April 02, 2009

Ancient mtDNA from La Palma (Canary Islands)

Related posts:
European Journal of Human Genetics doi:10.1038/ejhg.2009.46

The maternal aborigine colonization of La Palma (Canary Islands)

Rosa Fregel et al.

Abstract

Teeth from 38 aboriginal remains of La Palma (Canary Islands) were analyzed for external and endogenous mitochondrial DNA control region sequences and for diagnostic coding positions. Informative sequences were obtained from 30 individuals (78.9%). The majority of lineages (93%) were from West Eurasian origin, being the rest (7%) from sub-Saharan African ascription. The bulk of the aboriginal haplotypes had exact matches in North Africa (70%). However, the indigenous Canarian sub-type U6b1, also detected in La Palma, has not yet been found in North Africa, the cradle of the U6 expansion. The most abundant H1 clade in La Palma, defined by transition 16260, is also very rare in North Africa. This means that the exact region from which the ancestors of the Canarian aborigines came has not yet been sampled or that they have been replaced by later human migrations. The high gene diversity found in La Palma (95.2 +/- 2.3), which is one of the farthest islands from the African continent, is of the same level than the previously found in the central island of Tenerife (92.4+/-2.8). This is against the supposition that the islands were colonized from the continent by island hopping and posterior isolation. On the other hand, the great similarity found between the aboriginal populations of La Palma and Tenerife is against the idea of an island-by-island independent maritime colonization without secondary contacts. Our data better fit to an island model with frequent migrations between islands.

Link

7 comments:

Maju said...

IMO, it's hard to say for sure that the so-called "subsaharan African" mtDNA, that is L(xM,N) is necesarily of recent Tropical origin (i.e. slave trade). These clades are common enough in North Africa as to be considered native from that area or at least estabilished since long ago (possibly Capsian period).

In your other post on Tinerfeño aDNA it's said that 15.6% L meant the high importance of slave trade, but now this study on Canarian aborigines also shows a high incidence of L (albeit lower: 7%), what implies that at least a fraction of Canarian L was not directly from Tropical Africa. And this must be also applied to North African L in general.

The most abundant H1 clade in La Palma, defined by transition 16260, is also very rare in North Africa.

Is it? In a previous post in this blog it was mentioned that Tunisian H (not really "rare") is made up of H1 and H3 and that these lineages constitute a subset of the diversity found in Iberia.

It is true that H, and in general European lineages, are rarer in Morocco though. But Morocco is not all North Africa.

the indigenous Canarian sub-type U6b1, also detected in La Palma, has not yet been found in North Africa

This strongly suggests it is a local Canarian developement. The few cases spotted in Iberia and Cuba are probably original from the islands. Checking Maca-Meyer 2003, the ancestor nodes (there are two possible ones) seem to be shared by Berbers and Iberians in similar apportions (and this is the general trend for U6b). U6c also seems to stem from Morocco, even if it is more common in Canary Islands [Ian Logan mentions one case in Italy - have not seen any for Iberia though].

U6a is the only U6 subclade that shows a wider distribution, reaching as far east as West Asia, Ethiopia and Italy and as far south as Nigeria (probably Peul). The only exception are two Bedouins with U6b(xU6b1), that is in any case derived from the same North African/Iberian node mentioned before. [Ian Logan also mentions a rare U6d subclade found in Morocco, Ethiopia and Italy, probably Sardinia].

Kepler said...

Do you know if there is any chance to determine whether it was during recent history (last 6 centuries as opposed to more) based on mDNA material?
I mean: whether it is of recent slave trade origin or not?
I suppose there is no chance, data found in HVR1+HVR2
not being enough for that.
Or?
I say it because I have L1c3 and I am Latin American. My grand-grandmother did not look African, I know on the paternal side of her family it was Canary Islands. The maternal side could be some descendant of slave (from 1528 to 1810 slaves were imported and mixed very fast with the rest), but also there is this possibility of being L1 through indirect route.
Probably there is no way to find out.

Maju said...

At least it does overcome my limited knowledge of L(xM,N) haplogroups. Maybe a careful study of haplotypes could lead you to a more likely answer but this task overwhelms me.

What is clear is that it can't be said that paragroup L(xM,N) is strictly Tropical African but that some of it has been in North Africa for long.

Anonymous said...

I'll probably not get a response from this post, but I looked at the mt-DNA Mutations Database (http://www.familytreedna.com/mtDNA-Haplogroup-Mutations.aspx)and in the HV no 16260 comes up at all. However, 16260T is shown for L3i2, L4a, M1a2b, and R7. Is it possible that the author made a mistake by classifying 16260 as H, i.e., 'European' ? (Revised Cambridge Sequence) rCRS : 16260C.

Maju said...

You are right, Ann in that 16260 is not accepted as marker for H1 by PhyloTree either.

It is also found in R7, W1d, M1a2b, M38d, Z, M62b, M71b, M79, L3i2 and L4a. W and M1 might be candidates but the site is also too common and hypervariable.

So the authors did other tests (table 2) on markers 7025 AluI(-) and 3010 Tsp509I(+), which in their criteria mean H1.

These markers (insertions) are not listed in PhyloTree, which does not seem to use in/del markers for clade identification, relying instead on the much more certain (and expensive to test) coding region.

The same AluI test has been used in other cases to deny that some CRS-haplotype aDNA is mtDNA H (at least in Gravettian Italy and, unsure here, Magdalenian Swabia).

Charles said...

As with Kepler, my ancestors come from the Canary Islands. My maternal is L3b1a. I would love to know whether she arrived there as a convert fleeing the Inquisition or was an original Guanche wedded in to the Portuguese and Genovese. At any rate, according to this research article: Research article
Demographic history of Canary Islands male gene-pool: replacement of native lineages by European
Rosa Fregel1*, Verónica Gomes3,2, Leonor Gusmão2, Ana M González1, Vicente M Cabrera1, António Amorim2 and Jose M Larruga1; it is estimated that the original Guanche population had a total of up to 12% Mitochondrial L haplogroups. This has diminished over the centuries to less than half of that. I would say due to the forced migrations of thousands of families over the centuries to the New World. This is why the Canarian component is so important to many in Latin America. For places like Puerto Rico, Cuba and others, it formed the bulk of the white pheasant class.

Unknown said...

U8a U6b1 U6d including the H1 and 16260. I can appreciate Maju comments on the groups who charge extra - as I have paid willingly, to contribute to our minority group!
How though do we integrate Tunisia and/or Yemen if the full position is not matched for La Palma? Thanks.
e.g. I was skeptical for a very long time.. thinking it possibly a advertising tactic. Now I CAN include this as exclusive? I can I can't I can I can't?