Eastern Hamites & Nilotes -- No close relations

[Noah]

There's an important new genetic study published earlier this year, Babiker et al. (2011), which has some bearing on Hamitic ethnogenesis. It is an autosomal analysis on 15 STR markers of several populations in Northeast Africa and the Great Lakes region. Various Nilotic groups, as well as Egyptians, Somalis, Copts and Nubians were analysed. The Nilotic populations from both regions clustered together, whereas the Eastern Hamitic and Nubian samples generally clustered separately. It's an especially interesting study because it demonstrates a low genetic affinity between the Eastern Hamites on the one hand, and their Nilotic neighbors that don't have significant Cushitic admixture on the other (such as the Nuba). This divergence has also been confirmed by cranial studies, among other analyses.

Here are some key observations from the study; the essentially West Eurasian autosomal affinities of Egyptians have already been clearly demonstrated across various studies, so I'll focus a bit more here on the understudied Somalis:

• The Egyptian sample shared the greatest number of private alleles with the Coptic sample.
  
• Somalis had the highest mean number of private alleles among the populations studied.
  
• The second largest value for the number of private alleles for population pairs in the study was between 'Arabs' and Somalis. The authors suggest that this may indicate gene flow and/or shared ancestry:
  
  

  "The second largest value for the number of private alleles for population pairs in our study was for the Arab-Somali pair (0.036 ± 0.029; Figure 4), which may be a result of the influence of Arab groups in east Africa as the product of continuous migrations from the Arabian Peninsula across the Gate of Tears over the past three millennia [31]."
  

  
  
• The researchers estimated population structure using the STRUCTURE program, but dismiss the results as misleading mainly due to the low number of markers analysed (the importance of a high number of ancestry informative markers (AIMs) in an analysis is explained further in the Modern Nubians thread):
  
  

  "Although most individuals had substantial parts of their genomes assigned to more than one cluster, three main patterns could be distinguished (Figure 5). Individuals from northern Sudan and Egypt were (generally) more likely to fall within the ‘green’ cluster (P<0.001, Mann–Whitney U-test), these from southern Sudan and Uganda into the ‘red’ cluster (P<0.001,Mann–Whitney U-test), and those from Somalia into a (partly) separate ‘yellow’ cluster (P<0.001, Mann–Whitney U-test). The mixed ancestry of basically all individuals is probably a consequence of the limited number of markers (and of the set of markers not being particularly informative about ancestry) rather than an indication of recent admixture, a behavior of admixture analyses that has been reported previously (see Figure 4 in Rosenberg et al. [29]."

  
  
  The authors therefore ran a Principal Component Analysis (PCA) on the pairwise genetic distance between the study's populations. The first component, shown on the horizontal axis in the graphic below, contained 25% of the genetic variation. It differentiated the Coptic, Egyptian, Somali and Nubian populations from the Nilotic and Hausa populations. The other two components also differentiated the Hamitic and Nubian populations from the Sub-Saharan ones:
  
  
  
  

  "the first principal component explained 25.5% of the genetic variation, and distinguished the Coptic, Egyptian, Somali and Nubian populations from the others (Figure 6A). The second component (10.6%) distinguished the Egyptian, and to some extent, the Somali and the Nubian population from most of the others (Figure 6A), and the third component (9.9%) distinguished the Somali population from all others (Figure 6B). PCA indicated that the Egyptian, Coptic, Somali, and to some extent the Nubian groups form genetically distinct populations. The Nuba, Zagawa, Nilotic and Gemar groups clustered with the Karamoja group from Uganda, as was also indicated by the clustering using Structure (Figure 5)."

  
  

The one exception is the Beja sample, whose members here do cluster with the Sub-Saharan samples due to substantial Nilotic admixture in certain Beja clans. The comparatively unmixed Beja clans, however, ordinarily cluster with other Hamitic peoples (see the Caste systems amongst Hamites thread for a primer on the direct relationship between clan status and biology amongst Hamitic groups). As Carleton Coon explains:

"Of great importance from the standpoint of the history of Hamitic-speaking peoples in North Africa are the various tribal divisions of the great Beja people, who live to the east of the Nile from Eritrea north into Upper Egypt. Some of them now speak Tigré, others Arabic, but their original speech is Cushitic, and their racial relationship seems to be with the Somalis and Danakils for the most part. Some of them, such as the Haddendoa, have been largely mixed with Sudanese negroes; the less mixed, such as the Beni Amer in northern Eritrea, and the Bishari in the Egyptian desert, represent a fairly uniform type which Seligman compares to the predynastic Egyptians."

The authors also seem surprised by the general lack of affinities between their Somali sample and the studied Sub-Saharan groups. So they hypothesize at one point that this might be due to divergent Bantu admixture in the Somalis, citing Swahili as "one of the major languages in Somalia"! ;D Of course, this is a preposterous suggestion since:

a) Swahili is actually not a major language of Somalia; the Afro-Asiatic Somali and Arabic languages are. Swahili is a minority language spoken by some of the country's ethnic minorities in the southern part of the country.

b) The Sub-Saharan admixture in Eastern Hamitic peoples in general is mainly from Nilo-Hamitic females, not Bantu women (see the Skin color in Eastern & Western/Northern Hamites thread for a detailed explanation).

c) Autosomal tests have consistently shown Bantus as clustering closely with other Sub-Saharan populations, including Nilotes. Only the Fulani (who, at least many of the ones in Sudan, have significant Eurasian paternal ancestry) or sometimes Nilo-Hamitic groups like the Maasai (who have lots of Cushitic admixture per Tishkoff et al. (2009)) are positioned somewhat apart.

d) Y DNA, mtDNA, autosomal DNA, HLA antigens, cranial traits, dental traits and general morphology analyses have all demonstrated a low affinity between Somalis and Horners on the one hand, with Bantu groups on the other (that's not even counting diagnostically significant factors such as Duffy positivity, hair texture and the sickle cell trait).

Y DNA:

"The time of the eastbound Bantu expansion was estimated to be 3400plusminus1100 years ago.24 Bantu populations have high frequencies of E3a haplogroups.4 We have observed only a few individuals with the E3a haplogroup in our Somali population, thus, supporting the view that the Bantu migration did not reach Somalia.42 It has been suggested that a barrier against gene flow exist in the region.43 The barrier seems to be the Cushitic languages and cultures to which Somalis belongs. The Cushitic languages belong to the Afro-Asiatic languages that are spoken in Northern and Eastern Africa."

www.nature.com/ejhg/journal/v13/n7/full/5201390a.html

mtDNA:

"We analysed mtDNA variation in ~250 persons from Libya, Somalia, and Congo/Zambia, as representatives of the three regions of interest. Our initial results indicate a sharp cline in M1 frequencies that generally does not extend into sub-Saharan Africa. While our North and especially East African samples contained frequencies of M1 over 20%, our sub-Saharan samples consisted almost entirely of the L1 or L2 haplogroups only. In addition, there existed a significant amount of homogeneity within the M1 haplogroup. This sharp cline indicates a history of little admixture between these regions. This could imply a more recent ancestry for M1 in Africa, as older lineages are more diverse and widespread by nature, and may be an indication of a back-migration into Africa from the Middle East."

konig.la.utk.edu/AJPA_Suppl_40_web.htm

Autosomal DNA:

HLA antigens:

"HLA antigens of the Somali population are not categorised as well as those of other international ethnic groups. We analysed the HLA antigens of 76 unrelated Somalis who lived in the west of England. HLA -A, -B, -C and DRB1 typing was performed by polymerase chain reaction using sequence-specific oligonucleotide probes (PCR-SSOP) at a low-intermediate resolution level. Phenotype frequency, gene frequency and haplotype frequency were used to study the relationship between Somalis and other relevant populations. The antigens with highest frequencies were HLA -A1, A2, and A30; B7, B51 and B39; Cw7, Cw16, Cw17, Cw15 and Cw18; DR 13, DR17, DR8 and DR1. HLA haplotypes with high significance and characteristics of the Somali population are B7-Cw7, B39-Cw12, B51-Cw16, B57-Cw18. The result of HLA class I and class II antigen frequencies show that the Somali population appear more similar to Arab or Caucasoid than to African populations. The results are consistent with hypothesis, supported by cultural and historical evidence, of common origin of the Somali population."

onlinelibrary.wiley.com/doi/10.1111/j.1365-3148.2006.00694_52.x/abstract

Cranial traits:

Dental traits:

General morphology:

These rather silly speculations notwithstanding, the authors do offer an alternate, correct interpretation of the data by suggesting that Somalis, the Beja and other related populations are of Eurasian origin (Hamitic), but with Sub-Saharan admixture:

"Another explanation could be that the Somali population is of both Eurasian and sub-Saharan origin, as suggested by a recent study [33], potentially explaining the differentiation of this population from some east African groups, although many of the Sudanese populations, such as Arabs and the Beja, may also have mixed Eurasian and sub-Saharan origin."

All in all, the study's actual results are quite illuminating. They demonstrate, among other things, that Nilotes -- the most long-standing Negroid neighbors of Hamitic peoples -- are, like Bantus and other Sub-Saharan peoples, not particularly closely related to Hamitic folks. As with African Americans vis-a-vis White Americans, Nilotes/Bantus first have to have experienced Hamitic and/or West Eurasian admixture (either directly through intermarriage with Hamites or indirectly through interbreeding with a Hamitic/Eurasian admixed intermediary group) for there to be any significant ties between themselves and Hamitic people (or vice versa). This is precisely what one would expect from biologically distinct population groups.

By showing that Somalis (who mostly belong to the E-M78 paternal clade that is believed to have originated in Egypt/Libya) are autosomally distinct from neighboring Negroid peoples and instead share closer biological ties with Arabs -- just like their haplogroup J-carrying relatives in parts of Ethiopia and Sudan do (folks with whom they also share Eurasian mtDNA and group with maternally) -- the study also serves as yet another piece in the puzzle demonstrating the essentially West Eurasian affinities of the original haplogroup E bearers.

[amun]

Well, compared to the other 'Negro' types of Africa the Nilotes from Sudan are the closest to Eastern Hamites due to shared East African ancestry (relative to divergent West Africans and Paleo-Africans). The two groups are distant, but there's a link.

Many peripheral Nilotic groups are also heavily Hamitic mixed (especially the Maasai and Kunama).

The ones in South Sudan are rather pure though.

[Noah]

Yes, Nilotes with Cushitic admixture -- the so-called "Nilo-Hamites" like the Maasai and Samburu -- cluster somewhat apart from other Negroid groups. They possess some Eurasian autosomal affinities that are responsible for that divergence (~20% in the MKK Maasai sample below); and these Eurasian autosomal affinities were genetic contributions from Hamitic people.

Hamitic admixture is also the source of these Nilotes' less broad, mesorrhine nasal indices, lower prognathism, lighter skin color, etc..

"A few Negro tribes of northeast Africa, namely, the Baganda, Akamba, and Akikuyu, show a reduction of the platyrrhine condition which is characteristic of Negroes, especially the far western groups. Hamitic blood probably affected this trait, for in the Hamiticized Masai the nasal index is distinctly mesorrhine (76.2)."

books.google.ca/books?id=o6W0AAAAIAAJ&q=%22Hamitic+blood+probably+affected+this+trait,+for+in+the+Hamiticized+Masai%22&dq=%22Hamitic+blood+probably+affected+this+trait,+for+in+the+Hamiticized+Masai%22&hl=en&ei=p1_LTojzNsbv0gHpmoUx&sa=X&oi=book_result&ct=result&resnum=1&ved=0CC4Q6AEwAA

Had these Nilo-Hamites not experienced that Hamitic admixture/genetic contribution, they would be about as phenotypically and genetically related to Hamitic people as are their comparatively unmixed Nilotic kinsmen in south Sudan i.e. not particularly close at all.

Among all the Nilotes, the Maasai probably have the most Hamitic admixture. About 50% of them carry the paternal haplogroup E1b1b, which they acquired from Hamitic males. The other 50% belong to typical Sub-Saharan clades, especially the Nilotic A3b2 lineage. 80%-85% of the Maasai belong to a variety of Sub-Saharan L(xL3) mtDNA haplogroups. Only 15%-20% have proto-Eurasian (L3) or Eurasian maternal clades, which, again, they got from Hamitic people.

In terms of autosomal DNA, the Maasai largely belong to their own Nilo-Hamitic cluster. This cluster appears as an admixture component in many Hamitic and Middle Eastern populations because Nilo-Hamitic females are the main source of the Sub-Saharan admixture in said West Eurasian-descended populations.

We know from recent studies that the Sub-Saharan admixture component in most Hamitic and Middle Eastern populations is Nilo-Hamitic in affinities. For instance:

"Approximately 30% of mtDNA lineages in South Arabian samples are African L haplotypes, whose origin has usually been attributed to migration and assimilation of African females into the Arabian population over approximately the last 2,500 years. Few In contrast, few Y chromosome lineages of clear recent sub-Saharan African origin have been found in Southern Arabian populations. This bias in maternal and paternal lineages is in accord with historical accounts of the female bias in the Middle Eastern slave trade. In order to evaluate autosomal African ancestry, we collected high-resolution SNP genotype data from a geographically representative set of 62 Yemenis selected from a collection of 552 samples acquired in the Spring of 2007. The ancestry of chromosomal segments in the Yemeni population was estimated using a haplotype-based local ancestry estimation method, HAPMIX. The HAPMIX method is based on a two way admixture model that requires two phased reference populations; we used the HapMap Yoruba in Ibadan, Nigeria (YRI), Luhya in Webuye, Kenya (LWK), Maasai in Kinyawa, Kenya (MKK), and CEPH US residents with ancestry from northern and western Europe (CEU) samples. The three African reference populations include two Bantu-speaking groups (YRI and LWK) and one Nilotic-speaking group (MKK). We estimated local ancestry in the Yemeni sample with all three European-African reference population combinations (CEU-YRI, CEU-LWK, CEU-MKK). The correlations among African ancestry calculated using all three reference population combinations are high (r > 0.98 in all pairwise correlations). Furthermore, there is no significant difference between the average proportion of African ancestry in Yemenis calculated using either of the two Bantu-speaking reference populations: CEU-YRI (mean 0.062, sd 0.044) and CEU-LWK (mean 0.076, sd 0.049) (p=0.13, two-tailed Welch two sample t-test). However, the average African ancestry calculated using the Maasai reference population (CEU-MKK, mean 0.148, sd 0.060) is significantly greater from that calculated using either the Yoruba or Luhya reference populations (p less than 0.0001 in both comparison, two-tailed Welch two sample t-test). These data suggest that the source population for the African ancestry of the Yemeni population is more similar to the contemporary Maasai population than either the Luhya or Yoruba."

www.ashg.org/cgi-bin/2010/ashg10s?author=R.%20L.%20Raaum&sort=ptimes&sbutton=Detail&absno=20343&sid=189644

So how did that Nilo-Hamitic component get so diffused in the first place? How, for example, did it wind up at once in the Horn, North Africa and the Near East? Answer: through the pre-Islamic slave trade.

In the historical literature, the Nilo-Hamites are labeled "slaves of the better sort", and they were preferred for concubinage. The regular type of Sub-Saharan slaves were Bantu/West African and they mainly worked the fields, so they left a negligible genetic impact on their host populations (the men were also often castrated). The Barbar/Berber -- the immediate ancestors of the Somali, Afar, Saho, etc. of the Bilad al Barbar in the northern half of the Horn of Africa -- are described as extensively trading in these "slaves of the better sort". The Barbar/Berber merchants shipped the aforementioned Nilo-Hamitic peoples to many of their trading partners in North Africa and the Middle East, such as Egypt and Yemen.

"The Periplus of the Erythraean Sea, an anonymous Egyptian merchant's guidebook of the second or third century ad, carefully catalogues the products and the entrepots important in the commerce of the Indian Ocean. From Adulis, Axum's port on the Red Sea, to Berbera on the southern littoral of the Gulf of Aden, the indigenous products included ivory, tortoise shell suitable for inlaid furniture and woodwork, obsidian (a dark volcanic stone valued for statues and votive offerings), spices, and myrrh. Berbera supplied true frankincense, a hard cinnamon, Indian copal, macir (an aromatic bark), and, "but rarely," slaves. East of Berbera merchants could buy good cassia and frankincense, fragrant gums, spices, some tortoise shell, ivory, and myrrh. Beyond Cape Guardafui, the so-called Cape of Spices, Opone (probably near the modern Ras Hafun) was the first major port. Here the local merchants offered large quantities of cinnamon, tortoise shell, and "slaves of the better sort," who were sent to Egypt."

books.google.ca/books?id=nAhBAAAAIAAJ&q=%22East+of+Berbera+merchants+could+buy+good+cassia+and+frankincense%22&dq=%22East+of+Berbera+merchants+could+buy+good+cassia+and+frankincense%22&hl=en&ei=nmzLTsuoCMHs0gHWrsD7Dw&sa=X&oi=book_result&ct=result&resnum=1&ved=0CC4Q6AEwAA

Those same areas in the Red Sea and Indian Ocean region where the Nilo-Hamitic slaves were shipped from and to are the same areas where the Nilo-Hamitic element today constitutes most of the minority Sub-Saharan admixture component.

[amun]

The African (non-Eurasian) genetic affinities in the Horn are spread equally within populations. There's a minor Northeast-to-Southwest gradient, but within local populations there isn't much difference individually. This makes slave trade scenarios highly unlikely. The East African affinities were probably simply absorbed early on as Hamites moved into the Horn from either Egypt or the Arabian Peninsula.

I disagree that slave-trade related African ancestry in the Arabian Peninsula is predominantly Nilo-Hamitic. Several studies on Arabians have shown that there is a minor Bantu presence in their mtDNA, which matches sequences from Mozambique best. Particularly in places like Oman and Kuwait which had strong Swahili contacts it's obvious. However, you can still find many tribes in said region with 0% slave-trade related/Bantu ancestry, which is a classical characteristic of recent gene flow events (strong individual differences).



The West African cluster denotes Bantu ancestry, which is 0% in Afro-Asiatic Horners and also relatively low in Nilo-Hamites. While the Saudi outlying individual clearly has admixture from Southeast African Bantus. Those couple of Saudis who have stronger West Asian ancestry might be Syrian admixed, and those nearly completely composed of Southwest Asian are probably the true natives (most likely of isolated Bedouin origin).

[Noah]

Good observations. I agree that the relatively uniform distribution of the East African component within Horner sub-populations does not suggest recent admixture. Given that, it does seem more likely that the Hamites instead picked that element up as they entered the Horn rather than later on through involvement in the slave trade.

I also agree that the Sub-Saharan admixture in certain parts of the Middle East is Bantu-associated. However, in other areas like Yemen, a different admixture component more similar to that which characterizes Nilo-Hamitic groups like the Maasai seems to exist.

For instance, in the chart below, that admixture component constitutes almost all of the Maasai MKK25 sample's ancestry. But it also exists as a secondary admixture component in the Ethiopian, Yemeni and Saudi samples.

Since we already know that the Maasai have notable Eurasian affinities due to Hamitic admixture, that blue admixture component that characterizes them is thus also a mixed Sub-Saharan/Eurasian cluster.

That's the admixture component I'm referring to. You explained its presence in the Horn as being due to the Hamitic migrants absorbing an authocthonous East African population early on. This seems to be the case. However, this scenario doesn't really explain the component's presence in the Middle East. I think this is perhaps where the slave trade does factor in. Any thoughts on this?

[amun]

That particular run lacked many crucial samples. It gives the incorrect image that the Maasai are unadmixed (since there are no pure(r) Nilotes or Bantu included) and that the African ancestry in those Arabian samples is solely from the Horn region.

There are outliers among those Saudi and Yemeni samples (like the Saudi one I showed earlier) who have ancestry from Bantus. Finding Arabians with strong recent Horner-/Nilo-Hamitic-like ancestry is very rare. I haven't come across any so far among those public genomes. Most of the African slave-trade related ancestry in the Arabian Peninsula comes from Bantu or Southern Sudanese types. Just look up the 'Akhdam' in Yemen, it's very obvious that those people didn't get their African ancestry from the Horn nor Nilo-Hamites, but instead from Bantus or other broad types.

[Noah]

I agree that that run lacks some important samples, such as a Bantu/West African population for comparison. But even when the Yoruba are added and Bantu/West African admixture is detected in the Maasai, as in the run below, the Maasai Nilo-Hamites still, albeit by a very slim margin, have the highest proportion of that 'East African' component in the dataset:

It's clear from the Maasai column above that that 'East African' component is a stabilized mixture of earlier Sub-Saharan and West Eurasian elements since other autosomal analyses, like the table I posted above, have consistently shown the Maasai as being 20%-30% Eurasian. Some of those Eurasian elements are already contained within the mixed light purple 'East African' component. This can be confirmed by looking at the other populations in the run that also have that component. Like the Maasai, various autosomal analyses have shown all of these groups as possessing significantly higher Eurasian affinities than what superficially appears here i.e. those Eurasian affinities were likewise 'eaten up' by/stored within the East African component.

The 'East African' component is thus actually a consolidated fusion of Sub-Saharan and West Eurasian elements. As we know, discrete population elements can over time combine to form new stabilized components of their own. And this is what we're seeing here.

Regarding the Achdam, it's certain that their conspicuously Negroid physical characteristics have nothing to do with Horners since, generally speaking, Horners physiologically fit comfortably within the Mediterranean range on a variety of traits (low nasal index, orthognathism, Caucasoid craniofacial pattern, etc.). The Achdam, on the other hand, don't; they're also not native to the Middle East:

"We examined two groups of people in Southern Arabia: the Arabs themselves (iii), and the Achdam (I04) and we found the sickle cell trait among the latter at a high frequency[...] The Africans who live nearest to Arabia are the Somali. They are a very proud people who regard themselves as of Arab origin, and it is inconceivable that any of them, or their descendants, should have fallen to the low social level of the Achdam. They have no sickle cells (we examined fifty of them and found none) as one would expect from an African population of the Hamitic type. Thus the Achdam are unlikely to have descended either from imported African slaves or from the Somali. There is a possibility that they are the last remains of pre Mohammedan Abyssinian camp-followers. The armies of Abyssinia finally expelled in the early days of Mohammedan rule, had with them Sudanese and it is just possible that some of these might have been left behind. If the Achdam were derived and had inherited their sickling gene from such Sudanese Africans, they might be expected to resemble the Nilotic speaking tribes with high sickling frequency."

books.google.ca/books?id=BCTVAAAAMAAJ&q=%22We+examined+two+groups+of+people+in%22&dq=%22We+examined+two+groups+of+people+in%22&hl=en&ei=h4XNTpbJG6TX0QG73LAs&sa=X&oi=book_result&ct=result&resnum=1&ved=0CC8Q6AEwAA

The Sub-Saharan ancestry in authocthonous Middle Easterners therefore doesn't come from Horners (who, in any case, are originally descendants of Hamitic peoples of Proto-Mediterranean origin) or from the Achdam (who are admixed southern Sudanese Nilotes). Depending on the population studied, that admixture component is either a) Bantu-related, as you pointed out; or b) Nilo-Hamitic in structure, as I pointed out. With regard to indigenous Yemenis (not the Achdam), there is a presence of both Bantu and Nilo-Hamitic-linked uniparental lineages. However, when compared together, the Sub-Saharan influence in Yemen was more closely associated with the Nilo-Hamites.

"Phylogenetic analysis reveals that the origin of sub-Saharan African mtDNA variants in Yemenis is a mosaic of different episodes of gene flow. Three different passages can be outlined. The first is gene flow, likely mediated by the Arab slave trade from southeastern Africa, as evidenced by exact mtDNA haplotype matches. Such matches account for 23% of the total variation in Yemenis and occur in lineages and lineage groups that cannot be found in Ethiopia and northeastern Africa. Many of these can be traced to the Bantu dispersal; they have their origin in West Africa and supply thereby the upper time limit of 3,000–4,000 years for their departure from southeastern Africa toward Arabia[...] the low level (9%) of haplotype sharing within haplogroups L0–L6 between Ethiopians and Yemenis also shows that a directional eastward maternal gene flow, even during the peak of the bipartite Aksum kingdom, must have been relatively minor."

www.africandna.com/ScienPapers%5CEthiopian_Mitochondrial_DNA_Heritage.pdf

"Furthermore, a majority of the L1–L3A lineages in the Hadramawt—such as members of L2a, L2d, L3b, and L3d—trace back ultimately to West Africa, so that it is likely that they were delivered to East Africa by the Bantu dispersals. Supporting this suggestion, all of the L2a types in the Hadramawt occur at elevated frequency in the Bantu speakers of Mozambique (Pereira et al. 2001; Salas et al. 2002). Moreover, the chief L1a type in the Hadramawt also occurs at elevated frequency in Bantu speakers and is implicated in the Bantu dispersals, albeit having been picked up in East Africa en route (Salas et al. 2002)."

www.ncbi.nlm.nih.gov/pmc/articles/PMC1180338/

"Approximately 30% of mtDNA lineages in South Arabian samples are African L haplotypes, whose origin has usually been attributed to migration and assimilation of African females into the Arabian population over approximately the last 2,500 years. Few In contrast, few Y chromosome lineages of clear recent sub-Saharan African origin have been found in Southern Arabian populations. This bias in maternal and paternal lineages is in accord with historical accounts of the female bias in the Middle Eastern slave trade. In order to evaluate autosomal African ancestry, we collected high-resolution SNP genotype data from a geographically representative set of 62 Yemenis selected from a collection of 552 samples acquired in the Spring of 2007. The ancestry of chromosomal segments in the Yemeni population was estimated using a haplotype-based local ancestry estimation method, HAPMIX. The HAPMIX method is based on a two way admixture model that requires two phased reference populations; we used the HapMap Yoruba in Ibadan, Nigeria (YRI), Luhya in Webuye, Kenya (LWK), Maasai in Kinyawa, Kenya (MKK), and CEPH US residents with ancestry from northern and western Europe (CEU) samples. The three African reference populations include two Bantu-speaking groups (YRI and LWK) and one Nilotic-speaking group (MKK). We estimated local ancestry in the Yemeni sample with all three European-African reference population combinations (CEU-YRI, CEU-LWK, CEU-MKK). The correlations among African ancestry calculated using all three reference population combinations are high (r > 0.98 in all pairwise correlations). Furthermore, there is no significant difference between the average proportion of African ancestry in Yemenis calculated using either of the two Bantu-speaking reference populations: CEU-YRI (mean 0.062, sd 0.044) and CEU-LWK (mean 0.076, sd 0.049) (p=0.13, two-tailed Welch two sample t-test). However, the average African ancestry calculated using the Maasai reference population (CEU-MKK, mean 0.148, sd 0.060) is significantly greater from that calculated using either the Yoruba or Luhya reference populations (p less than 0.0001 in both comparison, two-tailed Welch two sample t-test). These data suggest that the source population for the African ancestry of the Yemeni population is more similar to the contemporary Maasai population than either the Luhya or Yoruba."

www.ashg.org/cgi-bin/2010/ashg10s?author=R.%20L.%20Raaum&sort=ptimes&sbutton=Detail&absno=20343&sid=189644

[amun]

Among common Yemenis, i.e. non-recently admixed ones, the African component is usually just strictly from the Horn region. Usually tribes who did not engage in the slave trade, like Temani Yemenites show this tendency of all their minor African affinity coming from the Horn and nowhere else. I believe those Yemenis sampled in the last study (Raaum et al.) you cited are of these types. While among certain sects who engaged more in the slave-trade you find strong Bantu or Southern Sudanese strains in some individuals.

What do you think of this cluster labeled 'Southern' in this latest admix run:

dodecad.blogspot.com/2011/10/eurasia7-calculator.html

Is it the unifying Hamitic cluster? It seems very high in Berbers and Arabians.

[Noah]

The Sub-Saharan mtDNA sequences in Arabia that trace back to the slave trade are typically found in urban centers. The isolated tribal Arab groups tend to carry fewer such lineages because they didn't engage nearly as much in that Peculiar Institution. Note that some of the Proto-Eurasian haplogroup L3's subclades likely also have non-African origins:

"With respect to the L sequences, it is widely accepted that they have a sub-Saharan origin, excepting some L3* lineages that, as analysis of Figure 4 suggests, might indeed have a non-African origin."

www.backintyme.com/admixture/pereira01.pdf

That said, exact mtDNA matches that Somalis and other Northeast Africans share with Yemenis in particular are mainly Eurasian:

"The most frequent haplotype in west coastal Yemen is 16126–16362, which is found not only in the Ethiopian highlands but also in Somalia, lower Egypt and at especially high frequency in the Nubians. The Tihama share some West Eurasian haplotypes with Africans, eg J and K with Ethiopians, Somali, and Egyptians."

www3.interscience.wiley.com/journal/117899911/abstract

I'm guessing you were perhaps confused by those two studies from Watson et al. from the mid-1990s, which show a high frequency of L haplotypes in their shared "Somali" sample from Kenya? Cause Watson et al. mislabeled as "L3" many Eurasian haplogroups that we now know that Somalis, like other Horners, do in fact carry at a high frequency. This is because the methodology that the authors used back then didn't detect many more downstream Eurasian M and N mtDNA haplogroups that descend from L3. Since haplogroup L3 in their sample represented well over half of the reported lineages, that's quite a few misrepresented mtDNA sequences. Watson et al.'s inaccurate data was also referenced by later studies using the same sample set; papers such as Cerny et al. (2004). It wouldn't be until 1999, two years after Watson's 1997 study was published, that these Eurasian lineages would first be detected in the Horn by Quintana-Murci et al.. Richards et al. (2006) describe the situation thusly:

"There was a complication. The analysis of Watson et al. (1997) was based largely on control-region sequence data, which fails to resolve many mtDNA haplogroups. By targeting newly identified coding-region variants, Quintana-Murci et al. (1999) distinguished two major clades in non-Africans, within haplogroup L3. One of these had already been identified as the Asian super-haplogroup M (Torroni et al. 1994a); Quintana-Murci et al. (1999) showed that all other non-African L3 lineages fell into a second major clade, later named haplogroup N. Haplogroups M and N appear to be almost identical in age, estimated from complete coding-region sequences to be about 63000–69000 (±5000) years old (Macaulay et al. 2005; see also Forster et al. 2001; Kong et al. 2003; Forster 2004). Moreover, haplogroup M was present at high frequency in Ethiopia and Somalia, in addition to its known distribution from the Near East to East Asia and Australasia—and amongst Native Americans"

books.google.ca/books?id=9b2XRmhC2ssC&pg=PA235#v=onepage&q&f=false

Luckily, though, there are at least three sound mtDNA studies on ethnic Somalis that do exist.

The first is Comas et al. (1999), which indicates the following:

"Somali, as a representative East African population, seem to have experienced a detectable amount of Caucasoid maternal influence[...] Our results agree with the hypothesis of a maternal influence of Caucasoid lineages in East Africa, although its contribution seems to be higher than previously reported in mtDNA studies."

www.nature.com/ejhg/journal/v7/n4/pdf/5200326a.pdf

The last part of the quote above is a reference to Watson et al.'s two earlier studies and their mislabeled haplogroup L3 sequences. Comas et al. quantify the number of Somalis in their sample set that carry Caucasoid maternal lineages at about half of them, which is pretty typical in the Horn. That's also not counting the others that carry Proto-Eurasian (and possibly non-African) L3 lineages. Comas et al. specify that among those Caucasoid lineages in their Somali sample set was the Cambridge Reference Sequence (CRS), which is linked with the West Eurasian haplogroup H that is especially frequent amongst the Hamitic Tuareg of Libya (and which descends from the Southwest Asian R0a clade that is also common in the Horn):

"the presence of CRS mtDNA haplotypes (with different lengths of Cs at position 309) are typical European mtDNA haplotypes"

The second legitimate mtDNA study on Somalis is Holden (2005). This paper likewise clusters the Hamitic Somali and Libyan samples together, separate from the Sub-Saharan sample:

"We analysed mtDNA variation in ~250 persons from Libya, Somalia, and Congo/Zambia, as representatives of the three regions of interest. Our initial results indicate a sharp cline in M1 frequencies that generally does not extend into sub-Saharan Africa. While our North and especially East African samples contained frequencies of M1 over 20%, our sub-Saharan samples consisted almost entirely of the L1 or L2 haplogroups only. In addition, there existed a significant amount of homogeneity within the M1 haplogroup. This sharp cline indicates a history of little admixture between these regions. This could imply a more recent ancestry for M1 in Africa, as older lineages are more diverse and widespread by nature, and may be an indication of a back-migration into Africa from the Middle East."

konig.la.utk.edu/AJPA_Suppl_40_web.htm

The third legit mitochondrial study on Somalis is from this year, Stefflova et al. (2011). Using SAMOVA genetic analysis, the paper in turn groups the Northeast African samples together (Somalia, Eritrea, Ethiopia, Sudan, Egypt) and apart from the various Sub-Saharan samples:

"Africa was divided based on geography combined in a few cases with ethnicity (in the case of Pygmy hunter gatherers and Bantu speakers). SAMOVA was used to eliminate the following outlier groups: West Pygmy (Cameroon, C.A.R., and Gabon), Khoisan speakers (South Africa and Botswana), East Pygmy group Mbuti (D.R.C.), and Moroccan sample (mainly Berbers from North Africa). After removing outliers, the remaining states were divided using SAMOVA into 4 groups: WC/SW Bantu speakers from Angola, Cameroon, Gabon and Equatorial Guinea, E/SE sample from Kenya and Mozambique Bantu speakers, NE/E sample from Egypt, Sudan, Eritrea, Ethiopia, and Somalia, and W/WC countries after excluding Bantu speakers and Pygmy hunter-gatherers[...] MDS plot showing the mtDNA variation-based genetic distances between African populations after the outliers[...] were excluded from the calculations. This plot shows the general structure of the remaining regions (with highlighted W/WC clustering) and their relationship to the admixed African American populations, depicted using only the African portion of their ancestry."

www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0014495

Unfortunately, the Watson et al. papers aren't the only misleading studies that have been published on Hamitic people in general or on Somalis in particular.

Like Watson et al., another mtDNA study, Stevanovitch et al. (2004), also mislabeled as "L3" what were in actuality various Eurasian haplogroup X matches shared between Tigrayans and fellow Hamites in Egypt. As Kivisild et al. (2004) explains:

"Two Ethiopian haplogroup X sequences from this study have been characterized elsewhere as belonging to North and East African–specific subclade X1 (Reidla et al. 2003). A control-region sequence similar to the Tigrai X1 haplotype was found recently in a Gurna sample from Egypt, though it was probably mislabeled as “L3” by the authors, since no coding-region markers specific to either haplogroup X or L3 were determined in that study (Stevanovitch et al. 2004)."

www.ncbi.nlm.nih.gov/pmc/articles/PMC1182106/

There was another study from 1987 by Sistonen et al., this time on blood groups, which claimed that its "Somali" sample showed few Caucasoid characteristics. However, the "Somalis" in that study refers to the general population of Somalia, not just to people from the Hamitic Somali ethnic group that the country was named after (the term "Somali" also denotes nationality, and Somalia is a multiethnic nation). That's why the frequencies that the study's authors reported for several blood group systems were sometimes off by as much as 20% relative to those observed in previous studies on both ethnic Somalis and their Hamitic kinsmen in the Horn. The authors do admit, though, that there results "are markedly in contrast with some earlier findings" -- hardly a revelation.

"The results of a population survey on blood group distribution in Somalia, East Africa, are presented. Over 1,000 subjects were tested for most blood groups included in the survey. The sampling covered the whole country and was well in accordance with the population density as estimated by the recorded birth places of the subjects."

content.karger.com/ProdukteDB/produkte.asp?Aktion=ShowAbstractBuch&ArtikelNr=153722&ProduktNr=237568

Here are some examples of such earlier blood group studies that, contra Sistonen et al.'s "Somali" national sample, did repeatedly observe notable Caucasoid affinities in their actual ethnic Somali samples.

"Two of the four Hamitic groups show considerable Negroid admixture, sufficient in fact to place the Hausa (15.78) and Tuareg (12.87) solidly among Negroid populations. This admixture appears to be very slight among the Somali (8.64) and altogether lacking in the Berber (2.87). The Berber sample is too small to be of much significance (4 genes), but the positions of the other Hamitic groups seem to be explained by Caucasoid origin and various degrees of secondary Negroid admixture."

www.jstor.org/stable/2739971

"In Rhesus and other systems, however, it is clear that none of the major contemporary Kenyan populations can be classified as Ethiopid, except for the Oromo and Somali. Neither Luo, nor Masai, nor Kikuyu, nor Kamba emerge as anything but ordinary members of Nilotic and Bantu clumps. Only Masai shows some tendency to deviate"

books.google.ca/books?id=OmZ1AAAAMAAJ&q=%22In+Rhesus+and+other+systems%22&dq=%22In+Rhesus+and+other+systems%22&hl=en&ei=lZjPTs3BE4eKgwfxnZXkDQ&sa=X&oi=book_result&ct=result&resnum=1&ved=0CC8Q6AEwAA

"On the basis of full grouping, the peoples of the Near East fall into two main classes: the Turks and Eti-Turks, and the Arabians. The Turks and Eti-Turks, seem to be related mainly with Europe, and especially with the Mediterranean area. They have a high A gene frequency, moderately high M gene frequency, along with a high R1, and fairly high rh-negative frequency. "Arabians" a sub-class of the Mediterranean race, include the Yemenite Arabs and Jews, the Zabidi Arabs, and the Socotrans, and with "reservation", the Somalis. They all have low A and high 0 gene frequencies, but high Rl, and rather lower rh-negative gene frequencies. They have also a very high M frequency which is distinct from the high M of Asia. It is significant to note that the Berbers, the "White" race of N. Africa, resemble the Arabs so far the high 0 frequency is concerned, but differ from them in having high N and Rh negative frequencies respectively."

books.google.ca/books?id=mfkZAAAAIAAJ&q=%22The+Turks+and+Eti-Turks,+seem+to+be+related+mainly+with+Europe%22&dq=%22The+Turks+and+Eti-Turks,+seem+to+be+related+mainly+with+Europe%22&hl=en&ei=0SXPTqWoHujt0gGOtPBL&sa=X&oi=book_result&ct=result&resnum=1&ved=0CC8Q6AEwAA"

"In two respects the Beja differ from most of the surrounding peoples, namely in their low frequencies of B and M. In their low frequency of B they resemble the southern Arabs and the Somalis, though they differ from both in their much higher frequency of all the typical African marker genes. In their lower frequency of M they resemble, on the other hand, some of the African peoples of north-east Africa who possess few of the typical Mediterranean genes, but also the Somalis, who possess many such genes."

www.jstor.org/pss/2798505

Now... back to the subject at hand. The "Southern" cluster in that Dodecad run is indeed a Hamitic cluster. However, it is not an all-unifying one because it only applies to the Later Proto-Hamite pastoralists (the gracile herders depicted on the Saharan & Horn rock art), not to the earlier Paleo-Hamite hunter-gatherers (the more robust, taller Kenya Capsians and Iberomaurusians). As explained in the Paleo-Hamites and Proto-Hamites and Near Eastern origin of the Paleo-Hamites & Capsian threads, there have been several waves of Hamitic/West Eurasian immigration into Africa, not just one great big burst of "Hamitic wanderers", as caricatured by anti-Hamitic commentators. Without this understanding, it is difficult to make sense of and correlate, on the one hand, the various uniparental lineages and autosomal DNA clusters with, on the other, the manifold prehistoric archaeological cultures and the skeletal remains associated with their makers. With that understanding, by contrast, it's surprisingly easy to do so.

Have a look at the Egyptians & Proto-Mediterraneans thread. It associates some of the various West Eurasian autosomal components in Behar et al.'s seminal 2010 paper with Mediterranean elements, citing some other studies along the way.