New blogs

Leherensuge was replaced in October 2010 by two new blogs: For what they were... we are and For what we are... they will be. Check them out.

Friday, April 3, 2009

Brief review of recent mtDNA H info



Haplogroup H (mtDNA) has been popping up at Dienekes' and Mathilda's blogs several times as of late. In order to keep a clear reference, here is the list of the papers:

1. V. Alvarez-Iglesias, A. Mosquera-Miguel et al. New Population and Phylogenetic Features of the Internal Variation within Mitochondrial DNA Macro-Haplogroup R0. PLoS-ONE, 2009. (open access)
2. H. Ennafaa, V. M Cabrera et al. Mitochondrial DNA haplogroup H structure in North Africa. BioMed Central Genetics 2009. (open access)
3. L. Cherni et al. Post-last glacial maximum expansion from Iberia to North Africa revealed by fine characterization of mtDNA H haplogroup in Tunisia. American Journal of Physical Anthropology, 2008. (abstract only - full paper behind paywall)
4. D. Caramelli et al. A 28,000 Years Old Cro-Magnon mtDNA Sequence Differs from All Potentially Contaminating Modern Sequences. PLoS-ONE, 2008. (open access)
5. H. Chandler, B. Sykes and J. Zilhao. Using ancient DNA to examine genetic continuity at the Mesolithic-Neolithic transition in Portugal. Actas del III Congreso del Neolítico en la Península Ibérica, 2005. (PDF - freely accessible)


Fig. 3 from paper #1
H subclades' distribution in Europe, West and Central Asia


(1) and (2) are just out of the oven and it is really nice that such important papers are open access. (1) covers Europe, West Asia and Central Asia, while (2) focuses in North Africa (with context references for West Asia and Iberia), being the perfect complement of (1), illustrating how West European H1 and H3 comprise the largest part of North African H (which in turn is 25% of all regional mtDNA). (3) is sadly behind a paywall but the main conclusion visible in the abstract: Although no signs of local expansion were detected, which would allow a clear dating of their introduction, the younger and less diverse Tunisian H1 and H3 lineages indicate Iberia as the radiating centre.

(4) is a most important research that has found CRS haplotype in the HVR-I (most likely H) in a 28,000 years old individual from Paglicci Cave, Puglia, Italy. This discovery clashes directly with he usual MCH hunches that make this subclade much recent and document that haplogroup H and at least some of its subclades were already common in the early Gravettian and are no modern developement at all. (5) is an older yet relevant paper that shows that haplogroup H was strongly dominant in both Epipaleolithic and Neolithic Portuguese, along with some U and less important clades.

Overall it seems reasonably clear that H expansion dates at least from Gravettian times (no other cultural group has been so pan-European ever) and that its expansion into North Africa suggest that Iberomaurusian (Oranian) culture was with all likehood derived from Southern Iberian Gravetto-Solutrean. It also seems very clear that the DNA "molecular clock" does not seem to be ticking at the usually speculated rythms at all.

____________________

Justify FullUpdate: Another paper on aDNA, including H, that I had forgotten is R. Kèfi et al. Diversité mitochondriale de la population de Taforalt (12.000 ans bp - maroc): une approche génétique a l’étude du peuplement de l’afrique du nord. Anthropologie 2005. Reviewed and translated in its essentials by Mathilda several months ago [note: previous link has gone broken but I found a Power Point direct download from Pasteur Institut].

The HVS-I haplotype sequencing of 23 bone remains from Taforalt cave, Morocco, yielded the following results:

· H or U (CRS, most likely all H): 11 (47.8%)
· H (not CRS): 4 (17.4%)
· JT: 3 (13%)
· U6: 2 (8.7%)
· V: 2 (8.7%)
· L3, M or N: 1 (4.3%)

JT means JT(xJ,xT), a very rare finding, exclusive nowadays of North Africa and Italy. The authors mention that this same mtDNA pool is the one found among modern Notheren Moroccan Berbers who live in that same area.
.

18 comments:

Manju Edangam said...

Overall it seems reasonably clear that H expansion dates at least from Gravettian times (no other cultural group has been so pan-European ever) and that its expansion into North Africa suggest that Iberomaurusian (Oranian) culture was with all likehood derived from Southern Iberian Gravetto-Solutrean.

So Cameroonians are the relic population of Oranian culture?

Maju said...

Cameroonians do not have mtDNA H AFAIK. Some north Cameroonians (Ouldeme basically) have strong apportions of Y-DNA R1b1c and R1b*. These may be different issues.

At the moment I am thinking in terms of R1b spreading from West Asia in three directions: Europe, Sahelian Africa and Central Asia. This spread may or not be related with the spread of H.

R1b and H may have some apparent distribution parallels but this requires excluding Sahel (where there is R1b and no H) and Eastern Europe (where there is aboundant H but little R1b). Another problem is that R1b appears to have to be "forced" to fit into Gravettian ages, while we know empirically that H subclades were already there by that time.

Just came from posting back at Mathilda's and she came up with the independent estimate of c. 45 kya for H, what would make it Bachokirian (or maybe related to the proto-Aurignacian of Central-West Asia). It is really hard to fit R1b in such deep chronologies, while I have no such problem in accepting them for H instead.

African R1b1c and R1b* do not seem in any case to derive from Europe but from a West Asian core spreading rapidly in several directions. In this sense, I'd rather relate it with some mtDNA M1 and maybe even U6a, if anything, though of course they could have been associated with some of the African-specific L(xM,N) lineages or even be a mostly male-mediated spread like that of Q in North Asia and Beringia.

Ebizur said...

Maju said,

"Some north Cameroonians (Ouldeme basically) have strong apportions of Y-DNA R1b1c and R1b*."

Actually, all that can be said for certain at this point is that some ethnic groups in Sub-Saharan Africa, particularly (but not exclusively) some ethnic groups who live around the border between the Sahel and the Sudan ecoregions, have notable frequencies of some sort of haplogroup R Y-DNA, but it is probably not R1b1c, which has been found only in a single individual in northeastern Turkey.

I shall now present some relevant data.

Fulvio Cruciani et al., "A Back Migration from Asia to Sub-Saharan Africa Is Supported by High-Resolution Analysis of Human Y-Chromosome Haplotypes," American Journal of Human Genetics 70:1197–1214, 2002:

Ethiopian Jews (n=22; Afro-Asiatic, Semitic)
1/22 = 4.5% J2b-M12/M102
1/22 = 4.5% T-M70

Cameroon (northern):
Fali (n=39; Niger-Congo, Adamawa)
9/39 = 23.1% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Tali (n=15; Niger-Congo, Adamawa)
1/15 = 6.7% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Mixed Adamawa (n=18; Niger-Congo, Adamawa)
10/18 = 55.6% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Fulbe/Cameroon (n=17; Niger-Congo, West Atlantic)
3/17 = 17.6% T-M70
2/17 = 11.8% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Ouldeme (n=21; Afro-Asiatic, Chadic)
20/21 = 95.2% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Daba (n=18; Afro-Asiatic, Chadic)
8/18 = 44.4% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Mixed Chadic (n=15; Afro-Asiatic, Chadic)
10/15 = 66.7% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Mixed Nilo-Saharan (n=9; Nilo-Saharan, Central Sudanic/Saharan)
1/9 = 11.1% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Cameroon (southern):
Ewondo (n=29; Niger-Congo, Benue´-Congo, Bantu)
1/29 = 3.4% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

J. R. Luis, D. J. Rowold, M. Regueiro, B. Caeiro, C. Cinniog˘lu, C. Roseman, P. A. Underhill, L. L. Cavalli-Sforza, and R. J. Herrera, "The Levant versus the Horn of Africa: Evidence for Bidirectional Corridors of Human Migrations," American Journal of Human Genetics 74:532–544, 2004:

Bantu/Southern Cameroon
2/14 = 14.3% R1-M173(xR1a1a-M17, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Wairak/Tanzania
2/43 = 4.7% T-M70

Hutu/Rwanda
1/69 = 1.4% R1-M173(xR1a1a-M17, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Elizabeth T Wood, Daryn A Stover, Christopher Ehret, Giovanni Destro-Bisol, Gabriella Spedini, Howard McLeod, Leslie Louie, Mike Bamshad, Beverly I Strassmann, Himla Soodyall and Michael F Hammer, "Contrasting patterns of Y chromosome and mtDNA variation in Africa: evidence for sex-biased demographic processes," European Journal of Human Genetics (2005) 13, 867–876:
West Africa
Wolof (Gambia/Senegal; n=34; Niger-Congo, Atlantic)
4/34 = 11.8% E1a-M33
1/34 = 2.9% E2b-M54
1/34 = 2.9% E1b1-P2(xE1b1a-P1, E1b1b1-M35)
23/34 = 67.6% E1b1a-P1(xE1b1a7-M191)
2/34 = 5.9% E1b1b1-M35(xE1b1b1a-M78, E1b1b1b-M81)
2/34 = 5.9% E1b1b1b-M81
1/34 = 2.9% K-M9(xL-M20, M1-M4, N1-LLY22g, O-M175, P-P27, T-M70)

Fante (Ghana; n=32; Niger-Congo, Kwa)
1/32 = 3.1% E1a-M33
1/32 = 3.1% E1b1-P2(xE1b1a-P1, E1b1b1-M35)
14/32 = 43.8% E1b1a-P1(xE1b1a7-M191)
13/32 = 40.6% E1b1a7-M191
1/32 = 3.1% E1b1b1-M35(xE1b1b1a-M78, E1b1b1b-M81)
1/32 = 3.1% R1b1-P25(xR1b1b2-M269)
1/32 = 3.1% R1b1b2-M269

Central Africa
Podokwo (Cameroon, North; n=19; Afroasiatic, Chadic)
18/19 = 94.7% R1b1-P25(xR1b1b2-M269)
1/19 = 5.3% E1b1b1a-M78

Mandara (Cameroon, North; n=28; Afroasiatic, Chadic)
4/28 = 14.3% A3b2-M13
1/28 = 3.6% B2a1a-M152
2/28 = 7.1% E2b-M54
3/28 = 10.7% E1b1a-P1(xE1b1a7-M191)
1/28 = 3.6% E1b1a7-M191
17/28 = 60.7% R1b1-P25(xR1b1b2-M269)

Uldeme (Cameroon, North; n=13; Afroasiatic, Chadic)
4/13 = 30.8% B2a1a-M152
9/13 = 69.2% R1b1-P25(xR1b1b2-M269)

Tupuri (Cameroon, North; n=9; Niger-Congo, Gur)
2/9 = 22.2% A3b2-M13
1/9 = 11.1% B2a-M150(xB2a1a-M152)
6/9 = 66.7% R1b1-P25(xR1b1b2-M269)

Bassa (Cameroon, South; n=11; Niger-Congo, Bantu)
6/11 = 54.5% E1b1a-P1(xE1b1a7-M191)
4/11 = 36.4% E1b1a7-M191
1/11 = 9.1% R1b1-P25(xR1b1b2-M269)

East Africa
Amhara (Ethiopia; n=18; Afroasiatic, Semitic)
3/18 = 16.7% A3b2-M13
1/18 = 5.6% E1b1-P2(xE1b1a-P1, E1b1b1-M35)
2/18 = 11.1% E1b1b1-M35(xE1b1b1a-M78, E1b1b1b-M81)
6/18 = 33.3% E1b1b1a-M78
6/18 = 33.3% J-12f2

Oromo (Ethiopia; n=9; Afroasiatic, Cushitic)
1/9 = 11.1% A3b2-M13
1/9 = 11.1% E1b1-P2(xE1b1a-P1, E1b1b1-M35)
1/9 = 11.1% E1b1b1-M35(xE1b1b1a-M78, E1b1b1b-M81)
2/9 = 22.2% E1b1b1a-M78
3/9 = 33.3% J-12f2
1/9 = 11.1% T-M70

South Semitic (Ethiopia; n=20; Afroasiatic, Semitic)
1/20 = 5.0% A3*-M32
1/20 = 5.0% A3b2-M13
2/20 = 10.0% E1b1-P2(xE1b1a-P1, E1b1b1-M35)
4/20 = 20.0% E1b1b1-M35(xE1b1b1a-M78, E1b1b1b-M81)
7/20 = 35.0% E1b1b1a-M78
3/20 = 15.0% J-12f2
1/20 = 5.0% K-M9(xL-M20, M1-M4, N1-LLY22g, O-M175, P-P27, T-M70)
1/20 = 5.0% T-M70

Dama (Namibia; n=18; Khoisan, Central)
1/18 = 5.6% A2b-P28
1/18 = 5.6% A3b1-M51
1/18 = 5.6% B2-M182(xB2a-M150, B2b-50f2(P))
1/18 = 5.6% E-SRY4064(xE1a-M33, E2-M75, E1b1-P2)
1/18 = 5.6% E2*-M75(xE2a-M41, E2b-M54)
1/18 = 5.6% E2b-M54
6/18 = 33.3% E1b1a-P1(xE1b1a7-M191)
4/18 = 22.2% E1b1a7-M191
1/18 = 5.6% J-12f2
1/18 = 5.6% R1b1b2-M269

Nama (Namibia; n=11; Khoisan, Central)
1/11 = 9.1% A2b-P28
6/11 = 54.5% A3b1-M51
1/11 = 9.1% E1b1a-P1(xE1b1a7-M191)
1/11 = 9.1% E1b1a7-M191
1/11 = 9.1% E1b1b1-M35(xE1b1b1a-M78, E1b1b1b-M81)
1/11 = 9.1% R1b1b2-M269

Herero (Namibia; n=24; Niger-Congo, Bantu)
38% (9/24) E1b1a-P1(xE1b1a7-M191)
33% (8/24) E1b1a7-M191
4% (1/24) F-P14(xG-M201, H1-M52, I-P19, J-12f2, K-M9)
4% (1/24) I-P19
4% (1/24) J-12f2
4% (1/24) R1a1-SRY10831b
4% (1/24) R1b1-P25(xR1b1b2-M269)
8% (2/24) R1b1b2-M269

Ambo (Namibia; n=22; Niger-Congo, Bantu)
1/22 = 4.5% E-SRY4064(xE1a-M33, E2-M75, E1b1-P2)
1/22 = 4.5% E2b-M54
11/22 = 50.0% E1b1a-P1(xE1b1a7-M191)
7/22 = 31.8% E1b1a7-M191
1/22 = 4.5% E1b1b1-M35(xE1b1b1a-M78, E1b1b1b-M81)
1/22 = 4.5% R1b1-P25(xR1b1b2-M269)

Sotho-Tswana (South Africa; n=28; Niger-Congo, Bantu)
2/28 = 7.1% A3b1-M51
5/28 = 17.9% B2a1a-M152
1/28 = 3.6% E-SRY4064(xE1a-M33, E2-M75, E1b1-P2)
1/28 = 3.6% E2b-M54
10/28 = 35.7% E1b1a-P1(xE1b1a7-M191)
6/28 = 21.4% E1b1a7-M191
2/28 = 7.1% E1b1b1-M35(xE1b1b1a-M78, E1b1b1b-M81)
1/28 = 3.6% F-P14(xG-M201, H1-M52, I-P19, J-12f2, K-M9)

Hisham Y. Hassan, Peter A. Underhill, Luca L. Cavalli-Sforza, and Muntaser E. Ibrahim, "Y-Chromosome Variation Among Sudanese: Restricted Gene Flow, Concordance With Language, Geography, and History," American Journal of Physical Anthropology (2008):
Borgu (Agriculturists; n=26; Nilo-Saharan, Maban)
9/26 = 34.6% A3b2-M13
10/26 = 38.5% E1b1b-M215(xE1b1b1a-M78)
3/26 = 11.5% E1b1b1a1b-V32
1/26 = 3.8% E1b1b1a3-V22
3/26 = 11.5% R1b1-P25

Masalit (Agriculturists; n=32; Nilo-Saharan, Maban)
6/32 = 18.8% A3b2-M13
1/32 = 3.1% B-M60
1/32 = 3.1% E1b1b1a-M78(xE1b1b1a1-V12, E1b1b1a2-V13, E1b1b1a3-V22, E1b1b1a4-V65)
17/32 = 53.1% E1b1b1a1b-V32
5/32 = 15.6% E1b1b1a3-V22
2/32 = 6.3% J-12f2(xJ2-M172)

Fur (Agriculturists; n=32; Nilo-Saharan, Fur)
10/32 = 31.3% A3b2-M13
1/32 = 3.1% B-M60
13/32 = 40.6% E1b1b1a1b-V32
6/32 = 18.8% E1b1b1a3-V22
2/32 = 6.3% J-12f2(xJ2-M172)

Nubians (Agriculturists; n=39; Nilo-Saharan, Eastern Sudanic)
3/39 = 7.7% B-M60
3/39 = 7.7% E1b1b-M215(xE1b1b1a-M78)
5/39 = 12.8% E1b1b1a1-V12(xE1b1b1a1b-V32)
1/39 = 2.6% E1b1b1a1b-V32
4/39 = 10.3% F-M89(xH1-M52, I-M170, J-12f2, K-M9)
2/39 = 5.1% I-M170
16/39 = 41.0% J-12f2(xJ2-M172)
1/39 = 2.6% J2-M172
4/39 = 10.3% R1b1-P25

Fulani (Nomadic Pastoralists; n=26; Niger-Congo, Atlantic)
3/26 = 11.5% E1a-M33
1/26 = 3.8% E1b1b1a1b-V32
8/26 = 30.8% E1b1b1a3-V22
14/26 = 53.8% R1-M173(xR1b1-P25)

Hausa (Agriculturists; n=32; Afro-Asiatic, Chadic)
4/32 = 12.5% A3b2-M13
5/32 = 15.6% B-M60
5/32 = 15.6% E1a-M33
4/32 = 12.5% E1b1a-M2
1/32 = 3.1% E1b1b1a1b-V32
13/32 = 40.6% R1b1-P25

Beja (Pastoralists; n=42; Afro-Asiatic, Cushitic?)
2/42 = 4.8% A3b2-M13
7/42 = 16.7% E1b1b-M215(xE1b1b1a-M78)
2/42 = 4.8% E1b1b1a1-V12(xE1b1b1a1b-V32)
13/42 = 31.0% E1b1b1a1b-V32
15/42 = 35.7% J-12f2(xJ2-M172)
1/42 = 2.4% J2-M172
2/42 = 4.8% R1b1-P25

Copts (Agriculturists; n=33; Afro-Asiatic, Ancient Egyptian > Semitic)
5/33 = 15.2% B-M60
2/33 = 6.1% E1b1b-M215(xE1b1b1a-M78)
5/33 = 15.2% E1b1b1a1-V12(xE1b1b1a1b-V32)
13/33 = 39.4% J-12f2(xJ2-M172)
2/33 = 6.1% J2-M172
1/33 = 3.0% K-M9(xL-M11, O-M175, P-M74)
5/33 = 15.2% R1b1-P25

Arabs/Gaalien (Agriculturists; n=50; Afro-Asiatic, Semitic)
3/50 = 6.0% A3b2-M13
3/50 = 6.0% E1b1b1a1-V12(xE1b1b1a1b-V32)
3/50 = 6.0% E1b1b1a1b-V32
3/50 = 6.0% E1b1b1a3-V22
5/50 = 10.0% F-M89(xH1-M52, I-M170, J-12f2, K-M9)
2/50 = 4.0% I-M170
18/50 = 36.0% J-12f2(xJ2-M172)
2/50 = 4.0% J2-M172
3/50 = 6.0% K-M9(xL-M11, O-M175, P-M74)
1/50 = 2.0% R1-M173(xR1b1-P25)
7/50 = 14.0% R1b1-P25

Arabs/Meseria (Nomadic Pastoralists; n=28; Afro-Asiatic, Semitic)
1/28 = 3.6% E1b1b1a1-V12(xE1b1b1a1b-V32)
3/28 = 10.7% E1b1b1a1b-V32
3/28 = 10.7% F-M89(xH1-M52, I-M170, J-12f2, K-M9)
2/28 = 7.1% I-M170
12/28 = 42.9% J-12f2(xJ2-M172)
7/28 = 25.0% R1b1-P25

Arabs/Arakien (Agriculturists; n=24; Afro-Asiatic, Semitic)
2/24 = 8.3% E1b1b1a1-V12(xE1b1b1a1b-V32)
1/24 = 4.2% E1b1b1a1b-V32
1/24 = 4.2% E1b1b1a3-V22
2/24 = 8.3% F-M89(xH1-M52, I-M170, J-12f2, K-M9)
16/24 = 66.7% J-12f2(xJ2-M172)
2/24 = 8.3% R1b1-P25

Sudanese Arab total:
3/102 = 2.9% A3b2-M13
6/102 = 5.9% E1b1b1a1-V12(xE1b1b1a1b-V32)
7/102 = 6.9% E1b1b1a1b-V32
4/102 = 3.9% E1b1b1a3-V22
10/102 = 9.8% F-M89(xH1-M52, I-M170, J-12f2, K-M9)
4/102 = 3.9% I-M170
46/102 = 45.1% J-12f2(xJ2-M172)
2/102 = 2.0% J2-M172
3/102 = 2.9% K-M9(xL-M11, O-M175, P-M74)
1/102 = 1.0% R1-M173(xR1b1-P25)
16/102 = 15.7% R1b1-P25

Maju said...

After checking my "references" you are probably right, Ebizur. It seems that talking of R1b1c as Sahelian African is a misconception based on a tree published in a private site. I could not find anything else that could support it and YSOGG has R1b1c marked as "private" what fits well with your descrption of a single individual in Turkey.

And thanks for your review of African R (and other clades). Let's recapitulate focusing on R alone:

North Cameroun:
Fali: 9/39 R1(xR1a, R1b1a, R1b1b1, R1b1b2)
Tali: 1/15 R1* (id.)
Mixed Adamawa: 10/18 R1* (id.)
Fulbe: 2/17 R1* (id.)
Ouldeme [Ouldeme 1]: 20/21 R1* (id.)
Daba: 8/18 R1* (id.)
Mixed Chadic: 10/15 R1* (id.)
Mixed Nilo-Saharan: 1/9 R1* (id.)
Podokwo: 18/19 R1b1(xR1b1b2)
Mandara: 17/28 R1b1* (id.)
Uldeme [Ouldeme 2]: 9/13 R1b1* (id.)
Tupuri: 6/9 R1b1* (id.)
Bassa: 1/11 R1b1* (id.)

South Cameroon:
Ewondo: 1/29 R1(xR1a, R1b1a, R1b1b1, R1b1b2)
Bantu: 2/14 R1* (id.)

Rwanda:
Hutu: 1/69 R1(xR1a, R1b1a, R1b1b1, R1b1b2)

Ghana:
Fante: 1/32 R1b1*, 1/32 R1b1b2 [IMO colonial origin for the latter, maybe both]

Namibia [IMO colonial]:
Dama: 1/11 R1b1b2
Herero: 2/24 R1b1b2
Ambo: 1/22 R1b1(xR1b1b2)

Sudan:
Nubians: 4/39 R1b1
Borgu: 3/26 R1b1
Haussa: 13/32 R1b1
Fulani: 14/26 R1(xR1b1)
Beja: 2/42 R1b1
Copts: 5/33 R1b1
Arabs: 1/102 R1(xR1b1), 16/102 R1b1

So what do we have? Apparently loads of R1b1*, with some R1* (notably among the Sudanese Fulani), through Sahelian Africa, particularly concentrated in N. Cameroon and Sudan. Possibly a distinct R1b1 subclade (call it "R1b1d").

It's noticeable that the groups with highest R1b1* apportion are also those with links to the West (Haussa, Fulani), so it's not impossible that Sahelian R1b1* may have spread from West to East. On the other hand, both Haussa and Fulani may have ultimate remote origins in NE Africa, being pastoralist peoples (Haussa are Chadic-speakers and the Fulani/Peul have often been linked with the spread of pastoralism in the Sahara).

Can you recall the distribution of Egyptian R1? From memory R1b was like 50% R1b1b2 and 50% R1b* but can't find the source right now. I guess that the R1b* is the same as Sudanese/Cameroonian R1b1*.

Ebizur said...

Maju said,
"Namibia [IMO colonial]:
...
Herero: 2/24 R1b1b2"

Actually, the data on haplogroup R in the Herero sample of Wood et al. (2005) are as follows:

Herero (Namibia; n=24; Niger-Congo, Bantu)
4% (1/24) R1a1-SRY10831b
4% (1/24) R1b1-P25(xR1b1b2-M269)
8% (2/24) R1b1b2-M269

You have missed the 1/24 R1b1-P25(xR1b1b2-M269). You might have intentionally left out the 1/24 R1a1-SRY10831b, but that is technically a subclade of haplogroup R, too.

Maju said,
"It's noticeable that the groups with highest R1b1* apportion are also those with links to the West (Haussa, Fulani), so it's not impossible that Sahelian R1b1* may have spread from West to East."

What I find very interesting about this sample of Hausa from Sudan is that, except for the moderate presence of A3b2-M13 and a single E1b1b1a1b-V32 individual, they lack Y-DNA haplogroups that one would typically find in East Africa, and rather have a Y-DNA pool whose composition resembles that of other Chadic-speaking populations, who live in west-central Africa, especially Nigeria, Cameroon, and Chad:

Hausa (Sudan)
Afro-Asiatic > Chadic > West Chadic
4/32 = 12.5% A3b2-M13
5/32 = 15.6% B-M60
5/32 = 15.6% E1a-M33
4/32 = 12.5% E1b1a-M2
1/32 = 3.1% E1b1b1a1b-V32
13/32 = 40.6% R1b1-P25

Note that the West Chadic language family, to which the Hausa language belongs, is distributed mainly in Nigeria, and, in fact, the Hausa language is one of the most important languages in that country. I don't know whether the Hausas in Sudan have migrated there recently from Nigeria or vicinity, but the composition of their Y-DNA pool does seem to be very "Chadic," and overall more similar to the Y-DNA pools of West-Central African populations than to those of East African populations.

By the way, most individuals in this sample of Fulani from Sudan are R1-M173(xP25), so they are either R1(xR1b1) or else R1b1 with a reversion of the P25 mutation (which actually is not too rare).

Fulani (Nomadic Pastoralists; n=26; Niger-Congo, Atlantic)
3/26 = 11.5% E1a-M33
1/26 = 3.8% E1b1b1a1b-V32
8/26 = 30.8% E1b1b1a3-V22
14/26 = 53.8% R1-M173(xR1b1-P25)

It seems like every study of African Y-DNA comes up with completely different results for its samples of Fulani/Fula/Peul Y-DNA. I think it is yet impossible to draw any conclusion about the origin and history of this very enigmatic population.

Maju said,
"Can you recall the distribution of Egyptian R1? From memory R1b was like 50% R1b1b2 and 50% R1b* but can't find the source right now. I guess that the R1b* is the same as Sudanese/Cameroonian R1b1*."

Luis et al. (2004) have found equal proportions of R1b1b2-M269 and R1-M173(xR1a1a-M17, R1b1a-M18, R1b1b1-M73, R1b1b2-M269) in their sample of Arabs in Egypt:
3/147 = 2.0% R1b1b2-M269
3/147 = 2.0% R1-M173(xR1a1a-M17, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Wood et al. (2005) have provided the following data on North Africans:

Egyptian (Egypt; n=92; Afroasiatic, Erythraic)
3/92 = 3.3% A3b2-M13
2/92 = 2.2% B2a1a-M152
1/92 = 1.1% E-SRY4064(xE1a-M33, E2-M75, E1b1-P2)
1/92 = 1.1% E1a-M33
2/92 = 2.2% E1b1a-P1(xE1b1a7-M191)
1/92 = 1.1% E1b1a7-M191
8/92 = 8.7% E1b1b1-M35(xE1b1b1a-M78, E1b1b1b-M81)
28/92 = 30.4% E1b1b1a-M78
4/92 = 4.3% E1b1b1b-M81
2/92 = 2.2% F-P14(xG-M201, H1-M52, I-P19, J-12f2, K-M9)
2/92 = 2.2% G-M201
1/92 = 1.1% I-P19
21/92 = 22.8% J-12f2
1/92 = 1.1% K-M9(xL-M20, M1-M4, N1-LLY22g, O-M175, P-P27, T-M70)
7/92 = 7.6% T-M70
1/92 = 1.1% R-M207(xR1-M173)
2/92 = 2.2% R1-M173(xR1a1-SRY10831b, R1b1-P25)
4/92 = 4.3% R1b1-P25(xR1b1b2-M269)
1/92 = 1.1% R1b1b2-M269

Tunisian (Tunisia; n=28; Afroasiatic, Semitic)
4/28 = 14.3% E1b1b1a-M78
10/28 = 35.7% E1b1b1b-M81
13/28 = 46.4% J-12f2
1/28 = 3.6% R1b1-P25(xR1b1b2-M269)

Hassan et al. (2008) have reported a much higher frequency of R1b1-P25 in their sample of Copts in Sudan:

Copts (Agriculturists; n=33; Afro-Asiatic, Ancient Egyptian > Semitic)
5/33 = 15.2% B-M60
2/33 = 6.1% E1b1b-M215(xE1b1b1a-M78)
5/33 = 15.2% E1b1b1a1-V12(xE1b1b1a1b-V32)
13/33 = 39.4% J-12f2(xJ2-M172)
2/33 = 6.1% J2-M172
1/33 = 3.0% K-M9(xL-M11, O-M175, P-M74)
5/33 = 15.2% R1b1-P25

Maju said...

You have missed the 1/24 R1b1-P25(xR1b1b2-M269). You might have intentionally left out the 1/24 R1a1-SRY10831b, but that is technically a subclade of haplogroup R, too.

Was not intentional, just an error.

Anyhow, while the R1b1* found in Namibia might be of African origin, the R1a1 and R1b1b2 are with all likehood colonial, like the R1b1b2 among the Fante (who were in close symbiosis with the European slave traders at Elmina and the other Gold Coast forts).

What I find very interesting about this sample of Hausa from Sudan is that, except for the moderate presence of A3b2-M13 and a single E1b1b1a1b-V32 individual, they lack Y-DNA haplogroups that one would typically find in East Africa, and rather have a Y-DNA pool whose composition resembles that of other Chadic-speaking populations, who live in west-central Africa, especially Nigeria, Cameroon, and Chad...

Kind of complex to analyze but guess they may well represent a west-east migration from the Niger/Chad area, getting Chad clades in the process.

Note that the West Chadic language family, to which the Hausa language belongs, is distributed mainly in Nigeria, and, in fact, the Hausa language is one of the most important languages in that country.

Chadic is Afroasiatic so probably the early Chadic migration was east-west, while Haussa ethnical adscription in Sudan surely comes from a minor west-east back-migration instead. Though admittedly I do not know either about the actual origins of Sudanese Haussa I'd suspect them arriving with the slave caravans leading to Egypt and Arabia (much like the Fulani, I guess).

By the way, most individuals in this sample of Fulani from Sudan are R1-M173(xP25), so they are either R1(xR1b1) or else R1b1 with a reversion of the P25 mutation (which actually is not too rare).

I did not think about that but YSOGG itself warns against using P25, while at the same time (inconsistently) keep the R1b1 category defined only by such unreliable mutation.

Just more confussion... :(

I think it is yet impossible to draw any conclusion about the origin and history of this very enigmatic population.

Not sure at all about the ultimate origins but it seems pretty clear that in the 17th century they became the dominant military and political force of West Africa, conquering the Sahelian part of it from West (Futa Toro and Futa Djallon, their homelands) to East, destroying Haussa historiography and creating a number of kingdoms from the Atlantic to lake Chad that persisted until the French conquest a century ago. Guess that the Sudanese Fulani are an offshot of this expansive wave, motivated by the slave trade and the Muslim religious interest for Mecca.

Another thing is to figure out the origins of the Fulani prior to this expansion.

...

And thanks a lot for reviewing North African R.

Most of it would seem to be the same kind of R1b1* found in Sudan and North Cameroon, what may hint that we are missing (for lack of research surely) an African subclade of R1b extended at both sides of middle and eastern Sahara. It also suggests its spread may be related with Afroasiatic migrations.

Anonymous said...

Maju, I have not read the entire paper, but do they say anything about the origin of H6a? That is my mt haplogroup. I noticed it covers the Iberian Peninsula, but, previous to this, I had always read H6a was very rare in Spain. In fact, I a study where it was only 1% found in Portugal, not Spain. My maternal line comes from the Canary Islands, which means the original bearer probably settled there from Spain, Portugal or perhaps any other nation in Europe.

I had read H6 is one the main H clades in the Arabian Peninsula, then it is also found in the Caucasus (Daghestan), then in some Russian Republic and in Europe in smaller percentages.

Maju said...

I presume you mean paper #1 (Alvárez-Iglesias, Mosquera-Miguel et al. 2009), even if this post mentions several papers.

According to the supplementary material table S3, haplogroup H6a is found in the following sampling areas (H6a/H/total - %H/%total):

1. Galicia 6/124/282 - 4.8/2.1
2. Cantabria 3/83/135 - 3.6/2.2
3. Catalunya 2/39/101 - 5.1/2.0
15. NW Caucasus 3/69/234 - 4.3/1.3
16. Daghestan 3/60/269 - 5.0/1.1
17. Ossetia 2/45/296 - 4.4/0.7
19. Syria 1/28/159 - 3.6/0.6
22. Arabian Pen. 2/52/493 - 3.8/0.4
23. Austria 25/859/2214 - 2.9/1.1

In all cases it's all H6a1, except in the large Austrian sample where there are 4 cases of H6a(xH6a1).

You have the largests apportions in relation with H in North Caucasus and North Iberia, and the largest apportions in relation with total sample in Northern Iberia specifically. But the differences are not extreme.

In paper #2, H6a is found in 3-4% of the following populations: Iberia, Sahara, Moroccan Berbers and Near East.

I'd say this clade looks like it's been hanging around for long and is quite homogeneously distributed through West Eurasia and North Africa. In your particular case, guess it can perfectly be of either Iberian or native Canarian (Guanche, Berber) origin with the highest likehood. No apparent logic in looking to Arabia, Austria or the Caucasus for your ancestral origins but impossible to tell wether it is immigrant from Iberia or native Guanche (Berber).

Maju said...

Correction: in paper #2 H6a is found in 3% in Arabia Peninsula, not "Near East" (where it's only 1%). The percentages in table 1 of this paper are in relation to all H and not to all the total sample.

Anonymous said...

Thank you, Maju, for the information on H6a.

Ebizur said...

To illustrate my previous comment about the lack of a constant pattern in the results of Y-DNA genotyping on samples of Fulbe, here are some more data:

Alexandra Rosa, Carolina Ornelas, Mark A Jobling, António Brehm
and Richard Villems, "Y-chromosomal diversity in the population of Guinea-Bissau: a
multiethnic perspective," BMC Evolutionary Biology 2007, 7:124 doi:10.1186/1471-2148-7-124.

Fulbe, Guinea-Bissau
1/59 = 1.7% A1a-M31
4/59 = 6.8% E1a-M33
1/59 = 1.7% E2-M75
43/59 = 72.9% E1b1a-M2(xE1b1a7-M191)
1/59 = 1.7% E1b1a7-M191
2/59 = 3.4% E1b1b1-M35(xE1b1b1a-M78)
6/59 = 10.2% E1b1b1a-M78
1/59 = 1.7% R1b1-M173+P25

Fulvio Cruciani, Piero Santolamazza, Peidong Shen, Vincent Macaulay, Pedro Moral, Antonel Olckers, David Modiano, Susan Holmes, Giovanni Destro-Bisol, Valentina Coia, Douglas C. Wallace, Peter J. Oefner,5 Antonio Torroni, L. Luca Cavalli-Sforza, Rosaria Scozzari, and Peter A. Underhill, "A Back Migration from Asia to Sub-Saharan Africa Is Supported by High-Resolution Analysis of Human Y-Chromosome Haplotypes," American Journal of Human Genetics 70:1197–1214, 2002.
(Samples originally from Scozzari et al. 1997, 1999)

Fulbe, Burkina Faso
18/20 = 90.0% E1b1a-DYS271(xE1b1a2-M116, E1b1a5-M155, E1b1a7-M191, E1b1a6-M10, E1b1a3-M149, E1b1a1-M58, E1b1a4-M154)
2/20 = 10.0% E1a-M33/M132(xE1a1-M44)

Fulbe, Cameroon
2/17 = 11.8% A3b2-M13/M219
1/17 = 5.9% E1b1-PN2(xE1b1a-DYS271, E1b1b1-M35)
9/17 = 52.9% E1a1-M44
3/17 = 17.6% T-M70
2/17 = 11.8% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Hisham Y. Hassan, Peter A. Underhill, Luca L. Cavalli-Sforza, and Muntaser E. Ibrahim, "Y-Chromosome Variation Among Sudanese: Restricted Gene Flow, Concordance With Language, Geography, and History," American Journal of Physical Anthropology (2008)

Fulani, Sudan
3/26 = 11.5% E1a-M33
1/26 = 3.8% E1b1b1a1b-V32
8/26 = 30.8% E1b1b1a3-V22
14/26 = 53.8% R1-M173(xP25)

Maju said...

Yah totally unrelated, except the two first ones. They are different populations, even if they share language and ethnic identity.

I guess only themselves can explain how that could happen - if anyone can.

My best guess is different founder effects by different subpopulations, plus a good deal of assimilation of people from outside, via adoption or whatever. I guess that the Fulani Empire(s) and slavery played a role in such absorption and change.

Whatever the case, they are not a pure simple ethnicity, stretching over such huge area, and their language is in fact many related ones, and don't seem mutually intelligible over long distances (Ethnologue estimates 7 different languages in fact).

Guess that, mutatis mutandi, we can compare them somewhat to the Turkic peoples, sometimes just called Turks, who spand over a similarly large area but are very different among themselves, especially in their genetics. The spread of the Fulani is much more recent than the Turkic one anyhow.

mathilda said...

..."references"...

In a small voice... I'm so embarrassed I just want to curl up and die.

On the other hand, at least the R1b distribution makes some sense if the m335 isn't in Cameroon. Your comments page is just a mine of information. Dang it, time for a rewrite.

Thank you Ebizur.

Maju said...

Actually my "reference" on that R1 tree was older than your post containing it, Mathilda. Blame on Wikipedia - and on myself for following that trail of pseudo-source.

On the other hand, at least the R1b distribution makes some sense if the m335 isn't in Cameroon.

Doesn't make such a big difference to me because whatever that Sahelian R1b is it's probably some unique specific subclade (not yet defined but just for lack of sufficient research, obviously). Call it "R1b1d" tentatively if you wish, though it cannot be discarded that it is some other known subclade of R1b1, like R1b1a or R1b1c itself.

What seems clear is that it's not "European" R1b1b2 nor "Asian" R1b1b1. It must be therefore another African-specific subclade, derived from the same West Asian core.

Your comments page is just a mine of information.

Ebizur is in fact the miner here. He often provides that kind of great relevant detailed info that we all appreciate.

Maju said...

Notice: I have updated the post to include another relevant paper on the ancient mtDNA of Taforalt (dug at Mathilda's), which includes aboundant H and some U6, JT* and V and is virtually the same to that of the modern Berber inhabitants of that same area.

Ebizur said...
This comment has been removed by the author.
Ebizur said...

Thanks for the compliments. I'm just trying to get people to stick to the raw data as much as possible, lest they should stray too far into the shadowy realm of speculation.

To that end, here is a copy of a recent post on DNA Forums in which all the data on the Y-DNA of Chadic-speaking populations from Cruciani et al. (2002), Wood et al. (2005), and Hassan et al. (2008) have been collected:

QUOTE (Jafety R1b-U152 @ Apr 17 2009, 10:33 AM)
I think Semitic language is the result of J1 contact with E1b1b people (who carried the Afro-Asiatic languages out of Ethiopia or Egypt).
If the Afro-Asiatic language family has originated in a prehistoric population of exclusively haplogroup E1b1b Y-DNA, then why do Chadic speakers (i.e. Hausa, Uldeme, etc.) have haplogroup R1b1-P25, haplogroup B2a1a-M109/M152, haplogroup A3b2-M13, and various subclades of haplogroup E (including E1b1a-M2 and E2b-M54/M98), but practically zero E1b1b? Associating the spread of Afro-Asiatic with haplogroup J would invite the same sort of problem; Cushitic populations have very little haplogroup J Y-DNA (according to some studies, at least), and Chadic populations appear to have none at all.

In my opinion, your hypothesis would only be plausible if we assumed that all of haplogroup E (or at least E1b1) originally spoke proto-Afro-Asiatic, and the E1b1b subset (or, more likely, a subset of the E1b1b subset) of haplogroup E spoke the Semitic subset of Afro-Asiatic, and these Semitic-speaking E1b1b-derived populations eventually mixed with J1-derived (or perhaps just J-derived) populations, who spoke some language(s) of indeterminable affinity, in some indeterminable location.

Here are the data on the Y-DNA of Chadic-speaking populations on which I have based my previous comments:

Hausa (Agriculturists; n=32; Afro-Asiatic, Chadic; Hassan et al. (2008))
4/32 = 12.5% A3b2-M13
5/32 = 15.6% B-M60
5/32 = 15.6% E1a-M33
4/32 = 12.5% E1b1a-M2
1/32 = 3.1% E1b1b1a1b-V32
13/32 = 40.6% R1b1-P25

Podokwo (Cameroon, North; n=19; Afroasiatic, Chadic; Wood et al. (2005))
1/19 = 5.3% E1b1b1a-M78
18/19 = 94.7% R1b1-P25(xR1b1b2-M269)

Mandara (Cameroon, North; n=28; Afroasiatic, Chadic; Wood et al. (2005))
4/28 = 14.3% A3b2-M13
1/28 = 3.6% B2a1a-M152
2/28 = 7.1% E2b-M54
3/28 = 10.7% E1b1a-P1(xE1b1a7-M191)
1/28 = 3.6% E1b1a7-M191
17/28 = 60.7% R1b1-P25(xR1b1b2-M269)

Uldeme (Cameroon, North; n=13; Afroasiatic, Chadic; Wood et al. (2005))
4/13 = 30.8% B2a1a-M152
9/13 = 69.2% R1b1-P25(xR1b1b2-M269)

Ouldeme (n=21; Afro-Asiatic, Chadic; Cruciani et al. (2002))
1/21 = 4.8% B2a1a-M109
20/21 = 95.2% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Daba (n=18; Afro-Asiatic, Chadic; Cruciani et al. (2002))
5/18 = 27.8% E1b1a-DYS271(xE1b1a1-M58, E1b1a2-M116, E1b1a3-M149, E1b1a4-M154, E1b1a5-M155, E1b1a6-M10, E1b1a7-M191)
1/18 = 5.6% E1b1b1a-M78(xE1b1b1a3a-M148, E1b1b1a1a-M224)
4/18 = 22.2% E2b1-M85(xE2b1a-M200)
8/18 = 44.4% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Mixed Chadic (n=15; Afro-Asiatic, Chadic; Cruciani et al. (2002))
1/15 = 6.7% A3b2-M13/M219
1/15 = 6.7% B2a1a-M109
1/15 = 6.7% E1b1a7-M191
1/15 = 6.7% E1b1a-DYS271(xE1b1a1-M58, E1b1a2-M116, E1b1a3-M149, E1b1a4-M154, E1b1a5-M155, E1b1a6-M10, E1b1a7-M191)
1/15 = 6.7% E1b1b1a-M78(xE1b1b1a3a-M148, E1b1b1a1a-M224)
10/15 = 66.7% R1-M173(xR1a1-SRY10831.2, R1b1a-M18, R1b1b1-M73, R1b1b2-M269)

Chadic total (Hassan et al. (2008) + Wood et al. (2005) + Cruciani et al. (2002))
9/146 = 6.2% A3b2-M13
12/146 = 8.2% B-M60, including at least 7/146 = 4.8% B2a1a-M152/M109
5/146 = 3.4% E1a-M33 (only in the Sudanese Hausa sample)
15/146 = 10.3% E1b1a-M2/P1/DYS271
4/146 = 2.7% E1b1b1a-M78
6/146 = 4.1% E2b-M54
95/146 = 65.1% R1-M173(xR1a1-SRY10831.2), including at least 57/146 = 39.0% R1b1-P25 and at least 44/146 = 30.1% R1b1-P25(xR1b1b2-M269)

So, which of these haplogroups has/have been imparted to these populations by the original speakers of the proto-Chadic language?

Maju said...

In my opinion, your hypothesis would only be plausible if we assumed that all of haplogroup E (or at least E1b1) originally spoke proto-Afro-Asiatic, and the E1b1b subset (or, more likely, a subset of the E1b1b subset) of haplogroup E spoke the Semitic subset of Afro-Asiatic, and these Semitic-speaking E1b1b-derived populations eventually mixed with J1-derived (or perhaps just J-derived) populations, who spoke some language(s) of indeterminable affinity, in some indeterminable location.Not exactly but close to my thoughts. E1b1b would be for me the original Afroasiatic speakers. Harifian early pastoralists of the Negev-Sinai area became influenced by them (and surely absorbed many of them too) but remained dominated by local J1. In due time they expanded as Semitic-speakers (via the circumarabian pastoral seminomadic complex of the PPNB period).

I am not sure of the role of E1b1a, which seems to have experienced an unrelated process, probably with a more recent timeframe. Being the most common clade in West Africa, I presume it became associated to the Niger-Kongo language family instead.

So, which of these haplogroups has/have been imparted to these populations by the original speakers of the proto-Chadic language? I'm not sure but it seems to me that African R1b1* may have been leading here (from a Sudanese origin with older, non-AA, West Asian roots). It seems a clear case of founder effect: R1b1 migrated to Sudan in an unclear Paleolithic epysode, it lurked there for many many generations (probably coalescing into a distinct haplogroup, yet to be identified) and, eventually, was "lucky" enough to lead the AA-Chadic expansion into Central Africa.

Other clades may have propagated in parallel too. Can't say for sure because haven't meditated enough in African Y-DNA but the Sudan area is very diverse. E1b1b was one of them almost for sure.

A distinct possibility would be that R1b1* was once common in West Europe (shrinking in diversity at the LGM) but that could have migrated to North Africa at the Oranian genesis epysode, going from there to Sudan and Chad areas. But I find this explanation far less convincing: more unnecesarily complicated.