Who We Are: #4 Denisovans

In Chapter 3, Reich talks about the discovery of the Denisovans, a sister archaic group to the Neanderthals that lived in eastern Asia. It all started out with a pinky bone found in a cave in southern Siberia. The DNA in the little bone was very well preserved – they got better info from that one bone than all previous Neanderthal DNA work. It’s an odd situation – we now know a lot about Denisovan genetics, but we don’t have a skeleton and have no idea what they looked like.

The Denisovans were closer to the Neanderthals than they were to AMH, but not by much. Apparently modern humans split with the common ancestors of Denisovans and Neanderthals about 700,000 years ago, while Neanderthals and Denisovans separated not much later. Almost a trichotomy. Something similar happened when AMH spread into Eurasia: quite early, maybe 50,000 years ago, we split into eastern and western branches. Probably it’s all geography.

With a high-quality Denisovan genome to work with, they tested whether any population seemed to be especially close. To their surprise, people from New Guinea were indeed significantly closer: they must have some Denisovan ancestry, somewhat surprising because New Guinea is a long way from Siberia. Although, since the Denisovans had hundreds of thousands of years to differentiate and spread over eastern Eurasia, far more time than we have had, it’s not really that odd.

In a sense, Reich’s team already knew this, although it sounds as if they didn’t know that they knew it. That same year, they had put out their big Neanderthal paper. Deep in the supplements, there was a table that showed genetic distance between a number of populations, including the Bushmen. Since the Bushmen had split off before other modern humans moved out of sub-Saharan Africa and settled the rest of the world, everybody in Eurasia should have been the same distance from Bushmen. And almost all were – except for people from New Guinea, who were noticeably more distant. Which mean that they had extra admixture from some archaic group, probably not Neanderthals, since there’s never been any sign of them that far east. My guess was Denisovan.

Sarah Phillips-Garcia, at the University of New Mexico, saw a sign of archaic admixture in people from New Guinea even earlier: she saw that something had introduced new variation in STRs (short tandem repeats) in New Guinea, variation you didn’t see in the rest of humanity.

Meanwhile Reich’s team were analyzing that Denisovan genome. They found that something like 5% of New Guinea ancestry is Denisovan, on top of the usual Neanderthal ~2% in Eurasians. This Denisovan admixure came from a Denisovan population that was fairly divergent from the genome found in Siberia. Some people in New Guinea have considered taking advantage of this high level of archaic ancestry: when facing potential criticism from the UN Human Rights Commission for some kind of intertribal strife, they were going to argue that it didn’t apply to them, because they weren’t human. I never saw that one coming. Similar amounts of Denisovan admixture are found in Australian Aboriginals, people in the Solomon Islands, and some Philippine Negritos.

You might think that this suggests that this major admixture maybe happened in Southeast Asia or nearby islands, but Reich accepted the argument of Yousuke Kaifu, an anthropologists who pointed to the lack of archaeological artifacts in the region that would support the notion of big-brained hominins in that region. But that’s silly. Bamboo. Probably it did happen in that region, likely in Sundaland. Very recent work suggests that there may have been two separate Denisovan admixture events: the first one contributing to Melanesians, the second to the much smaller amount of Denisovan admixture found in mainland Asia.

Linkage analysis showed that the main Denisovan admixture occurred a bit more recently than Neanderthal admixture, around 50,000 years ago. The possible second admixture seems to have happened more recently.

Reich talks briefly about some of the advantageous genes we picked up from Neanderthals and Denisovans. It was predictable [ John Hawks and I predicted it] that at least some of their gene variants would be useful – after all, they’d lived in Eurasia for hundreds of thousands of years and ought to be well-adapted to local conditions. Reich mentions the high frequency of Neanderthal-origin keratin gene variants in Eurasians, variants that affect hair and skin. He also mentions a particular Denisovan variant of the EPAS1 gene that has become common in Tibetans and helps them deal with life at high altitude. Since Tibetan adaptations to high altitude are considerably more effective than those of Andean Indians, more like those of species that had spent a long time at high altitude, there was an obvious possibility that Tibetans had picked up variants from archaic populations that had spent hundreds of of thousands of years in that environment. And lo, it was so.

The oldest-found human DNA is from ancient fossils at the Sima de Los Huesos caves in Spain, more than 400,000 years old. These skeletons looks like ancestors of Neanderthals, but they carry mtDNA that is much closer to that found in the Denisovan sample. There is an idea that Neanderthals may have mixed a bit with the ancestors of modern humans, say a quarter of a million years ago, and picked up a new form of mtDNA. Which they may have needed: Neanderthal population was small enough to potentially cause trouble in the long run ( because of inefficient selection) but that’s even more true for mtDNA, whose effective population size is four times smaller.

It also looks as if the Denisovans may have mixed with an unknown but very archaic population in East Asia, possibly descendants of homo erectus, which goes back 2 million years in Eeurasia.

One growing suspicion, originating from ancient DNA work and recent fossil finds, is that anatomically modern humans may not necessarily have originated in sub-Saharan Africa. The oldest probably-modern skeleton we have was recently found in Morocco. There is evidence that African populations, too, mixed with unknown archaic hominids (evidence for Pygmies, Bushmen, and also more typical West Africans such as the Yoruba, which implies that at least some parts of sub-Saharan Africa were occupied relatively recently by non-AMH populations. Homo Naledi, a small-brained homonin identified from recently discovered fossils in South Africa, appears to have hung around way later that you’d expect (up to 200,000 years ago, maybe later) than would be the case if modern humans had occupied that area back then. To be blunt, we would have eaten them.

Deep prehistory was always complicated: we just didn’t know much about it before. Ancient DNA analysis is the path forward.

This entry was posted in Archaic humans, Book Reviews, Denisovans. Bookmark the permalink.

64 Responses to Who We Are: #4 Denisovans

  1. So, our first known address was Morocco?

  2. Eff says:

    These are really interesting to read, Greg. I’m just a lay person trying to understand human origins — just curiosity. It seems like a complicated mystery to solve. Do you recommend Reich’s book for a person like me who wants to learn more on the subject?

    • gcochran9 says:

      Probably the most current book on the subject, and quite good except where politics interferes ( contemporary stuff). There will be other books out on this eventually.

      • Eff says:

        Well, it seems like it is impossible to write honestly about the sensitive topics of human nature. As you’ve noted in previous blog posts, even Reich has been attacked despite doing his darndest to stay within acceptable grounds (And unfortunately, Reich himself doing some attacking to distance himself from those “evil” biologists). I can filter out the parts where it is obvious that he is just being politically correct, if its not too excessive and the rest is good.

        Regarding other books to come, are you thinking of any in particular? Or just a general prediction?

  3. Hallie Scott Kline says:

    Any chance of DNA from the proto-Moroccan skeleton, or from Homo naledi?

    • GAGCAT says:

      Probably not – DNA degrades quickly in warm climates. That’s why the Denisovian DNA is from Siberia, and ancient DNA people love the Yukon (those guys are constantly digging up tundra looking for gold)

      Ancient DNA is usually short, so you can match (allowing a few base changes) the small fragments against the modern human reference genome and see the differences, but they’re not long enough to assemble a genome from scratch (de novo assembly). This means you’re not able to see the parts of the genome that are significantly different because they won’t match anywhere in a human.

      Actually this is a general problem – the human reference genome was built at great cost using long read sequencing technology, and now we genotype people cheaply using microarrays or short reads. Both of these are blind to large changes between the sample and the reference genome (built from the DNA of only a few white males).

      Given enough money, or waiting a few years for the price to come down, we can start assembling the genomes of different human ethnic groups from scratch, and check out any larger changes. The Koreans did it: https://www.nature.com/articles/nature20098

  4. Yudi says:

    “when facing potential criticism from the UN Human Rights Commission for some kind of intertribal strife, they were going to argue that it didn’t apply to them, because they weren’t human.”

    I want to see links to this story now!

    • Seconded. That’s too hilarious to be true.

      • gcochran9 says:

        It’s true – heard it from an anthropologist that spent most time in New Guinea. That’s nothing – back in the 70’s a bunch of locals offered to consume the corpses in the city morgue as a tourist draw: See Port Moresby, the Cannibal Capital ! In order to help out the local economy, you understand. However, local law enforcement didn’t buy it – all those volunteers were known former cannibals. Not disinterested local citizens.

      • gcochran9 says:

        You fundamentally misunderstand the world. Many of the funniest events are ones we could never have imagined. For example,

        [ Shahbal Shabpareh and his band Black Cats — a premier Iranian American pop group — have performed American hits with a Persian twist at upper-crust Iranian celebrations almost weekly for years.

        They’ve seen lots of lavish weddings, but one stands out as the most over-the-top.

        As guests enjoyed hors d’oeuvres outside the banquet hall, the bride was placed in a glass coffin. The groom fitted on a white half-mask. Then, the carefully planned Phantom of the Opera theme devolved into chaos.

        Condensation formed inside the coffin as guests delayed filtering in. When the groom finally took his cue to present the bride, the lid wouldn’t budge. Before long, he was slamming the glass trying to break through as the bride wailed inside, her makeup running down her face. It would be an hour before she was finally freed.
        For Shabpareh, the night crystallized the breakneck rise in ostentation at weddings hosted in recent years by L.A.’s wealthiest Iranian Americans. For some, party hosting can be a competitive sport, with spending used as a yardstick for status. Weddings boasting guest lists almost a thousand deep with price tags nearing half a million dollars are not unheard of. ]

  5. RCB says:

    Wild stuff. One wonders if we’re only scratching the surface.
    A Haldane quote comes to mind: “My own suspicion is that the Universe is not only queerer than we suppose, but queerer than we can suppose.”

    • Paul Morel says:

      Great quote. It reminds me, in it’s turn, of Vance’s Green Magic.

      “What realms of magic lie beyond the green?” asked Fair.

      “I can’t give you an exact answer,” replied the sprite, “because I don’t know. There are at least two more, corresponding to the colors we call rawn and pallow, and very likely others.”

  6. To make sure I understand, rather than AMH developing in Sub-Saharan Africa and then a batch leaving and settling the world, mating with some Neanderthals and Denisovans on the way, there is the possibility that the batch was not quite AMH – Homo sophomorus or something – which subsequently became AMH? The spread was then not only to Ultima Thula, Atlantis, and the Seven Cities of Gold, but back “In To Africa” as well?

    And we thought we were in trouble last time.

    • Jm8 says:

      I don’t believe so. It means AMH might have originated in North Africa (or possibly East Africa near Ethiopia, where early evidence of AMH is also found, or somewhere between/a broader zone including some of both regions) and then migrated from there to elsewhere in Africa, as well as later to Eurasia. AMH/homo sapiens likely did not originate in Southern Africa (where Naledi lived), but do seem to have arrived there by at least around 164,000 bc (which is when modern human artifacts begin to appear at sites like Pinnacle Point.

  7. Pau says:

    Citation for New Guinea anecdote? It’s an incredible one

  8. Ilya says:

    I’m pretty certain he meant it merely as sarcastic device…

  9. Pau says:

    It’s certainly possible that Yoruba carry only, say, 10% archaic ancestry, but from a MUCH more archaic line (ie, more “chimp-like”– think of possibly the quite archaic “homo naledis” specimen from South Africa). Alternatively if the archaic input was from a “later” specimen (only “slight drift” from BasalHuman to BasalAfrica) — maybe “BasalHuman” could be an intermediate specimen representing an “early sapiens-late heidelbergensis” hybrid pop. In that case the archaic admixture may indeed be as high as 31%.

    Neanderthals are the most “advanced” archaic population (from a phylogenetic standpoint– they are both genetically and culturally the most closely related population to sapiens). So 2-3% archaic admix in Eurasians is one thing. But we don’t quite know what kind of hominid we are dealing with here in sub-Saharan Africa.

    The percentage of Archaic ancestry and the source of archaic ancestry is up for debate still. What is most significant, however, is that formal statistics upon which QpGraph is based, explicitly REJECTED any type of model that didn’t involve sizeable archaic input into Yoruba. So there is definitely something at play.

    • capra internetensis says:

      Maybe the kind of archaic human that was around 15 000 years ago in Yorubaland? Iwo Eleru isn’t Homo naledi level archaic, but why the heck would it be.

  10. Archaic admixture for West Africans, but Eastern Africans lack it?

  11. GAGCAT says:

    If AMH developed in the Saharan savanna or more North, why didn’t some end up in North Africa as well as Sub Saharan Africa? If they did – wouldn’t we see: very old haplogroups, large genetic diversity between Moroccans, and a large genetic distance between Moroccans and everyone else? Do we?

    • Sahara became desert since. There could be major turnover and replacement

      • Jm8 says:

        Also, (in the case of the the Maghreb region) it was later repopulated primarily by (waves of) Eurasians (the majority of the ancestors of modern Moroccans and other Maghrebians). The original population would have been largely absorbed overtime (to the extent that it remained). Though a recent study indicates that the mesolithic Iberomaurusians of Taforalt (in the Northern Maghreb) were about one third subsaharan genetically (or rather one third descended from native African peoples who had never left Africa) and two thirds Eurasian (from back migrants originating in the Middle East). Later Eurasian migrations from Eurasia since increased the Eurasian genetic amount in North Africans to the higher level it typically is today (though the southern parts of North Africa/twhat is now the sahara had settlement waves from subsaharan Africa to the southas well during the meso-neolithic). The one native African third component in the Northern Iberomaurusians had affinities to West Africans but also to the genetically distant (and distinct) Hadza people indigenous to Tanzania, which suggests that the population it came (was not directly descended from either region but) from had diverged fairly early (was a basal branch) and have been established in North Africa for a long time. Some of the genetic diversity might have been pushed south or mixed with other waves from the savannah/sahel region (but as mentioned, modern North Africans have little of that component today due to later migrations/turnover/replacement, as mentioned—though some might remain in Africans the savannah and sahel regions, and to a lesser extent the southern parts of North Africa.

        • Jm8 says:

          “…(was a basal branch) and have been established in North Africa for a long time”

          the above should be “… and had been established in North Africa for a long time.”

      • Jm8 says:


        “…from had likely diverged in Africa fairly early—from the early homo sapiens stock/population (was a basal branch)…”

  12. epoch2013 says:

    The Ceprano cranium has recently been reconstructed and afterwards “its morphological features suggests that the specimen belongs to an archaic variant of H. heidelbergensis, representing a proxy for the last common ancestor of the diverging clades that respectively led to H. neanderthalensis and H. sapiens.”

    That is really surprising as Ceprano is in Italy and the cranium has been dated “to the beginning of MIS 11, providing a date bracketed between 430 and 385 ka5”, which means it is for all practical purposes contemporary to Sima de los Huesos, from Spain, which show clear cranial evidence for some Neanderthal influence (cf. Stringer). As is Tautavel man, by the way, even closer to Sima de los Huesos, and often called a Homo erectus sample. I somehow can not separate that from the odd mtDNA find there. Or could these Heidelbergs somehow be the source for Denisovans? Even if it is an enormous distance from the Denisova cave?

    Whatever the speculation, I think it is interesting to see Halfway Neanderthals existing so close to vanilla Heidelbergs.

  13. dermachine says:

    When you say we would have eaten them, do you mean that literally? If so, what is the best prior evidence for such behaviour?

    • Space Ghost says:

      Name one animal that humans don’t eat.

    • gcochran9 says:

      “Monkey brains” – see Indiana Jones and the Temple of Doom. Also bushmeat.

    • epoch2013 says:

      There is a large amount of evidence for paleolithic cannibalism showing it was a common practice.

      • Anuseed says:

        I’ve often wondered if we’ve actually evolved to like the taste of human flesh. People who have eaten it usually say it tastes like pork but a bit stronger. Does pork taste good? Yes it does!

        • epoch2013 says:


          I don’t think cannibalism is men hunting humans for prey. Calculations show that ice age Europeans were far better of hunting big game. But cannibalism may very well be a byproduct of warfare. In an ecology were life is hard, what else should one do with the slain bodies of the adversaries?

          • Jim says:

            Isn’t cannibalism more often found on islands having little in the way of large game?

            • gcochran9 says:

              Was pretty common in the Americas.

              • Jim says:

                Yes, in North America it was particularly common on the Gulf Coast. The Karankawa who lived in the Houston area where I used to live were notorious cannibals. It was also reported in the nineteenth century as being practiced by other Texan tribes.

                In the good old days of the eighteenth century the Texas Gulf Coast had cannibals, pirates, swamps and alligators. What more could you ask for?

              • gcochran9 says:

                In grad school, when asked for ethnicity, I said Karankawa.

              • Jim says:

                The Karankawa were pretty cool. They had longbows so powerful they supposedly would go clear through a person.

          • Anuseed says:

            LOL I’m not suggesting that we regularly hunted other humans, just that if cannibalism did have evolutionary purpose it would make sense for us to evolve to enjoy doing it.

            I think one function cannibalism had was intimidating other people. Cannibalism is always associated with depravity and badassary. Kill someone, you’re a badass. Kill someone then cut them up and eat them, you’re a real badass.

  14. MawBTS says:

    It’s an odd situation – we now know a lot about Denisovan genetics, but we don’t have a skeleton and have no idea what they looked like.

    This is the weakness with a morphological definition of species, like you mentioned earlier. Sometimes we don’t have a complete skeleton!

    And this is only going back a few hundred thousand years. Our knowledge of dinosaurs is probably so wrong it isn’t even funny. Argentinosaurus is believed to be the biggest dinosaur species that ever lived, but we only have a tiny part of a skeleton. How do we know it even existed? Maybe the bones belonged to a huge Apatosaurus. It’s hard to know much when you have only a few fossils, no DNA, and a sample size of one.

  15. Karl Zimmerman says:

    I wonder if the Red Deer Cave people are Denisovan – or at least a hybrid population with far more Denisovan than modern-day Papuans. They’re way too close to the present to be the source of the admixture of course, but they’re archaic looking, and in roughly the right region of Asia. They had pretty big molars too, which is similar to Denisovans.

    Unfortunately, the cochlear region of the skull doesn’t look like it was preserved, otherwise we’d have a fairly good chance of extracting DNA, given the time frame. Of course, as Reich notes, you can’t take human remains out of China, so it will be likely up to a Chinese lab to investigate this.

    Also, googling around, I found this article from last year, which reports that partial skulls have been found in China dating to 105,000-125,000 which look very much like Neandertals (and are large brained), but aren’t as robust. I have to wonder, however, if they really represent some hybrid population between the apparent failed OOA migration of AMH and Altai Neandertals instead.

    • dave chamberlin says:

      As Cochran said ancient DNA analysis is the path forward. The Denisovan cave has been at a constant temperature of 1 degree above freezing for hundreds of thousands of years. It is a reasonable assumption to think there are similar caves out there at lower latitudes but higher elevations that were homes to Denisovans and have tiny bone fragments that still preserve their DNA because they are refrigerators. The Denisovans covered a huge territory and we already know they thrived at high elevations and in places as cold as Siberia.

      The Reich lab has done an amazing job of working with tiny bone fragments and yielding huge results. There are an uncounted number of cold caves out there that could have housed Denisovans in Asia. It will take a different approach by field workers in the future.

      • Karl Zimmerman says:

        As Reich made clear in his new book, discovering that the petrous bone around the cochlea retains DNA at a much, much higher rate has opened the doors to ancient DNA in warmer climates. So long as we can find a reasonably intact skull, we would have a good chance of getting usable DNA even at lower elevations in Southeast Asia.

        • Yudi says:

          He also says they only learned that in 2015. This makes me wonder if there are other bones that have as much or even more DNA…

          • dearieme says:

            “He also says they only learned that in 2015”: golly, I thought I’d known it for longer. It’s funny how quickly you can come to believe “everyone knows that ….”.

      • The Denisova cave is at extreme south of Siberia. If they “thrived” wouldn’t they make a visit to America too?

      • dave chamberlin says:

        There is a very slow reaction from human evolution field workers to this new science that the Reich lab is leading the way in developing. We don’t know diddley squat about Denisovans and we won’t until field workers start a methodical search in those locations that Denisovans possibly lived and left their DNA in a preserved state.

        100,000 year old refrigerators. We don’t look for inner ear bones, we don’t look for skulls, we don’t even use our eyes. Sample the dirt floor of 10,000 freezing cold caves in locations they possibly lived for hominid DNA and when you have a hit then and only then is patient and methodical search in that location conducted.

        Will some rich old geezer thoughiiay disgusted with his spoiled stupid children please put Greg Cochran in charge of a search that actually does this. Humanity will thank you later.

  16. Pingback: Chicago Boyz » Blog Archive » Ancient DNA, genetics and race,

  17. Greying Wanderer says:

    fascinating stuff

  18. John Bull says:


    What do you make of ‘Garstang’s Hypothesis’ that is the theory that vertebrates, chordates, ultimately evolved from sessile filter-feeders resembling extant sea-squirts?

    Can’t help thinking of a strange link between this and all the hominid DNA work.
    Since I first read of Garstang’s Hypothesis it has haunted and fascinated me in equal measure.
    I can’t help thinking that, if true, it’s the most mind-blowing fact in biological science. On a par with quantum theory in physical science.

  19. dearieme says:

    “sessile filter-feeders resembling extant sea-squirts”: I’m sure we’ve all met someone who fits that description.

  20. thomas says:

    please write a textbook about evolutionary genetics before you croak.
    I will be your first customer, need at least 50 copies.

  21. dataintelligence says:

    Primitive girl’s parents were two different human species https://www.theinforadar.com/2018/09/primitive-girls-parents-were-two.html

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s