Location, Location, Location: Did Humans Start Out in Southern Africa?

by Mike Macpherson, Shirley Wu, and Joanna Mountain

A young Hadza hunter

From just flecks of bone and specks of DNA, scientists have been able to piece together a story of the origin of our species. That story, by most accounts, begins in eastern Africa 100-200,000 years ago. After all, multiple lines of evidence point to eastern Africa — it’s where the oldest recognizably human bones have been found, where the linguistic diversity is richest, and it fits the available genetic data. But in this week’s issue of PNAS, researchers, including those from Stanford, UCSF and 23andMe, present much more extensive genetic evidence suggesting that humans may have originated not in eastern, but in southern Africa.

How can genetics tell you where a species originated? A general principle is that you tend to find the highest levels of genetic variation in the portion of the population that’s been around the longest. That’s because sub-populations that split off from an original population start out with just a sample of the original population’s gene pool, and it takes a very long time for new genetic variation to build up in the new sub-population.

Using this principle, one strategy for identifying the location a species originates is to measure genetic variation in several places the species lives, and look for the place the genetic variation is highest. For this study, the authors gathered new genome-wide data from three click-speaking peoples from southern Africa that have been living for millennia in relative isolation, and seven northern African peoples. They also assembled existing genome-wide data from 14 further African populations, covering eastern, western, and central Africa, resulting in a continent-wide representation of genetic variation.

With this much coverage, the researchers would have been in a position to find evidence for human origins in any part of Africa; eastern Africa would be the front-runner, but researchers have argued that northern, central and southern Africa are also plausible candidates. But they found that among all the populations they looked at, the highest levels of genetic variation were found in the southern Africa. This suggests, the authors conclude, that each of us may trace our ancestry (ignoring any contributions from Neandertals for now!) back to southern Africa, sometime less than 100,000 years ago.

Although this new genetic evidence is intriguing, it’s not the end of the story. There’s some archaeological evidence to support a southern African origin, but for now it’s fairly limited. And these new data are also consistent with a scenario where humankind arose in eastern Africa, but later migrated to southern Africa. Studies sampling genetic diversity in further African populations may help clarify the early history of our species.

Dr. Mountain in Africa during her time in the Peace Corps

The insights of this study were enabled by genome-wide genetic data of the sort that 23andMe generates. In fact, they used exactly the kind of data that 23andMe generates; the southern African Sandawe, Hadza, and Khomani individuals were all genotyped on our 580,000-SNP v2 platform, just like many of our customers. 23andMe was proud to participate in this study, which was actually launched almost 15 years ago by 23andMe’s Senior Director of Research, Joanna Mountain. Dr. Mountain collected samples for the Hadza in Tanzania and initiated the recent sample collection in South Africa. We are also proud to note that the study’s lead and second author, Brenna Henn and Christopher Gignoux, are former members of Team 23andMe. Brenna and Chris made many contributions to 23andMe’s product, most visibly to the Maternal and Paternal Line features.






  • Eze

    Will 23andMe incorporate these interesting new African samples in the global similarity tool? Considering they are genotyped on similar chips as most customers it should be doable.

    For such a diverse continent there are only few populations.. The Sandawe could help discern East African ancestry better than the Kenyan Bantus currently used in the global similarity tool.

    Also, adding the HapMap Maasai samples would be great.

  • Ruby Stanton

    How can I contact you to make suggestions? I would very much like to be able to return to the same page on Ancestry finder when I am scanning for relatives.

    Ruby Stanton

  • joe17

    “A general principle is that you tend to find the highest levels of genetic variation in the portion of the population that’s been around the longest. That’s because sub-populations that split off from an original population start out with just a sample of the original population’s gene pool, and it takes a very long time for new genetic variation to build up in the new sub-population.”
    I have heard this idea many times and it makes sense but I am still kind of skeptical. Modern 100% Spanish people in Latin America even if their family has been living there for 500 years will have probably the exact same genetic makeup as Spaniards in Spain. The Same variation of Y DNA and mtDNA haplogroups. Why cant sometimes when there is a migration to a new area that the genetic variation doesn’t change from the ones that stayed in the source area and the ones who left.
    I do agree that it seems the maternal and paternal lines all humans today trace back to began in Africa it is probably very hard to say which part. Different non African families like west Eurasians aka Caucasians started to migrate back to Africa specifically North Africa 10,000′s of years ago probably mainly around 40,000 years ago. They became the only people in all of middle east, Europe, and north Africa(some mixing with sub Saharan Africans). So just because genetic studies of modern Arabians doesn’t show any evidence humans began there doesn’t mean they didn’t because modern Arabians ancestry came in migrations in the last 50,000 years. I think that modern Sub Saharan Africans do descend from humans that never left Africa because they are not apart of macro maternal and paternal lines all non Africans are apart of. I don’t think people can assume that it is so simple. That humanity began in Africa and modern sub Saharan Africans have 100% African ancestry. I have heard many pretty good hypothesis that Y DNA E dominate sub Saharan African y DNA haplogroups is from a back migration to Africa and that most of their mtDNA haplogroups and all non Africans are from migrations of Neanderthals into Africa and archaic humans in Africa. There are so many possibilities and genetic history of humans is extremely extremely extremely complicated.
    The dates of when the common paternal and maternal ancestor’s of all modern humans lived may be wrong I would just guess with the estimates they have at 200,000-400,000 years ago. The age estimates that humans left Africa(if they originally came from Africa) just 50,000-60,00ybp is way to recent. Because there are multiple human remains in Europe that are dated as over 41,000 years old and also multiple ones in Asia. There is a very well preserved what looks like to me a modern human skull in Israel that is dated as about 100,000 years old. I haven’t heard of any for sure modern humans skulls in Africa that are over 30,000 years old. I get annoyed when I hear archaeological evidence shows humans began in Africa. Is that because that is the main place people look and many assume that is where humans began. I have seen some skulls they say is evidence but they are very modern like but still not modern there isn’t a modern human remain from what I know in Africa that is over 30,000 years old. I agree it seems humans ancestral family probably began in Africa but I think it will be a mistake for people to assume that and not consider other possibilities.

  • joe17

    I have always wondered why did our family dominate the world and no other related breeds survived? There is evidence there was mixing between humans and Neanderthals(what exactly is a Neanderthal?) but still it would be very little so we mainly stayed pure human. Our family dominated everywhere we went there has to be constant reasons for why. I know there are very little remains from the past compared to what there once was. But I know that in Europe from about 30,000-40,000 years ago you see extremely accurate statues of humans and animals, flutes, very accurate cave paintings. It is obvious humans had arrived by that time. I dis agree with the theory the ancestor of all modern non Africans left Africa just 50,000-60,000ybp.
    The “Cro magnon” who definitely were not a united people but probably from multiple migration of West Eurasians into Europe from the middle east. Where not “our ancestors” meaning all modern humans ancestors like many say. They were pretty late in human history much closer to our time than the first humans in probably Africa 200,000-400,000ybp. There are 4 over 30,000 year old mtDNA samples from Cro magnon all had were in haplogroup U which is today exclusively west Eurasian aka Caucasian. One from Kostenki Russia about 33,000 years old had U2, three from Czech republic 31,155 years old one had U8 and two had pre U5. Also there is a dated 42,000 year old mtDNA sample from northeast China and had B4’5 haplogroup B today is exclusively Mongoloid meaning east Asian and Native American.
    mtDNA U5 is the dominate haplogroup of European hunter gathers from Mesolithic and Neolithic age. Today it is most popular and varied in Europe. Autosomal DNA from mtDNA U5 and U4 dominated European hunter gathers from Mesolithic and Neolithic age. Show that the what has been called North European like ancestry in modern Europeans probably originated in them. So at least some of the Cro magnon where partly ancestral to hunter gathers in Europe over 20,000 years later and to modern Euro’s. And I am not even mentioning the other west Eurasians middle easterns and north Africans. Showing this non African genetic family I think is over 50,000 years old. Same for the 42,000 year old mtDNA B4’5 in northeast China that person was In probably the ancestral family of modern Mongoloids which probably had been in central-north-east asia for over 50,000 years. I think many people by now know about 24,000 year old Y DNA R and mtDNA U in Siberia and autosomal DNA showing the individual had mixed west Eurasian and native American like ancestry. Ancient DNA is already showing evidence that there is now ay the ancestral family of all non Africans lived just 50,000-60,000ybp. I would say at the very least 70,000-80,000 years ago.

Return to top