PDA

View Full Version : Genetic signatures of the Hellenic colonisation in southern Italy and Sicily



Angela
15-07-15, 21:24
The Greeks in the West: genetic signatures of the Hellenic colonisation in southern Italy and Sicily, Tofanelli et alhttp://www.nature.com/ejhg/journal/vaop/ncurrent/abs/ejhg2015124a.html

I would have been very grateful for this once, but all the ancient dna has spoiled me. Also, it's only uniparental markers. Still...

"Greek colonisation of South Italy and Sicily (Magna Graecia) was a defining event in European cultural history, although the demographic processes and genetic impacts involved have not been systematically investigated. Here, we combine high-resolution surveys of the variability at the uni-parentally inherited Y chromosome and mitochondrial DNA in selected samples of putative source and recipient populations with forward-in-time simulations of alternative demographic models to detect signatures of that impact. Using a subset of haplotypes chosen to represent historical sources, we recover a clear signature of Greek ancestry in East Sicily compatible with the settlement from Euboea during the Archaic Period (eighth to fifth century BCE). We inferred moderate sex-bias in the numbers of individuals involved in the colonisation: a few thousand breeding men and a few hundred breeding women were the estimated number of migrants. Last, we demonstrate that studies aimed at quantifying Hellenic genetic flow by the proportion of specific lineages surviving in present-day populations may be misleading."

Hopefully, the full paper will be released soon, but until then, the Supplementary Info is here:
http://www.nature.com/ejhg/journal/vaop/ncurrent/suppinfo/ejhg2015124s1.html?url=/ejhg/journal/vaop/ncurrent/abs/ejhg2015124a.html

Hauteville
16-07-15, 13:26
Aren't there already y-dna frequency table available?

Aaron1981
16-07-15, 15:15
Similar YDNA results to what we've seen before but the R1b seems much lower in Albania and Turkey from previous results I have seen.

Some takeaways:

-Northern Greece seems heavily Slavic in the modern era. This is seen with high rates of I2-M423 and R1a1a which are at extremely low frequency in the Greek colony regions of Italy. (as well as an earlier study in former Greek territories in Turkey)
-Sicily yields a substantial amount of R1b overall. Impressive is the P312* (DF27?) in Eastern Sicily.
-E-V13 was no doubt involved in the spread of classical Greeks but it's unlikely to be the origin of the IE language itself. Having an origin in the northern Levant or possibly western Turkey, it just wouldn't make a whole lot of sense all other things considered.

I still suspect a mixed bag of R1b, leaning towards the earlier xL51 type brought the IE Greek language to the territory.

Hauteville
16-07-15, 21:17
R1b P312 in Eastern Sicily was imported by mainlander Italians repopulation in the middle ages who replaced the muslims.

Angela
16-07-15, 22:09
It's interesting to see the contrast in R1a/M423/R1b levels in Sicily versus Corinthia and Euboea. I just did this in my head, so it may not be exact...

Euboea 11/4/9=24

Corinthia 17/12/10=39*

West Sicily l/1/33=35

East Sicily 3/2/34=39* 9
(If the four samples of R1(xM17, xM269 are added, R1b is 35%)

I'm not sure all that R1a1 is necessarily "Slavic", although it's possible. That would make both the Mycenaeans and the Dorians R1b, I suppose, or could the Dorians at least have been R1a1? Those R1b levels are really low to have brought all that language change.

*Original percentages were incorrect.

Hauteville
17-07-15, 01:28
We can say that Sicily has a western european input and Greece an eastern euro.
Because it seems that M423 have similar percentages.

LeBrok
17-07-15, 01:41
It's interesting to see the contrast in R1a/M423/R1b levels in Sicily versus Corinthia and Euboea. I just did this in my head, so it may not be exact...

Euboea 11/4/9=24

Corinthia 17/1/11=29


Very interesting. Either original Slavs were very high in R1a and low on M423, meaning I2 Dinaric was already in Balkans, or R1a already existed in Greece pre Slavic expansion, possibly IE Greek brought it, as you suggested. If the later is right, that could mean that either Greeks didn't do much of genetic impact on Italy, or they were all from south Greece with low R1a.
We should know in few years.

Aaron1981
17-07-15, 05:03
I think you guys copied those numbers down incorrectly

Corinthia

I2-M423 = 13/104
R1a1 = 18/104

Euboea

I2-M423= 4/93
R1a1= 10/93

Croatia
I2-423 = 46/93
R1a1= 24/93

Ionian Italy as compare:

I2-423 = 3/125
R1a1= 5/125

Aaron1981
17-07-15, 05:11
R1b P312 in Eastern Sicily was imported by mainlander Italians repopulation in the middle ages who replaced the muslims.

There is plenty of L2, L20 to account for that in these results. S116* is more in line with western French YDNA rather than Italian, although there is definitely overlap here.

Angela
17-07-15, 05:18
We can say that Sicily has a western european input and Greece an eastern euro.
Because it seems that M423 have similar percentages.

I agree with that. However, I think that perhaps the up-stream R1b clade, which is about 6-7% of the total, may also have come from the east as part of the original Indo-European migrations from the steppe, perhaps with steppe groups coming down the west side of the Black Sea and then into Greece, and from there into Italy, or perhaps into Italy from the western Balkans.

The upstream clades might have gone into Italy from central Europe and Gaul, and then from there into Sicily, with a good chunk of it coming in the Middle Ages, as you suggested.

Unfortunately, the R1a1 wasn't differentiated, so we don't know if a similar dynamic was at work in Greece.

Angela
17-07-15, 06:30
I think you guys copied those numbers down incorrectly

Corinthia

I2-M423 = 13/104
R1a1 = 18/104

Euboea

I2-M423= 4/93
R1a1= 10/93

Croatia
I2-423 = 46/93
R1a1= 24/93

Ionian Italy as compare:

I2-423 = 3/125
R1a1= 5/125

There was a typo in the I2-M423 percentage for Corinthia where I typed 1 instead of 12. I've corrected it in the original post.

Angela
17-07-15, 07:00
http://cdn.eupedia.com/forum/images/misc/quote_icon.png Originally Posted by Hauteville http://cdn.eupedia.com/forum/images/buttons/viewpost-right.png (http://www.eupedia.com/forum/showthread.php?p=462556#post462556)
We can say that Sicily has a western european input and Greece an eastern euro.
Because it seems that M423 have similar percentages.


Angela;462567]I agree with that. However, I think that perhaps the up-stream R1b clade, which is about 6-7% of the total, may also have come from the east as part of the original Indo-European migrations from the steppe, perhaps with steppe groups coming down the west side of the Black Sea and then into Greece, and from there into Italy, or perhaps into Italy from the western Balkans.

The upstream clades might have gone into Italy from central Europe and Gaul, and then from there into Sicily, with a good chunk of it coming in the Middle Ages, as you suggested.

Unfortunately, the R1a1 wasn't differentiated, so we don't know if a similar dynamic was at work in Greece.

@Hauteville, I misled you about M423 because the percentage for Corinthia was incorrect. Hopefully this is now correct. :)

R1a/M423/R1b

Euboea 11/4/9=24

Corinthia 17/12/10=39

West Sicily l/1/33=35

East Sicily 3/2/34=39

I'm still not sure what to make of I2-M423. Is it "Slavic" or is it Neolithicized WHG?

Look at Croatia:

R1a1- 26%
I2-M423- 49%
R1b 5%
J2 3%
E-V13 5%
No G

EEF/WHG/ANE
56/29/14

Hauteville
17-07-15, 10:21
There is plenty of L2, L20 to account for that in these results. S116* is more in line with western French YDNA rather than Italian, although there is definitely overlap here.



Also Franks, Bretons and Normans moved to Sicily and South Italy of course.

Hauteville
17-07-15, 10:26
.

@Hauteville, I misled you about M423 because the percentage for Corinthia was incorrect. Hopefully this is now correct. :)

R1a/M423/R1b

Euboea 11/4/9=24

Corinthia 17/12/10=39

West Sicily l/1/33=35

East Sicily 3/2/34=39

I'm still not sure what to make of I2-M423. Is it "Slavic" or is it Neolithicized WHG?

Look at Croatia:

R1a1- 26%
I2-M423- 49%
R1b 5%
J2 3%
E-V13 5%
No G

EEF/WHG/ANE
56/29/14


Western Sicily has 1.5% of R1a which is lower than other studies, if my table is correct.

Aaron1981
17-07-15, 17:25
I2-M423 is probably neolithicized WHG, but arriving from around Belarus with Slavic speakers to places like Czech Rep/Croatia in west and as far south as Macedonia with South Slavic speakers.
I made a mistake in my earlier post, none of the sampled Greek regions are in the north, they all appear south. It would be interesting to see the levels of the same haplogroups in the northern regions.

Other studies have found a large chunk of the R1a in Greece as M458 which appears common in Poland/Belarus area.

Hauteville
17-07-15, 17:45
There is undoubtedly a balkanic and east european input in Greece and an italic and west euro in Sicily and South Italy like I said above.
I always think that honestly.

Hauteville
18-07-15, 11:33
Angela was all the combined R1b in eastern Sicily 36%?
About I, the combined I in West Sicily is around 18-19% right?I'm without pc in this moment so I don't remember very well the right percentages.

Angela
18-07-15, 16:09
Angela was all the combined R1b in eastern Sicily 36%?
About I, the combined I in West Sicily is around 18-19% right?I'm without pc in this moment so I don't remember very well the right percentages.

There are 246 total samples from eastern Sicily. This is the breakdown for R1b:

M269* (18)
M269 (1
S116* (26)
S127* (2)
S139* (13)
S144* (1)
S145* (1)
S167* (2)
S21* (3)
S28* (13)
SRL2627 (3)

I didn't include the four samples of R1(xR1a1,R1b3) or the three P(xR1).

As for the "I":
M223 (6)
M253 (4)
M26 (6)
M423 (4)
M438 (3)
I don't see how you would get 18%.

For completeness, there are also 8 samples of R1a1 so 3%.

Hauteville
18-07-15, 19:19
For I i said for west Sicily.

Yetos
18-07-15, 20:12
Angela is there any special in Dorian colonizations?

Corinth was pelasgian,
Euvoea was 50-50,
Anyway R1a has high peek in Greece in North/Central,
land of Makedonia Epirus Thessaly, the homeland of Dorians, and drops as moving southern, but exists in every known Dorian colonisation, and is almost missing in Pelasgian and Ionian lands

anyway I am interesting in
Epizephyreian Locroi
Agrigento
Crotone
lipari
Iblea
Taranto
Selinunta
Gela
ancona

I had wrote about that, before years but everybody laughed,
time for some to recognise the possible

Hauteville
18-07-15, 20:24
Gela was also a Dorian colony.

Angela
18-07-15, 23:34
Y Dna "I" in West Sicily, with a total sample of 68:

M223 (5)
M253 (2)
M26 (4)
M423 (1)
M438 (1)

Of course, M26 is the "Sardinian" lineage.

@Yetos,
According to this, Corinthia was in a Doric speaking area, but not Euboea. If this is incorrect, someone please correct the record.


https://upload.wikimedia.org/wikipedia/commons/thumb/d/de/AncientGreekDialects_%28Woodard%29_en.svg/2000px-AncientGreekDialects_%28Woodard%29_en.svg.png

Yetos
19-07-15, 00:19
@ Angela

Corinth adopted Doric, just spoke Doric
it was not a Dorian city,
their culture was not Dorian,
they had other customs, architecture, etc



anyway I know that in Locroi in Italy R1a is about 8% while in Locroi in Greece is 12% and in North is 24%
from the begining in the forum, that was my observation,
the R1a in N Greece, might be responsible for many strange linguistic and possible Dorian

Angela
19-07-15, 01:16
Someone mentioned that a large chunk of the R1a1 in Greece is M458, a "Slavic" clade. Does anyone have a citation for that? I think it would be important to get a fix on the break down.

If no clade of R1a was involved in the initial Indo-Europeanization of Greece, then it was obviously R1b. Interesting, in that case, that it's only 10% and 9% in these two samples. It would have been nice to have gotten this kind of yDna data for Thessaly further north. Does it exist anywhere?

The results are also interesting in light of all the conversations I remember in the past where people vehemently insisted that in the ancient world you needed mass migration to effect language change.

It occurred to me to take a look at the figures for Yamnaya in Greeks in Haak et al. For Thessaly, which is where those samples originate, it seems to be about 20%, yes? This correlates pretty well with these R1b numbers, especially if we assume they would be a little higher in the north.

7373

Italy is different. The total for R1b and R1a in northern Italy is very high in comparison, and doesn't correlate with the number estimated for Yamnaya ancestry.

@Yetos,

Cumae (near Naples) Founded 740 BCE, Chalcis (Euboea)
Sybaris founded 720 BCE Achaean
Croton founded 710 BCE, Achaean
Tarentum founded 706 BCE, (some sources say Achaean, some say Sparta)
Rhegium, founded 720 BCE, Chalcis
Elea founded 540 BCE, Phocaea
Thurri founded 443 BCE, Athens
Heraclea founded 433 BCE Tarentum

On Sicily...
http://www.timediver.de/Sizilien/Magna_Graecia.jpg
Siracusa, founded 733 BCE, Corinth
Naxos, founded 734 BCE, Chalcis, Euboea
Gela, founded 688 BC, Rhodes and Crete
Selinunte, founded 630 BCE, Megara (Corinth)
Himera 630 BCE, Chalcis (Euboea) and Siracusa
Akragas (Agrigento), founded 580 BCE, Gela

If Wiki is to be believed, "Himera was a colony of Zancle, but Thucydides tells us that, with the emigrants from Zancle, who were of Chalcidic origin, were mingled a number of Syracusan (https://en.wikipedia.org/wiki/Syracuse,_Italy) exiles, the consequence of which was, that, though the institutions (νόμιμα) of the new city were Chalcidic, its dialect had a mixture of Doric (https://en.wikipedia.org/wiki/Doric_Greek)."
https://en.wikipedia.org/wiki/Himera

The samples for East Sicily were taken from Siracusa and nearby areas, and the samples for West Sicily from Trapani, Caccamo etc. I'm not sure why they used samples from Trapani.

Hauteville
19-07-15, 02:12
Click on "Colonie della Magna Grecia" and "Colonie greche in Sicilia" there is probably the full list.

https://it.m.wikipedia.org/wiki/Magna_Grecia

About Trapanese samples, indeed Trapani was an Elymian city with a Greek name (Drepanon) but for better study they must collected samples near ancient Selinunte like Castelvetrano or ethnic people in Mazara del Vallo.

Hauteville
19-07-15, 02:16
Y Dna "I" in West Sicily, with a total sample of 68:

M223 (5)
M253 (2)
M26 (4)
M423 (1)
M438 (1)

Of course, M26 is the "Sardinian" lineage.



So i'm right to the percentages in the west.

Imo M26 rather than specific Sardinian is one of the most ancient haplogroups diffused in modern italian nation and western Mediterranean.
I think that in the early ages the percentages of M26 were more higher in Italy, similar to modern Sardinia.

Arame
19-07-15, 07:12
Yetos

I believe that R1a is involved in the spread of Doric. But someone on Anthrogenica insisted that it is impossible. So I am confused on this question. Did anybody tried to calculate the TMRCA of Greek R1a?

Maciamo
19-07-15, 08:50
I am surprised that this new study still uses very old-dated haplogroup names like E3b (dropped by ISOGG in 2008 in favour of E1b1b) or K* instead of T (listed by ISOGG since 2008). The sample sizes are very small of each region, so the percentages are not very indicative. I will integrate the data into my Y-DNA table for Italian regions and for Greece. At least it won't be totally useless.

Hauteville
19-07-15, 10:58
I am surprised that this new study still uses very old-dated haplogroup names like E3b (dropped by ISOGG in 2008 in favour of E1b1b) or K* instead of T (listed by ISOGG since 2008). The sample sizes are very small of each region, so the percentages are not very indicative. I will integrate the data into my Y-DNA table for Italian regions and for Greece. At least it won't be totally useless.
Eastern Sicilian samples are 256, which is more interesting.

Angela
19-07-15, 16:10
The more samples the better, of course, but how big are the samples for individual towns in any country in most of these studies or even countries themselves? I see that our Eupedia yDna table for all of England is based on 250-500 samples? Posters here spent days discussing the results from Boattini et al 2013, and they had 984 samples for all of peninsular Italy plus the islands, and only 45 samples for Agrigento. Here we have 556 samples for southern Italy alone, 68 for Trapani, and 256 for Siracusa and a few small nearby areas. That seems pretty darn good to me, although I'd love it if Italian researchers had the money to do a POBI type study for Italy. Meanwhile, I'll take what I can get.

As for why the dated nomenclature, I can't be sure until I see the study itself. I'm wondering if these are new samples. See Table 2.

Aaron1981
19-07-15, 17:20
R1(xM17, xM269) is most certainly a branch of R1b, either R1b-M73 or R1b-P25. The latter is found in both Italy and Greece, and the former has turned up in Italy.

Hauteville
19-07-15, 17:51
The more samples the better, of course, but how big are the samples for individual towns in any country in most of these studies or even countries themselves? I see that our Eupedia yDna table for all of England is based on 250-500 samples? Posters here spent days discussing the results from Boattini et al 2013, and they had 984 samples for all of peninsular Italy plus the islands, and only 45 samples for Agrigento. Here we have 556 samples for southern Italy alone, 68 for Trapani, and 256 for Siracusa and a few small nearby areas. That seems pretty darn good to me, although I'd love it if Italian researchers had the money to do a POBI type study for Italy. Meanwhile, I'll take what I can get.

As for why the dated nomenclature, I can't be sure until I see the study itself. I'm wondering if these are new samples. See Table 2.


Indeed, for example between Boattini, Sarno, Brisighelli, this paper, the paper of Latium Appennines cities, Francalacci with 1200 Sardinians and the last of Arbereshe of Calabria and Sicily we have got a really high number of samples for the whole Italy.
So I think that we are the nation with higher number of samples.

Angela
19-07-15, 18:15
Angela was all the combined R1b in eastern Sicily 36%?
About I, the combined I in West Sicily is around 18-19% right?I'm without pc in this moment so I don't remember very well the right percentages.

As I said, I didn't include the four samples of R1(xM17, xM269) in the total for R1b in eastern Sicily (Siracusa etc.) If I add them, I would get to 35% if my math is correct.

I've made a notation in the original post.

Angela
20-07-15, 18:57
I haven't had much success getting hold of studies on Greek yDna.

Here are the results from our table here at Eupedia:
http://www.eupedia.com/europe/european_y-dna_haplogroups.shtml
The R1a definitely shows a north to south cline of decreasing frequency. The R1b is more variable, with the highest R1b in the south. The numbers for "Center" roughly correlate with those for Euboea and Corinth in the subject paper.

I did find this table on Dienekes' blog where he aggregated samples from three studies: DiGiacomo et al 2003, King et al 2008, and Martinez et al 2007. The plus is that it's done by pretty specific area. The minus is that it dates back to 2008, so there's no subclade resolution. For what it's worth:
7376
http://dienekes.blogspot.com/2008/05/exploring-y-chromosome-haplogroup.html

Just looking at some of the highest R1b, this is the picture:
Lashithi plateau 41.5
Chios 26.2
Nea Nikomedia 19.3
Larisa 19
Lasithi 18
Ioannina 16.7

The rest are much lower. I don't know how much value to put on the Lasithi plateau as it was depopulated and then repopulated under Venetian rule. The Chios result is interesting, but I'm never quite sure how to interpret results from islands, given the stronger founder effects. That would hold true for the other Crete samples. What remains seems to show a decreasing cline from north to south, which we had already surmised.

This is the picture for R1a:
Thessaloniki 25
Rethymnon 22.7
Nea Nicomedia 21.1
Kardhitsa 20
Lashithi Plateau 19.5

Again, if we take Crete out of the equation as perhaps unduly subject to founder effect, it's north to south on a decreasing cline. Without a subclade breakdown I don't know how much else we can say.

Kardhitsa and Patrai are perhaps closest to Corinth, with Agrinion and Larissa north of, but still approaching Euboea? I don't know why they didn't take a lot of samples from the Peloponnese and a some from Euboea while they were at it. Y Dna lineages in Greece seem very variable. Might it be fair to say that in the areas which provided colonists to Greece there was little R1a, no matter the reason? So, even if it was there from migrations before the Slavic era, it might not have been a part of the migrations to Sicily, or even if it was, it might have diminished into the low single digits just by chance?

Hauteville
21-07-15, 10:52
Maybe R1b in Lasithi is due to Eteo-Cretans?for one recent study that language was related to the Latin.

moore2moore
14-12-15, 01:10
There are a lot of deficiencies in this study, as Maciamo noted. And several more, too numerous to list here.

I for one am always a little shocked that no one seems to make a pretense of multi-disciplinary approaches anymore. Few geneticists bother to try to learn history, and incorporate it into their work.

For example, we will likely never have much ethnic Roman DNA. By ETHNIC ROMAN, I mean NOT Roman as a political unit, but Republican era (pre-44 BC), Latini DNA. Because the Romans cremated their dead mostly.

But by reading history, we can reconstruct Roman DNA. How?

(1) Take a tiny Roman suburb that was populated by ethnic Romans, before imperial times, which then kind of became a backwater.

(2) Take towns in Spain or South Italy or France that were colonized by pre-imperial ethnic Romans from Italy.

(3) Run a scrub to remove any of the towns that were subsequently depopulated or invaded.

(4) Compare the DNA of said towns.

Voila, you have the Roman signature. I am a Roman history lecturer. I could rattle of 10 such candidate towns in a heartbeat. We want isolated towns far enough apart settled by Romans before the empire could make it polyglot, who have become rural backwaters, untouched ever since.

What is this relevance to the thread at hand? The same principles apply to the obsessive chase for purported Greek signatures. Are they from Neolithic farmers? Or are they truly Ancient Greek matches?

In some of these towns, we have clear documentation that an Italic tribe (or other force) came in and wiped the Greek colonists off the map. Sadly, this happened a lot in Lucania and Cosenza provinces, for instance. The Lucani and Bretti (Italic tribes) would invade, and completely destroy a Greek colony.

Or towns like Thuri, which lost its entire population and had to be resettled by poor urban Romans. https://en.wikipedia.org/wiki/Thurii

What I am saying is: why don't they look for towns that are better candidates? It just seems so random and scattershot an approach right now, which means the conclusions (this is Greek DNA, versus this is merely EEF DNA) are invalid.