New map of mtDNA haplogroup U5

[Maciamo]

Haplogroup U5 was the most common maternal lineage among European hunter-gatherers, not just during the Palaeolithic and Mesolithic, but until much later in North and Northeast Europe, notably with the Sami people. U5 is absent from Southwest Asia and very low in most of the Middle East, where its presence could be attributed to foreign invasions and settlements from the Bronze Age onwards (Hittites, Indo-Iranians, Phrygians, Armenians, Greeks, etc.).

U5 correlates mostly with Y-DNA haplogroups I, N1c1 and R1a.

Oddly enough southern Italians (data from Boattini et al. 2013) have even less U5 than the Greeks, Albanians and Maghrebians. This would tend to confirm my theory that E-V13 arrived in Europe in the Palaeolithic from North Africa, then crossed over to Albania and Greece.




UPDATE: a detailed page about the origins, history, distribution and subclades of haplogroup U5 is now available here.

[Angela]

Haplogroup U5 was the most common maternal lineage among European hunter-gatherers, not just during the Palaeolithic and Mesolithic, but until much later in North and Northeast Europe, notably with the Sami people. U5 is absent from Southwest Asia and very low in most of the Middle East, where its presence could be attributed to foreign invasions and settlements from the Bronze Age onwards (Hittites, Indo-Iranians, Phrygians, Armenians, Greeks, etc.).

Oddly enough southern Italians (data from Boattini et al. 2013) have even less U5 than the Greeks, Albanians and Maghrebians. This would tend to confirm my theory that E-V13 arrived in Europe in the Palaeolithic from North Africa, then crossed over to Albania and Greece.

Thank-you for all these maps. They're extremely helpful.

As to the U5 in Italy, I know that Brisighelli et al found total U in the south at up to 20%, and up to 28% total U in the southeast. Were you able to use their data and break it out? Did other studies perhaps have far lower levels?

It generally seems to me as if the U5 was pushed north and north east by the farmers, with perhaps some back migration from Slavic areas into the Balkans perhaps? That might explain some of the numbers in Greece as well, although as I said, Brisighelli gave the impression that U was much stronger in southern Italy.

Some of the refugia areas seem to make sense...Brittany as sort of a finisterre...one near the Caucasus...Sardinia.
I don't quite understand the slice around Perpignan(?)... Narbonne)?)... unless that's a bleed over from the Pyrennees.

Speaking of Sardinia, I don't know if the models that see it as totally Neolithic are correct...there are some signs of Paleolithic presence there. These levels of mtDNA U5, and the y dna I2a1(?) if that is indeed Mesolithic, would seem to indicate otherwise. That's certainly the view of Francalacci.

[albanopolis]

Had E-V13 arrived in the Ballkans through Anatolia, J1 in Ballkans should have been in line with Turkey. Instead its insignificant in Albania. The 1-2% that is actually present can be atributed to Italian or Ottoman conquest of Albania. So I think it makes sense that E-V13 arrived by boat to Sicily and from there spread all over Europe by land or boat. I don,t know if there is any argument that J1 was not present in middle east or Turkey when E-V13 crossed the region.

[Jackson]

Haplogroup U5 was the most common maternal lineage among European hunter-gatherers, not just during the Palaeolithic and Mesolithic, but until much later in North and Northeast Europe, notably with the Sami people. U5 is absent from Southwest Asia and very low in most of the Middle East, where its presence could be attributed to foreign invasions and settlements from the Bronze Age onwards (Hittites, Indo-Iranians, Phrygians, Armenians, Greeks, etc.).

U5 correlates mostly with Y-DNA haplogroups I, N1c1 and R1a.

Oddly enough southern Italians (data from Boattini et al. 2013) have even less U5 than the Greeks, Albanians and Maghrebians. This would tend to confirm my theory that E-V13 arrived in Europe in the Palaeolithic from North Africa, then crossed over to Albania and Greece.

Great map, thanks! What sort of area is roughly represented by elevated levels of U5a1? I had heard it was more eastern European but i don't know.

[Maciamo]

As to the U5 in Italy, I know that Brisighelli et al found total U in the south at up to 20%, and up to 28% total U in the southeast. Were you able to use their data and break it out? Did other studies perhaps have far lower levels?

The total for U is not very useful for the U5 map. Some of the highest frequencies of U observed outside Lapland is actually in the Middle East (Iraq, Iran), but that is U1, U3, U7 and U9.

According to Boattini et al. South Italy has 11.7% of U. The highest frequencies are for U1 (2.4%) and U3 (2.4%), both linked to Y-haplogroup J. Sicily has 9.8% of U (mostly U1, U2 and U5).

It generally seems to me as if the U5 was pushed north and north east by the farmers, with perhaps some back migration from Slavic areas into the Balkans perhaps?

That's also how I see it.

I don't quite understand the slice around Perpignan(?)... Narbonne)?)... unless that's a bleed over from the Pyrennees.

The data for France and northern Spain comes from Garcia et al. 2011 (supplementary table S3). All western France from Loire-Atlantique to Pyrénées-Atlantique was very low, but the Mediterranean coast of France (Hérault, Var) was higher than the Western European average.

Speaking of Sardinia, I don't know if the models that see it as totally Neolithic are correct...there are some signs of Paleolithic presence there. These levels of mtDNA U5, and the y dna I2a1(?) if that is indeed Mesolithic, would seem to indicate otherwise. That's certainly the view of Francalacci.

Sardinia is not purely Neolithic but about 40% Mesolithic too. This is expressed in the high percentage of I2a1a and mtDNA U5, H1 and H3. I am not sure whether I2a1a people mixed with Near Eastern newcomers already during the Neolithic period or if they lived side by side like in Central Europe. But it would have been easier for the two populations to mix in Sardinia than in Central Europe, mostly because Cardium Pottery people were at first essentially goat and sheep herders who also hunted too to complement their diet (as still did Ötzi much later). Nomadic herders therefore had a more compatible lifestyle with nomadic hunter-gatherers than with cereal farmers settled in villages. That may be why in Southwest Europe Near Eastern lineages are typically found alongside I2a1 lineages, and that was already the case at the Cardium Pottery site in Treilles, in Languedoc-Rouissillon.

[PaschalisB]

Thanks a lot for this map! I was surprised to discover I belong to this mtdna haplogroup. My maternal grandmother's grandmother belonged to the Sarakatsani population, a group said to be one of the most ancient in Greece. Maybe they have more U5 than the rest of the greek populations?

[Angela]

Great map, thanks! What sort of area is roughly represented by elevated levels of U5a1? I had heard it was more eastern European but i don't know.

It looks like the highest concentration is in the Saami areas in the far north of Scandinavia, but those would be low population density areas. After that, Finland, far northern Scandinavia, and then the Baltics, with some bleed down into Poland. That darker swathe into Russia might be the Udmurt type areas? Someone with more expertise in that area of Europe would know more about it. Croatia seems to have pooled a bit of it.

I'm surprised by how low the levels are in most of Europe. There couldn't have been all that many Mesolithic fishermen inhabiting the landscape with the Neolithic farmers, unless the ancient y dna tells a very different story when we get it.

Well, there's also U4 to consider...although isn't most of that later?

I'm still chewing over all the pages in the Brandt et all supplement, so I might have to edit this at some point. :)

[ElHorsto]

Enlightening maps as usual, thanks!
Looks like a very good match with the north_east euro autosomal component from K12.
I don't expect perfect local matches because haplogroups have much more volatile distributions than autosomals, yet mtDNA is perhaps still less volatile than Y-DNA.
Regarding Sardinia I still think they almost lost their autosomal legacy from the north-euro-like hunter-gatherers, despite U5 and I2 frequencies, because in all autosomal distance measures I've seen Sardinians appear most distant from north-east europeans.
Although admittedly the increased U5 in Sardinia raises some questions.

[Wilhelm]

The map seems wrong to me. According to the study of García et. al 2011, the north of Spain (the Vasco-Cantabrian area) should have one of the highest levels in Europe (besides Finns and Lapps) :

Cantabria : 17.3%
Pasiegos : 19.5%
Gipuzkoa Basque : 17.7%

[Jackson]

It looks like the highest concentration is in the Saami areas in the far north of Scandinavia, but those would be low population density areas. After that, Finland, far northern Scandinavia, and then the Baltics, with some bleed down into Poland. That darker swathe into Russia might be the Udmurt type areas? Someone with more expertise in that area of Europe would know more about it. Croatia seems to have pooled a bit of it.

I'm surprised by how low the levels are in most of Europe. There couldn't have been all that many Mesolithic fishermen inhabiting the landscape with the Neolithic farmers, unless the ancient y dna tells a very different story when we get it.

Well, there's also U4 to consider...although isn't most of that later?

I'm still chewing over all the pages in the Brandt et all supplement, so I might have to edit this at some point. :)

Well if i remember correctly from a recent lecture, in many parts of Europe hunter gatherers and farmers were contemporary for many hundreds of years or more.

Thanks for the information on U5a1, i remember an earlier study (and shown on 23&me) found in most common in Norway and northern Germany, but thought there would be more information about it's general distribution by now.

[Maciamo]

The map seems wrong to me. According to the study of García et. al 2011, the north of Spain (the Vasco-Cantabrian area) should have one of the highest levels in Europe (besides Finns and Lapps) :

Cantabria : 17.3%
Pasiegos : 19.5%
Gipuzkoa Basque : 17.7%

I have averaged the data for Cantabria to 10.7% (n=242) and the Basque country to 11.7% (n=618). However I just ran into new data about North Navarre (n=318) that gives 17% for U5b1f only, so certainly over 17.5% for all U5.

[bicicleur]

during the last ice age, E1b1 was in Ethiopia, there was nobody in the Sahara, only few A1a in the Atlas mountains and Iberia, E1b1b must have come 6-8000 years ago, when the Sahara turned into a desert again

[bicicleur]

Had E-V13 arrived in the Ballkans through Anatolia, J1 in Ballkans should have been in line with Turkey. Instead its insignificant in Albania. The 1-2% that is actually present can be atributed to Italian or Ottoman conquest of Albania. So I think it makes sense that E-V13 arrived by boat to Sicily and from there spread all over Europe by land or boat. I don,t know if there is any argument that J1 was not present in middle east or Turkey when E-V13 crossed the region.

in answer to above

[Maciamo]

I have added a higher frequency shade for northern Navarre, Bipuzkoa and the Pas-Miera region.

[Maciamo]

during the last ice age, E1b1 was in Ethiopia, there was nobody in the Sahara, only few A1a in the Atlas mountains and Iberia, E1b1b must have come 6-8000 years ago, when the Sahara turned into a desert again

Except that the Sahara was green and luxuriant during a good part of the last Ice Age and it might have had one of the highest human population density on earth.

[bicicleur]

Except that the Sahara was green and luxuriant during a good part of the last Ice Age and it might have had one of the highest human population density on earth.

The Sahara was dead and empty 18-20.000 years ago.
That's when I think some A1a fled to Europe via Iberia, and E1b1 retreated to the Ethiopian heights.
The Sahara may have been green 13-16.000 year ago, and it certainly was 10-6.000 years ago.
But I see no reason why E1b1 would flee to Europe, that was allready well populated at that time.
I believe some E1b1b fled to Europe at the end of that period, e.g. like Almagro culture 7.500 years ago, Andalucia.

[LeBrok]

The Sahara was dead and empty 18-20.000 years ago.
.

It goes like this. The warmer the Earth is the more water evaporates from oceans, therefore more rain falls and greener our globe is.. It means that during cold Ice Ages there is less evaporation and precipitation. Sahara will be the dryest and biggest during Ice Ages. I think Sahara was the greenest in Neolithic, in holocene maximum, when it was warmer than today.
Having said that the end of Ice Age could be wetter too. When moisture builds up in the air and it is drastically cooled when close to edge of Ice Sheets. Also Ice would keep moist air, and gulfstream more south, making South Europe and even Sahara wetter.

[Maciamo]

It goes like this. The warmer the Earth is the more water evaporates from oceans, therefore more rain falls and greener our globe is.. It means that during cold Ice Ages there is less evaporation and precipitation. Sahara will be the dryest and biggest during Ice Ages. I think Sahara was the greenest in Neolithic, in holocene maximum, when it was warmer than today.
Having said that the end of Ice Age could be wetter too. When moisture builds up in the air and it is drastically cooled when close to edge of Ice Sheets. Also Ice would keep moist air, and gulfstream more south, making South Europe and even Sahara wetter.

Actually it has to do with the air and sea currents that vary with climate change. The Sahara was green until the Last Galcial Maximum (24,500 to 17,000 BCE), when it turned into a desert again. Rain came back around 12,500 BCE and the Sahara was green again for three millennia. The desertification re-started around 9500 BCE (at the same time as the start of the Neolithic in the Levant). Then there was a renewed humid period from 7500 to 3000 BCE, after which time the desertification started again and lasted to this day.

The Sahara therefore alternated from dry to wet for hundred of thousands of years and each period lasted several millennia. Humans have lived in the Middle East for over 60,000 years and could have inhabited the Sahara for a much longer period, as attested by the presence of Sahelanthropus tchadensis, our oldest presumed ancestor since the split with the Chimpanzees, who lived between 7 and 6 million years ago in the Sahara (which was obviously not desert at the time).

I do not think it is a coincidence that the Neolithic started in the Levant just as the Sahara started drying up. I believe it could have started earlier in eastern Libya (which would have been criss-crossed by rivers) and Egypt. This would also explain why haplogroup E1b1b had spread all over northern Africa and to the Levant before 9500 BCE.

[Wilhelm]

I have added a higher frequency shade for northern Navarre, Bipuzkoa and the Pas-Miera region.

Ok where can I see the updated map ?

[Maciamo]

Ok where can I see the updated map ?

Did you refresh the page ?

[bicicleur]

I do not think it is a coincidence that the Neolithic started in the Levant just as the Sahara started drying up. I believe it could have started earlier in eastern Libya (which would have been criss-crossed by rivers) and Egypt. This would also explain why haplogroup E1b1b had spread all over northern Africa and to the Levant before 9500 BCE.

There was a theory that agriculture was brought into the Levant from Africa.
Since then it has been discovered that people around lake Tiberias (or lake of Galilee) allready consumed wild grains 20.000 years ago.
Other technologies which were supposed to have been imported from Africa now appear to have existed in the Levant long time before.
I believe that agriculture started in the Levant and southern Anatolia 11.500 years ago by the same people that settled in the Levant allready 44.000 years ago (Ahmarian culture - haplo J2), and I think that it was the other way around : agriculture was brought into Africa from the Levant. Of course it spread very rapidly in the Sahara because of the favourable climate at that time. Indeed most people in the Sahara must have been E1b1, but don't forget, some T and even a branch of R1b also came into Africa, maybe around that time.

[Degredado]

Maciamo, what is your understanding of U5b1? Would it have first pooled in NE Europe during the Paleolithic/Mesolithic and then spread to Western/SW Europe, or the other way around?

[LeBrok]

Actually it has to do with the air and sea currents that vary with climate change. The Sahara was green until the Last Galcial Maximum (24,500 to 17,000 BCE), when it turned into a desert again.

I'm unable to find information showing that Sahara was greener during Ice Age. All my searches say it was very dry during cool periods.

During the last glacial period, the Sahara was even bigger than it is today, extending south beyond its current boundaries.^[17] The end of the glacial period brought more rain to the Sahara, from about 8000 BC to 6000 BC, perhaps because of low pressure areas over the collapsing ice sheets to the north.^[18]

http://en.wikipedia.org/wiki/Sahara

The Sahara therefore alternated from dry to wet for hundred of thousands of years and each period lasted several millennia. Humans have lived in the Middle East for over 60,000 years and could have inhabited the Sahara for a much longer period, as attested by the presence of Sahelanthropus tchadensis, our oldest presumed ancestor since the split with the Chimpanzees, who lived between 7 and 6 million years ago in the Sahara (which was obviously not desert at the time)

.
Yes, it was warmer, wetter and greener on Earth up to 2.5 million years ago when climate really cooled.

http://en.wikipedia.org/wiki/Geologi...erature_record

Ice Ages are last 2 million year fenomenon. Quite recent in geological terms. This might mean that Sahara was Savana like with many lakes till 2 million years ago.

[bicicleur]

last 8 ice ages came in cycles of +/- 100.000 years
last ice age was 20.000 years ago, the one before 128.000 years ago
between 130.000 - 55.000 years ago there seem to have been several (4) wet periods in the Sahara and Arabia
then it became gradualy dryer, and 20.000 years ago it became extremely dry
maybe it was wet 13-16.000 years ago, and it was wet again 6-10.000 years ago
12,7 - 11,6 years ago there was a cold spike (the 'youngest dryas') resulting in the tundra coming back all over Europe and a dry period in the Sahara and Arabia
after that agriculture started in the Levant and southeast Anatolia
wet periods in the Sahara seem to be coming with shifting monsoon winds

[Angela]

Enlightening maps as usual, thanks!
Looks like a very good match with the north_east euro autosomal component from K12.
I don't expect perfect local matches because haplogroups have much more volatile distributions than autosomals, yet mtDNA is perhaps still less volatile than Y-DNA.
Regarding Sardinia I still think they almost lost their autosomal legacy from the north-euro-like hunter-gatherers, despite U5 and I2 frequencies, because in all autosomal distance measures I've seen Sardinians appear most distant from north-east europeans.
Although admittedly the increased U5 in Sardinia raises some questions.

Sorry, I'm playing a little catch up here...I don't recall and can't seem to find any thread where Dienekes deconstructed the North East European component of K-12 in order to show it's relationship to other clusters, but, with the caveat that we're not talking about the same exact cluster, I think it still might be informative to consider what he says about the K=12b North European component.
See:

The North European in that analysis is about 2/3 Atlantic Med, a little less than 1/3 Gedrosia, and a slice of Siberian. Atlantic-Med itself is 90% Caucasus (which we know has a big "Southern" component along with North Euro), with about another 10% North Euro.

If I had to guess, the North East Euro would have less Atlantic Med, less Gedrosia, and more Siberian.

Just for comparison, these are the scores for Germans and Poles in the two analyses:
North Euro/North East Euro
Germans: 48/25.3
Poles: 63%/44.9
Lithuanians: 77/59

This is my round about way of saying that I don't think that North East Euro equates to northern European mesolithics either. :)

That component is mixed as well, just like the Northern European one. There is no remaining North Eurasian Mesolithic population...that's why they keep saying it's outside the range of modern variation. That's also why, I think, Skoglund said that the Northern Europeans are slightly more related to these people than are southern Europeans. We're all picking apart what is essentially not that much variation in Europe any longer, however stark the differences may have been in the early Mesolithic.

As to Sardinians, I posted a study on another thread about the fact that in the Balkans the authors saw assimilation between the foragers and the farmers within a few hundred years. I don't know if that study will stand up in the face of the new results that hopefully will be coming from the Balkans soon. If they are correct, however, and even if I2a1 is "mesolithic", all the wives that the paper indicates they seem to have taken from the newcomers might indeed have changed the autosomal picture for them before some of them set out to colonize the western Mediterranean. On the other hand, should mtDNA "H" turn out to have a Mesolithic presence in Mediterranean Europe, then the calculus would be more in line with what Maciamo has suggested. Of course, if people then wanted, for whatever reason, to differentiate between Northern or North Eastern Mesolithic, and Southern European Mesolithic that could be done as well.