How did I2a-Din get to the Balkans?

[sparkey]

I really dont see any connection with I2a1b and Slavs. For one thing I2a1b is mainly in southeast Europe not slavic speaking areas.

...Eh? Those aren't mutually exclusive. Southeast Europe is largely Slavic speaking, and I2a-Din is higher in Slavic-speaking areas of Southeast Europe (for example, its frequency dips in Albania and Greece).

I suppose you're trying to argue that I2a-Din is not as common in the Slavic countries without Illyrian influence. But it's really not so much of a dip that it invalidates a Slavic connection to it: 21% in Ukraine, 18% in Belarus, 11% in Russia, 9% in Poland (including out members of the haplotype most common in the Balkans), etc. These are not particularly Illyrian places, but quite Slavic.

This, of course, leads to the question: Why do West Slavs and East Slavs have lower I2a-Din : R1a ratios than South Slavs? To me, the obvious answer is regional differences that were amplified during expansion. We also see regional differences among different Germanic peoples, and among different Celtic peoples, so why not among different Slavic peoples?

I've repeated this line of argument several times now, although I can forgive you if you don't have time to read through this whole thread. :embarassed:

The Balto Slavic language was spread by Corded ware culture there is no connection with I2a1b Slavs, or corded ware culture.

I agree that there is no obvious connection between I2a-Din and Corded Ware culture, other than that its initial launching point was once within Corded Ware culture. But as I just said, the Balto-Slavs were likely R1a dominant, and I suspect that the drift (or amplification) of I2a-Din within their population happened after (or as) they differentiated into Balts and Slavs. So this line of argument doesn't contradict me.

In my opinion I2a1b is a Paloithic haplogroup that was able to survive migrations and invasions.

This doesn't contradict me either.

&,000ybp it probably dominated all of eastern Europe except Russia which is where proto Indo Europeans would have been.

Then why do the 2 currently known subclades of I2a1b to have existed around that time not converge on Eastern Europe? One has its highest diversity in Western and Central Europe (with an interesting outlier in Iraq), and the other is evenly split between Western and Eastern Europe. If anything is clear about I2a1b, it's that we can't pin it down yet. Eastern Europe seems like far from a safe assumption at this point, much less speculating that it "dominated all of eastern Europe except Russia."

I1 is not yound at all i get sick of these young haplogroup predictions. The people that make these predictions dont even consider ancient Cultures like where they where at what time. They would realize there is no way I1 is less than 10,000 years old. The Scientits who figure all of the DNA stuff out are terrible at making conclusions on how it spread to figure that stuff out u need to be like a detective not a sceintits.

So they need to start from conclusions about how haplogroups map to ancient cultures, then work backwards to come up with dates, and, uh... ignore STR and SNP dating converging to give a young age for I1?

What about STR and SNP dating makes you think they're so inaccurate? Just that they don't match the conclusions you've already made about I1?

how do u explain how spread out it is in Scandinavia

Because I1 and Scandinavia were both in the equation during the Germanic ethnogenesis.

and why the rest of Europe has diff subclades than Scandinavia

They don't have different subclades entirely, at least of the common higher-level subclades, just different frequencies of the same ones. This is due to normal regional variance patterns that you get with every haplogroup.

and how I1 has so many subclades.

Having many subclades all expanding around the same time is evidence for rapid expansion out of a bottleneck. I would suggest its drift into, and expansion via, the early Germanic peoples is the main culprit.

 

[gyms]

...Eh? Those aren't mutually exclusive. Southeast Europe is largely Slavic speaking, and I2a-Din is higher in Slavic-speaking areas of Southeast Europe (for example, its frequency dips in Albania and Greece).
There is 20% I2a-Din in Székelys from Bukovina and 25% I2a-Din in Jász people from Jászság,Hungary.Gagauzes are not slavs.They have 24-32% I2a-Din.

 

[sparkey]

There is 20% I2a-Din in Székelys from Bukovina and 25% I2a-Din in Jász people from Jászság,Hungary.Gagauzes are not slavs.They have 24-32% I2a-Din.

The 15% as a whole in Hungary doesn't really contradict that these groups could be impacted by neighboring groups. It borders Serbia (35%) and Croatia (37%) after all. Besides, Hungary is usually counted as Central Europe rather than Southeast Europe.

The Romanians and Moldovans are more challenging, I admit, although it's worth noting that they are heavily I2a-Din-N, as opposed to the I2a-Din-S that is more prominent in the Slavic expansion on the Balkans. Could I2a-Din-N in Romanians and Moldovans be a leftover from the I2a-Din source population(s) before it drifted into, and expanded with, the Slavs? Maybe. If anybody has that, it's the Romanians and Moldovans.

On a related note, has anybody followed the discovery of the "I2a-Din-cousins" branch? Basically, a more ancient sample on the I2a-Din branch than I2a-Din-N has been found in a family in southern Poland. See Cullen. This may place the Dinaric branch more solidly in Central Europe, and imply that the early I2a-Din-N is an Eastern split off it, with perhaps I2a-Din-S a backmigration to Poland before its expansion on the Balkans. It may not help us solve which populations it belonged to, but it's an interesting indication that there may be more yet to find among modern populations...

 

[gyms]

"You're making an argument from modern frequency again. Haven't you read yet about why this is a poor line of reasoning?"

 

[Sile]

Why do you think I2a-Din migrated "in BC times" ?

It would be because if the Illyrians are predominately R1b and that the I2a-Din arrived with the slavs in 500AD, then R1b would be more dominant than it is now in the areas of Croatia, Serbia etc. Instead I2a-Din dominates

 

[Shetop]

It would be because if the Illyrians are predominately R1b and that the I2a-Din arrived with the slavs in 500AD, then R1b would be more dominant than it is now in the areas of Croatia, Serbia etc. Instead I2a-Din dominates

I don't know if this is also your opinion, but I am not following the logic there. Is it because Illyrian language was Indoeuropean (and R1b should be Indoeuropean also)? My opinion is that there are no real arguments for such scenario.

I mean, look at the Hungarians - the dominant haplogroups among them and the language they speak. That simple example knocks the above theory down. Both language and high frequency arguments are not those that should be considered as strong in that kind of analysis.

 

[sparkey]

"You're making an argument from modern frequency again. Haven't you read yet about why this is a poor line of reasoning?"

Are you trying to make a point by quoting me here?

 

[gyms]

Yes sparkey.Your arguments are based on modern frequencies and languages. "The 15% as a whole in Hungary doesn't really contradict that these groups could be impacted by neighboring groups. It borders Serbia (35%) and Croatia (37%) after all. Besides, Hungary is usually counted as Central Europe rather than Southeast Europe."

 

[sparkey]

Yes sparkey.Your arguments are based on modern frequencies and languages. "The 15% as a whole in Hungary doesn't really contradict that these groups could be impacted by neighboring groups. It borders Serbia (35%) and Croatia (37%) after all. Besides, Hungary is usually counted as Central Europe rather than Southeast Europe."

My principal objection is with using modern frequencies to determine ancient frequencies, as Fire Haired was doing. But you brought up modern frequencies only. Using neighboring modern frequencies to determine the source of modern frequencies is less shaky ground.

That said, of course, performing a diversity analysis would tell us more. We need to look at the haplotype distribution of I2a-Din in Hungary, and which direction its cline points. Verenich sees a cline coming from the direction of Bosnia:

Two "arrows" - one, connecting Bosnia to Hungary, and another, linking Montenegro to Serbia - may represent more recent I2a2a population migrations of XV-XVII centuries.

Is that better?

 

[gyms]

"Is that better?" Who knows?We don`t know anything for sure. http://www.fsigenetics.com/article/S1872-4973(13)00102-6/abstract

 

[sparkey]

"Is that better?" Who knows?We don`t know anything for sure. http://www.fsigenetics.com/article/S1872-4973(13)00102-6/abstract

So you're not even willing to comment on an analysis until we know things for sure? How boring. I'll be looking at the evidence we have available while you wait on that.

Interesting citation by the way, looks like a follow up to the Wei 2012 paper Dienekes discussed here. In that one, their GENETREE estimates had a +/- range of 8,100 years (!) for R1b estimation, as opposed to the 1,000 years of the Rho method, which matched pretty much exactly with older STR+SNP methods. So I'm not entirely clear why this 2013 update is making it sound like they are now taking GENETREE out to be the most correct method, and suggesting that we need to shift all of our old STR+SNP estimations to be older just so that they line up with it. It'll be interesting to see if this is convincing enough to have other experts change their methodology, and what comes from that if they do. I wouldn't get my hopes up about that, though.

 

[Garrick]

According to Regueiro et al in Gene (2012) high levels of paleolithic Y-chromosome lineages characterize Serbia (approximately 58%), in this percent was not included R1b (approximately 8% according to authors). Near eastern neolithic lineages (E and J subclades) are about 26,5%. It can be assumed that the people who inhabited Serbia in paleolithic, also inhabit today.
[h=1][/h]

 

[sparkey]

According to Regueiro et al in Gene (2012) high levels of paleolithic Y-chromosome lineages characterize Serbia (approximately 58%), in this percent was not included R1b (approximately 8% according to authors). Near eastern neolithic lineages (E and J subclades) are about 26,5%. It can be assumed that the people who inhabited Serbia in paleolithic, also inhabit today.

Without reading the paper in-depth, they seem to have selected Y lineages as representative of the "Paleolithic" based on how old they seem to be in Europe (counting the Steppes in the case of R1a), not based on how old they seem to be in Serbia. High levels of Paleolithic European Y-DNA in Serbia does not imply that these lineages have been in Serbia since the Paleolithic, just that they were around Europe somewhere during the Paleolithic (probably) and are in Serbia now.

 

[Garrick]

Without reading the paper in-depth, they seem to have selected Y lineages as representative of the "Paleolithic" based on how old they seem to be in Europe (counting the Steppes in the case of R1a), not based on how old they seem to be in Serbia. High levels of Paleolithic European Y-DNA in Serbia does not imply that these lineages have been in Serbia since the Paleolithic, just that they were around Europe somewhere during the Paleolithic (probably) and are in Serbia now.

No, authors are explicit (quote):

"The relatively old expansion time (14.0±3.3 KYA) (Supplementary Table 3), associated mean variance (0.384) and high haplotype diversity (0.9905±0.0178) (Table 1), also evident in the phylogenetic network (Supplementary Fig. 1C), among Serbian R1a1a-M198 carriers, are consistent with previous studies (Peričić et al., 2005; Semino et al., 2000; Wells et al., 2001) that suggest that the common ancestor for all R1a1a-M198 individuals in the Balkans existed in Paleolithic times. "

According to authors R1a (precisely R1a1a-M198) is very old in Serbia, and older from I branches (I1 and I2a) and all other today's haplogroups in the Serbia and Balkans.

 

[sparkey]

No, authors are explicit (quote):

"The relatively old expansion time (14.0±3.3 KYA) (Supplementary Table 3), associated mean variance (0.384) and high haplotype diversity (0.9905±0.0178) (Table 1), also evident in the phylogenetic network (Supplementary Fig. 1C), among Serbian R1a1a-M198 carriers, are consistent with previous studies (Peričić et al., 2005; Semino et al., 2000; Wells et al., 2001) that suggest that the common ancestor for all R1a1a-M198 individuals in the Balkans existed in Paleolithic times. "

According to authors R1a (precisely R1a1a-M198) is very old in Serbia, and older from I branches (I1 and I2a) and all other today's haplogroups in the Serbia and Balkans.

They may be right about R1a, although I'm worried about it being a possible pooling point. Serbia is fairly close to the highest diversity areas of R1a, and expansion of R1a into Serbia quite possibly could have been by those who had high R1a diversity already. I won't say much more about R1a because I haven't studied it as close as many others.

Serbia is not close to having the highest diversity of I1 or I2a, though. Neither can be traced to the Paleolithic there. It's possible that regional Haplogroup I is a backmigration; in fact, I'd argue that it's very likely. But the temporal separation there must be quite large to match with phylogenetic patterns.

 

[Garrick]

They may be right about R1a, although I'm worried about it being a possible pooling point. Serbia is fairly close to the highest diversity areas of R1a, and expansion of R1a into Serbia quite possibly could have been by those who had high R1a diversity already. I won't say much more about R1a because I haven't studied it as close as many others.

Serbia is not close to having the highest diversity of I1 or I2a, though. Neither can be traced to the Paleolithic there. It's possible that regional Haplogroup I is a backmigration; in fact, I'd argue that it's very likely. But the temporal separation there must be quite large to match with phylogenetic patterns.

I read that you wrote that I2a was 8,500 years ago. And Regueiro et al. attempted to determine the age of I2 in Serbia (quote):

"The high haplotype diversity of I2a2-P37.2/M423 lineages in Serbia (0.9977±0.0094) also supports the hypothesis that the P37.2 mutation has been present in the Balkans before the LGM (Rootsi et al., 2004; Semino et al., 2000). Moreover, the age of I2a2-M423 chromosomes in Serbs based on accumulated Y-STR variation, is ~9000 years ago (Table 1) and is consistent with previously reported time frames in the Early Holocene, during the European Mesolithic period (Battaglia et al., 2009; Peričić et al., 2005; Rootsi et al., 2004; Underhill et al., 2007). Our data corroborate the previous assertion by Battaglia et al. (2009) that the autochthonous haplogroup I2a2-M423 is indicative of the adoption of farming by Mesolithic hunter–gatherers in the Balkans."

 

[sparkey]

I read that you wrote that I2a was 8,500 years ago. And Regueiro et al. attempted to determine the age of I2 in Serbia (quote):

"The high haplotype diversity of I2a2-P37.2/M423 lineages in Serbia (0.9977±0.0094) also supports the hypothesis that the P37.2 mutation has been present in the Balkans before the LGM (Rootsi et al., 2004; Semino et al., 2000). Moreover, the age of I2a2-M423 chromosomes in Serbs based on accumulated Y-STR variation, is ~9000 years ago (Table 1) and is consistent with previously reported time frames in the Early Holocene, during the European Mesolithic period (Battaglia et al., 2009; Peričić et al., 2005; Rootsi et al., 2004; Underhill et al., 2007). Our data corroborate the previous assertion by Battaglia et al. (2009) that the autochthonous haplogroup I2a2-M423 is indicative of the adoption of farming by Mesolithic hunter–gatherers in the Balkans."

OK, I found the paper, and found the culprit for these unusually high estimates for I2a-Din in the Balkans:

The age of microsatellite variation within haplogroups E, I and R was estimated as reported earlier by Zhivotovsky et al. (2004) and modified according to Sengupta et al. (2006).

I'm not familiar with the Sengupta method, but the Zhivotovsky method basically recommends tripling age estimates, so this paper actually comes pretty close to corroborating the Nordtvedt estimates if we reverse the Zhivotovsky adjustment. Additional difference (there's little) could come from the low number of STRs in this paper and the older mutation rate estimates.

Anyway, this doesn't invalidate the fact that Poland has the highest I2-L147.2 diversity. If it turns out that Zhivotovsky is right, then the likely backmigration happened earlier. But I think acceptance of Zhivotovsky is still a minority. Non-Zhivotovsky calculations have come closer to newer SNP-only estimates, for example. Some have been higher than Nordtvedt estimates, but usually at adjustments of about 1.1x, not 3x.

Edit: This also means that their conventional R1a age estimate is ~4700 YBP.

 

[Garrick]

OK, I found the paper, and found the culprit for these unusually high estimates for I2a-Din in the Balkans:

I'm not familiar with the Sengupta method, but the Zhivotovsky method basically recommends tripling age estimates, so this paper actually comes pretty close to corroborating the Nordtvedt estimates if we reverse the Zhivotovsky adjustment. Additional difference (there's little) could come from the low number of STRs in this paper and the older mutation rate estimates.

Anyway, this doesn't invalidate the fact that Poland has the highest I2-L147.2 diversity. If it turns out that Zhivotovsky is right, then the likely backmigration happened earlier. But I think acceptance of Zhivotovsky is still a minority. Non-Zhivotovsky calculations have come closer to newer SNP-only estimates, for example. Some have been higher than Nordtvedt estimates, but usually at adjustments of about 1.1x, not 3x.

Edit: This also means that their conventional R1a age estimate is ~4700 YBP.

It is not easy to grasp movements that occurred thousands of years ago. Your contribution to these complex topics deserves high respect.

A lot of scientists claim that R1a in Serbia and Balkans is Paleolithic, and their serious scientific research cannot be neglected. Their papers were published in relevant scientific journals. Certainly on the methods can be argued but it is clear that strict scientific findings cannot be ignored. Is R1a in Serbia and Balkans 20,000 years ago, 10,000 years or younger, the fact is that it is very old.

Reguiero et al. (2012) write (quote):

“Three major episodes of gene flow have been described to explain the high R1a haplotype diversity in the Balkans: (1) re-colonization from the refugium in the Ukraine (early post-LGM, ~20–12 KYA) (Passarino et al., 2001; Semino et al., 2000); (2) migrations from the Pontic steppe associated with the Indo-European Kurgan culture (3000–1000 B.C.) (Rosser et al., 2000; Semino et al., 2000); and, more recently, (3) the massive Slavic migration (5th–7th centuries) (Barać et al., 2003; Peričić et al., 2005).”

According authors first expansion R1a in the Balkans was between 20,000 and 12,000 years ago. Prior to the researches performed in the Balkans and published studies about haplogroups there were those who thought that the R1a first time came to the Balkans in the fifth century BC with the expansion of Slavs. We can all agree that it was a mistake.

 

[sparkey]

It is not easy to grasp movements that occurred thousands of years ago. Your contribution to these complex topics deserves high respect.

And thank you for bringing the study to my attention and challenging my assumptions. This is productive.

A lot of scientists claim that R1a in Serbia and Balkans is Paleolithic, and their serious scientific research cannot be neglected. Their papers were published in relevant scientific journals. Certainly on the methods can be argued but it is clear that strict scientific findings cannot be ignored. Is R1a in Serbia and Balkans 20,000 years ago, 10,000 years or younger, the fact is that it is very old.
...
According authors first expansion R1a in the Balkans was between 20,000 and 12,000 years ago. Prior to the researches performed in the Balkans and published studies about haplogroups there were those who thought that the R1a first time came to the Balkans in the fifth century BC with the expansion of Slavs. We can all agree that it was a mistake.

As mentioned, non-Zhivotovsky estimates using this study has a lower than 10k year estimate by a lot, but not so much to suggest that the R1a in the Balkans is the result of an expansion out of a single lineage from the Slavs. The R1a in the Balkans isn't necessarily Paleolithic (I can't construct that scenario with the younger estimate), but it seems clear that it's older than the Slavs, layered, or pooled.

 

[Noman]

Let me say this, I'm personally quite torn regarding I2a-Din. Generally, the idea that Haplogroup I as a whole is Paleolithic/Mesolithic in Europe is almost certain at this point. However, wether the I2a peak in the western Balkans is ancient or the result of a more recent founder effect, I'm quite divided on and I have come to no conclusion to this for myself.



Well Dale, I am terribly sorry, I do not know what your source is, but this view regarding Haplogroups R1a/R1b is hopelessly outdated. We know for sure now that both R1a and R1b entered Europe significantly later than the end of the last ice age. The oldest occurence known thus far of R1a in Europe is from a site near Eulau, Germany, which dates back to the Corded Ware Culture (circa 2600 BC). The oldest find of R1b in Europe (thus far!) is from Lichtenstein Cave in northern Germany, which dates into the Urnfield Culture (1000 BC). Both the Neolithic sites of France (Treilles) and Germany (Derenburg) yielded no R1a or R1b what so ever.

Evidence says to me r1a has always been right where it is. How can you say it's not always been there without providing an example of something there in the same place before it was. What a joke.