Modern populations and neolithic farmers?

[Hauteville]

Angela maltese are the europeans with higher neolithic farmer affinities (circa 93%) followed by Ashkenazi jewish. No samples from south and island Greece though in Lazaridis study.

 

[Sile]

Have you guys read this ?

http://www.pnas.org/content/111/25/9211.abstract

 

[Hauteville]

Yes but it is different from Lazaridis study however.

 

[Angela]

Angela maltese are the europeans with higher neolithic farmer affinities (circa 93%) followed by Ashkenazi jewish. No samples from south and island Greece though in Lazaridis study.

As far as the Ashkenazim are concerned, I think that part of it may have to do with the fact that they have been mostly endogamous for the last approximately 2000 years, which means that they have preserved many autosomal (as well as uniparental)l signatures lost not only in Europe but in the Near East. The other, somewhat contradictory part of the equation is that I think the mercantile, coastal Mediterranean "Jewish" population of the Classical era which provided the "founder" groups of the non Near Eastern "Jewish" population was already pretty "mixed" away from its strictly Levantine roots. However, I don't want to derail the thread into a long discussion of Ashkenazi, in particular, ethnogenesis.

As to the Greeks, there is a dearth of samples. We don't know the origin of the samples in Dodecad. The Lazaridis study didn't use samples from the Peloponnese or the islands, to my knowledge, but their sample, which I believe is from far northern mainland Greece (Thessaly?), is 79% EEF.

The Paschou et al study does provide some academic samples for these areas, but doesn't provide an EEF/WHG/ANE analysis. As you say in your PM, looking at all their data and graphics it does seem to show that Sicilians are between the mainland and the island Greeks, so it's not unreasonable to speculate that island Greeks have more EEF, and mainland Greeks less than Sicilians. In those graphics, although a few mainlanders overlap with far southern Tuscans(probably the Thessaly group), most mainland Greeks are south of central Italians. The people of Firenze who provide the sample for the Tuscans are 75% EEF. Thessaly is 79%. By the time you get to Lazio I don't think it would be outlandish to speculate that you'd be close to the 83% levels you see in much of Spain. So I'd say that upwards of 83%, would be in the ballpark for most mainland Greeks and certainly for the Peloponnese. I wonder if they would be a more ANE shifted version of Campanians or Lucanians perhaps? However, someone should do a study on the Greeks which is careful to screen samples for long history in certain areas, at least to great grandparent level, especially given the population upheavals in Greek history in the last 1000 years and even very recently.

 

[Hauteville]

As far as i know the samples of Lazaridis et al are from Thessaloniki.

 

[oreo_cookie]

The Paschou et al study does provide some academic samples for these areas, but doesn't provide an EEF/WHG/ANE analysis. As you say in your PM, looking at all their data and graphics it does seem to show that Sicilians are between the mainland and the island Greeks, so it's not unreasonable to speculate that island Greeks have more EEF, and mainland Greeks less than Sicilians.

The islands sampled in that study are Cretans and Dodecanese, and from having seen many 23andme, and Gedmatch results for different Greek islanders it is likely that these Greeks have the most Near Eastern ancestry of all the islands, and other islanders seem to be more similar to the mainland, either due to proximity or due to the smaller pre-Greek populations as compared to Crete (which was heavily populated by Minoans).

Peloponnesians are very similar to other mainland Greeks. In fact some of them on the 23andme PCA plot fall near me, and I am half southern Italian, a quarter Polish and a quarter Portuguese. I suspect most mainland Greeks are a more Balto-Slavic influenced version of Abruzzese and Campanians.

Anyway, since the vast majority of the Greek population is mainland (and centered toward the north and center of the country), it makes most sense to represent Greece in studies with a mainland sample.

 

[Angela]

The islands sampled in that study are Cretans and Dodecanese, and from having seen many 23andme, and Gedmatch results for different Greek islanders it is likely that these Greeks have the most Near Eastern ancestry of all the islands, and other islanders seem to be more similar to the mainland, either due to proximity or due to the smaller pre-Greek populations as compared to Crete (which was heavily populated by Minoans).

Peloponnesians are very similar to other mainland Greeks. In fact some of them on the 23andme PCA plot fall near me, and I am half southern Italian, a quarter Polish and a quarter Portuguese. I suspect most mainland Greeks are a more Balto-Slavic influenced version of Abruzzese and Campanians.

Anyway, since the vast majority of the Greek population is mainland (and centered toward the north and center of the country), it makes most sense to represent Greece in studies with a mainland sample.

Sorry, 23andme results, like the results of any private testing firm, are self-selected samples from people interested in the topic and with sufficient income to indulge their interest. Any analysis based on these results is therefore not dispositive. (The same thing happened with y Dna testing by the private companies, heavily skewed toward people of British Isles descent, which obscured the source and routes of migration of many R1b clades. Of course the results when properly understood have been very helpful in other ways.)

In addition, there is no way of knowing if any individual's "collection" of results is an "accurate" reflection of all 23andme results for a certain group as a whole. Of course, I'm excluding the situation where some hacker could have gotten into the 23andme data base, but in that case, such a person would be unlikely to present the date honestly any way, and ultimately the 23andme me samples are still not necessarily representative of any "national" group.

Therefore, only academic samples chosen using standard protocols should form the basis for any solid conclusions. Of course, such results then have to be analyzed carefully. There are academics, and then there are academics. None of their conclusions should be taken at face value. At least, however, we have scientifically selected samples most of the time as a start point.

Your decision to ignore Greek island samples is novel to say the least. Perhaps we should ignore Basque samples as well? How about the decision in Lazaridis et al to divide the Spanish sample into a "northern" group and a "southern" group? Was that a bad idea as well? Or, in a highly structured country like Italy, is it a bad idea to divide the samples into northern, Tuscan and southern Italian results?

The fact that modern Greeks (like modern Spaniards, or Italians, or anyone else) currently live in large urban areas with more economic opportunities is irrelevant for population genetics purposes. The purpose of this testing is to see if the results from "modern" populations can give us any clues about ancient migrations, which looks increasingly as if it is very problematic indeed. Unless, of course, your analysis has another purpose entirely, of which I'm unaware?

Also, you might want to refresh your recollection of Paschou et al. Samples from the Peloponnese were included in the study.

 

[oreo_cookie]

Your decision to ignore Greek island samples is novel to say the least.

It's not "ignoring" it, Just stating that Crete and the Dodecanese might not be genetically representative of the islands in the North Aegean (like Lesbos) or Ionian islands (like Lefkada). And based on the differences in results I have seen for people in Crete and Dodecanese compared to other islands, they would not be representative as these are the most genetically outlying islands.

 

[Hauteville]

I am curious to see what kind of look the neolithic farmers had. It's must be interesting.

 

[Angela]

It's not "ignoring" it, Just stating that Crete and the Dodecanese might not be genetically representative of the islands in the North Aegean (like Lesbos) or Ionian islands (like Lefkada). And based on the differences in results I have seen for people in Crete and Dodecanese compared to other islands, they would not be representative as these are the most genetically outlying islands.

Let me rephrase: Your opinion that "it makes most sense to represent Greece in studies with a mainland sample" is novel to say the least. I've never before heard anyone interested in population genetics opine that characterizing the population of a highly genetically structured population by using one sample from one extreme of that spectrum (far northern Greece) is a good idea. I will also repeat: the population in all developed countries has moved to larger cities for work. However, in order to accurately represent the population structure of a country one must sample people from all areas. This is even more true if one is attempting to trace migration flows in order to answer questions about the peopling of Europe. However, perhaps your choices and focus are dictated by other motivations.

That may or may not be the case. Perhaps I was unclear, so I will repeat. Any collection of results you may have assembled through sharing with people on 23andme, even if they are honestly presented, are not necessarily representative of the entire body of Greek results at 23andme, without even taking into consideration the fact that the samples at 23andme are not a scientifically chosen representative sample, so I'm afraid I find your conclusions to be on the level of anecdotal evidence.

When you are able to provide scientifically chosen samples from these areas that could be interpreted in such a way, I would, of course, be happy to discuss further what these would then say about migration flows and the peopling of Europe. I'm afraid I'm uninterested in anthrofora games on the level of who has the most North African, who has the most Near Eastern, who has the most SSA, and who has the most Asian ancestry in order to determine who is more "European" so I will not be participating in any such discussions.

 

[oreo_cookie]

There was a y-dna study of the islands that showed that Crete was genetically distinct and that Chios and Lesbos were closer to the mainland; it doesn't directly translate into autosomal DNA but supports what I have seen so far, I can link you to it if you'd like.

 

[Angela]

There was a y-dna study of the islands that showed that Crete was genetically distinct and that Chios and Lesbos were closer to the mainland; it doesn't directly translate into autosomal DNA but supports what I have seen so far, I can link you to it if you'd like.

I've read all the major studies about the dna of the Mediterranean so I doubt it's new to me. However, FYI, if you're going to make a claim based on a scientific study it's your responsibility to provide the link or at least the name of the study. Otherwise, I, at least, am not going to take it very seriously.

Also, as you say, we are discussing autosomal analysis here, and y Dna haplogroups are only very speculatively tied to autosomal "clusters".

 

[Aberdeen]

If I was going to do a study of Greek genetics, I'd want to find out where they came from a few generations back, at least. After all, about 1.5 million people were transferred from Turkey to Greece as a result of the Chanak crisis of 1922 and that must have affected modern Greek genetics quite significantly. And of course people from all over rural Greece and nearby countries have been moving to Athens looking for jobs for several decades now, although the current economic crisis seems to be causing a small reversal of that trend. As for islands, some of them may have populations that have remained relatively stable for generations while others have had complete population turnovers, so understanding the history of the island whose population one studies would be quite important, I think.

 

[Hauteville]

I always thought that the modern greeks who are more closer to ancient greeks are the maniotes. It's my opinion eh.

 

[oreo_cookie]

I always thought that the modern greeks who are more closer to ancient greeks are the maniotes. It's my opinion eh.

Lakonia region? If so, it's between them and Cretans (or somewhat overlapping with both) that Sicilians plot on the study you mentioned earlier.

 

[Hauteville]

A greek girl i know in my personal life (she is from Rhodes and she has studied here) said me that about maniotes. She thinks that maniotes descendent from ancient spartans and remained isolated. I also think that southern calabrians from Bovesia have manteined very good their greek genetic heritage.

 

[oreo_cookie]

A greek girl i know in my personal life (she is from Rhodes and she has studied here) said me that about maniotes. She thinks that maniotes descendent from ancient spartans and remained isolated. I also think that southern calabrians from Bovesia have manteined very good their greek genetic heritage.

You'd have to see if southern Calabria and Mani plot together on a PCA plot. Calabria might end up more like Crete though.

 

[Alan]

Alright, but from what I know one of the components that make up EEF is a Mesolithic component that is WHG like, please correct me if I'm wrong though.

EEF is basically ~75% Early Near Eastern Farmer and ~15 to 20% WHG like.


Sardinians come closest to Early European Farmers but perfect examples of early Near Eastern farmer don't exist anymore. But I would imagine something close to Cypriots. Since they have very low ANE (~7%) and almost no SSA (~2%) admixture.

 

[Alan]

I will here speak only for myself. I get the distinct feeling at times that this obsession with quantifying the amount of "WHG" in EEF may, in some people, stem from an attempt to nail down exactly how "European" a group or a person is by taking that figure and adding it to the "regular" WHG in the figures given for modern populations in Lazaridis et al. Apparently, ANE is considered "European" enough not to raise concern, despite its eastern affinities. The short answer is that we are all 100% European, whether we come from France, or Finland, or southern Italy, and whether we can be modeled best with two or three of these ancestral EEF/WHG/ANE populations.

The funny thing. People questioning this want to be the least Near Eastern most WHG admixed but at the same time "lightest" of all, not knowing or ignoring the basic fact that genes for light skin were brought to Europe by farmers.

 

[Greying Wanderer]

Questions concerning the amount of "WHG" in EEF keep being asked and again. For suggesting that it is a distraction in terms of the discussion of the peopling of Europe, a respected, published author has been attacked on another forum.

I will here speak only for myself. I get the distinct feeling at times that this obsession with quantifying the amount of "WHG" in EEF may, in some people, stem from an attempt to nail down exactly how "European" a group or a person is by taking that figure and adding it to the "regular" WHG in the figures given for modern populations in Lazaridis et al. Apparently, ANE is considered "European" enough not to raise concern, despite its eastern affinities. The short answer is that we are all 100% European, whether we come from France, or Finland, or southern Italy, and whether we can be modeled best with two or three of these ancestral EEF/WHG/ANE populations.

That may well be true but there is also another more fundamental historical reason for getting to the bottom of it imo i.e the two distinct paths East and West Asia took caused imo by the East Asian farmer expansion being more total.

The dominant theory until very recently was European replacement by neolithic farmers and that appears not to be the case probably (again imo) because the process was interrupted by the I-E. So East Asia had an almost complete farmer expansion (apart from a few refuge zones) whereas in Europe it was stalled. I'd say that was possibly quite a big deal in historical terms. For example the earlier spread of civilization in East Asia and southern Europe.

edit:

After reading through the rest of the comments

EEF is defined as 44% plus or minus 10% "Basal Eurasian". (for some reason I keep remembering it as 50 +/- 10)

Combining Basal and WHG this way and then comparing the composite to WHG separately, conceptually fits the neolithic expansion model better but when you separate the two components then it looks more like mesolithic survival so which one it actually is matters. If it's mesolithic survival (not necessarily directly but by cousin) then how come given the extent of the initial farmer expansion? The I-E are the most likely answer to how come but either way if the farmer expansion into Europe was stalled in the north did the slower arrival of farming (crop-centric version as opposed to herding version) to at least northern and central Europe have any significant consequences? I'd say the answer is likelyto be yes.

#

"Basal Eurasian peaks among Arabs. Israeli Beduins..."

If it's Bedouin then it's possibly so high because the desert was a refuge so that component could have originally been much more widespread i.e. Basal may have been in some parts of Europe (south and west) before the farmers as well as coming with the farmers.

#


"The import of the Lazaridis paper is that there were three migrations into Europe [my notes in brackets]:

1. From the Asian crossroads/Middle East in the Palaeolithic. [mtDNA U and Y-DNA IJ and F]
2. From the Middle Eastern Neolithic heartland in the Neolithic. [mtDNA U3, H, I, J, V etc and Y-DNA G, with a bit of E]
3. From the Asian steppe in the Copper Age. [Y-DNA R]"
(Jean Manco)

The following are just guesses for my own entertainment but personally I think E is likely to be or be connected to Basal, with a widespread coastal distribution including north Africa, southern and western Europe and Arabia and earlier than G - whether before or after IJ I guess time will tell although my gut feel is IJ are both the product of Basal + archaics.

I also think the fertile crescent was likely a big swamp before farmers drained it so the spread of neolithic farming was more likely from the surrounding highlands or possibly the coast.

There was probably a significant geographic / climactic feature that split R1b and R1a.