Friday, February 6, 2015

DNA - "Grass Fire Pattern"

The map below will go a long way for understanding the puzzling presence of R1b in the under-gird of North Africa.  Taken with a host of other artifacts, such as lithics, pottery, metallurgy and domestics, it offers at least one explanation for the changes in Western Iberia in the late 4th millennium B.C.

North African climate change is a slow and steady process punctuated by sudden collapses.  One collapse began in the early 4th millennium (5.9 kiloevent) and possibly flooded the Upper Nile valley with immigrants from the Acacus (c3,600 B.C.).  At the same time, this process pushed immigrants to Western Morocco (also archaeologically detectable) and eventually Iberia (IMO) which also experience dramatic social change and traditions. 

The important point here is to note the population peak of grassy Sahara during its pastoral/dairy phase, followed by a burnout in which the population center begins shifting South and West, to the Sub-Sahel and the Oases.  This burnout progressed since then; the Western Desert (Morocco, not Egypt) once respectably populated, is now in modern times completely uninhibited -completely-.
Still frames from the kernel density animation showing the key stages of demographic development across time and space. A shows initial Early Holocene incursions into the Sahara between 14,000 and 11,000 years BP, B–C show gradual population increase across the Sahara between 11,000 and 10,000 years BP, especially in the eastern Sahara and northern Air mountains, D shows exponential growth especially in the eastern Sahara, E shows the early Holocene population peak around 7500 years BP with most regions, except the western littoral being populated, F shows population retraction, especially in the central Sahara and the start of occupation along the western littoral, G shows renewed population expansion in the Middle Holocene, H shows retraction in all areas coinciding with the termination of the AHP, and I shows the de-population of the Sahara and southwestward movement of late Holocene populations.

The presence of R1b in Northern Cameroon, Chad, Niger, Mali, etc, is the most likely the result of those immigrant populations looking for greener pastures.  This is important in light of the modern emergence of Afro-Asiatics like the Berbers (at least the male half) that developed a nomadic life-style in the (now) desert based on browsing domesticates.  (They apparently coming from a more easterly direction)

One thing worth checking out is Andrew's blog where he discusses the dichotomous situation between the genetics of Western Europe and Northern Africa.  I think he has laid some of the problems out a little better than I've tried to.

The demographic response to Holocene climate change in the Sahara, Katie Manning, Adrian Timpson, (2014)  [Link]


  1. I don't think there's any "mystery" in R1b (V88) being found in Chadic peoples. Why? Because there's also a lot of R1b found in Sudan (and to lesser extent Egypt), from where it no doubt originated. Chadic is a branch of Afroasiatic and the whole family appears to have expanded in the earliest Holocene from the Nile. Other branches brought E1b and J1 clades to NW Africa or West Asia but this one had an R1b-V88 founder effect instead. No mystery, even if some extra research on Sudanese genetics is surely desirable to be sure.

    Also it's important to emphasize that R1b-V88 has nearly no relationship with the mainline European subclade of R1b (M412): both come from a common West Asian origin, probably deep in the Upper Paleolithic but that's about it. It's only somewhat closer than the relationship with R2, Q or even NO. We should not be deceived by simplified nomenclature (i.e. "R1b").

    There is in Europe some R1b-V88, notably in Sardinia but generally scattered in the Eastern Mediterranean arch, also in West Asia. The exact relationship with Chadic R1b-V88 remains a bit obscure but in general terms it can be assumed that R1b-V88 spread from West Asia to (a) Sudan → Chad and (b) the Balcans → Italy → Sardinia with separate pathways - although an African origin can't be fully discarded at this stage of the research on V88.

    Otherwise the archaeological reconstruction of the Green Sahara is interesting indeed. But I do not see in it any "arrow" pointing to possible migration routes. I am aware that migrations from the Nile (Afroasiatic family particularly) are likely at the beginning of the Holocene but that's not something apparent in the density maps.

    1. It may indeed be Late Paleolithic connections. On the other hand, other individual traits seem to suggest a more recent connection, especially when coupled with lithics and metallurgy.
      Of course, every option has problems. If option A is Paleolithic, then another phenomenon must explain more recent similarities in lithics, cattle, metallurgy and LP. Same for Mesolithic and Early Neolithic. If option B is maybe Middle Neolithic to Bronze Age then directionality is a problem and obvious paternal mismatches, with the exception of a more diverse cross section of R1b at one time.

    2. Which similarities in "lithics, cattle, metallurgy and LP" (and what is LP)?

      All I can think now is about cattle and African cattle has its own founder effects, dispelling any illusion of Neolithic unity with Europe.

    3. I'd echo Maju's points here.

    4. This comment has been removed by the author.

  2. Excellent visual and a great post, cheers. The R1b R-V88 clade is not the only part of former Y DNA haplogroup K2b found in African populations. So-called 'European' R1b R-M269, differentiated from R-V88, is found in (at minimum) North Africa, Nigeria, Sudan, Chad, Horn of Africa, Equatorial Guinea, Khoisan, Pygmies; R1a in North Africa, Sudan, Horn of Africa, Namibia, Sao Tome, Zambia, Mozambique; R2 in Egypt, Sudan, Biaka Pygmies; Q-MEH2 in Tanzania Sandawe. That is from reading a cross section of papers. But Africa is chronically under-sampled as I keep saying.

    1. R1b and K2 (now T) are not the same thing at all: they are different haplogroups under F - i.e. the only thing they have in common is that both are "Asian" lineages.

      Their distribution patterns are also rather different: T seems to have an Indian Ocean centrality, with important centers in Pakistan, Arabia and Somalia. R1b instead has a more Mediterranean tendency and lacks significant presence in the Indian Ocean arch. Both represent, along with J1, backflows from (Eur-)Asia into Africa but they surely did so in two different processes with their respective founder effects.

    2. I thought that 'KLT' ('K') splits into 'LT' ('K1'); and 'K' ('K2'). 'LT' ('K1') comprises L and T. 'K' ('K2') comprises 'MPS' ('K2b'); and 'X' ('K2a'). 'MPS' ('K2b') comprises 'MS' (or M and S); and 'P' which consists of 'QR' (Q and R). While 'X' ('K2a') comprises 'NO' (N and O).

    3. I know that anything other than Y DNA haplogroups A, B, and E within Tropical Africa are always regarded as Eurasian backflow or colonial era European admixture. But some of these really have me vexed. For instance: H-M69(xM82) and R-M124 (R2) both in Biaka Pygmies. Q-MEH2 in Tanzania Sandawe. N1c-tat in Equatorial Guinea. MtDNA haplogroup R7 in Rwanda Tutsi. R-M269 in Khoisan and Pygmies. MtDNA haplogroup A in Sierra Leone and Mozambique. I can't help but have suspicions that at least *some* of these so-called 'erratics' have either been in Africa a very long time (even if they originated in West Asia) or actually evolved in Africa in the first place. I am probably one of the few non-Afrocentrists willing to consider this possibility. Was there really a migration from India to the Biaka pygmies?

    4. Pygmization appears to me to be consistent with island dwarfing. It's an unusual trait for a region where gigantism rules.

      I read something recently on whole genome comparisons between the s. Pacific pygmies and African pygmies. Maybe there is a possible connection after all.

    5. You're right, Chris. I was not considering (nor really aware of) the latest revision of the K macro-haplogroup nomenclature. My confusion comes because until some years ago, what is now known as T was called K2.

      So let's start all over. By K2b you mean the macro-haplogroup that includes P (incl. Q, R), S and M, as well as some other K-derived lineages of Australasia (all these now gathered under K2b1), right? And you say that some other P sublineages (= K2b2, rather than K2b in general) are also found in various African populations, what is correct, interesting and somewhat intriguing.

      The big question is not whether those P-derived lineages are found, mostly at very low frequencies, scattered in Africa, but whether some of them can be considered pre-colonial. In most cases, it is assumed that they are colonial erratics from Europe, whose settlers, traders, soldiers and missionaries have been prowling Africa for some 500 years now.

      The lineages are not always well described in the papers, so it's difficult to be sure which are their actual affiliations outside Africa or even inside it. In the case of Egypt and Sudan it's possible that some of the R1b(xV88) belongs to lineages that migrated with V88 (and J1) to that part of Africa long ago (LSA genesis possibly). As far as I know, there's no data to claim they are part of the mainline European sublineage M412 and they are probably not, just relatives within the wider M269 and hence surely of West Asian origin.

      The R2 in Egypt and Sudan is also noticeable but again it surely belongs to a continuum of this lineage (most common in India) through West Asia, where it's found at low frequencies almost everywhere. Sadly R2 remains ill researched.

      The case of Q in East Africa is also intriguing but I'd imagine it to be related to Iranian Q (highest basal diversity of this lineage) or maybe some Arabian derivates (Yemeni Jews have some of that too).

      In general I'd advocate for a West Asian origin of the lineages found in East Africa (incl. Nile basin) and a European origin of the rest (colonial south of the Sahara and probably quite older in NW Africa).

      "Was there really a migration from India to the Biaka pygmies?"

      Surely not. Biaka have absorbed various influences and what we should consider is whether those lineages are privative of them or rather found also in nearby populations. Also there's no need to attribute Indian origin to those lineages: a West Asian one is much more likely. In any case, you may want to consider the issue of Africa-India crop exchange c. 4000 years ago, which could be related if anything (but then again neither the Biaka nor the Sandawe are farmers).

      As for N1c, it's a European lineage found often enough among Dutch or Scandinavians to explain its presence in Equatorial Guinea.

    6. "Pygmization appears to me to be consistent with island dwarfing. It's an unusual trait for a region where gigantism rules".

      Seems to be more related to forest life. There may be several advantages to small size in jungles, where some nutrients may be lacking and where small size helps for climbing (a lot of resources are up the trees rather than on the ground).

      "I read something recently on whole genome comparisons between the s. Pacific pygmies and African pygmies. Maybe there is a possible connection after all".

      Not really. Autosomally, as well as by other markers, African Pygmies are quite unique and, if anything, they cluster with other Africans. SE Asian small sized populations ("Negritos") are at least three different populations (Andamanese, Orang Asli and Filipino Negritos), all them stemming from the main Asian (out-of-Africa) trunk. They do not even have any particular relation among each other.

  3. The dessication of the Sahara really seems to have scattered populations over large distances. There are dozens of HLA haplotypes found in European populations which are shared with North Africans, West Asians, and populations in the Sahara and Sahel or points southward in tropical Africa. But the haplotypes are in dis-quilibrium in the European populations indicating a more recent entry [& not at high enough frequency to indicate that they have been selected for], while the haplotypes and their components are at peak frequency and diversity of recombinants in various African and to a lesser extent Middle Eastern populations. One example is A29-Cw16-B44-DR7-DQ2, notable in Maghreb, Iberia, Basque Country, Western France, Western British Isles. The haplotype is also found in the same and slightly different forms in Cameroon [both Yaounde and rainforest], Zambia Bantu, Mozambique Bantu, South Africa Bantu. On closer inspection the haplotype and many of its components peak in frequency and diversity in West or Central Africa. A*29:02 - Cameroon Yaounde/Baka Pygmies/Bamileke/Sawa/Beti, Zimbabwe, South Africa; Zambia; C*16:01 - Mali Bandiagara, Burkina Faso Fulani, Burkina Faso Mossi, Ghana Ga-Adangbe, Senegal Niokholo Mandenka, Burkina Faso Rimaibe, Equatorial Guinea, Zambia, Kenya, Zimbabwe, South Africa; B*44:03 - Cameroon Beti, South Africa, Cameroon Bakola Pygmy, Ghana Ga-Adangbe, Cameroon Bamileke, Central African Republic Mbenzele Pygmy, Zimbabwe; etc.
    This particular haplotype is #2 in Spanish; #3 in Tunisia Pop.1; #3 in Morocco Settat Chaouya; #3 in Portuguese; #3 in French; #5 in UK; #7 in Italy; #7 in Germany; #8 in Austria; #12 in Netherlands; #14 in Romania; #17 in Poland; #27 in Bosnia; #43 in Croatia; #97 in Greece; #139 in Turkey.

    1. I agree. Thanks for a thorough comment, Chris. It's hard to fully refine this view just yet (directionality), but this problem uniquely has lateral limits. It's difficult to dismiss these connections in the usual ways, to simply relegate it to the Paleolithic. Obviously many of the sub-clades among these peoples are simply too young, in other cases the presence of LP traits that at one time were supposed to evolved with LBK are too young.
      The emergence of Afro-Asiatic peoples in the Bronze Age puts its right lateral limit on the timeframe in which connections could occur. For whatever reason, this simple logic escapes so many people.

    2. HLA haplotypes aren't good ways to trace historical migrations because they are not selectively neutral and very, very thin levels of population exchange can produce dramatic levels of fitness enhancing genes in the receiving population. For example, a disproportionate share of Neanderthal DNA involves fitness enhancing HLA genes.