September 18, 2004

The African mtDNA landscape

In the comments of a previous post, I mentioned that the present-day populations of East Africa are only partially descended from the prehistoric East Africans which gave rise to all non-African groups. Apart from the physical anthropological evidence for a lack of craniometric continuity in the region, we also have the findings of population genetics.

Most non-African mtDNA belongs to two macro-groups, called N (frequent in West Eurasians) and M (frequent in East and South Asians and Americans). Both M and N are derived from a particular African group called L3. Thus, it appears that non-Africans are descended from a prehistoric African population which had L3-type mtDNA.

A comprehensive study [1] has studied the distribution of mtDNA haplogroups in present-day African populations:



As we can see, modern East Africans have less than 50% of L3-group mtDNA, with the remainder being either M1 (whose origin is disputed, with some considering it Asian and others African in origin), a little bit of U6 (which is of North African origin), and the remainder (~50%) belongs to L-groups other than L3.

It is clear that this latter component was not present in the East African source of non-African populations in the past. If it were, then it would be present outside Africa. But, even in the Near East, African-specific mtDNA are a minority, mostly confined to Arab populations and attributed to the slave trade [2].

Thus, it appears that a large fraction of present-day East African mitochondrial ancestry is derived from different populations than the ones that spawned non-Africans. This element is probably responsible for the introduction of the Negroid type in the region, which now forms a major element in the population, together with the pre-Negroid East Africans and more recent Caucasoid arrivals from across the Red Sea.


[1] Am. J. Hum. Genet., 71:1082-1111, 2002
[2] Am. J. Hum. Genet., 72:1058-1064, 2003

No comments: