Talk:Y-DNA haplogroups in populations of Sub-Saharan Africa

Nilo-Saharans & E1b1b in Hassan et al. (2008) & Wood et al. (2005)

I think there has been a misunderstanding as to why it is inappropriate to include a joint figure for the Nilo-Saharan groups that Hassan and his associates studied in their 2008 paper. It is not inappropriate because the Nubians are Nilo-Saharan speakers, since they of course are. It is inappropriate because Hassan et al. never actually studied a lumped sample of Nilo-Saharans. They only studied individual ethnic groups (see figure 2 of study [1]), unlike Wood et al. (2005) who do study samples grouped in broad linguistic categories (contrast with figure 2 of this paper [2]). Hassan et al. don't do this because the Nilo-Saharan groups in Sudan are much more heterogenous than the ones in southeastern Africa. In the table in this Wikipedia article, for instance, it has been suggested that Nilo-Saharans in Sudan have an average of 40.5% of the E1b1b haplogroup [3]. However, this is highly misleading since the authors clearly indicate that it's actually only a few Nilo-Saharan ethnic groups in Western Sudan that have high frequencies of the haplogroup, while the other Nilo-Saharan groups have rather low frequencies. And that this high frequency in the Western Sudanese Nilo-Saharan speakers is specifically due to genetic drift (a recent population bottleneck); otherwise, per the authors, the modern frequencies of E1b1b here too probably would've been a lot lower and frequencies of other Y DNA haplogroups in these Western Sudanese populations in turn likely would have been significantly higher.

• Regarding E-M78: "it has exceptionally high frequency among populations like those of western Sudan (particularly Darfur) and the Beja in eastern Sudan. The analysis of M78 subclades among Sudanese suggests that two subclades, E-V12 and E-V22, which are very common in northern African (Cruciani et al., 2007), might have been brought to Sudan from North Africa after the progressive desertification of the Sahara around 6,000–8,000 years ago. Sudden climate change might have forced several Neolithic cultures/people to shift northwards to the Mediterranean and southwards to the Sahel and Nile Valley (Dutour et al., 1988; Rando et al., 1998). E-V32 is the most frequent subclade among Sudanese. The Masalit possesses by far the highest frequency of the E-M78 and of the E-V32 haplogroup, suggesting either a recent bottleneck in the population or a proximity to the origin of the haplogroup."

Also note that the authors specifically distinguish Sudan's Nilotic Nilo-Saharan-speaking groups from its non-Nilotic ones (see tables 1 & 5), a distinction that this Wikipedia article also fails to make. Hassan et al.'s Nilotic Nilo-Saharan category includes the Dinka, Shilluk and Nuer, and is marked by high frequencies of haplogroups A and B; the authors' non-Nilotic Nilo-Saharan category includes the Masalit, the Borgu and the Fur, and is characterized by high frequencies of haplogroup E1b1b, and to a lesser extent, haplogroup A.

Adding all of the figures together to arrive at an average for all Nilo-Saharans in Sudan is therefore highly misleading since it's actually only one specific group of Nilo-Saharan speakers that have high frequencies of the clade: the non-Nilotic ones in the western part of the country (this fact also appears to have been already partly noted in the row of this article's table titled 'West Sudan (Darfur)'). 174.94.117.104 (talk) 11:54, 12 November 2010 (UTC)

Against

Sorry, I have taken my time to answer because I was reading about Nilo-Saharians.

Fundamentally I desagree. In Wood et al. the Nilo-Saharan groups are much more heterogenous than the ones in Sudan, in Wood we see mainly in Alur E2-M41, Massai E1b1b-M35, Luo E1b1a-M192 and Pygmies Mbuti B2. Besides Pygmies Mbuti are more relatives to other Pygmies and to Hadza people, since they lost their own language by influence from others Nilo-Saharian people. That's why linguistics can be different to genectics many times. Therefore using only Wood's work can be misunderstanding.

Instead, Hassan studies good samples from Sudan, and Sudan is the principal region and the center of the major diversity in Nilo-Saharan languages. That's the main point. And here E1b1b is always common, precisely you say that the Masalit possesses by far the highest frequency of the E-M78 and of the E-V32 haplogroup, maybe because there is proximity to the origin of the haplogroups. This means these haplogroups are Nilo-Saharians (probably).

In genetics, Nilotics show difference with non-Nilotics. But they all are indisputably Nilo-Saharans. Probably it is the same case with pygmies, Nilotics are maybe relatives to khoisans 90.000 years ago, and later they lost their antique language by influence from other Nilo-Saharians.

If we want a better vision about Nilo-Saharians, we need more sources about people from other countries, specially Chad, for a better comprehension of them.--Maulucioni (talk) 04:29, 25 May 2011 (UTC)

This isn't a vote, so there's no need to state that one is against the remarks above. Either the remarks are correct or one attempts to prove that they're not. With that said, I agree that genetics don't necessarily correlate with language. The fact remains, though, that it sometimes does, and it's this correlation that Wood et al. (2005) attempted to test in their study. The authors found that E1b1b was observed in only 14.3% of their combined Nilo-Saharan sample, not anywhere near the ~40% indicated in this article. Only their Maasai sample had a significant frequency of the clade, and that's because the Maasai, as Nilo-Hamites, are known to have interbred with many Cushitic speakers over the years.

Hassan et al. (2008) [4] indicate that only three Nilo-Saharan-speaking ethnic groups from Western Sudan in their study have high frequencies of the haplogroup, while the other Nilo-Saharan-speaking groups actually have quite low frequencies. They also do not at any point state that E1b1b is Nilo-Saharan. On the contrary, the authors quite clearly indicate that the haplogroup "might have been brought to Sudan from North Africa after the progressive desertification of the Sahara around 6,000–8,000 years ago" and that the clade's unusually high frequency in the Masalit is probably due to a population bottleneck.

As for Tishkoff et al. (2007) [5], they didn't test a combined Nilo-Saharan sample either. In fact, all they did was report the findings of other studies (including Wood et al. 2005), as well as one Datog sample that they did actually study.

If a person is not familiar with the existence of mixed Nilo-Hamitic populations such as the Maasai, Samburu, Kalenjin, Datog/Tatog and Turkana (that is, Nilotes who have assimilated many Cushitic peoples [6]), they are liable to believe that E1b1b is associated with Nilotes. It isn't, and there is no combined sample from any study that indicates it is. Per Wood et al. (2005) and other studies such as Gomes et al. (2010) [7], haplogroups A and B -- which are found at high frequencies in Nilotes, such as the Dinka, who have not mixed to a great extent with Cushitic or North African groups -- are the clades that are associated with Nilotes. Besides North Africans, haplogroup E1b1b is associated with the Cushitic groups that those particular Nilo-Hamitic populations mixed with, just as haplogroup J is found at high frequencies in Afro-Arab populations due to admixture with Arab peoples (where that clade is most frequent). 76.65.175.96 (talk) 05:19, 26 May 2011 (UTC)

OK, this isn't a vote, it's just a statement. No one of the authors I know has deeply devoted to Nilo-Saharan speakers, however, given the limited evidence, it can be seen that haplogroups A3b2 (M13), B2 (M182) and E1b1b1 (M35) are characteristic of Nilo-Saharans. All of these haplogroups are old, even E1b1b1 with 20,000 years old. E1b1b1 reached great development and diversity in East and Northeast Africa; it is totally African, specially at the beginning.

Nilo-Saharan languages have probably originated in Sudan and Afroasiatic ones probably in Ethiopia. They both would be 10,000-12,000 years old, for it is absolutely possible that E1b1b1 is involved with the origin and/or development of both groups. Proof of this is the presence of E1b1b1e (M293) in Datoga people and its spread towards South (Khoisan, Sandawe) may be related to Nilotic expansion. I don’t mean that all of this is for sure, but it’s certainly probable. The unity of Nilo-Saharans languages is not accepted by all linguists, so maybe they are heterogeneous and so is their genetics. Let’s be aware that we don’t have enough genetic studies to go into conclusions, we don’t know the haplogroups of main Nilo-Saharan groups as Songhay, Saharians, Central Sudanics, West-Sudanics-non-Nilotics, etc. Little or nothing we know about them. But, what we do know is that E1b1a is Niger-Congo. You trust only in Wood et al.’s results, but you have not realized that, out of 4 populations and 91 Nilo-Saharan samples, 47 are Mbuti pygmies: This is 52%!!, and we haven’t considered that Mbutis are not actually Nilo-Saharians!! Pygmy languages became extinct thousands of years ago, Efé and Asuá Mbutis speak Nilo-Saharan languages, but Sua and Kango Mbutis speak bantu languages, and even so they are considered one only ethnic and cultural group. Then the significant presence of E1b1a dues unquestionably to bantu expansion and the majority presence of B2b (instead of B2a) dues to typical pygmy linage, and consequently the result by Wood et al. is, in this case, strongly misleading and not representative.

To me this is an absolutely clear matter. Please, let’s be constructive, let’s respect the data we have until we find new sources and don’t delete the work and effort of the others. Regards. --Maulucioni (talk) 01:36, 29 May 2011 (UTC)

While numerous studies indicate that Y DNA haplogroups A and B are associated with Nilotes, there is unfortunately none extant that associates E1b1b with Nilotes. That's because only Nilotes that are known to have mixed with North Africans and/or Cushitic peoples have significant frequencies of E1b1b (Nilo-Hamites such as the Maasai, not the Dinka, Nuer, etc.; I see you have added the Maasai's individual frequency to the table, which is a much better course of action), just as haplogroup J is common in Afro-Arabs due to interbreeding with Arabs, but not at all in Bantus/Nilotes without Arab admixture. And even here, these Nilo-Hamitic peoples have lower frequencies of the clade than the North African and Cushitic average, with a lower diversity too. I also don't see anyone suggesting on the Y-DNA haplogroups by populations of Near East and North Africa page that haplogroup J ought to be associated with Bantus and Nilotes simply based on the existence of said mixed Afro-Arab populations. And even if they had, there's no existing study here either that supports such an association. Also please note that Tishkoff discusses this assimilation of Cushitic peoples by such Nilotes in her more recent autosomal study from 2009 [8]; so this admixture process is quite well-established.

With that said, I agree that the Pygmy sample in Wood et al. (2005) is not representative of Nilo-Saharan ancestry. That doesn't, however, change the fact that even with the Pygmy sample discounted, Nilotes as a whole do not have particularly high frequencies of E1b1b per Hassan et al. (2008). That's why the table in this article is better as it is since it clearly indicates that, with regard to Hassan et al.'s study, it's only a few mixed tribes in Western Sudan that have notable E1b1b frequencies, not the Southern Sudanese Nilotes. Speculating on the Y DNA of the Songhay, Central Sudanics, etc. is also, with all due respect, irrelevant since there is no published paternal DNA data on these groups. If and when that material is published, then of course we can and should add it to the table. But for the moment, we can only go with what we actually know. If you want to dispute this (which, in principle, I am not opposed to), you first need to produce an actual study that explicitly associates the clade with Nilotes as a whole and not as a consequence of admixture with other peoples. A study like Gomes et al. (2010), which associates a particular haplogroup B sub-clade with Nilotes: "it was also possible to identify a new Y-SNP apparently specific to Nilotic groups, as well as the presence of particular haplogroups that characterize Nilotic populations. The detection of a new haplogroup B2a1b defined by G1, could be, therefore, important to differentiate Nilotes from other groups, helping to trace migration and admixture events that occurred in eastern Africa." 174.88.66.168 (talk) 23:16, 29 May 2011 (UTC)

Also, I think we should consider splitting up all the lumped ethno-linguistic groups into their constituent entities if the studies that they are cited to only present the data in terms of individual population samples rather than combined ones. By the same token, we should keep samples that are lumped in their aggregated form. This is the only way we can avoid making broad but inaccurate generalizations that give the impression that clades are more or less frequent in certain populations than they really are. 174.88.66.168 (talk) 23:34, 29 May 2011 (UTC)

Okay, I'm just confused now. The table indicates two different sets of frequencies for Wood et al. (2005)'s Nilo-Saharan groups across various haplogroups (one dubbed "Nilo-Saharan" and the other "Nilotics"). However, Wood et al. [9] only have a combined sample for Nilo-Saharan speakers (Figure 2), not for Nilotics. Where exactly is this data coming from and why is it there? Is it intended to show the NRY frequencies for Nilotes alone in that study as opposed to all Nilo-Saharan speakers, including those whom aren't ethnically Nilotic (such as the Mbuti)? Because the Maasai were the only Nilo-Saharan speakers (Nilotic and non-Nilotic alike) in Wood et al. that carried E1b1b sub-clades; please refer to the study's Appendix A [10]. This extra "Nilotic" category in the table therefore seems superfluous and even somewhat misleading since, at a glance, it suggests that on average, 29% of Nilotes belong to E1b1b sub-clades. This of course isn't the case; take away the Cushitic-influenced Maasai, and 0% of the Nilotes in Wood et al. carried the haplogroup. 174.88.66.168 (talk) 00:26, 30 May 2011 (UTC)

The presence of E1b1b, A3 and B2a are ancient in this region (Ethiopia and Sudan), and I recognize that E1b1b has more diversity in Afroasiatics than Nilosaharians, however, I still believe that these haplogroups could be a pool related to the origin and/or development of both linguistic groups. I don’t deny the interbreeding, any way that is irrelevant because we can’t either deny the existence of E1b1b in most of Nilo-Saharian groups, whatever the reason is. Since you recognize that the Wood’s sample is not representative of Nilo-Saharan ancestry, then you shouldn’t delete other different samples, because it is only by the comparison between them that we will have an actual idea on the Nilo-Saharian gene pool. Because of this, I have made an average of the highest number of samples that I’ve got. --Maulucioni (talk) 23:55, 3 June 2011 (UTC)

Sudan does not have a higher diversity of E1b1b sub-clades than North Africa or the Horn nor is there any study that indicates it does. In fact, E-M78 represented 75% of all haplogroup E lineages in Hassan et al. (2008), including E1b1b. The researchers also indicate that the significant frequencies of the clade found today in a few Western Nilotic groups is a consequence of a population bottleneck that occurred fairly recently. So it's only really the presence of E1b1b in Sudan's Cushitic groups, such as the Beja, where the clade's ancient local presence has any real significance (since the modern frequencies of the clade in these Cushitic groups is not due to a population bottleneck). Woods et al. 2005 is indeed also representative of Nilotic ancestry. All of the study's samples that are labeled "Nilo-Saharan" are ethnic Nilotic groups, except of course for the Mbuti Pygmies. But even when the Pygmies are removed from the analysis, it's still only the Maasai (who have absorbed Cushitic peoples for generations) who carry the clade. The rest of the actual Nilotes in the sample set don't carry the haplogroup at all i.e. a 0% frequency; so one cannot credibly argue that the only reason that the study's overall E1b1b frequency is low is due to the presence of Pygmies in the sample set. In any event, the 33.9% average looks more accurate, though it would've been preferable if it had been taken from an actual study. 70.53.210.148 (talk) 02:16, 4 June 2011 (UTC)

R1b1a

The link atop the R1b1a column needs to be changed to point to the more upstream R1. As it currently stands, the percentage table fails to capture the full diversity of the lineages found in the Sub-Saharan region since it omits the very high individual frequencies of R1 observed in several tribes in West and Central Africa, such as the Ouldeme of northern Cameroon. Wood et al. (2005) [11], for example, show clade frequencies as high as 95% in the Podokwo and Cruciani et al. (2010) [12] found the clade in 95.5% of its Ouldeme sample. 70.53.210.148 (talk) 02:16, 4 June 2011 (UTC)

Renaming proposal

Interested/knowledgeable parties, please see Talk:Y-chromosome haplogroups by populations#Renaming for a renaming proposal that would affect this article and 10 others. Please comment over there to keep the discussion centralized. Thanks. - dcljr (talk) 00:20, 20 April 2012 (UTC)

External links modified

Hello fellow Wikipedians,

I have just modified one external link on Y-DNA haplogroups by populations of Sub-Saharan Africa. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive http://web.archive.org/web/20130615030452/http://dna.xyvy.info to http://dna.xyvy.info

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 19:04, 21 July 2016 (UTC)