April 1, 2011

Canarians, NW Africans, Iberians, etc. from the viewpoint of autosomal DNA

This new paper has several points of interest: on one side it studies Canarians in some detail, on the other it compares NW Africans and Iberians (and others) in a way that I cannot recall being done previously. All through the use of autosomal markers (those that represent best overall ancestry, regardless of historical accidents of gender bias).



The overall comparison of Canarians, mainland NW Africans, Iberians and other European populations is probably easiest to appreciate in fig. 1:

FST-based multidimensional scaling plot

Avery similar graph is achieved with a different technique (PCA) in figure 2, with samples plotted individually. 

Some overlap between North Africa and Iberia can be appreciated, however, as the authors note, this is smaller that what could have been led to believe based on autosomal markers such as Y-DNA. The greater (9%) North African influence in the Western half of the peninsula in comparison to the Eastern half (2%), also apparent in autosomal markers, is confirmed here. 

Figure 3. STRUCTURE [K=2] results based on EuroAIMs.
[Note: EuroAIMs are one specific subset of autosomal markers, chosen because they vary a lot between populations. CAN and CBN are both Canarians but from different datasets].

The authors also noticed that, in spite of the appearance created by some mtDNA lineages and the historical evidence of slave trade, specially in the Canary Islands, no West African (YRI) influence could be detected:

Fig. S1


Different levels of persistence of Guanche blood

While it is clear that Castilian (Spanish) colonization of the Canary islands was very intense, there is still some very noticeable background of original Guanche blood (represented surely by the North African, green, component). There are important differences however among the various Canary Islands (table 2):
  • La Gomera retains by far the greatest apportion of Guanche blood, showing a 43% of North African affinity.
  • Fuerteventura, La Palma and El Hierro also retain important Guanche blood (20-22%).
  • Lanzarote (16%) and the larger islands of Tenerife (14%) and Gran Canaria (12%) seem the ones that have received the greatest Iberian input, however they still retain some aboriginal genetics as well.
The Canary Islands

The island of La Gomera is also the one to have retained best some of the ancestral Guanche customs, notably the whistling language known as Silbo Gomero.


See also:

21 comments:

  1. Very interesting!

    But I can't believe eastern Iberia has only 2% in common with North Africa, sorry. And what's the Iberian/European input in North Africans? I think it's a lot higher.

    It seems western Europeans are somewhat closer to NW Africans than the Polish, Italians and Greeks.

    What's even more interesting is they didn't find any West African DNA in Iberians, NW Africans nor Canarians. That contradicts the vast majority of studies I've seen, where NW africans show a noticeable % of sub-saharan blood.

    It could have been even more interesting if other populations, such as East Africans and the Southern French were included.

    ReplyDelete
  2. "I can't believe eastern Iberia has only 2% in common with North Africa"

    It has 2% of (clear) genetic input from North Africa, always according to this study. The figure is almost in perfect agreement with the Y-DNA marker E1b1b1b1-M81 (former E3b2) in Adams 2008 (discussed here).

    That is also the case for West Iberian 9%.

    I imagine that your judgment is intutitive based on phenotypes. However when I have discussed with you such phenotypes you always show very europid faces, not people who look typically North African (i.e. like Hassan II maybe).

    "It seems western Europeans are somewhat closer to NW Africans than the Polish, Italians and Greeks".

    I do not see that in the data. The horizontal and vertical axes in fig. 1 are not in the same scale. I estimate that Italians are c. 2 points from the NW Africans, while the English are at some four points and even the Spanish are at 3 points (you can do it yourself using Pythagoras' theorem). The vertical axis is almost trivial, measuring only 1.2 points altogether (while the horizontal axis has 5 points).

    Maybe in this case, you visualize better in fig. 2 where both axes are more to scale in relation to each other.

    "What's even more interesting is they didn't find any West African DNA in Iberians, NW Africans nor Canarians. That contradicts the vast majority of studies I've seen, where NW africans show a noticeable % of sub-saharan blood".

    Not really. As soon as you allow the North African-specific cluster(s) to show up: it takes over very clearly and minimizes other components to virtually zero. You can see it very clearly in Behar 2010: supp. fig. 4a (at K=10) among Mozabites (Moroccans still seem to be asking for their own other specific component at that depth, what is normal considering the world-wide scale of that analysis).

    However the authors here caution that the use of the EuroAIMs markers over-emphasizes differences among populations: it reduces "noise" but maybe that "noise" is important too.

    "It could have been even more interesting if other populations, such as East Africans and the Southern French were included".

    I listed Athaniasadis 2010 (its review) in the "see also" section: it includes Tolousaines.

    Comparing with Egyptians, and other NE Africans would be interesting indeed, as they are supposed to be a major parent population of NW Africans. However in Behar 2010, Egyptians clustered mostly with West Asians and Ethiopians did mostly with West Africans and West Asians, rather than NW Africans (also but weaker connection). So maybe there's not so much to discover by that line either.

    ReplyDelete
  3. "It has 2% of (clear) genetic input from North Africa, always according to this study. The figure is almost in perfect agreement with the Y-DNA marker E1b1b1b1-M81 (former E3b2) in Adams 2008 (discussed here). "

    Maybe these markers are not capable to identify ancient gene flow (+15.000) only some recent migrations, but many north-africans have European mtDNA.

    "I imagine that your judgment is intutitive based on phenotypes. However when I have discussed with you such phenotypes you always show very europid faces, not people who look typically North African (i.e. like Hassan II maybe). "

    Yes, there are many NA who look completely European.
    With people who look typically
    north African, do you mean this?

    http://www.danheller.com/images/Africa/Morocco/Berbers/Slideshow/img16.html#img19

    He/she looks Asian to me. It's a very diverse region, and most N.A. seem to be heavily admixed.

    "I do not see that in the data. The horizontal and vertical axes in fig. 1 are not in the same scale. I estimate that Italians are c. 2 points from the NW Africans, while the English are at some four points and even the Spanish are at 3 points (you can do it yourself using Pythagoras' theorem). The vertical axis is almost trivial, measuring only 1.2 points altogether (while the horizontal axis has 5 points)."

    Surpisingly, Italians and Greeks are closer to NA than Spaniards, althought I read that the sample labelled as "Spaniards" included the French Basque!

    "Not really. As soon as you allow the North African-specific cluster(s) to show up: it takes over very clearly and minimizes other components to virtually zero. You can see it very clearly in Behar 2010: supp. fig. 4a (at K=10) among Mozabites (Moroccans still seem to be asking for their own other specific component at that depth, what is normal considering the world-wide scale of that analysis)."

    That's pretty confusing, because many NA show considerable negroid features as well as many sub-saharan mtDNAs (L1,L2,L3). supposedly due to the trade of sub-saharan slaves.

    It's really weird that some Canarians and Spaniards as well showed 0% and others 96% of NA contribution!

    "Comparing with Egyptians, and other NE Africans would be interesting indeed, as they are supposed to be a major parent population of NW Africans. However in Behar 2010, Egyptians clustered mostly with West Asians and Ethiopians did mostly with West Africans and West Asians, rather than NW Africans (also but weaker connection). So maybe there's not so much to discover by that line either."

    I'm wondering what could have happened if the authors sampled Central or East North Africans instead of NWA. Most of these studies are focused in finding connections between NW Africa and "Iberia" (with a clear cut around the Pyrenees), but in my opinion other Europeans such as Italians and Greeks have connections with NA as well.

    ReplyDelete
  4. Who are these non-Iberian Spaniards in the PCA chart?

    Or alternately, who are the non-Spaniard Iberians?

    If Iberian=Portugese, one would think that one would use that label. Does Iberian mean Portugese + Basque + Catalan? What about the Southern French Basque or otherwise?

    ReplyDelete
  5. " As soon as you allow the North African-specific cluster(s) to show up: it takes over very clearly and minimizes other components to virtually zero."

    Interesting point. In other words, the West African component of Canarians and Iberians is the same share of the non-European component as it is in North Africa. Thus, if Northwest Africans are 5% West African in origin, Canarians are 40% non-European, and Iberians are 10% non-European (to use round numbers to illustrate the math only) then Canarians are 3% West African and Iberians are 0.5% West African.

    This implies, but doesn't definitively show, that Canarians and Iberians receive almost all of their West African ancestry through the filter of North Africa, which makes sense given that recent immigrants to these places from Africa would probably be excluded from the regional samples on the ground that they are not indigeneous to the area. Indeed, my guess would be that recent West African DNA contributions in the 20th and 21st centuries probably exceed all prior direct West African DNA contributions (as opposed to DNA contributions from West Africa via Northwest African admixture with West Africa).

    ReplyDelete
  6. The W-E differences in Iberia are clearly a Neolithic or older influence from North Africa. IMO they are from the Solutrean period, so this method is detecting well this influence that has some 20-7 Ka of age (i.e. old).

    This influence is clearly not recent: it is not from the Muslim period (should be the opposite: more NAf influence to the East, or transversal: more NAf influence to the South) nor it is from the Phoenician period (similar patterns to expect, if any). It must be from an older stage.

    When you see residual West African influence in North Africa it is mostly noise from the algorithms rather than true admixture. Or worse: lack of sufficient resolution because the algorithm is not allowed to go as deep as needed.

    "With people who look typically north African, do you mean this?"

    Yeah. That's a good pseudo-Khoisanid phenotype. She's probably from the Marrakesh area, where that type is somewhat common.

    "He/she looks Asian to me".

    Appearances are often misleading. She looks transitional Khoisanid-Australid to me but that's probably because my mind just don't have a well defined slot for that typology yet.

    "It's a very diverse region"...

    It is an ill researched region. I'm not sure it is so diverse (it's not that big). We should expect some affinities to SW Europeans and NE Africans, yet to be also different themselves.

    ReplyDelete
  7. "Italians and Greeks are closer to NA than Spaniards"...

    Not sure. In fig. 2 both Greeks and Iberians contact (barely overlap if at all) with North Africans. Italians do not. This may indicate the strong (but old) North African influence in Greeks via Y-DNA E1b-V13.

    "... I read that the sample labelled as "Spaniards" included the French Basque!"

    No it says (text S1) that the initially estimates of such differences was found between the three populations using Northern Basques, Mozabites and YRI. But this corresponds to other papers (Price, Li and Lao).

    "... many NA show considerable negroid features"...

    Not so much Negroid as Capoid (your example) or, more commonly, Ethiopid. These may indicate (my best guess) a very old layer from the Aterian period and a more recent layer from the Capsian (Afroasiatic) period respectively.

    "It's really weird that some Canarians and Spaniards as well showed 0% and others 96% of NA contribution!"

    Indeed. As we cannot explore the individual cases, considering their particular origins, we are surely missing some very valuable info in these differences. Probably the high admixture Iberians and Canarians originate from communities where such apportions are common.

    "... other Europeans such as Italians and Greeks have connections with NA as well".

    Greeks and Albanians specially (permeating surely to Southern Italy too), indeed, as is evident in Y-DNA. This is some sort of Neolithic or Epipaleolithic migration that we have not yet fully understood (we see the genetics but we lack the archaeology). Theirs is probably original from Egypt, maybe via Palestine and Southern Turkey. It is also a founder effect and hence probably pre-Neolithic (or earliest Neolithic).

    ReplyDelete
  8. Andrew: the Iberian, Canarian and North African samples are from this paper (see Samples section), the Spanish and other European samples (except CEU) are from Price 2008.

    In spite of label the "Iberian" sample is all from the state of Spain, mostly from the Southern half (and no Basques were included).

    ReplyDelete
  9. @Andrew: that is not what I think. What I understand is that, after all, Bayesian components (such as those generated with Structure, Admixture, etc.) are just an indicator of affinity.

    At K=2 all populations are forced to be akin to A or B. Populations around 50% are probably not akin to either one (unless we know for a fact that they are an admixed descendant). Typically at K=3 and so on, new clusters replace the appearance of admixture. I understand that it is informative to find a good number of such clusters before claiming any real admixture.

    For example we know that some North Africans and some Iberians are (most likely) product of admixture because we have first identified the North African and Iberian clusters.

    Instead, at K=2 in Bauchet 2007 we get an appearance of admixture in all West Europe between a NE and SE components but this appearance rapidly vanishes as the three real West European (NW Euro, Iberian and Basque) clusters show up at K=3, K=4 and K=5. However no Italian-specific component was discovered in spite of Italians being rather well sampled, what suggests that Italians are largely a mixture of the other European components (though there may be Italian specifics hidden at lower levels, they are not clearly dominant, as the others are).

    Appearance of admixture in such analysis is not necessarily true admixture. We must be very very careful on that matter: it only says "best affinity when forced to fit K clusters" - nothing else. Depending on samples and other factors you may need a rather high K to find certain important components. (Typical examples are tiny Bushmen or Papuan samples not showing up as distinct because they are just to small to "compete").

    "This implies, but doesn't definitively show, that Canarians and Iberians receive almost all of their West African ancestry through the filter of North Africa"...

    Not really. It implies that often, when Iberians are compared with YRI, they may show some such "West African" component but that only indicates a hidden (and larger) NW African component. Here YRI acts as a (very poor) proxy for North Africans.

    ReplyDelete
  10. "This influence is clearly not recent: it is not from the Muslim period (should be the opposite: more NAf influence to the East, or transversal: more NAf influence to the South) nor it is from the Phoenician period (similar patterns to expect, if any). It must be from an older stage. "

    Unforntunately that's unknown for the moment, but I think some European admixture in NA might date to the times of Ibero-Maurisian (+20.000). Gene flow between the strait might have happened since remote times, though.

    "When you see residual West African influence in North Africa it is mostly noise from the algorithms rather than true admixture. Or worse: lack of sufficient resolution because the algorithm is not allowed to go as deep as needed. "

    Yes, in fact in some studies NWA seem to have their own cluster, and Iberians have an higher % than other Europeans.

    "Yeah. That's a good pseudo-Khoisanid phenotype. She's probably from the Marrakesh area, where that type is somewhat common."

    Althought you might not believe me, I still insist I've seen some Catalans who look much like her: slanted eyes, yellowish skin, flat and wide noses, very dark hair, high cheekbones, etc. I have recognized her Khoisanid features because I'm relatively familiar with them, but I admit these types are relatively uncommon here. Once you told me they can be seen sometimes in Andalusia. Giving the fact we're relatively close to North Africa, I'm not suprised at all.

    "Appearances are often misleading. She looks transitional Khoisanid-Australid to me but that's probably because my mind just don't have a well defined slot for that typology yet. "

    That's quite weird, because to my knowledge no Khoisanid, Australoid nor Asian DNA has been found among these Moroccans, but that doesn't mean they don't have any, because North Africa has been quite poorly sampled. I hope this will change someday.

    ReplyDelete
  11. "It is an ill researched region. I'm not sure it is so diverse (it's not that big). We should expect some affinities to SW Europeans and NE Africans, yet to be also different themselves."

    Most north Africans I see every day at the street don't look like that woman at all. I think many of them are caucasoids and very close to south-western Europeans, althought others show negroid features as well.

    "No it says (text S1) that the initially estimates of such differences was found between the three populations using Northern Basques, Mozabites and YRI. But this corresponds to other papers (Price, Li and Lao). "

    Now I see, thanks for the correction :)

    "Not so much Negroid as Capoid (your example) or, more commonly, Ethiopid. These may indicate (my best guess) a very old layer from the Aterian period and a more recent layer from the Capsian (Afroasiatic) period respectively."

    I've seen some Aterian skulls (or associated with this industry) dated 25.000-60.000 BP (Dar es Soltan, Temara), and to me they don't look like present day North Africans. Well, in fact, they don't even look Homo sapiens!! I like to think the dates are wrong, and these skulls are more than 100.000 years old. Do you know if any other skulls associated with Aterian have been found? I read somewhere some Aterian skulls are very similar to these of Ibero-Maurisians, but I'm not sure.

    "Theirs is probably original from Egypt, maybe via Palestine and Southern Turkey. It is also a founder effect and hence probably pre-Neolithic (or earliest Neolithic)."

    I've seen some blue-eyed and big-nosed North-east africans that remind me a lot of many people living in southern Europe. Maybe they arrived during the Epipaeolithic or Neolithic? On the other hand, Southern Italians, Sicilians and Sardinians may have north African input as well, but it's impossible to date this gene flow, it could be very ancient,at least in the case of Sardinians.

    ReplyDelete
  12. "If Iberian=Portugese, one would think that one would use that label. Does Iberian mean Portugese + Basque + Catalan? What about the Southern French Basque or otherwise?"

    In Spain, you learn at school that the Iberians were an ancient people who lived in the east of the Iberian peninsula and some areas of south-eastern France until the arrival of the Romans. This culture started around 500 BC and lasted with the Roman empire.

    The other half of the peninsula was inhabited, not by Iberians, but CELTS, like those of Ireland and Scotland. There were many other tribes/ethicies there before the arrival of the Romans.
    Today, many people living in the Iberian peninsula don't even know who/what are the Iberians. It's a word invented by Americans to describe the peoples living down the Pyrenees.
    These peoples are not the same, and many, like Catalans, Occitans and Basques live in both sides of the Pyrenees and show noticeable differences in their cultures and languages.
    Then we have the official contries like Portugal and Spain, but what I mean is that no one is Iberian here. I mean, this designation is quite incorrect, because the Iberians became extinct 2000 years ago, so, why shouldn't we call it better "Hispania" "Al-Andalus" or "Visigothia" or "Frankia"?

    ReplyDelete
  13. "Althought you might not believe me, I still insist I've seen some Catalans who look much like her"...

    It's not that I do not believe you but that the pictures you have provided as references were not at all that type but pretty much normal Europeans.

    "That's quite weird, because to my knowledge no Khoisanid, Australoid nor Asian DNA has been found among these Moroccans"...

    But it makes perfect sense if you think these types (this range of variation rather) as common in the time of the OoA, maybe even dominant then. I have seen Pygmies who have Papuan faces, Wallaceans (East Indonesians) who look pretty much like that Moroccan lady, Indians who look like Australians (in spite of not sharing any genetics since the Eurasian scatter), etc.

    It's reasonable to think that some types that look both archaic and more or less shared across the globe like these are relics from our shared ancestry and probably approach better how (most) humans were at the OoA time-frame than modern regional variants ("races").

    Have you ever tried to blend a Nigerian, European and Chinese face? The mix might well look like that - not sure because I have not tried but it's a good guess. But that does not mean that this real face is a blend but a distinct shared relative instead, another less common, less regionally specific, and hence less archetypal branch.

    ReplyDelete
  14. "I like to think the dates are wrong, and these skulls are more than 100.000 years old".

    That's more or less the age for Aterian considered today (though there seems to be much uncertainty).

    "Do you know if any other skulls associated with Aterian have been found?"

    I'm not aware at the moment.

    "... they don't even look Homo sapiens!!"

    To me and most paleoanthropologists they do.

    "I read somewhere some Aterian skulls are very similar to these of Ibero-Maurisians, but I'm not sure".

    Skulls from the Iberomaurusian (Oranian) period are typically related to the European Cro-Magnon type.

    "Southern Italians, Sicilians and Sardinians may have north African input as well"...

    Why do you say that? How would that be different from Greek admixture, which is probably high in South Italy?

    Whatever the case, I do not think that there is any indicator of direct Italy-North Africa gene flow. All (and is not much in any case) seems mediated by either SW Europe (Iberia) or SE Europe (Greece and Albania).

    You are anyhow making claims without providing any evidence, I'm not sure how seriously I must take them. Apparent affinity in a bidimensional plot is not the same as admixture, it may just indicate an ancient common origin in West Asia, for instance.

    "Iberians were an ancient people"...

    Pointless. In this paper Iberian is used in the same sense as the Iberian Anarchist Federation: i.e. Spain by another name. See the samples section.

    ReplyDelete
  15. "It's not that I do not believe you but that the pictures you have provided as references were not at all that type but pretty much normal Europeans."

    I recognize that these Catalan actresses I put vaguely resemble the Moroccan woman, except perhaps for the skin tone, but unfortunately I can't take photos from these people without their permission, and it would take a week to explain why and convince them :( but they remind me a lot Hassan II and that woman.

    "But it makes perfect sense if you think these types (this range of variation rather) as common in the time of the OoA, maybe even dominant then. I have seen Pygmies who have Papuan faces, Wallaceans (East Indonesians) who look pretty much like that Moroccan lady, Indians who look like Australians (in spite of not sharing any genetics since the Eurasian scatter), etc."

    That's really interesting I think, althought no one has found a link between NW Africa and OoA, for the moment.

    "Have you ever tried to blend a Nigerian, European and Chinese face? The mix might well look like that - not sure because I have not tried but it's a good guess. But that does not mean that this real face is a blend but a distinct shared relative instead, another less common, less regionally specific, and hence less archetypal branch."

    LOL no, but I understand what you mean.
    Other weird-looking NA:

    http://s1.zetaboards.com/anthroscape/topic/3780844/1/

    "To me and most paleoanthropologists they do."

    Apparently not to E. Trinkaus. See here, they have obvious archaic features:

    http://s1.zetaboards.com/anthroscape/topic/2629102/1/

    In any case, I don't see any connection between these skulls and the pseudo-Khoisanid NAs, but that's my opinion.

    "Why do you say that? How would that be different from Greek admixture, which is probably high in South Italy?"

    http://mathildasanthropologyblog.wordpress.com/2008/10/12/sicilian-y-chromsomes-show-a-link-to-north-africa/

    I'm not an expert, but there was a Moorish occupation of the island.

    As for Sardinia, they have a very high % of an old North-African haplotype, and no one knows how and when it arrived there.

    http://en.wikipedia.org/wiki/A30-Cw5-B18-DR3-DQ2_%28HLA_Haplotype%29

    ReplyDelete
  16. I have discussed at Leherensuge (and linked in the "see also" section of this article, which you may find interesting to browse) that some L(xM,N) lineages appear to be not "Sub-Saharan" but Arabian (L0a1c, L0f2, L0k2, L6, L4a, L3i) or North African (L2a1j, L3b1, L3d1c, L3k).

    These lineages appear to have an age (my estimates) close to that of M and N and hence would have coalesced about the time of the OoA, which by now is pretty clear that happened at least c. 90 Ka ago (roughly the age of Aterian).

    "In any case, I don't see any connection between these skulls and the pseudo-Khoisanid NAs, but that's my opinion".

    I'm not saying that these skulls should render that phenotype (faithfully reconstructing the soft parts from a skull is almost impossible) but that these archaic H. sapiens (they are) are probably in part ancestors of the modern H. sapiens who display these features.

    ...

    Regarding Mathilda's article (link to source is now broken), it's just a haplotype (no haplogroup was identified) and hence surely a false alarm by some obsessive Nordicist (my guess but typical). I do not think it's relevant.

    "... there was a Moorish occupation of the island".

    But for only 100 years. It was surely a very shallow military occupation without any long term effects. Spain occupied Fernando Poo (Equatorial Guinea) for centuries and nobody expects to find any Spanish lineages in the island right now. There may be one or two... sure.

    Similarly we do not find Spanish lineages in the Philippines either and it was an occupation of several centuries too. It is in fact easier to find Filipino lineages in Mexico as result.

    The HLA haplotype is also very much unreliable. Some people just see things.

    Of course, half of Sicily and all Sardinia were once part of the Carthaginian empire (and, as you mention, were later occupied by the Muslim forces briefly - and they were part of the Roman Empire and later of the Vandal state of North Africa as well, etc.). So it's not impossible and I would be very much surprised if you cannot find any connection at all looking with magnifying glasses as you do... but I consider that pretty much trivial, minor, pointless, nitty-picky...

    No offense meant, just that the connection is so fuzzy, so weak, that it doesn't really matter.

    ReplyDelete
  17. "I have discussed at Leherensuge (and linked in the "see also" section of this article, which you may find interesting to browse) that some L(xM,N) lineages appear to be not "Sub-Saharan" but Arabian (L0a1c, L0f2, L0k2, L6, L4a, L3i) or North African (L2a1j, L3b1, L3d1c, L3k). "

    Very interesting indeed, but these data suggests a southern route for the OoA via Ethiopia --> Arabia, that's what has been proposed recently. Maybe Skhul/Qafzeh dataed +90.000 carried these mtDNA haplogroups?

    "I'm not saying that these skulls should render that phenotype (faithfully reconstructing the soft parts from a skull is almost impossible) but that these archaic H. sapiens (they are) are probably in part ancestors of the modern H. sapiens who display these features."

    Unfortunately, we just don't know. Some say it's possible these Aterians got extinct with the extremely arid conditions of the area by 60 Ka, while others think they migrated to south. An interesting paper from 2006 found 5% of archaic admixture in West Africans, maybe it has something to do these Aterians, althought the continuity we see with Aterian in the region, since 100.000 until 30.000 might suggest some sort of cultural and genetic continuity as well.
    We just don't know what happened to these people and to what degree they're related with modern humans. DNA tests on these ancient fossils could clarify it.

    "No offense meant, just that the connection is so fuzzy, so weak, that it doesn't really matter."

    As I said above, I'm not an expert. I've seen lots of studies who contradict each other with this issue, some claiming there's no contribution from north Africa, and others it was very important (20%) at least in Spaniards.

    mtDNA/Y-chromosome can't tell us what was the level of gene flow, because they just account for a tiny fraction of our genomes.
    The 1000 genomes project could clarify better what are the relationships between Europe and North Africa, and the degree of admixture.

    I have a question: how did you know these Moroccan person I put is a woman and not a man? For me it's very difficult to tell...

    ReplyDelete
  18. I did not meant the OoA as such but the processes that include the OoA. That time, that expansiveness... not necessarily the same exact migration.

    Actually I do not find anything that can be directly attributed to the Palestinian sites. The lineages appear to be concentrated in Arabia Peninsula and North Africa. So it fits best with the coastal migration model (with whatever technical caveats you or Petraglia wish).

    What these lineages suggest to me is that at the time of the OoA there were peoples migrating across the Sahara (Nile or the other river that existed for several millennia and ended up in Libya) and also across the Red Sea into Arabia (and beyond).

    As for Aterians, I am almost certain that they subsisted... because their characteristic back-tipped points do even when other peoples arrive from Egypt or Iberia. However their genetic share was diluted more and more (but it is still there).

    "mtDNA/Y-chromosome can't tell us what was the level of gene flow, because they just account for a tiny fraction of our genomes".

    MtDNA is quite faithful. You are right in principle but more when applied to individuals. When we consider populations, haploid markers tell us a lot, notably the mtDNA.

    ReplyDelete
  19. "I have a question: how did you know these Moroccan person I put is a woman and not a man? For me it's very difficult to tell..."

    In truth I do not know. My first impression was of a woman but maybe not.

    ReplyDelete
  20. "This influence is clearly not recent: it is not from the Muslim period (should be the opposite: more NAf influence to the East, or transversal: more NAf influence to the South) nor it is from the Phoenician period (similar patterns to expect, if any). It must be from an older stage"

    ... the could it be form the first Iberians who according to some historians originally came from North Africa?

    ReplyDelete
  21. Uh? Which are those "some historians"?

    It is a most unlikely origin because Iberian (nor Tartessian for all that is known) is not related to Berber.

    Also the genetic influence is greater and more homogeneous from Iberia into NW Africa than vice versa, and even that last one is oddly concentrated in the Western half of the peninsula, unrelated to any known historical episode (nor "Iberians")

    ReplyDelete

Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (sorry, too many trolls).