Southern Illyrians & Mycenean Greeks on a PCA plot

[enter_tain]

Given the Slavic admixture in the Balkans, the Eurogenes Blogspot created this nice chart comparing the ancient EEF-heavy Balkans all the way to Ukrainians. As you can see there is a natural cline from Minoans to Ukranians, with Slavic ancestry increasing right on the line/slope.

If you look at Croatian IA and Helladic MBA they are very similar and more Eastern-European like. This is obviously less EEF-admixed populations.

When they got to the Balkans, they mixed with EEF-heavy populations, and became more "Mycenean-like". Then more Slavic admixture in the Middle Ages increase the steppe ancestry/EE-like admixture.

But this is where users like Aspurg who are getting a little desperate, try to say that modern Albanians are Bulgarian IA + Slavic. That can't be true, because Bulgarian IA is way too eastern shifted to begin with. And this makes sense. They were from the Eastern Balkans.

Adding Slavic ancestry to Bulgaria IA, gets some random new population that doesn't exist.

The actual southern Illyrians (Taulanti, Bylliones, Albanoi, Dardanians) and other northern Hellenized people like Epirotes and Macedonians would be somewhere around here

Adding Slavic ancestry to Mycenean Greeks you get a perfect fit to modern Greeks/Albanians. Southern Illyrians and Epirotes/Macedonians + Slavic ancestry are a perfect model for both northern Greeks and Macedonians. Certainly not Thracian/Dacian or any other already-eastern shifted people.

 

[enter_tain]

When we get more samples out of ancient Macedonia, Epirus, or southern Illyria, they will 100% fall in that red box.

 

[Illyria]

Is your East and West flipped? Seems like it. I remember the Croatia_IA sample being North-West of Albanians, and Bulgaria_IA being South-West of Albanians, not more East. In fact I believe most of the ancient Balkan samples we have are more Western-shifted than Albanians.

I do agree that Southern Illyrians were probably more Mycenaean-like, assimilating more EEF populations on their way down the Balkans. Seems likely that they will plot in that red box once we get samples.

 

[Morina]

Yeah the PCA Map is completely wrong. The PCA map shows Albanians as more Southern shifted than Greeks while many Albanians are actually more North and more West.

The Iron Age Croatians are closer to many Northern Albanians than they are to Croats or Serbs as shows by other calculators and PCA Maps. It was very Western shifted.

The R1b Vucedol did not plot to far away but interesting enough had a weird admixture compared to some of the J2b2 that plotted further away but actually appears to be closer.

 

[Morina]

Also I definitely do not believe his theory that Southern Illyrians mainly plotted there and therefore more North = More Slavic Ancestry. We can clearly see some of the North shift in some Albanians is North-Italian like derived like some of these Illyrian samples and not from Slavic.

I also remember a Bronze Age sample from Montenegro clustered like an Iberian. In fact, I wouldn't be surprised if even the Southern and Central Illyrians end up clustering like some Iberians and Italians. Some PCA Map I saw showed there is almost an overlap with Iberia, Italy and some Albanians. But most Albanians do not cluster that West.

Also Hallstatt samples clustered like Italians and Iberians if I recall.

PCA plotting can also be the result of a bottle neck effect.

 

[torzio]

When southern Illyrian term or Illyrian proper existed as a term was only in the time of Roman historian Pliny and the roman province of Illyricum

The Roman province of Illyricum stretched from the Drilon River (the Drin, in modern Albania) in the south to Istria (modern Slovenia and Croatia) in the north and to the Savus (Sava) River in the east; its administrative centre was Salonae (near present-day Split) in Dalmatia. The Illyrians, bearers of the Hallstatt culture, were divided into tribes.

In the Hallstatt phase II period (c.800-c.500 BCE), we can identify the first Illyrian kingdoms. The Vače situla, a bronze vessel found in Slovenia.

Illyrii proprie dicti[122] were the Illyrians proper, so called by Pliny (23–79 AD) in his Natural History. They later formed the Docleatae. They were the Taulantii, the Pleraei or Pyraei, the Endirudini, Sasaei, Grabaei, Labeatae.


Illyrii proprie dicti tribes where the only illyrian tribes which had not become celtinized since the late bronze-age when celts started moving south from Central and Southern Germany.

between these tribes above and the dardanian people stood a big celtinized Illyrian tribe called the Autariatae.

 

[enter_tain]

Yeah the PCA Map is completely wrong. The PCA map shows Albanians as more Southern shifted than Greeks while many Albanians are actually more North and more West.

The Iron Age Croatians are closer to many Northern Albanians than they are to Croats or Serbs as shows by other calculators and PCA Maps. It was very Western shifted.

The R1b Vucedol did not plot to far away but interesting enough had a weird admixture compared to some of the J2b2 that plotted further away but actually appears to be closer.

How can a PCA be wrong? This is just data.

 

[enter_tain]

Is your East and West flipped? Seems like it. I remember the Croatia_IA sample being North-West of Albanians, and Bulgaria_IA being South-West of Albanians, not more East. In fact I believe most of the ancient Balkan samples we have are more Western-shifted than Albanians.

I do agree that Southern Illyrians were probably more Mycenaean-like, assimilating more EEF populations on their way down the Balkans. Seems likely that they will plot in that red box once we get samples.

This particular plot was made by Eurogenes to show a cline between "Pelasgians" and Slavs and the impact of Slavic migrations.

Croatian IA has more steppe ancestry than Albanians that's why it plots like that. Southern Illyrians definitely had more EEF-admixture/"Pelasgian" admixture maternally.

 

[enter_tain]

I took a closer look at this, and I urge others to do as well. It seems that the very south of the Balkans and eastern Balkans had an uptick in "Iran Neolithic" ancestry that the western Balkans don't have. In the Western Balkans, Iran Neolithic either doesn't exist or is always paired with ANE.

That is probably why that Bulgarian Iron Age is so eastern shifted even though it's so far south and certainly not ancestral to Albanians. Eastern shifted would make sense in very steppe heavy people. I think Mycenean Greeks largely had more "Western Balkans Neolithic" than the CHG/Iranian-heavy farmer DNA. This would abide by Aristotle's account that Proto-Greeks formed somewhere in Epirus.

The main difference between Illyrians and Thracians will be that Thracians have an unbalanced Iranian/CHG component that Illyrians don't have.

 

[Constantine]

So you're saying Helladic MBA and similar ilk just sashayed into the Balkans and only managed to become "Mycenaean-like"? Whereas only the noble Slavs, ages later, were able to "penetrate" those stubborn Myceneans and make them a "perfect fit" to modern Greeks?

 

[enter_tain]

So you're saying Helladic MBA and similar ilk just sashayed into the Balkans and only managed to become "Mycenaean-like"? Whereas only the noble Slavs, ages later, were able to "penetrate" those stubborn Myceneans and make them a "perfect fit" to modern Greeks?

Yeah. If you look at some samples from Montenegro, you have pure EEFs with no CHG/Iranian Neolithic still during Late Bronze Age, and then an incursion of a 50% steppe/50% EEF person. That's essentially the Western Balkans. It was very EEF heavy until the Late Bronze Age. It's insane that we still had pure EEFs during that age.

Eastern Balkans like Thracians/Dacians had similar Proto-people, but the "natives" were heavy in CHG/Iranian Neolithic. Not surprising since Anatolia/Aegean were becoming more CHG-like during the Bronze Age, even without steppe incursions.

The "native Balkan" components of Greeks and Albanians is more Western Balkans-like, while for Cypriots and population like Thracians/Dacians it was more "Bronze Age Aegean/Anatolian"-like

 

[mount123]

Yeah. If you look at some samples from Montenegro, you have pure EEFs with no CHG/Iranian Neolithic still during Late Bronze Age, and then an incursion of a 50% steppe/50% EEF person. That's essentially the Western Balkans. It was very EEF heavy until the Late Bronze Age. It's insane that we still had pure EEFs during that age.

Eastern Balkans like Thracians/Dacians had similar Proto-people, but the "natives" were heavy in CHG/Iranian Neolithic. Not surprising since Anatolia/Aegean were becoming more CHG-like during the Bronze Age, even without steppe incursions.

The "native Balkan" components of Greeks and Albanians is more Western Balkans-like, while for Cypriots and population like Thracians/Dacians it was more "Bronze Age Aegean/Anatolian"-like

Absolutely not true.

 

[enter_tain]

Absolutely not true.

Open up a map of the ancient Balkans will you? All the samples are there. Dacian/Thracian and Aegan/Anatolian are HEAVY in Neolithic/CHG. Bulgarian IA is like 25% CHG/Iran. I saw some Dacian ones that were upwards of 30% percent.

Western Balkans had very little "extra" CHG, if none at all. And when they do it's balanced with ANE from steppe invaders. That Late Bronze Age Montenegrin in Illyria was like fully EEF.

 

[ihype02]

So you're saying Helladic MBA and similar ilk just sashayed into the Balkans and only managed to become "Mycenaean-like"? Whereas only the noble Slavs, ages later, were able to "penetrate" those stubborn Myceneans and make them a "perfect fit" to modern Greeks?

Yes that is exactly what happened. Native Peloponnesians in particular before the Proto-Mycenaean invasion were similar to Minoan or EBA Euboea and after the Greek language was brought end up with a Mycenaean-like genetic profile which was largely preserved till at least the Classical Period. Classical Attica received some Anatolian and Thracian admixture mostly through slavery.

But Greeks before the Slavic migration as the Marathon sample revealed were like Dodecanese Islanders if they were homogenous enough but I suspect (not know for sure) more Western shifted towards Mycenaeans compared to Rhodes.

Eastern Cretans are very still close to Rhodes and overlap with some more Dodecanese Island.

 

[enter_tain]

Yes that is exactly what happened. Native Peloponnesians in particular before the Proto-Mycenaean invasion were similar to Minoan or EBA Euboea and after the Greek language was brought end up with a Mycenaean-like genetic profile which was largely preserved till at least the Classical Period. Classical Attica received some Anatolian and Thracian admixture mostly through slavery.

But Greeks before the Slavic migration as the Marathon sample revealed were like Dodecanese Islanders if they were homogenous enough but I suspect (not know for sure) more Western shifted towards Mycenaeans compared to Rhodes.

Eastern Cretans are very still close to Rhodes and overlap with some more Dodecanese Island.

Yeah this shows that division

Western Balkan farmers = More EEF

Eastern Balkan/Aegean/Anatolian = More CHG/Iranian Neolithic

Proto-Greeks and pre-Proto-Albanians mixed with Western Balkan farmers. Thracians, Dacians, and Cypriotes mixed with more CHG/Iranian Neolithic farmers.

 

[Bekas]

Myceneans and Minoans have Iranian_N related (Ganj-Dareh here as a proxy) admixture, as it's very easily produced and verifiable in a 3-way qpAdm run.

The language of the "proto-Greeks" (lol) is not proven to have come from the north, it's one of the two plausible pathways.

qpAdm Minoan 2-way, 91.1% Greece_Neolithic + 8.9% Iranian_Neolithic.

qpAdm: parameter file: parqpadm.txt
### THE INPUT PARAMETERS
##PARAMETER NAME: VALUE
fstatsname: fstatsa.txt
popleft: left.txt
popright: right.txt
details: YES
## qpAdm version: 1520
seed: 1891837379
*** recommended that inbreed be explicitly set ***
inbreed set NO

left pops:
Greece_Minoan_Lassithi
Greece_Peloponnese_N
Iran_GanjDareh_N

right pops:
Mbuti.DG
Russia_Ust_Ishim.DG
China_Tianyuan
Goyet_Neanderthal.SG
Turkey_N_published
Georgia_Kotias.SG
Russia_MA1_HG.SG
Georgia_Satsurblia.SG
Papuan.DG
Israel_Natufian_published
Russia_DevilsCave_N.SG
ONG.SG

codimension 1
f4info:
f4rank: 1 dof: 10 chisq: 15.117 tail: 0.127863692 dofdiff: 12 chisqdiff: -15.117 taildiff: 1
B:
scale 1.000
Russia_Ust_Ishim.DG 0.173
China_Tianyuan -0.007
Goyet_Neanderthal.SG 0.028
Turkey_N_published 2.683
Georgia_Kotias.SG -0.080
Russia_MA1_HG.SG 0.102
Georgia_Satsurblia.SG -0.174
Papuan.DG 0.020
Israel_Natufian_published 1.920
Russia_DevilsCave_N.SG 0.175
ONG.SG 0.072
A:
scale 359.708
Greece_Peloponnese_N 0.137
Iran_GanjDareh_N -1.408


full rank
f4info:
f4rank: 2 dof: 0 chisq: 0.000 tail: 1 dofdiff: 10 chisqdiff: 15.117 taildiff: 0.127863692
B:
scale 2412.295 248.383
Russia_Ust_Ishim.DG -0.152 -0.200
China_Tianyuan 1.564 0.100
Goyet_Neanderthal.SG 0.220 -0.031
Turkey_N_published 2.427 -2.627
Georgia_Kotias.SG -0.282 0.065
Russia_MA1_HG.SG -1.186 -0.173
Georgia_Satsurblia.SG -0.796 0.136
Papuan.DG -0.528 -0.064
Israel_Natufian_published 0.142 -1.987
Russia_DevilsCave_N.SG -0.135 -0.196
ONG.SG 0.394 -0.056
A:
scale 1.414 1.414
Greece_Peloponnese_N 1.414 0.000
Iran_GanjDareh_N 0.000 1.414


best coefficients: 0.911 0.089
totmean: 0.911 0.089
boot mean: 0.911 0.089
std. errors: 0.015 0.015

error covariance (* 1,000,000)
235 -235
-235 235


summ: Greece_Minoan_Lassithi 2 0.127864 0.911 0.089 235 -235 235

fixed pat wt dof chisq tail prob
00 0 10 15.117 0.127864 0.911 0.089
01 1 11 42.393 1.38468e-05 1.000 0.000
10 1 11 1706.226 0 0.000 1.000
best pat: 00 0.127864 - -
best pat: 01 1.38468e-05 chi(nested): 27.276 p-value for nested model: 1.76368e-07

coeffs: 0.911 0.089

## dscore:: f_4(Base, Fit, Rbase, right2)
## genstat:: f_4(Base, Fit, right1, right2)

Mycenean 3-way: 81.7% Minoan + 5.2% Iranian_Neolithic + 13% Yamnaya_Samara.

qpAdm: parameter file: parqpadm.txt
### THE INPUT PARAMETERS
##PARAMETER NAME: VALUE
fstatsname: fstatsa.txt
popleft: left.txt
popright: right.txt
details: YES
## qpAdm version: 1520
seed: 1858803783
*** recommended that inbreed be explicitly set ***
inbreed set NO

left pops:
Greece_BA_Mycenaean
Greece_Minoan_Lassithi
Iran_GanjDareh_N
Russia_Samara_EBA_Yamnaya

right pops:
Mbuti.DG
Russia_Ust_Ishim.DG
China_Tianyuan
Goyet_Neanderthal.SG
Russia_Sunghir3.SG
Russia_Kostenki14.SG
Russia_MA1_HG.SG
Georgia_Satsurblia.SG
Russia_DevilsCave_N.SG
Papuan.DG
Switzerland_Bichon.SG
Israel_Natufian_published
Russia_Steppe_Eneolithic
ONG.SG

codimension 1
f4info:
f4rank: 2 dof: 11 chisq: 16.213 tail: 0.133421934 dofdiff: 13 chisqdiff: -16.213 taildiff: 1
B:
scale 1.000 1.000
Russia_Ust_Ishim.DG 0.516 -0.067
China_Tianyuan 0.572 -0.505
Goyet_Neanderthal.SG 0.126 -0.012
Russia_Sunghir3.SG 1.048 0.773
Russia_Kostenki14.SG 0.848 0.449
Russia_MA1_HG.SG 1.852 -1.092
Georgia_Satsurblia.SG 0.451 -0.574
Russia_DevilsCave_N.SG 0.641 -0.105
Papuan.DG 0.439 -0.170
Switzerland_Bichon.SG 1.842 0.785
Israel_Natufian_published 0.237 2.999
Russia_Steppe_Eneolithic 1.601 -0.822
ONG.SG 0.568 -0.309
A:
scale 519.538 651.690
Greece_Minoan_Lassithi -0.141 0.248
Iran_GanjDareh_N -1.104 -1.399
Russia_Samara_EBA_Yamnaya 1.327 -0.990


full rank
f4info:
f4rank: 3 dof: 0 chisq: 0.000 tail: 1 dofdiff: 11 chisqdiff: 16.213 taildiff: 0.133421934
B:
scale 1756.254 332.414 320.679
Russia_Ust_Ishim.DG -0.290 -0.319 0.440
China_Tianyuan -1.565 -0.178 0.587
Goyet_Neanderthal.SG -0.914 -0.206 0.000
Russia_Sunghir3.SG 0.542 -1.227 0.517
Russia_Kostenki14.SG 1.501 -0.704 0.670
Russia_MA1_HG.SG -1.706 -0.538 2.017
Georgia_Satsurblia.SG 0.346 0.227 0.764
Russia_DevilsCave_N.SG -0.748 -0.401 0.528
Papuan.DG -0.903 -0.262 0.361
Switzerland_Bichon.SG -0.621 -1.872 1.101
Israel_Natufian_published 0.548 -2.540 -1.478
Russia_Steppe_Eneolithic -1.098 -0.490 1.728
ONG.SG -0.886 -0.241 0.549
A:
scale 1.732 1.732 1.732
Greece_Minoan_Lassithi 1.732 0.000 0.000
Iran_GanjDareh_N 0.000 1.732 0.000
Russia_Samara_EBA_Yamnaya 0.000 0.000 1.732


best coefficients: 0.817 0.052 0.130
totmean: 0.817 0.052 0.130
boot mean: 0.819 0.051 0.130
std. errors: 0.059 0.049 0.041

error covariance (* 1,000,000)
3462 -2126 -1336
-2126 2440 -314
-1336 -314 1651


summ: Greece_BA_Mycenaean 3 0.133422 0.819 0.051 0.130 3462 -2126 -1336 2440 -314 ...
1651

fixed pat wt dof chisq tail prob
000 0 11 16.213 0.133422 0.817 0.052 0.130
001 1 12 25.948 0.0109172 0.926 0.074 0.000
010 1 12 17.311 0.13827 0.862 0.000 0.138
100 1 12 208.217 0 0.000 0.634 0.366
011 2 13 27.905 0.00932871 1.000 0.000 0.000
101 2 13 281.570 0 0.000 1.000 0.000
110 2 13 435.639 0 0.000 0.000 1.000
best pat: 000 0.133422 - -
best pat: 010 0.13827 chi(nested): 1.098 p-value for nested model: 0.294619
best pat: 011 0.00932871 chi(nested): 10.594 p-value for nested model: 0.00113426

coeffs: 0.817 0.052 0.130

## dscore:: f_4(Base, Fit, Rbase, right2)
## genstat:: f_4(Base, Fit, right1, right2)

 

[peloponnesian]

So you're saying Helladic MBA and similar ilk just sashayed into the Balkans and only managed to become "Mycenaean-like"? Whereas only the noble Slavs, ages later, were able to "penetrate" those stubborn Myceneans and make them a "perfect fit" to modern Greeks?

You won't see anything like Helladic_MBA in LBA, archaic, classical or Hellenistic Greece, trust me.

 

[peloponnesian]

It's not perfect of course, Cypriots need an additional Anatolian and/or Levantine source (hence the higher distance).

 

[Jovialis]

https://www.eupedia.com/forum/threads/42211-Modeling-with-aDNA?p=638821&viewfull=1#post638821
With my model, you can see it chooses Yamnaya and Minoans for the Balkans and Greece. That matches Lazaridis et al. 2017 and Clemente et al. 2021. There's also a small Anatolia_BA component which probably comes from the Aegean during later periods. Then there's the Coded Ware which matches the modeling of Slavic countries.

 

[Jovialis]

^also small amounts of Hallstatt and Hungary BA in the model increasing as you go north in the Balkans.