Croatian Migration Era breakdown

[ph2ter]

Isn't the Germanic component a bit high for the North Western group? Looks as high as in some Slovenians and Hungarians. Do the Kajkavians have known German-Hungarian settlement? How many Kajkavian samples did you use?

The western Slovenians are more native pre-Slavic than Kaikavians and northeastern Slovenians.
The Germanic admixture is ancient, not modern.

[Bukva_]

I collected unmixed regional Croat G25 samples (real coords only, criteria as following: 4 ethnic Croat grandparents from same area) and created averages based on their regional dialect.

Dialect map: Croatia has 3 major dialects


Green = Štokavian, basies of standard Croatian
Purple = Kajkavian
Blue = Čakavian

Now the results:



Samples used for averages were taken from last Balkan paper on South Slav ethnogenesis. Illyro-Roman average was composed of mixed Balkan_IA and Roman_Anatolian genetic profiles from present-day Croatia, barbarian averages contain Eastern_European like genetic profiles from the paper plus some other suitable samples like Avar_Szolad 2 - also north European like genetic profiles from early medieval Germany and finally, elite Avar samples from Panonnia with East Asian profile for the steppe nomad input.

Conclusion is štokavian and čakavian speaking Croats have essentianly no differences worthy of note while kajkavian speaking Croats do have considerable differences to the rest with more Germanic than old Balkan ancestry while other two on average don't contain any. Kajkavians also have somewhat lower eastern European input than the other two averages although difference isn't by nay means significant.

If you wish to comment my model and conclusions you are very welcome.

I would heavily disregard the "Illyro" as a prefix before the Roman because of the archaeogenetic contexts of these CE samples. Given the uniparental distribution of Croats it is rather evident that BA-IA Illyrian Y-DNA and mtDNA are very rare. Imperial Roman era ancestry, some southeast Urnfield, perhaps also some Daco-Thracian-like source pop would fit the overall picture much better.

Also, on a side note "Balkan_IA" or "Old Balkan" as toponyms are extremely generic considering the diverse genetic makeup of the Paleo-Balkans, as in there is not some sort of homogenous bunch.

What would be an example of southeast Urnfield in G25?

[Leeloo]

Target: Bukva_father
Distance: 1.8274% / 0.01827435
67.2 Slav_proxy
32.4 Illyro-Roman
0.4 Asian

Your Slav_medieval doesn't fit to us that well. This is one I made is more Balto-Slavic like. From 3.17 to 1.82 reduction in distance, and German is gone.

Code:

Slav_proxy,0.13100482,0.127428115,0.076713868,0.070202767,0.037995783,0.031810557,0.011687291,0.013470371,-0.002809884,-0.024040643,-0.006156002,-0.005388771,0.018035942,0.029634356,-0.010994435,-0.008493723,-0.010030366,-0.003818951,0.002675593,-0.00874243,-0.002410523,-0.002930757,0.011274889,-0.003391698,-0.003268445

Average isn't made to fit specific individuals, but everyone. It is based on EE-related Slavs from Balkan peper (medieval Czechia, Croatia, Slovakia, Austria and Avar_Szolad 2)

Goal was to estimate Slavic ancestry based on genetic makeup of Slavs from migration era.

It is pointless to look for better fit for individual samples, averages matter. There is no model which produces tight fit for every indivudual and tighter fit does not equal being historically accurate.

[Leeloo]

What samples or how did you model Croatian_Chakavian and Stokavian?

I personally tested quite a few people with deeply rooted regional ancestry and collected some other samples online, since I am into genetics for quite long time.

[Leeloo]

Target: Bukva_father
Distance: 1.8274% / 0.01827435
67.2 Slav_proxy
32.4 Illyro-Roman
0.4 Asian

Your Slav_medieval doesn't fit to us that well. This is one I made is more Balto-Slavic like. From 3.17 to 1.82 reduction in distance, and German is gone.

Code:

Slav_proxy,0.13100482,0.127428115,0.076713868,0.070202767,0.037995783,0.031810557,0.011687291,0.013470371,-0.002809884,-0.024040643,-0.006156002,-0.005388771,0.018035942,0.029634356,-0.010994435,-0.008493723,-0.010030366,-0.003818951,0.002675593,-0.00874243,-0.002410523,-0.002930757,0.011274889,-0.003391698,-0.003268445

My average is closest to NW Ukrainians from Rivne who are not very southern shifted and Volynia is likely as close to proto Slavic homeland as one could get. I believe proto Slavs were extremely northern, on Belarussian-south Lithuanian cline, but early migrating Slavs being like NW Ukrainians makes sense.

[Mik]

I gave it a try:

Target: Mik
Distance: 1.7705% / 0.01770530
64.6 Slavic_medieval
35.4 Illyro-Roman

Target: Mik_Mom
Distance: 1.8938% / 0.01893774
57.8 Slavic_medieval
42.2 Illyro-Roman

Target: Mik_Dad
Distance: 2.7812% / 0.02781229
61.2 Slavic_medieval
38.8 Illyro-Roman

Target: Mik_Cousin
Distance: 1.3690% / 0.01369035
63.4 Slavic_medieval
36.4 Illyro-Roman
0.2 Asian

[Leeloo]

Isn't the Germanic component a bit high for the North Western group? Looks as high as in some Slovenians and Hungarians. Do the Kajkavians have known German-Hungarian settlement? How many Kajkavian samples did you use?

It's the reality. Kajkavian Croats are more northern shifted than Slovenes and Hungarians on average and many cluster closest to academic average of Czechs from Prague.

Germanic ancestry in the region seems medieval since it is evenly spread in the rural population. There were some German Saxon miners in some towns, but far too little to impact autosomal genetics in such a way.

It is quite interesting since historiography did not record a major Germanic expansion in the area as far as I am aware. I used 7-8 samples with deeply rooted ancestry, which should be enough since they all have rather uniform scores and come from various locations. Only one sample is different to others, and that one is Kajkavian speaker from Gorski kotar which is region cut from main Kajkavian region in north Croatia and with complex settlement history. His score is more Slovene-like than Kajkavian like and it actually correlates with dialect spoken in his region that some lignuists classify as Slovene-proper influenced.

[Leeloo]

I gave it a try:

Target: Mik
Distance: 1.7705% / 0.01770530
64.6 Slavic_medieval
35.4 Illyro-Roman

Target: Mik_Mom
Distance: 1.8938% / 0.01893774
57.8 Slavic_medieval
42.2 Illyro-Roman

Target: Mik_Dad
Distance: 2.7812% / 0.02781229
61.2 Slavic_medieval
38.8 Illyro-Roman

Target: Mik_Cousin
Distance: 1.3690% / 0.01369035
63.4 Slavic_medieval
36.4 Illyro-Roman
0.2 Asian

Are you a Croat or Bosniak?

[Mik]

I gave it a try:

Target: Mik
Distance: 1.7705% / 0.01770530
64.6 Slavic_medieval
35.4 Illyro-Roman

Target: Mik_Mom
Distance: 1.8938% / 0.01893774
57.8 Slavic_medieval
42.2 Illyro-Roman

Target: Mik_Dad
Distance: 2.7812% / 0.02781229
61.2 Slavic_medieval
38.8 Illyro-Roman

Target: Mik_Cousin
Distance: 1.3690% / 0.01369035
63.4 Slavic_medieval
36.4 Illyro-Roman
0.2 Asian

Are you a Croat or Bosniak?

Croat, West-Herzegowina, all ancestors 4 generations back exclusively from Grude, Široki Brijeg and Posušje counties.

[Leeloo]

I collected unmixed regional Croat G25 samples (real coords only, criteria as following: 4 ethnic Croat grandparents from same area) and created averages based on their regional dialect.

Dialect map: Croatia has 3 major dialects


Green = Štokavian, basies of standard Croatian
Purple = Kajkavian
Blue = Čakavian

Now the results:



Samples used for averages were taken from last Balkan paper on South Slav ethnogenesis. Illyro-Roman average was composed of mixed Balkan_IA and Roman_Anatolian genetic profiles from present-day Croatia, barbarian averages contain Eastern_European like genetic profiles from the paper plus some other suitable samples like Avar_Szolad 2 - also north European like genetic profiles from early medieval Germany and finally, elite Avar samples from Panonnia with East Asian profile for the steppe nomad input.

Conclusion is štokavian and čakavian speaking Croats have essentianly no differences worthy of note while kajkavian speaking Croats do have considerable differences to the rest with more Germanic than old Balkan ancestry while other two on average don't contain any. Kajkavians also have somewhat lower eastern European input than the other two averages although difference isn't by nay means significant.

If you wish to comment my model and conclusions you are very welcome.

I would heavily disregard the "Illyro" as a prefix before the Roman because of the archaeogenetic contexts of these CE samples. Given the uniparental distribution of Croats it is rather evident that BA-IA Illyrian Y-DNA and mtDNA are very rare. Imperial Roman era ancestry, some southeast Urnfield, perhaps also some Daco-Thracian-like source pop would fit the overall picture much better.

Also, on a side note "Balkan_IA" or "Old Balkan" as toponyms are extremely generic considering the diverse genetic makeup of the Paleo-Balkans, as in there is not some sort of homogenous bunch.

No, I disagree. Illyro-Roman is used historiographic term to label Romanized population of Dalmatia and Panonnia. I used both native Balkan like samples (from Croatia only) and Roman_Anatolian like samples (also from Croatia only). Our pre-Slavic part is certainly not fitting to Imperial Roman.

Urnfield sound irrelevant to describe pre-Slavic late antiquity/early medieval population of Croatia, as does Daco-Thracian.

Label Illyro-Roman is generic and describes exatly what it states - Romanized native population. This average is most similar to Tuscans, and it makes sense since Illyrians from Croatia were North Italian like and Anatolian Romans from same area were much more southern, thus a Romanized native mix will come out central-Italian like.

[Leeloo]

I gave it a try:

Target: Mik
Distance: 1.7705% / 0.01770530
64.6 Slavic_medieval
35.4 Illyro-Roman

Target: Mik_Mom
Distance: 1.8938% / 0.01893774
57.8 Slavic_medieval
42.2 Illyro-Roman

Target: Mik_Dad
Distance: 2.7812% / 0.02781229
61.2 Slavic_medieval
38.8 Illyro-Roman

Target: Mik_Cousin
Distance: 1.3690% / 0.01369035
63.4 Slavic_medieval
36.4 Illyro-Roman
0.2 Asian

Are you a Croat or Bosniak?

Croat, West-Herzegowina, all ancestors 4 generations back exclusively from Grude, Široki Brijeg and Posušje counties.

Nice. We have similar results.

[Bukva_]

What samples or how did you model Croatian_Chakavian and Stokavian?

I personally tested quite a few people with deeply rooted regional ancestry and collected some other samples online, since I am into genetics for quite long time.

Could you post them?

[Riverman]

Isn't the Germanic component a bit high for the North Western group? Looks as high as in some Slovenians and Hungarians. Do the Kajkavians have known German-Hungarian settlement? How many Kajkavian samples did you use?

It's the reality. Kajkavian Croats are more northern shifted than Slovenes and Hungarians on average and many cluster closest to academic average of Czechs from Prague.

Germanic ancestry in the region seems medieval since it is evenly spread in the rural population. There were some German Saxon miners in some towns, but far too little to impact autosomal genetics in such a way.

It is quite interesting since historiography did not record a major Germanic expansion in the area as far as I am aware. I used 7-8 samples with deeply rooted ancestry, which should be enough since they all have rather uniform scores and come from various locations. Only one sample is different to others, and that one is Kajkavian speaker from Gorski kotar which is region cut from main Kajkavian region in north Croatia and with complex settlement history. His score is more Slovene-like than Kajkavian like and it actually correlates with dialect spoken in his region that some lignuists classify as Slovene-proper influenced.

Well, thank you, but there is a contradiction. If its Medieval, but there were just a few Saxon miners with no big impact, what else could it have been? Must be Pre-Medieval.

I thought its particularly odd that other Croatians get zero Germanic, even though I know that some components can leap, like low admixture being sometimes overlooked by the algorithm.

Such a high number should be reflected in uniparentals too by the way. Is that the case?

[Bukva_]

Target: Bukva_father
Distance: 1.8274% / 0.01827435
67.2 Slav_proxy
32.4 Illyro-Roman
0.4 Asian

Your Slav_medieval doesn't fit to us that well. This is one I made is more Balto-Slavic like. From 3.17 to 1.82 reduction in distance, and German is gone.

Code:

Slav_proxy,0.13100482,0.127428115,0.076713868,0.070202767,0.037995783,0.031810557,0.011687291,0.013470371,-0.002809884,-0.024040643,-0.006156002,-0.005388771,0.018035942,0.029634356,-0.010994435,-0.008493723,-0.010030366,-0.003818951,0.002675593,-0.00874243,-0.002410523,-0.002930757,0.011274889,-0.003391698,-0.003268445

Average isn't made to fit specific individuals, but everyone. It is based on EE-related Slavs from Balkan peper (medieval Czechia, Croatia, Slovakia, Austria and Avar_Szolad 2)

Goal was to estimate Slavic ancestry based on genetic makeup of Slavs from migration era.

It is pointless to look for better fit for individual samples, averages matter. There is no model which produces tight fit for every indivudual and tighter fit does not equal being historically accurate.

I know that you can't make it fit great for everyone, thats not the point I made. Not all Slavic populations were identical, even if you put Lithuania_Marvele sample, some Slovenian and Croatian samples lose your slavic sample entirely.

Just wanted to add to discussion, point out that for some South Slavs and Ukranians slavic ancestry is much more Baltic shifted, not that your slavic sample is wrong.

[Leeloo]

Isn't the Germanic component a bit high for the North Western group? Looks as high as in some Slovenians and Hungarians. Do the Kajkavians have known German-Hungarian settlement? How many Kajkavian samples did you use?

It's the reality. Kajkavian Croats are more northern shifted than Slovenes and Hungarians on average and many cluster closest to academic average of Czechs from Prague.

Germanic ancestry in the region seems medieval since it is evenly spread in the rural population. There were some German Saxon miners in some towns, but far too little to impact autosomal genetics in such a way.

It is quite interesting since historiography did not record a major Germanic expansion in the area as far as I am aware. I used 7-8 samples with deeply rooted ancestry, which should be enough since they all have rather uniform scores and come from various locations. Only one sample is different to others, and that one is Kajkavian speaker from Gorski kotar which is region cut from main Kajkavian region in north Croatia and with complex settlement history. His score is more Slovene-like than Kajkavian like and it actually correlates with dialect spoken in his region that some lignuists classify as Slovene-proper influenced.

Well, thank you, but there is a contradiction. If its Medieval, but there were just a few Saxon miners with no big impact, what else could it have been? Must be Pre-Medieval.

I thought its particularly odd that other Croatians get zero Germanic, even though I know that some components can leap, like low admixture being sometimes overlooked by the algorithm.

Such a high number should be reflected in uniparentals too by the way. Is that the case?

Perhaps I should have clarified better. It should be from migration era, not from later Saxon arrivals. There were Langobards and Goths in the region, but I believed they just passed trough the area without significant settlement. Well, genetics say otherwise.

There is pretty sharp cut between Kajkavian and non Kajkavian speakers that I find rather surprising. There are some more transitional zones indeed, like southern part of central Croatia where all three major Croatian dialects once met (so modern štokavian speaking Croats from there can have kajkavian-like substrata, I saw some examples like that, on gedmatch though).

Never studied haplogroups of the region much tbh. I know they have notably more R1b than other Croatian regions, but I think most of it is of Celtic variety, if I am not mistaken. There is some I1 too but overall I can't say I noticed large presence of Germanic-related haplogroups. Ph2ter should know more, he is from the region.