August 04, 2009

Ancient Y chromosomes from the Canary Islands

On the aboriginal remains (ABO):
Aboriginal remains were clearly pre-conquest for all the analyzed islands: Tenerife (2210 ± 60 to 1720 ± 60 BP), Gomera (1743 ± 40 to 1493 ± 40 BP), Hierro (1740 ± 50 to 970 ± 50 BP) and Gran Canaria (1410 ± 60 to 750 ± 60 BP) [33].
It is clear that the aboriginal population was dominated by haplogroups E-M81, E-M78, J-M267. In the historical period (a few centuries ago) new haplogroups make their appearance (e.g., R1a) and a massive increase in the frequency of R1b is observed.

UPDATE (Aug 5):

There are several interesting observations one could make based on these results:
  1. The idea of European-descended fair-haired Guanches has taken a hit, as the aboriginal population looks largely like North African Berbers in terms of their Y-chromosomes. No real need to invoke mythical "Nordic" tribes as some have attempted to do.
  2. The common view about the dispersal of J-haplogroup in the West has been of early Neolithic dispersal of J2 agriculturalists, followed by J1 dispersal of Arabs, Jews, etc. This paper pretty much destroys that picture.
  3. The two most conspicuous "missing" haplogroups in the clearly pre-Indo-European population of the Canary Islands are J2 and R1a,.

BMC Evolutionary Biology 2009, 9:181 doi:10.1186/1471-2148-9-181

Demographic history of Canary Islands male gene-pool: replacement of native lineages by European

Rosa Fregel et al.

Abstract (provisional)

Background

The origin and prevalence of the prehispanic settlers of the Canary Islands has attracted great multidisciplinary interest. However, direct ancient DNA genetic studies on indigenous and historical 17th-18th century remains, using mitochondrial DNA as a female marker, have only recently been possible. In the present work, the analysis of Y-chromosome polymorphisms in the same samples, has shed light on the way the European colonization affected male and female Canary Island indigenous genetic pools, from the conquest to present-day times.

Results

Autochthonous (E-M81) and prominent (E-M78 and J-M267) Berber Y-chromosome lineages were detected in the indigenous remains, confirming a North West African origin for their ancestors which confirms previous mitochondrial DNA results. However, in contrast with their female lineages, which have survived in the present-day population since the conquest with only a moderate decline, the male indigenous lineages have dropped constantly being substituted by European lineages. Male and female sub-Saharan African genetic inputs were also detected in the Canary population, but their frequencies were higher during the 17th-18th centuries than today.

Conclusions

The European colonization of the Canary Islands introduced a strong sex-biased change in the indigenous population in such a way that indigenous female lineages survived in the extant population in a significantly higher proportion than their male counterparts.

Link (provisional pdf)

36 comments:

Maju said...

Thanks for posting this, Dienekes. It's really fascinating.

As you say well: "it is clear that the aboriginal population was dominated by haplogroups E-M81, E-M78, J-M267". The latter (16.7%) is important because it really demonstrates that North African J1 (M267) is mostly pre-Arabic.

Anyhow, I'm surprised to find that that R1b1b2 was as high as 10% (I'm pretty sure is not that high in North Africa) and, if we add the P(xR1a,R1b1b2), which is probably some other R1b*, it's as high as 13.3%, almost as much as J1. In any case, pre-colonial R1b was higher than I would have expected - though I'm unsure of what this may mean.

I is also found at 6.7%, probably higher than in modern North Africa as well. Whatever the case, it means that these two typically European haplogroups were present in the islands (and therefore also in North Africa) before the historical period with all likelihood.

Maju said...

Correcting myself: if we add the P(xR1a,R1b1b2), which is probably some other R1b*, it's as high as 13.3%.

This is wrong: fig.2 clearly says P(xR1) so it must be something else (what? no idea).

Gioiello said...

Probably is P, and also this is meaningful. These ancient haplogroups are preserved in the Islands (like Sardinia in Italy)and demonstrate, I think, an ancient diffusion overall of these haplogroups, which many are thinking present only in Central Asia (see P, ancestor of R): it would be interesting to test them for new SNPs, but it won't be easy to have them for testing. Anyway they haven't generate subclades (at leat known), but they can have generated them in Italy or in Spain.

Maju said...

There's some R2 in Eastern Europe and West Asia and also some Q. P(xR,Q) has been detected at minimal ammounts in India too but obviously is not underived but just something else. It could also be R*, which I think exists at minimal levels in some places.

But, whatever the case, it brings me to that of K*. It could be T but I am thinking in Cabo Verdean genetics (Gonçalves 2003) where rather high amounts of (possibly North African) E1b1b1 and J1 were found, along with almost 10% of K(xT,R). AFAIK, K(xR) has never been tested for the defining SNPs of T in may places including North Africa. So I'm wondering if there is more than just R and T within West Eurasian/North African K. The possibility of Canarian founder effects in Cape Verde should not be overlooked, as the Canary Islands briefly belonged to Portugal in the conquest period.

Gioiello said...

Anyway the most ancient haplotypes of an haplogroup tend to disappear (bad for me, who are R1b1b2a) as time passes and survive only the most recent ones. Then every speculation on the origin and on the time of an haoplogroup (see Vizachero) is "blowing in the wind". The Hg. P is found on a dead man of perhaps 2 thousand years ago and probably there are no P there now. We all are asking for an aDNA exam also to resolve the problem of the European Refugia. I bet yet on Italy for R1b1b2/L23- and +.

Maju said...

All haplogroups evolve. If there's some P(xQ,R) it will not be for sure the same as P was when it got detached from NOP and K. It will be probably something like P1 or P2, so to say, just that at the moment we do not know enough and therefore we can just call it P*.

There is some P(xQ,R) among certain group of Indian Brahmins (at ridiculously low levels) and I guess some of the other P* also found in the area may be that too. But that doesn't mean it is P in a way R or Q sublineages are not. R and Q are just the most common, by large, P variants.

Gioiello said...

But there was one P* who generated Q and R. The SNPs are a work in progress, as from a R1b1 came R1b1a and R1b1b, and from an R1b1b came R1b1b1 and R1b1b2 etc. R1b1b2/L23+/L150+ (like me) arose from one R1b1b2/L23+/L150- and the unique we know so far has been found in Italy. R1b1b2/L23+ arose not from one R1b1b2/M269, but from one R1b1b2/M269/S3/S10/etc. Then there was in that time a bottleneck, etc.

Maju said...

You are in a misunderstanding, Gioello. Pure P (as it was in the depths of history, the ancestral P or Q and R and whatever other P) does not exist anymore for sure. Any P* extant now (including R and Q sublineages) must have dozens of derived SNPs, just like R and Q do. That we don't know (or even do not research) them does not mean that they are not there in the real chromosomes.

Also the asterisk mark only means always "other". The exact meaning may vary contextually, depending on what downstream SNPs have been tested. In this case it is P(xR1), in the Indian case P(xQ,R) but even in that case it is not "ancestral P", just some other derived lineage(s) of which we know nothing else.

argiedude said...

Maju said:
"Thanks for posting this, Dienekes. It's really fascinating."

It's priceless. Good to see you again, Maju, by the way.

Regarding P(xR1)...

I didn't look at North Africa, though I should have, but here's the data for y-dna Q in Iberia:

Adams, 2008..........found 2 Q out of 1000+ Spanish samples (Andalucía and Basque Country)
Beleza, 2005..........found 0 P(xR1) out of 600 Portuguese samples
Flores, 2004...........found 3 P(xR1) out of 500 Iberian samples (all north Portugal)
Goncalves, 2005.....found 1 Q out of 550 Portuguese samples (south Portugal)

Thus, rate of P(xR1) in Spain is 0,1%, and in Portugal it's 0,3%. Note the 2 Spanish samples were negative for Q3, so they aren't due to American Indian ancestry, so Q may have existed in the general region for quite some while back. How likely or possible is it that the lineage is actually R1b1b2 but its M269 marker has back-mutated (degradation after 1,000+ years)?

PS: I'm really glad you're posting here, Maju. I received your mass-PM before you left that place.

PS 2: These people desperately have to retest those samples, for downstream M269 SNPs and for haplotype info (especially the M81 samples).

Ebizur said...

argiedude said,

"Note the 2 Spanish samples were negative for Q3, so they aren't due to American Indian ancestry, so Q may have existed in the general region for quite some while back."

Not all American Indian Q individuals belong to the subclade Q1a3a-M3. Many of them belong to Q-M242(xQ1a3a-M3), which a recently published study has suggested should be Q1a3-M346(xQ1a3a-M3). However, the proportion of Q(xQ1a3a) to Q1a3a Y-DNA is much higher among North American natives than it is among South American natives, and any Native American influence in Iberia is likely to have originated from Latin America, where most of the indigenous males belong to Q1a3a-M3.

Maju said...

Ebizur: Latin America is not just South America. In fact the Spanish connection to the New World was specially intense with the Caribbean and Mexico. South America yes but North America as well Plus, while Caribbean peoples were probably original from the area of Venezuela, Aztecs and all the other Mesoamerican ruler ethnicities were immigrants from what is now the USA. The Spanish had some meaningful presence in SW USA and Florida, as well, in the case of Basques specially, in Newfoundland and eastern Canada.

So for me, if it can be American native, then it probably is. More intriguing are the two North African cases but we cannot discard an Iberian connection (Oran, Djerba, Tangiers, Ceuta, Melilla...), or maybe more likely a Turkish one (where Q2 has some presence). The most famous Moroccan conqueror of the 16th century, Joder Pasha, was a Spanish renegade, whose trans-Saharan expedition against Songhay was largely made up of Spaniards as well.

Maju said...

@argiedude:

I'm really glad you're posting here, Maju. I received your mass-PM before you left that place.

Seems I can't PM anymore. I have written at my blog on all that sick issue anyhow.

As for Q among Iberians, IMO it's Native American, same as Q among Norwegians or English.

What I said above to Ebizur is relevant for you as well.

Maju said...

Correcting myself:

More intriguing are the two North African cases...

Ok, there are no such cases, sorry. Not sure what I was thinking. :/

Ebizur said...

Maju said,

"Latin America is not just South America."

That doesn't affect the truth value of my statement: the proportion of Q-M242(xQ1a3a-M3) to Q1a3a-M3 in Latin America as a whole is very low, and any Native American influence in Iberia would likely have been derived from indigenous populations of Latin America. Therefore, to find Q(xQ1a3a) Y-DNA in Iberia without finding any Q1a3a-M3 Y-DNA in Iberia would suggest that the source of the haplogroup Q Y-DNA in Iberia is probably not Native American, at least in large part.

Maju said...

Canada, Florida, New Mexico...

Not sure. It could be from somewhere else but only two Iberian erratics at coastal areas...

Maju said...

How likely or possible is it that the lineage is actually R1b1b2 but its M269 marker has back-mutated (degradation after 1,000+ years)?-

Impossible. Some SNPs like P25 are known to have a back-mutation rate of 1-2%. But we would not be talking of one SNP but many dozens!

terryt said...

"Tenerife (2210 ± 60 to 1720 ± 60 BP), Gomera (1743 ± 40 to 1493 ± 40 BP), Hierro (1740 ± 50 to 970 ± 50 BP) and Gran Canaria (1410 ± 60 to 750 ± 60 BP)."

Presumably these dates do not represent the time of arrival of the first settlers. Does anyone know when the Canaries were first settled?

Maju said...

I just told you at my blog: c. 1000 BCE.

Gioiello said...

Maju writes: ”You are in a misunderstanding, Gioiello. Pure P (as it was in the depths of history, the ancestral P or Q and R and whatever other P) does not exist anymore for sure. Any P* extant now (including R and Q sublineages) must have dozens of derived SNPs, just like R and Q do. That we don't know (or even do not research) them does not mean that they are not there in the real chromosomes”.
Of course, but what I was saying was that when, for instance, L150+ arose in a man like Romitti (R1b1b2/L23+/L150-), there were many L150- and only one L150+. To-day there are hundreds of millions L150+ and his descendants and only 1 (known) L150-. If there weren’t Romitti, we wouldn’t have this haplotype, though we know it existed anyway, as there were R/M269 but not S3 or S10 etc. that now we haven’t ever found. In conclusion we have a few lines survived and we can only hypothesize other lines extinct, but when we find these lines, they are very interesting to understand what has happened and probably where those haplotypes lived. Among the survived P* elsewhere in the world there is a location where from one P was born R. Perhaps in that time all the P* were the same, but the ancestor of R was that P* with all the SNPs of R except M269. From one P* was born R, and that P* cannot having had descendants, but if we find P* not only in Central Asia but in Europe too, we can think that R was born here and not in Central Asia, but it is also possible that this European P* had come from Asia with its descendants. But all this is matter of study.
Coming t hg. R, what I was saying is that where we find R/L23+ but L150-, probably there lived the ancestor of all the millions of to-day L150+. Is it clear? Have I expressed my idea in an understandable way?

Maju said...

But they have other SNPs that are not named, not known, not tested. To know for sure the whole detail you'd need to sample every single man and sequence the whole Y chromosome and then analyze all that data...

It's a practical impossibility.

Btw, there's no L150, you must mean L51.

L51- people have, must have, other SNPs instead. Just that we don't know which ones: it's an unresearched area. Maybe 99% of L23+, L51- belong to some single unnamed haplogroup or maybe they belong to a dozen or hundred different small lineages.

Maybe some day someone would bother studying that and we'll all be happier. By the moment the only thing we know is that L23 includes L51, or in the other case that P includes R and Q. But the larger set has a remnant of others that cannot fit in the named subsets. Those are signified with the asterisk. They can be a single subset or mostly so, or they can be many different subsets.

What is clear is that they are not the ancestor, at best a signal of greater diversity (but unconfirmed).

Gioiello said...

If you go on the Adriano's spreadsheet, 682/ rs9785831 C/T
position 10618791, you can see that L150 exists, and Romitti is so far the unique L150-. I and all the others tested from R1b1b2/L23+ are derived.

Ponto said...

Interesting paper. It does not speculate on the appearance of the aborigines. The Blond hair may be fanciful but it it observed everywhere where Caucasoids inhabit including North Africa and the Near East. So some blond headed Guanches are not unexpected, they were Caucasoids, just the blonds were a small percentage of the population.

Some of the dated remains of the aborigines are close in date to the arrival of putative Arabians from Yemen, high J1 zone, in North Africa. It would be easier to state the lack of historically recent J1 Arabian input into North Africa, and hence the Canary Islands had the remains been much older.

I have always maintained that J1 in Europe and North Africa predate any Jewish or muslim Arabians "immigrating" to those places whether in Roman times or after Muhammad's death. Saying that, the results of the study do not disprove that those immigrations of Near Easterners did not affect the populations in North Africa which contributed to the population of the Canary Islands before the European rediscovery of the islands and subsequent colonisation.

With the common European haplogroups in the I and R groups found in the aborigines, it is likely that cross Mediterranean movements of people contributed to both sides of the Mediterranean Basin and later to the Canaries. I and R have made a western movement from the East, in the case of I from Eastern Turkey and R from Central Asia, so it is not unlikely that some men carrying those haplogroups landed in the Canaries among the Aborigines. The movement of I and R men into Europe proper occured many thousands of years back, older than the settlement of the Canaries.

It is interesting that the Moroccans are lower in J1 than the aborigines of the Canary islands yet they came from there. North Africans further east, Tunisians and Algerians are higher in J1. Higher European admixture in Morocco could be the answer.

argiedude said...

The 652 contemporary Canary samples are from Flores, 2003:

http://www.ncbi.nlm.nih.gov/pubmed/12675689

That study also includes the haplotype data (unfortunately only 5 STRs). It would be very interesting to see the haplotype data for the M81 samples, because I have a later Canary y-dna haplotype study, from Zurita 2004, and there's an unusually high incidence of M81 samples with 392=12, an extremely rare value amongst the M81 of the rest of the world... with one exception, Western Sahara, which is of course, right next to the Canary Islands. Western Sahara M81 (Bosch, 2001), has an even higher incidence of 392=12 (plus other features which clearly mark it as a distinct regional cluster of M81). So it would be great to observe even more M81 haplotype data for the Canary Islands, and confirm if there's yet again an unusual incidence of M81 with 392=12.

argiedude said...

In the Zurita study there's 1 sample that's a possible Q. Interestingly, of the hundreds of ysearch Q samples, the top 4 matches to this haplotype are from India, Pakistan, India, and Iran. Only about 3% of the ysearch Q samples are from those 3 countries (combined). I really should look at the North African samples, once and for all.

Maju said...

Gioello: maybe I'm wrong but it's not listed in http://www.isogg.org/tree/ISOGG_YDNA_SNP_Index09.html . That's why I thought you meant L51.

There's M150 that defines B2a and P150 that defines E. :?

Ponto:

I have always maintained that J1 in Europe and North Africa predate any Jewish or muslim Arabians "immigrating" to those places whether in Roman times or after Muhammad's death. Saying that, the results of the study do not disprove that those immigrations of Near Easterners did not affect the populations in North Africa which contributed to the population of the Canary Islands before the European rediscovery of the islands and subsequent colonisation.

I think it does. Guanches were not Muslim nor spoke Arabic. In fact they were stuck in a technological stage prior to metallurgy. J1 is not just some erratic among them, but the third most important lineage, just as in the mainland.

It is interesting that the Moroccans are lower in J1 than the aborigines of the Canary islands yet they came from there. North Africans further east, Tunisians and Algerians are higher in J1. Higher European admixture in Morocco could be the answer.

No: because Europeans are lower than Moroccans in J1. I understand that the levels of J1 found among Guanches are reasonaly comparable to those of Morocco and West Sahara.

More intriguing are R1b1b2, K(xP) and I, which could be a founder effect but seem to demand otherwise a European source (or some other maybe in the case of K).

Gioiello said...

Maju writes:
"Gioiello: maybe I'm wrong but it's not listed in http://www.isogg.org/tree/ISOGG_YDNA_SNP_Index09.html . That's why I thought you meant L51.

There's M150 that defines B2a and P150 that defines E. :?"

You can find L150 also among the new SNPs tested by FTDNA.

terryt said...

"You are in a misunderstanding, Gioello".

Not really. Sure, Q and R are simply two versions of P* (amoung many). But we know that all these P*s descend from just one single ancestor, who must have lived in a particular place. His descendants have subsequently become widely spread around the world, especially through the Northern Hemisphere. And further back P shares a single ancestor with NO.

To expand on this idea: most of these individual P* haplogroups are nowhere near as numerous, or as widely spread, as are just two of them, Q and R. That's precisely why these two haplogroups have been assigned letters. So it does seem that, as the population containing P's descendants expanded, different versions of the Y-chromosome became concentrated in different regions within the expansion. But ultimately the expansion become confined to just Q and R: basically Q to the east of the Aral Sea and R to the west and south. Several versions of P* may have accompanied them some of the way, but they seem to have eventually been left behind.
However we struggle to find any Y-chromosome haplogroups other than P involved in the relevant expansion. Perhaps T and, less likely, IJ and G. But it will be less complicated if we ignore these, at least for now.

Anyway R expanded around much of Western Eurasia and even into Africa. As you are well aware R has in turn periodically produced other haplogroup expansions derived from the single original R.

Interestingly Q is not very common south of the Central Asian mountains, such as the Altai and the complex of mountains south of Lake Baikal. So Q may have reached America through a route north of this region. Makes sense. Perhaps they were hindered in any movement south of this region by the presence of Y-hap C, and P's relations N and O. Subsequent expansions north of these Y-haps have reduced the proportion of Y-hap Q in populations along the route from Central Asia to America.

Maju said...

However we struggle to find any Y-chromosome haplogroups other than P involved in the relevant expansion.

P's "brother" NO for instance.

At the same phylogenetic level, though very different geography and origins, is also the very successful E1b1b. Less impressive maybe but still quite productive are I1, I2, J1 and J2, which are also at the very same phylogenetic level (4 nodes derived from F)

Anyway R expanded around much of Western Eurasia and even into Africa.

I'd say that's mostly limited to R1b, R1a1 too to some extent but probably with a more recent chronology. In any case you'd should talk of R1 at most.

So Q may have reached America through a route north of this region.

That's pretty obvious, isn't it?

What's your point? The same that you find very little P*, you find very little NO* and very little E1b1b*. And even I2* or J2* are rare too, while I* is totally unheard of (like NOP*). These nodes belong to a very old timeframe when drift was still most active, probably the early UP.

terryt said...

"At the same phylogenetic level, though very different geography and origins".

That was my point. The various migrations you mention are not really connected at all. Haplogroup expansions have been pretty much continuous. There has never been a moment when a whole bunch of haplogroups expanded from a single region.

Aaron said...

"The idea of European-descended fair-haired Guanches has taken a hit, as the aboriginal population looks largely like North African Berbers in terms of their Y-chromosomes. No real need to invoke mythical "Nordic" tribes as some have attempted to do."

--It's not a matter of being "Nordic", rather quite the opposite. The mutations that define the recessive traits were possibly, or likely spread throughout most of Europe in paleolithic, and exist today despite the onset of the neolithic. Keep in mind the "Nordic" regions were the last and least affected by the neolithic and its gene flow, thus a higher rate of blondness should be expected. There are many semi-isolate communities (ie:mountainous regions) throughout Europe including southern regions that have higher incidences of recessive traits.

Maju said...

The mutations that define the recessive traits were possibly, or likely spread throughout most of Europe in paleolithic...

I agree with this but (in most regions and notably in the central FC region) at minority levels. The traits are not necesarily recessive anyhow: that's an oversimplification: the genetics involved are not simple Medelian ones but involve many different genes, each one (or rather each allele within each of the genes) adding or substracting likelihood for this and/or that trait.

Keep in mind the "Nordic" regions were the last and least affected by the neolithic and its gene flow...

That's simply not true. Denmark for instance has a much older Neolithic than the Basque Country, maybe by 1000 years.

It is also quite clear that the Neolithic migrational flow hit much harder Central Europe than the West Mediterranean, and it is also clear that Chalcolithic flows from Eastern Europe hit Northern Europe (excluding British Islands), much harder and earlier than anywhere else except maybe the Eastern Balcans.

While I do not discard that Scandinavia may keep some Paleolithic fossil genetics, I don't think there is any reason to consider that area particularly conservative, much less in comparison with the Atlantic, where demic replacement of any meaningful dimension is likely to have never really happened at all.

However all the mythology of blond Guanches and blond Tuaregs is essentially a Nordicist construct born of the still racist mentality that dominated the early 20th century. Same that the unfounded claims that ancient Greeks and even Egyptians were somehow "Nordic" and not Eastern Mediterranean. It's not fundamentally different than the incredibly influential rants of Donelly on Atlantis.

Aaron said...

"I agree with this but (in most regions and notably in the central FC region) at minority levels. The traits are not necesarily recessive anyhow: that's an oversimplification: the genetics involved are not simple Medelian ones but involve many different genes, each one (or rather each allele within each of the genes) adding or substracting likelihood for this and/or that trait."

--Perhaps the incorrect terminology, but I realize it's actually a moot point to my argument. The idea is that these traits, light skin, hair, eyes may pre-date the onset of neolithic gene flow into Europe. I realize you are a Basque but you can't honesly sit here and tell me what the pre-Neolithic inhabitants of Spain, or Europe for that matter looked like prior to Neolithic gene flow. It's far more likely that these traits are rather old, rather than young, and objectively are widely distributed throughout Europe and areas of near east and Asia. It's also unlikely, due to the nature of these traits that they occurred on multiple occasions. Of course, I would need a geneticist to examine this. We must also keep in mind that the original mutations have also undergone further mutation.

"That's simply not true. Denmark for instance has a much older Neolithic than the Basque Country, maybe by 1000 years."

--Denmark isn't the best example of what I termed "Nordic". Again, I was referring to mostly North East Europe including Sweden. That being said, we can't expect a perfectly smooth gradient.

"It is also quite clear that the Neolithic migrational flow hit much harder Central Europe than the West Mediterranean, and it is also clear that Chalcolithic flows from Eastern Europe hit Northern Europe (excluding British Islands), much harder and earlier than anywhere else except maybe the Eastern Balcans."

--There are many other people to have landed and settled on the Spanish peninsula, especially within the last few thousand years. I am not suggesting the movement of neolithic was the only event to eliminate the "minority" traits, since you don't agree with the term recessive, but the first big change. Also, I was very adamant in stressing neolithic gene flow and not suggesting the aboriginal cultures were necessarily in the neolithic age, or were comprised completely of non-aboriginal European people, or that the originals had been displaced.

"While I do not discard that Scandinavia may keep some Paleolithic fossil genetics, I don't think there is any reason to consider that area particularly conservative, much less in comparison with the Atlantic, where demic replacement of any meaningful dimension is likely to have never really happened at all."

-- With small populations, like the Baltic and Finland we could have a process of sexual selection, in the case of a certain shade of blonde. Size does play a role, as the likelihood of those traits being passed to the next generation increase drastically. Also, I would add that Britain and Ireland do boast the lightest shades of skin and have the highest incidence of red hair. The quantity of light eyes also matches levels in the Baltic, if not exceeds them.

Aaron said...

Light eyes might live on, where as there may have been disadvantages to having light hair or skin. Keep in mind the strong process of sexual selection as well. (not to mention strong selection of women who likely didn't carry the traits- Neolithic onwards) The point is, these traits are old rather than young, and there is no reason to state that they were not more frequent in regions where they are rare today.

Aaron said...

Also, if anyone has the haplotypes, especially the R1b ones for this study it would be greatly appreciated. Please just post the link, I think the results would be very telling.

Maju said...

-Denmark isn't the best example of what I termed "Nordic". Again, I was referring to mostly North East Europe including Sweden. That being said, we can't expect a perfectly smooth gradient.

Eastern Europe does appear, to my somewhat limited knowledge, rather continuous before IE Kurgan migrations and mostly parallel Uralic ones in the far north. However, for the Baltic area the situation is much more complex, AFAIK with:

1. Post-Ice Age colonization mostly of Magdalenian derivation
2. Early chalcolithic arrival of the regressive foragers that seem derived from Eastern European Neolithic (Dniepr-Don) and not truly natives (Pitted Ware and related cultures)
3. The said Indoeuropean Kurgan migrations

So IMO the non-Uralic Baltic area, including Sweden, has two main components: Epipaleolithic which should not be too distinct from other areas just south of it and Chalcolithic from the southern parts of Eastern Europe (Ukraine and Southern Russia).

Is it European aboriginal, in the sense of pre-Neolithic? I'd say that, excepting the Kurgan component that I do not consider clearly European for lack of evidence of any Paleolithic continuity east of the Volga, it surely is. However it represents a blend of western (Epi-Magdalenian) and eastern (Eastern Epigravettian) components, what makes it hard to compare with other populations.

The Kurgan migration may have also caused a greater impact in said area that was scarcely populated even in the Chalcolithic, as it seems evident from the high apportions of R1a.

I rather think that if you want to know how were Northern Europeans before the arrival of Neolithic and Indoeuropeans (and Pitted Ware too), you have to look at Scots, as they seem much less mixed in general and also original from the Doggerland area, following the archaeological evidence. However Scots may have got some unusual founder effects, like the aboundance of red hair. But still they seem to me the best "living fossil" in that area.

There are many other people to have landed and settled on the Spanish peninsula, especially within the last few thousand years.

Iberian peninsula. The use of the term Spain for the geographical region fell out of use several centuries ago as it became synonym of an specific state.

I know where you want to reach but from Bauchet 2007, you can easily see that Spaniards have their own cluster and Basques their own one too. Spaniards also have a quite high apportion of Eastern Mediterranean autosomal DNA but this one is a distinct component. Basques have very low apportions of any other autosomal component.

The only doubt I have is which is the influence of North African DNA, as North Africans were not sampled by Bauchet (but are known to have some influence on the Iberian DNA pool, as seen in Y-DNA for example).

Also, Romans apart, the historical influence of Phoenicians was clearly small (it's evident if you compare with Tunisian DNA for instance, as they have loads of J1 but almost no J2, while in Iberians is exactly the opposite) and Greek one was virtually non-existent (only two Massillian trading posts are known for sure to have existed in Northern Catalonia).

In any case, genetics clearly show that the Eastern Mediterranean genetic flow through the ages existed but was limited. Among Basques was not important at all.

Maju said...

Keep in mind the strong process of sexual selection as well.

This is highly hypothetical and has never been proven in any meaningful way.

Much more likely is that the various depygmentation alleles, that we know that mostly add up, were selected because of vitamin D adaptation. Hence light hair and eyes are secondary effects of the adaptative selection in favor of skin depygmentation. This was surely more intense in some areas than in others already in the Paleolithic (the Central and Eastern provinces are some 5º north of the Franco-Cantabrian region on average and those are the populations where depygmentation is more common) and was intensified when the Far North was colonized in the Epipaleolithic.

As I said before Basques form a distinct genetic cluster within Europe that does not look Neolithic however you look at it. But while there are lots of blond Basques with blue eyes, there are also many more that are brunette with curly hair. Actually more in proportion than Spaniards/Portuguese it seems to me (though we normally have lighter skin color than Mediterranean Iberians, probably because they have that important transmediterranean influence that we don't).

If you follow Coon or similar analysis you'll get lost: because they are so extremely Nordocentric that they miss completely the Mediterranean (or other) variance. The emphasis on pigmentation in such analysis is also very much misleading, as this is a minor rather irrelevant (though admittedly very apparent) trait.

The point is, these traits are old rather than young, and there is no reason to state that they were not more frequent in regions where they are rare today.

There are reasons when the genetic continuity is very apparent and the outlier input virtually null, as is the Basque case. You seem to think that UP Europeans were all blond with blue eyes. But that is highly unlikely - I'd say that impossible on light of the available evidence.

I'd agree though that blondisms were present most likely as minority traits already in the UP (and maybe as majority traits in some provinces but not all Europe). However that may not be the case for the most extreme evolutions of these traits, which may well have only appeared in the Epipaleolithic, when they became necessary within the context of the colonization of the Far North probably.