May 13, 2010

mtDNA and trans-Saharan slave trade (Harich et al. 2010)

The main trans-Saharan slave routes are shown in Figure 1 (on the left). The paper also contains quite useful interpolation maps of the main Sub-Saharan African mtDNA haplogroups, who should be useful for future reference.

BMC Evolutionary Biology 2010, 10:138 doi:10.1186/1471-2148-10-138


The trans-Saharan slave trade - clues from interpolation analyses and high-resolution characterization of mitochondrial DNA lineages

Abstract

Background

A proportion of 1/4 to 1/2 of North African female pool is made of typical sub-Saharan lineages, in higher frequencies as geographic proximity to sub-Saharan Africa increases. The Sahara was a strong geographical barrier against gene flow, at least since 5,000 years ago, when desertification affected a larger region, but the Arab trans-Saharan slave trade could have facilitate enormously this migration of lineages. Till now, the genetic consequences of these forced trans-Saharan movements of people have not been ascertained.

Results
The distribution of the main L haplogroups in North Africa clearly reflects the known trans-Saharan slave routes: West is dominated by L1b, L2b, L2c, L2d, L3b and L3d; the Center by L3e and some L3f and L3w; the East by L0a, L3h, L3i, L3x and, in common with the Center, L3f and L3w; while, L2a is almost everywhere. Ages for the haplogroups observed in both sides of the Saharan desert testify the recent origin (holocenic) of these haplogroups in sub-Saharan Africa, claiming a recent introduction in North Africa, further strengthened by the no detection of local expansions.

Conclusions
The interpolation analyses and complete sequencing of present mtDNA sub-Saharan lineages observed in North Africa support the genetic impact of recent trans-Saharan migrations, namely the slave trade initiated by the Arab conquest of North Africa in the seventh century. Sub-Saharan people did not leave traces in the North African maternal gene pool for the time of its settlement, some 40,000 years ago.

Link

15 comments:

Andrew Oh-Willeke said...

"Also can anyone explain logically how all of these slaves could come from west Africa?"

Fourteen hundred plus years of historically documented trans-Saharan slave trade (e.g. the memoir of Ibn Battuta in the 14th century) involving an estimated four million people that still continues at a low level today (in addition to two million people taken as slaves on the Red Sea maritime route and millions more taken as slaves on the Atlantic route).

"The slavery theory contradicts itself because those undeniably black people in north Africa do not look like west Africans."

The paper suggests that almost all of the gene flow is maternal and that the average sub-Saharan African contribution is about 33%, hence an average total genetic contribution of about 1/6th. This is about the same amount of genetic contribution as the Asian DNA contribution found in the Volga-Ural Tartars, which isn't blatantly visible.

frenchy said...

Any 23andme's customer can see and confirm that most North Africans who took the Ancestry testing show 0% or < 0.5% Sub-saharan ancestry. Only the Mozabite (saharan) sample show 12% but Mozabites are saharan people.

Therefore L haplogroups among North-Africans cannot be recent and due to slave trade otherwise 23andme would detect it very easily.

L haplogroups among North-Africans are without any doubt very, very old... At this time there were no "black" or "white"... and this explain why 23andme is not able to detect it...

frenchy said...

Any 23andme's customer can see and confirm that most North Africans who took the Ancestry testing show 0% or < 0.5% Sub-saharan ancestry. (Only the Mozabite sample show 12% african ancestry but Mozabites are a saharan ethnic group not mediterranean)

Therefore L haplogroups among North-Africans cannot be recent and due to slave trade as otherwise 23andme would detect it very easily as they do for African ancestry among South Americans for example.

L haplogroups among North-Africans are without any doubt very very old... At this time there were no "black" or "white"...

Belenos said...

Mal, the map shows various trans Saharan routes, including from Central and East Central Africa. As to your subjective opinion of what black North Africans look like, if you look at the population of Mali and Senegal you will see a massive variety of physical traits and types, botth across and within ethnic groups. Black (coastal) Libyans do tend to look very much like Chadians, whereas black Tunisians are more mixed.

The paper wasn't looking at Europran mtDNA, so it provides little information in that regard.

Maju said...

IMO most of the mtDNA L(xM,N) in North Africa and West Asia is pre-modern. I have discussed this paper here and we have the Guanche aDNA to confirm it, so not really worth extending further.

IMO a good deal of the L lineages we now see in North Africa and West Asia arrived at the same time or 'soon' after the OOA migration. Other lineages may have scattered with the spread of Capsian culture and all the likely Afroasiatic expansion.

Onur Dincer said...

Any 23andme's customer can see and confirm that most North Africans who took the Ancestry testing show 0% or < 0.5% Sub-saharan ancestry. Only the Mozabite (saharan) sample show 12% but Mozabites are saharan people.

Coming to the phenotypes, non-Saharan North Africans don't look Black at all, only some Saharan people (mostly the ones in its southern regions) seem to have some Black admixture, however small.

Jack said...

What on earth does "...most North africans..." mean? If the world went by 51%s then reality would be fancier than a dream.
Have you ever seen a bunch of North Africans? Sometimes I even doubt they are mostly caucasian (whatever that means in this case).

Maju said...

They don't stand up in a crowd more than your standard German or British tourist. I see all types of people daily, including loads of North Africans (privilege of living in the mosque street), and these are only diffusely differentiated. You can have many doubts when identifying anyone as North African while you almost never do when identifying someone as Chinese (East Asian), for example. So they enter (as group) in the "other Caucasoid" category just like Nordics and Pakistanis: somewhat different on average but not clear cut.

The safest way to identify them is always by the accent... but then they could be Syrians or whatever else. Moroccans are more "exotic" on average possibly (but mostly in a very unique "Berber" way that at most may remind of Horners somewhat and never of West Africa) but as you head to Tunisia this changes towards more standard Europid types (Zidane for instance).

I understand that the formation of that (very varied) range of phenotypes is because a long and complex differentiated coalescence in the region, with occasional important inputs from other areas, namely: NE Africa, West Asia and Iberia (in no particular order).

That's what the genetic pool says and I don't see how the phenotype contradicts this.

I suspect that your impression is heavily biased because you may have a markedly Nordic or at least European archetype of West Eurasians but a much more correct reference type has to be found in West Asia. After all that's the core of the scatter of West Eurasians.

Malgoukas said...

Fourteen hundred years.

My goodness. How are those people still around?


What did those people do? Give away all of there architects, metal workers, doctors, security specialists, etc. etc? Yes that makes perfect sense.


Or maybe it was constant war of Fourteen hundred years. If that is the case then now you have to add in the deaths and crippling injuries that would most certainly occur in any attempts to obtain them. How could any population survive that?


You really need to go back to the drawing board with your analysis. It doesn't hold water much less make sense.

Malgoukas said...

Looking at the map, and where these supposed slave trades began. Those countries are not sub-saharan Africa nor west Africa. This paper is wrong.


Going by the map, who exactly were the slaves and where did they come from? This paper is extremely vague and not very scholarly.

mathilda said...

They have L haplotypes in Guanche mtDNA, so they can't all date back to slavery. Did the authors not bother to check the ADNA? Remiss of them.

I'd put a minority of the L haplotypes in NA down to WL pottery cultures in the Sahara and some from a spread into NA by the expansion from Nubia about 22k ago. I'm not so sure about the ages Maju is gving them, they seem too young in NA for that.

Structure said...

Maju said : "IMO a good deal of the L lineages we now see in North Africa and West Asia arrived at the same time or 'soon' after the OOA migration. Other lineages may have scattered with the spread of Capsian culture and all the likely Afroasiatic expansion."

An important new study (not yet discussed by Dienekes) seems to confrim that indeed L hgs are very ancient in North Africa and mostly not linked to slave trade...

See
"Ancient Local Evolution of African mtDNA Haplogroups in Tunisian Berber Populations", Frigi 2010

http://muse.jhu.edu/login?uri=/journals/human_biology/v082/82.4.frigi.pdf

Abstract:

"Our objective is to highlight the age of sub-Saharan gene flows in North Africa and particularly in Tunisia. Therefore we analyzed in a broad phylogeographic context sub-Saharan mtDNA haplogroups of Tunisian Berber populations considered representative of ancient settlement. More than 2,000 sequences were collected from the literature, and networks were constructed. The results show that the most ancient haplogroup is L3*, which would have been introduced to North Africa from eastern sub-Saharan populations around 20,000 years ago. Our results also point to a less ancient western sub-Saharan gene flow to Tunisia, including haplogroups L2a and L3b. This conclusion points to an ancient African gene flow to Tunisia before 20,000 years BP. These findings parallel the more recent findings of both archaeology and linguistics on the prehistory of Africa. The present work suggests that sub-Saharan contributions to North Africa have experienced several complex population processes after the occupation of the region by anatomically modern humans. Our results reveal that Berber speakers have a foundational biogeographic root in Africa and that deep African lineages have continued to evolve in supra-Saharan Africa."

jonaos said...

""Fourteen hundred years.

My goodness. How are those people still around?


What did those people do? Give away all of there architects, metal workers, doctors, security specialists, etc. etc? Yes that makes perfect sense.""


Do you also deny the Trans Atlantic slave trade ? That lasted nearly five hundred years. DNA is DNA, and clearly the slave trade brought Black Africans into North Africa. It is also obvious to anyone on a phenotype level, Black ancestry is obvious amongst many North Africans. Even Gaddafi [leader of Libya] has obvious black ancestry, atleast 25%.

Unknown said...

Not true in the sense that they all fall under this category. Some do, but most don't. DNA doesn't change; however, the interpretation of the results can signifantly alter history. The Fulani (who are Amazigh) say they migrated south and mixed in with the W. Africans. This is the more likely reason for the DNA matches, not a slave trade.

Maju said...

Fulani are NOT Amazhigh (Berber): they speak an Atlantic Niger-Congo language. They do seem to have noticeable West Saharan admixture however.