Out of South Africa?

By Razib Khan | March 7, 2011 9:25 pm


Credit: Ian Beatty

Apparently there’s a new paper on Bushmen genetic diversity coming out in PNAS. It reports that this group is the most diverse in the world in terms of intra-population variance. This is not a great surprise after the much heralded Nature paper from last year. But the bigger point seems to be that modern humans may derive from a southern African populations, not an eastern African one, as has been the predominant assumption for the past few generations. The reasoning is that the most diverse population has had the greatest amount of time to accumulate genetic variation since the emergence of modern humanity, other groups having been periodically purged of variance because of population bottlenecks during the process of migration. African populations are the most diverse of all, and of African populations the Bushmen are among the most diverse.

I think Sarah Tishkoff has the right of it: “African populations have had complex demographic histories and there is no a priori reason to believe that populations evolved in situ in the regions where they exist today. Some could have migrated from other regions.” In plain English she’s saying that Bushmen themselves may have come from elsewhere. Since the paper is not on the PNAS website I can not evaluate it, but this seems like a very strong objection to me. The first author counters:

Henn admits that migration could certainly be a possibility, but counters that when a population migrates, typically only a subset moves to a new area, and this subset is less genetically diverse than the parent population. She argues that if a group left eastern African for southern Africa it would be expected to have less diversity in the south. “This is not what we find in the data,” she says.

This is not addressing what I think is implicit in in Tishkoff’s critique: the Bantu expansion may have resulted in a near replacement of the original hunter-gatherer substrate of eastern Africa. Therefore, the presumably most genetically diverse source population from which the Bushmen derive may no longer exist. Of course there’s no reason to assume that anatomically modern humans do come from eastern Africa. I simply think that there’s plenty of historical money “left on the table” which could immediately question the plausibility of an “Out of South Africa” thesis. Hunter-gatherer groups like the Bushmen have likely had larger long term effective population sizes than farming groups, who may have expanded rather rapidly from a small initial group.

CATEGORIZED UNDER: Genetics, Genomics
MORE ABOUT: Genetics, Genomics
  • Ian

    I know this has changed over time, but doesn’t current thinking put the click-speaking East Africans closer to southern Africans? I wonder if there’s any reason to assume that pre-Bantu southern and eastern Africa shouldn’t be considered a single population. Or, for that matter, pre-Bantu central African populations as well.

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    but doesn’t current thinking put the click-speaking East Africans closer to southern Africans?

    there’s hardly any of the former. just the hadza and sandawe. see tishkoff’s paper. i think you are right for the hadza.

  • http://www.rishon-rishon.com David Boxenhorn

    Isn’t it interesting that the tiny Bushman population is the one with the largest long-term effective population size? It just goes to show how easily (as a result of compound population growth) long-term effective population size can be different from current population size.

    I think that agriculture always results in compound population growth of the agriculturists, which wipes out the diversity of the hunter-gatherer substrate. In addition to the Bantu expansion, there was an older agriculturist expansion in Ethiopia.

  • Brenna

    http://www.pnas.org/content/early/2011/03/01/1017511108.abstract

    Link to the PNAS paper. Please note the authors also sample eastern African click speaking populations as well as southern and central.

  • Eurologist

    There is also the where and when.

    The oldest sites attributed to culturally AMHs have been in South Africa for a long time – so, I never understood the focus on East Africa for the origin of AMHs (not OOA, of course). On the other hand, it depends on what time you are talking about. This is how I see it:

    After 800,000 and before 300,000 ya? I see strong indications of gene flow between Europe, West Asia, And North Africa. 300,000 to 200,000 ya? Climatic isolation from the rest-of-the-world, and evolution probably all along the African East Coast. 200,000 – 130,000 ya? Center in S and SE Africa. 130,000 – 100,000 ya? Extreme northward expansion with re-unification of separated humans, into Ethiopia, all of NA, the Nile valley, the Levant, Arabia, and beyond. First OOA. 100,000- 60,000 ya? contraction from much of the above. 60,000 -45,000 ya? Second OOA. Later, bi-directional flows.

  • Georg

    East Afrika (rift valley) has a extremly “friendly” geology to produce (cover)
    corpses and as well to uncover them by erosion later. I think we should say
    “ex Afrika” for humans, not ex east Afrika. Its not impossible humans developed
    there, but there is such a bias by geology, that one has to be cautious.

  • http://www.kinshipstudies.org German Dziebel

    Dienekes blogs about a new paper on Y-DNA presented at AAPA. Apparently, it directly contradicts the PNAS study, as it argues that the earliest human Y-DNA lineages are found in Central and East Africa and not in South Africa.

    http://dienekes.blogspot.com/2011/03/aapa-2011-abstracts.html

    Geneticists should probably learn how to better coordinate reports coming out of different labs. Plus the idea that a parent population is always more genetically diverse than a daughter population looks very simplistic and can hardly account for the vagaries of history.

  • Henry Harpending

    Razib do you have a handle on SNP ascertainment issues here? They used several standard chips on their samples, then pooled the results, but these chips certainly don’t have the unique SNPs from the Bushman guy sequenced last year. They say in the paper that they also used a ‘custom’ chip from 23andme but they don’t say what the ascertainment was. I would hesitate to make much of anything of these SNP papers without a better handle on ascertainment.

    Sarah is certainly right about geography: those Khoisan groups in southern Africa are what the Bantu expansion grinder has not yet ground. Cavalli-Sforza claims there are Khoisan-looking groups still in Ethiopia, and the common view of archaeologists is that Khoisan speakers occupied the whole eastern Africa.

    Henry

  • Sandgroper

    If the Khoisan once occupied the whole of eastern and southern Africa, and that is certainly the impression I had, then the migration/loss of diversity argument doesn’t hold at all.

  • http://www.kinshipstudies.org German Dziebel

    “Cavalli-Sforza claims there are Khoisan-looking groups still in Ethiopia, and the common view of archaeologists is that Khoisan speakers occupied the whole eastern Africa.”

    There’re no Khoisan-looking populations in Ethiopia. Even if there were, the Khoisan physical morphology is less than 36,000 years old and may be only 10,000 yo (Stynder, “Early to mid-Holocene South African Later Stone Age human crania exhibit a distinctly Khoesan morphological pattern”). Some peculiarities of their outward appearance (epicanthus, Mongolian spot, light skin), unless convergent, ties them back to Asia if anything. Some linguists consider East African Hadza and Sandawe click-based phonetic systems to be the oldest among all click-speaking languages, with South African Khoisans acquiring them through borrowing or common descent. If Hadza and Sandawe are counted as the two most divergent languages of the Khoisan or Macro-Khoisan family, then, from the point of view of glottochronology, its age is some 10,000 years old. Without Sandawe and Hadza, it’s 5,000 years old (see Starostin, “From Modern Khoisan Languages to Proto-Khoisan: The Value of Intermediate Reconstruction”). According to Guldemann (” A linguist’s view: Khoe-Kwadi speakers as the earliest food-producers of southern Africa. Southern Afr. Hum., 20: 93-132, 2008), Khoe-Kwadi languages separated from Sandawe some 2,000 years ago and moved to South Africa where they encountered older Khosan languages such as Juhoansi and Tuu. According to Blench, there may be some startling similarities between Omotic-speaking Dorze in Ethiopia and Pygmy and Khoisan music.

    Archaeologists see microlithic cultures (Nachikufan, etc.) in East Africa as potentially of Khoisan provenance. They are some 18,000 years old. This is roughly comparable with linguistic and paleobiological dates and with some published genetic dates (e.g., for Sandawe)

    All evidence combined, it’s rather likely that the age of Khoisans is greatly exaggerated in genetic literature, as there are no pedigree studies of the rate of lineage divergence among Khoisans but their basal place in the human tree is inferred from top-down, taxonomic reasoning. Archaeology and linguistics work from bottom up instead. It’s also likely that there were several migrations from East Africa to South Africa during the 20,000-10,000 years that Khoisan populations existed as a distinct historical entity.

  • Perahu

    I really wish there were more whole genome sequences of Africans besides the current Bushmen / Bantu ones. They ought to get some (non-Bantu) East Africans fully sequenced to see how they compare.

  • http://washparkprophet.blogspot.com ohwilleke

    German Dziebel (7) beat me to it. Out of East Africa lives.

    Also, South Africa is not rich in the mtDNA clades of haplogroup L3 from which all Eurasian clades of it descend, nor are the Bushmen, who are most strongly associated with Y-DNA haplogroup A particularly close to genetically to the CF Y-DNA haplogroups of most Eurasians, or the DE Y-DNA haplogroups found in Eurasia, North Africa, most of West Africa, a very large share of East Africa, and people with male Bantu lineages. The uniparental lines phylogenetically and in their estimated ages strongly support intermediate steps between Bushmen and Eurasians.

    Bushmen may be the closest on average and in average diversity and in lifestyle (and quite possibly linguistically) to the orginal modern human population, but the abstract cited at Dienekes strongly suggests that they are a relict exile population that preserved only a subset of the original East African population’s diversity and did not even capture the oldest lineages in that population.

  • pconroy

    I’ve a question: does anyone know if Bushmen type populations ever made it to Madagascar, or if the island was truly uninhabited when the Austronesians arrived?

  • http://www.kinshipstudies.org German Dziebel

    @pconroy: Mikea foragers are the likely descendants of a pre-Austronesian population in Madagascar. They carry a basal M lineage. No connection to Khoisan or any other African population.

    http://www.biomedcentral.com/1471-2164/10/605

  • Pingback: Where in the world did anatomically modern humans come from? | Gene Expression | Discover Magazine()

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    Razib do you have a handle on SNP ascertainment issues here? They used several standard chips on their samples, then pooled the results, but these chips certainly don’t have the unique SNPs from the Bushman guy sequenced last year. They say in the paper that they also used a ‘custom’ chip from 23andme but they don’t say what the ascertainment was. I would hesitate to make much of anything of these SNP papers without a better handle on ascertainment.

    agreed.

  • onur

    @pconroy: Mikea foragers are the likely descendants of a pre-Austronesian population in Madagascar. They carry a basal M lineage. No connection to Khoisan or any other African population.

    http://www.biomedcentral.com/1471-2164/10/605

    But also according to the paper you linked to, that basal M lineage likely originated in Africa or in an African-influenced adjacent region, in line with the African origin of M and modern humans as a whole.

  • http://www.kinshipstudies.org German Dziebel

    Yes, onur, you’re right, as the following passage indicates: “The identification of four individuals of African and Southwest Asian origin who share the 13 diagnostic control region mutations for M23 pinpoints these regions as potential sources for M23. Whilst, the data does not allow us to make clear phylogeographic inferences regarding M23 origin, our results may provide some evidence of ancient contacts across the Indian Ocean involving Africa, Madagascar and South Asia. The deep-rooted topology of M23 and its age estimate coupled with its very restricted distribution within Madagascar, makes unlikely its presence in the island as a result of recent contacts, and is more in agreement with the patterns of human contacts across the Arabian Sea and the Indian Ocean, which pre-dated the Austronesian expansion into Madagascar…”

    However, it’s rather well established that African M1, the only African lineage within the M cluster, represents a back migration into Africa. See, e.g., p. 75 in http://books.google.com/books?id=x6o4XLIKN0UC&pg=PA76&lpg=PA76&dq=gonzalez+M1+mtDNA&source=bl&ots=QN_kA4FEe1&sig=WfBnbLvN-j41c7FRXJGQcY7Yarc&hl=en&ei=n852TaCtBeWG0QG18MXWBg&sa=X&oi=book_result&ct=result&resnum=11&ved=0CGMQ6AEwCg#v=onepage&q=gonzalez%20M1%20mtDNA&f=false

    I therefore interpret the above statement as suggestive that, in addition to M1, there was at least one more M lineage, namely M23, that may have expanded west of India. In any case, we would expect to find L0 or L1 in Madagascar if Mikea were related to Khoisans – likely the earliest inhabitants of Africa. As a side note, it’s noteworthy that Andamanese also have basal M lineages, M31’32. This may be a trace of an ancient circum-Indian Ocean population, with ultimate roots in East Asia, IMO, not Africa.

  • http://www.riverellan.blogspot.com Tom Bri

    German Dziebel, the ‘Mongolian spot’ is not confined to Asians, any dark-complected baby may have it.

  • http://www.kinshipstudies.org German Dziebel

    I know this, Tom. I was referring – in a tongue and cheek way – to Harpending & Eller’s paper “Human Diversity and Its History” in which the authors listed Mongolian spot and other peculiarities of the Khoisan phenotype among traits that look East Asian: “They resemble, to European eyes at least, east Asians. They have yellowish rather than black skin, epicanthic folds, shovel-shaped incisors, and many newborns have “Mongoloid spots” at the base of the spine.”

  • http://www.kinshipstudies.org German Dziebel

    @pconroy and onur,

    As Terry Toohill informs me, the sister clade of mtDNA M23, namely M75 was detected in Southeast Asia (Laos). You can see it on PhyloTree. The paper is here. http://mbe.oxfordjournals.org/content/28/1/513.abstract. This suggests that Mikea, after all, may not be a pre-existing non-Austronesian population of Madagascar. This lineage may have been brought to Madagascar by Austronesians.

  • pconroy

    @German,

    So who were the Vazimba do you think?

  • http://www.kinshipstudies.org German Dziebel

    @pconroy
    A subset of Austronesians who separated from the main body early on and developed in isolation until they were reabsorbed in the 17th century.

  • onur

    Then I guess there is no – whether genetic, archaeological, paleoanthropological or linguistic – evidence of a pre-Austronesian human population in Madagascar.

  • http://www.kinshipstudies.org German Dziebel

    “Then I guess there is no – whether genetic, archaeological, paleoanthropological or linguistic – evidence of a pre-Austronesian human population in Madagascar.”

    Definitely no early African populations. It’s still possible that M23 arrived to Madagascar before Austronesians as part of a separate migration from Southeast Asia.

  • onur

    Definitely no early African populations. It’s still possible that M23 arrived to Madagascar before Austronesians as part of a separate migration from Southeast Asia.

    I don’t think there was any culture in SE Asia capable of such long distance seafaring before the Austronesian expansion.

  • http://www.kinshipstudies.org German Dziebel

    “I don’t think there was any culture in SE Asia capable of such long distance seafaring before the Austronesian expansion.”

    It’s a valid point, onur, and I’m somewhat puzzled by the situation myself. It’s noteworthy that Andamanese have mtDNA M31 and M32 and a linguistic link between Ongan-Jarawan and Austronesian was recently proposed by a very respectable linguist (see Blevins’s work). mtDNA M27 was recently detected in the Solomon islands (see http://massey.genomicus.com/publications/Ricaut_2010_JArchaeolSci_v37_p1161.pdf). M28 and M29 are known from Vanuatu, etc. If M27, M28, M29 are pre-Austronesian in Oceania, then M23 should also be considered pre-Austronesain in Madagascar, but in the case of Madagascar is requires two events of distant seafaring, which is unlikely, as you note. So, it’s possible that the colonization of Madagascar did occur in one wave but it took place earlier than we think, in “pre-Neolithic” (Hoabinhian?) times when Austronesians formed a more balanced continuum of different phenotypes and different foraging and agricultural adaptations along the western coast of SE Asia and down to Wallacea. Just like Papuans/Australians these early Austronesians were subdivided genetically between mtDNA R- and M- lineages and between Y-DNA C and MNOPS lineages. This interpretation, however, would fly in the face of the out of Taiwan theory of Austronesian dispersals. It does face a lot of criticism lately, though.

  • Pingback: Sightseeing « phknrocket1k()

  • onur

    Terry writes in Dienekes’ blog that not only M75 but also M23 is found in SE Asia (in Laos), so there is no reason to attribute the presence of M23 in Madagascar to a date earlier than the archaeologically established dates for the Austronesian arrival to Madagascar. A pre-Neolithic date is especially implausible given the technology, or rather, the lack of technology then.

  • onur

    so there is no reason to attribute the presence of M23 in Madagascar to a date earlier than the archaeologically established dates for the Austronesian arrival to Madagascar

    so also there is no reason to attribute the Austronesian arrival (or the first human arrival, indeed) to Madagascar to a date earlier than the archaeologically established dates for the Austronesian arrival to Madagascar

  • http://www.kinshipstudies.org German Dziebel

    Yes, this is likely. One odd thing that remains to be explained is the absence of M23 in Merino and in Borneo – the presumed source of the Austronesian colonization of Madagascar. The concentration of M23 in Mikea (with B4 evenly distributed between the two), who are also distinct economically and culturally, hint at an ancient population subdivision within the Austronesian colonists. This is conceivable if the source for the Austronesian colonists was in mainland SE Asia (e.g., Laos where both B4a1 and M23 are found, see Bodner, “Southeast Asian diversity: first insights into the complex mtDNA structure of Laos,” 2011). Borneo hasn’t yet shown M23.

  • onur

    Further genetic tests of the presumed source area(s) of the Austronesian colonists and environs is required to make strong judgements about the absence issue. The Mikea case is interesting indeed. I think hunter-gatherer (including fishing) populations are good candidates for detecting rare lineages and giving some clues about ancient population structure.

  • onur

    Further genetic tests of the presumed source area(s) of the Austronesian colonists and environs is required to make strong judgements about the absence issue.

    This is especially so as the lineage in question (M23) is particularly rare in the countries where it is found.

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at http://www.razib.com

ADVERTISEMENT

See More

ADVERTISEMENT

RSS Razib’s Pinboard

Edifying books

Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »