot right now. But I want to process the files, I'm disappointed some of these samples have too small to analyze, like the EBA UZZ057.

PRJEB50762

Jovialis said:

I'm juggling a lot right now. But I want to process the files, I'm disappointed some of these samples have too small to analyze, like the EBA UZZ057.

PRJEB50762

Click to expand...

If someone wants to take a crack at them, that would be great.

Seems like they all get errors in WGSextract, what a shame.

The WGSExtract Beta version processes them, but the results make no sense, and do not reflect the study.

They are bad, but can be used. The admixture results cannot be taken serious in most cases. Eurogenes K15 Extended Oracle:

UZZ069

Components %
Atlantic 36,26
Baltic 7,22
West_Med 24,37
Northeast_African 32,16

Using 1 populations approximation
1 100% Spanish_Andalucia @ 37,203
2 100% Spanish_Castilla_La_Mancha @ 37,489
3 100% French_Basque @ 37,603
4 100% Spanish_Aragon @ 37,956
5 100% Southwest_French @ 37,991
6 100% Spanish_Cantabria @ 38,316
7 100% Spanish_Valencia @ 38,458
8 100% Spanish_Castilla_Y_Leon @ 38,781
9 100% Spanish_Murcia @ 39,399
10 100% Spanish_Extremadura @ 39,642

UZZ079

Components %
North_Sea 17,90
Atlantic 13,29
Baltic 42,23
Eastern_Euro 16,02
Oceanian 3,50
Sub-Saharan 7,05

Using 1 populations approximation
1 100% Lithuanian @ 11,908
2 100% Latvian @ 12,474
3 100% Belorussian @ 15,310
4 100% Ukraine_East @ 16,120
5 100% Estonian_Polish @ 16,434
6 100% Russian_Smolensk @ 17,279
7 100% Southwest_Russian @ 17,471
8 100% Estonian @ 17,765
9 100% Ukrainian_Belgorod @ 17,903
10 100% Polish @ 18,979

UZZ080

Components %
North_Sea 46,60
Baltic 30,25
Eastern_Euro 21,88
Southeast_Asian 0,62
Sub-Saharan 0,66

Using 1 populations approximation
1 100% Estonian @ 26,339
2 100% Finnish @ 26,769
3 100% North_Swedish @ 28,578
4 100% East_Finnish @ 28,883
5 100% Ukrainian @ 29,911
6 100% Aland_Sweden @ 30,014
7 100% Southwest_Finnish @ 30,045
8 100% Ukrainian_Lviv @ 30,341
9 100% Latvian @ 30,518
10 100% RU_Pinega @ 31,003

UZZ082

Components %
North_Sea 40,35
Atlantic 37,52
Baltic 9,51
Eastern_Euro 7,95
Sub-Saharan 4,67

Using 1 populations approximation
1 100% West_Scottish @ 11,770
2 100% Irish @ 12,228
3 100% Orcadian @ 12,682
4 100% Scotland @ 13,484
5 100% Southeast_English @ 14,103
6 100% England @ 14,474
7 100% North_Dutch @ 14,494
8 100% Danish @ 15,062
9 100% Dutch @ 15,217
10 100% West_Norwegian @ 16,157

UZZ096

Components %
Baltic 67,33
Sub-Saharan 32,67

Using 1 populations approximation
1 100% Lithuanian @ 55,984
2 100% Latvian @ 56,098
3 100% Ukraine_East @ 58,058
4 100% Belorussian @ 58,892
5 100% Estonian_Polish @ 59,605
6 100% Ukrainian_Belgorod @ 59,802
7 100% Russian_Smolensk @ 59,813
8 100% Southwest_Russian @ 60,012
9 100% Ukrainian_Lviv @ 61,000
10 100% Ukrainian @ 61,633

UZZ034

Components %
West_Asian 50,31
East_Med 29,07
Oceanian 20,61

Using 1 populations approximation
1 100% Turk_Meskhetian @ 22,952
2 100% Georgian_Laz @ 23,416
3 100% Turkish_Trabzon @ 24,488
4 100% Georgian @ 24,924
5 100% Turkish_East @ 25,342
6 100% Abhkasian @ 25,468
7 100% Armenian_East @ 26,630
8 100% Kurd @ 26,790
9 100% Greek_Trabzon @ 27,603
10 100% Azeri_Dagestan @ 28,299

UZZ074 (Very poor)

Components %
Red_Sea 100,00

Using 1 populations approximation
1 100% Saudi @ 79,993
2 100% Bedouin @ 89,547
3 100% Moroccan @ 90,884
4 100% Yemenite_Jewish @ 90,986
5 100% Ethiopian_Tigray @ 92,553
6 100% Egyptian @ 92,557
7 100% Algerian @ 92,576
8 100% Mozabite_Berber @ 92,829
9 100% Ethiopian_Amhara @ 92,896
10 100% Tunisian @ 93,390

UZZ075

Components %
North_Sea 3,94
Atlantic 12,74
West_Med 52,13
West_Asian 4,22
East_Med 22,60
Red_Sea 2,82
Southeast_Asian 1,07
Oceanian 0,48

Using 1 populations approximation
1 100% Sardinian @ 11,723
2 100% France_Corsica @ 31,644
3 100% Italy_Lazio @ 35,670
4 100% Mozabite_Berber @ 35,823
5 100% Moroccan @ 35,874
6 100% Tuscan @ 36,022
7 100% Italy_Marche @ 36,218
8 100% Italy_Tuscany @ 36,384
9 100% West_Sicilian @ 36,707
10 100% Algerian_Jewish @ 36,781

UZZ087 (Very poor)

Components %
Southeast_Asian 37,96
Oceanian 0,46
Sub-Saharan 61,58

Using 1 populations approximation
1 100% Bantu_N.E. @ 46,473
2 100% Biaka_Pygmy @ 46,630
3 100% Luhya @ 46,805
4 100% Bantu_S.E. @ 47,045
5 100% Bantu_S.W. @ 47,422
6 100% Mandenka @ 48,006
7 100% Mbuti_Pygmy @ 48,741
8 100% Yoruban @ 51,209
9 100% San @ 54,677
10 100% Sudanese @ 62,814

The samples look like this:

# rsid chromosome position genotype
rs147404388 1 564654 .. ./. ./. G/G
rs9283150 1 565508 .. ./. ./. G/G
rs369202065 1 569388 .. ./. ./. ./.
rs199476136 1 569400 TT ./. ./. ./.
rs3094315 1 752566 .. A/A ./. A/A

^^Thanks, I was able to do it with an old version of WGSextract, but the results looked wacky.

UZZ069 is 32% "Northeast" Africa? I sincerely doubt it.

That's due to poor sample quality, and lack of SNPs.