Denisovan autosomal DNA was found archaic hominin remains in a Siberian cave a few years ago and is a match for traces of archaic admixture found in modest percentage in modern Australian aborigines and Melanesian populations (and populations derived from these populations), but in no other modern human populations, as shown in the following map (via John Hawks blog).
Recent analysis has determined that the Denisovan DNA is admixed with Neanderthal DNA (who were contemporaneous with the population whose remains were found in Siberia from ca. 30,000 years ago) and not too far to west of them and in the same cave at a later date), just as non-African modern humans are today. More intriguingly, there is also evidence in the Denisovan autosomal genome of admixture with another unidentified archaic hominin species. They were 17% Neanderthal and 4% unknown archaic hominin.
The $64,000 question is which species of archaic hominin account for the core of the Denisovan DNA and the unidentified archaic admixture in it, respectively.
Wednesday, May 1, 2013
Some Important Open Issues In Human Prehistory . . .
1. When and where did Neanderthal admixture take place?
One, two and three admixture event scenarios are all plausible (and, of course, more complex scenarios are also possible).
The simplest model assumes that admixture took place ca. 100 kya to 75 kya at a time when the archaeology shows the two species overlapping in the Levant and the effective population size would have been smallest, and not thereafter. But, this would mean that a schism into West Eurasian and East Eurasian populations would have had to happen very early on to prevent the admixed Neanderthal genes from coming closer to fixation between the proto-West Eurasians and proto-East Eurasians than differences in which Neanderthal genes are found in each population reflect.
Another one admixture model puts the date closer to 50-75 kya when the modern human population is at a post-Toba low point after an initial Out of Africa surge, perhaps in an Arbian or Persian Gulf refugium. This reduced effective population size, and allows for a sustained period of declining population during which genetic drift can loose some of the Neanderthal inheritance and give rise to Founder effects not likely to be seen in a rapidly expanding population that has reached fixation.
In a simple two event model, West Eurasians and East Eurasians split before admixture takes place and Neanderthal admixture takes place in parallel processes that produce similar overall levels of admixture in each clade of Eurasians in different places - perhaps one in Anatolia, and another in Arabia, Persia or South Asia. This also has the virtue of driving down effective population sizes in each source population.
A three event model combines these two models, which some admixture taking place early on and pre-schism and some taking place later in parallel.
An example of a more complex model would be one with a common admixture event, additional East Eurasian admixture, and then additional West Eurasian admixture that is diluted in the Upper Paleolithic to Neolithic transition by West Asian populations that did not experience the additional West Eurasian admixture experienced by Europeans with prolonged co-existence with Neanderthals.
2. Were there cases of archaic admixture in Africa?
Preliminary population genetic analysis and possibly one Y-DNA haplogroup A00 that looks older than the species make this look plausible and may have happened in two or three separate cases there quite recently.
The case for Neanderthal and Denisovan admixture where observed is in my opinion rock solid and not adequately explained by any other mechanism (such as deep population structure in Africa within modern humans).
The window of time in which additional non-African, non-Neanderthal, non-Denisovan admixture could be discovered in modern humans alive today hasn't completely closed but is well on its way to doing so. It is quite clear that this isn't present outside modern relict populations of Eurasia and the Americas.
6. What archaic hominin species is Denisovan DNA associated with?
Denisovan DNA was found in a Siberian cave. Denisovan admixture is found from roughly the Wallace line and beyond. At least two know archaic hominin species were present in between: Homo Erectus, in most of that region, and Homo Florensis on Flores and perhaps a couple of neighboring islands only. Theories about the existence as a separate species and range of Homo Heidlebergus is also sketchy.
There is no solid indication of Neanderthals beyond South Asia, but they aren't entirely ruled out, particular via a Northern route to the Denisovan cave. Some clade analysis of the Denisovan DNA suggests a clade shared with Neanderthals apart from Homo Erectus. . . .
9. How did Homo Erectus go extinct?
While the evidence is very thin indeed on this point, extinction due to warfare and/or inability to complete for food and territory with modern humans upon first contact, outside of Flores where a cooperative mode emerged, possibly boosted by the Toba erruption or a climate shift seems most plausible.
10. When, if ever, did Homo Florensis go extinct?
The earliest possible date for the extinction is about 10,000 years ago, which is the date of the oldest skeletal evidence, but there is good reason to think that they persisted even after first contact with Europeans and that there may even be a tiny population of this species extant on one known part of one Indonensian island. . . .
Posted 3:29 PM
"There is no solid indication of Neanderthals beyond South Asia"As far as I'm aware there is actually no evidence for Neanderthals anywhere within South Asia apart from some disputed presence in the far northwest. . ."How did Homo Erectus go extinct?"I think that technically it didn't become extinct. Collectively we probably have many genes picked up by 'modern' humans as they passed by and mixed with Homo erectus populations.
I am defining extinct to mean "cease to exist as a separate species if the resulting species is predominantly the other part of the hybrid." Also, it doesn't appear that there was any admixture in mainland Asia where most H. Erectus must have lived. They went completely extinct in most of their range.
"it doesn't appear that there was any admixture in mainland Asia where most H. Erectus must have lived. They went completely extinct in most of their range".
I actually believe our evolution has been far more complicated than the simple story usually presented.
"6. What archaic hominin species is Denisovan DNA associated with?"
I don't know if you've noticed John Hawks' latest post, but just in case you haven't here it is:
Some interesting aspects:
"And the comparison revealed another surprise: Four percent of the Denisovan genome comes from yet another, more ancient, human—'something unknown,' Pääbo reported".
So we have yet another ancient Homo species that could interbreed at least with another species that could interbreed with 'modern' humans. In fact the whole thing gets even more complicated:
"With all the interbreeding, 'it's more a network than a tree,' points out Carles Lalueza-Fox" . . .
Re the Denisovan admixture with yet another unknown archaic hominin species (i.e. not modern human and not Neanderthal), this really is exciting and mysterious. Fitting two non-Neanderthal, non-modern human archaic species into Eurasia (presumably pre-OOA), is not an easy task. I am inclined to think that one of Asian H. Erectus (probably Denisovan) and the only other real candidate I can come up with as the second archaic hominin species would be H. Ergaster (sometimes called African H. Erectus and believed to be ancestral to all known Eurasian hominin species), or Homo heidelbergensis aka Homo rhodesiensis which was antecedent to Neanderthals in essentially the same range and were probably antecedent to modern humans in Africa. There is no attested instance of H. Ergaster outside Africa, although arguably this is due to geography influenced taxonomy since not everyone agrees that H. Ergaster and H. Erectus are the same species.
If Denisovans are a late evolutionary phase of H. Erectus that first admixed with H. Heidelbergensis to the west of their range and then late admixed with Neanderthals who replaced them to the west of their range (at least amongst N. Asian Denisovans) (with H. Flores being a pygmy species of H. Erectus due to island dwarfism that probably lacked Neanderthal or H. Heidelbergensis admixture), then you start to have a pretty coherent story.
The plot would really thicken if the trace unidentifical archaic admixture in the Denisovans and the trace archaic admixture in either certain Khoisan or certain pygmies in Africa were a match - pointing strongly to the possibility that there were relict populations of H. Heidelbergensis both in and outside Africa through the Upper Paleolithic (with the other of the two separate unaccounted for archaic admixture in African populations possibly being a trace of H. Ergaster (I'd guess Ergaster for the Khoisan and Heidelbergensis for the Pygmies in light of the probably archaic admixed Y-DNA haplogroup found in West Africa near Pygmy territory and based on the need for Heidelbergensis to be closer to Eurasia than Ergaster - also we know that Ergaster's range extended to Southern Africa based on fossil evidence but have no evidence of Ergaster in jungle settings - perhaps due to preservation conditions, but perhaps because they never lived there). If H. Heidelbergensis evolved north of the Congo and never migrated to Southern Africa, then H. Ergaster would be the most evolved archaic hominin species in Southern African upon first contact with modern humans and might have survived for some time as a relict population in marginal environments that the Khoisan would eventually themselves being pushed into.
"Fitting two non-Neanderthal, non-modern human archaic species into Eurasia (presumably pre-OOA), is not an easy task".
I don't see a problem there. After all Denisovans and Neanderthals were very close neighbours, leaving a considerable amount of Eurasia for other species.
"the only other real candidate I can come up with as the second archaic hominin species would be H. Ergaster (sometimes called African H. Erectus and believed to be ancestral to all known Eurasian hominin species)"
My vote goes towards an East Asian separate species, perhaps descended from 'Pekin Man'. When the information on the EDAR370 variant came out the authors suggested one possibility was introgression from some unknown East Asian species as well as a possible mutation within Homo sapiens. I thought the former scenario quite likely.
"If Denisovans are a late evolutionary phase of H. Erectus that first admixed with H. Heidelbergensis to the west of their range and then late admixed with Neanderthals who replaced them to the west of their range (at least amongst N. Asian Denisovans) (with H. Flores being a pygmy species of H. Erectus due to island dwarfism that probably lacked Neanderthal or H. Heidelbergensis admixture), then you start to have a pretty coherent story".
Possible. But I think you are hugely influenced by your desire to fit the Australasian Denisova genetic trace with SE Asian Homo erectus. At this stage we have no evidence at all that SE Asian H. erectus had any significant genetic connection to Denisovans in the Altai. That is a huge distance and for most species with the apparent limited mobility of pre-modern humans would be sufficient to give rise to separate species. The distance involved certainly seems sufficient to have given rise to separate species within other genera.
I am influenced by the desire to figure out why traced of autosomal DNA sequences found in remains in Siberia are found in Australasian populations, but nothing in between. Some species or closely related group of hominin species necessarily had a range that extended from the far side of the Wallace line to Siberia sometime in the time frame from 100,000 years ago and 30,000 years ago. Despite the fact that the distance seems incredible, that part of the story is pretty much an undisputable fact. We can't say with certainty this species was indeed H. erectus, but no other known archaic hominin species had a range even vaguely that great in that region.
If you want to argue for a new species, my inclination is to say "show me a body."
There is archaeological evidence of S. Asian H. erectus from the island of Java to China from close to 1,000,000 years ago to 100,000-200,000 years ago or so (I've seen better dates on that than I have provided, and I'm sure that this isn't exactly right, but the material points are (1) that there is no archaeological evidence of H. erectus in Siberia or NE Asia other than the Denisova remains (to the best of my knowledge), (2) that H. erectus remains are present over an expansive range in Asia over a long time span of many hundreds of thousands of years, and (3) there is no clear overlap of dates in hominin remains between archaic and modern human populations in Asia, apart from H. Florensis on the island of Flores itself. Furthermore, H. Florensis is the only other species of archaic hominin for which we have clear evidence in the form of hominin remains east of India.
The map below (via For what they were . . . we are) sums up the existing evidence of archaic hominins in Asia prior to the Toba explosion.
The conventional wisdom about Peking Man is that this is a representative of H. erectus (the type fossil of which is Java man), both of which are more than 700,000 years old, and that they are not separate species. The distance between these two sets of remains is greater than the distance from the Peking man location to Denisova, so the possibility that a single species of archaic hominin could spread so far is well supported.
The correct species for some late East Asian hominin remains (ca. 100,000 years ago) such as the Llujiang remains (139,000 to 168,000 years ago) and Zhirhendong remains (100,000 to 110,000 years old) are not definitively determined. This is highly relevant because the Denisova remains and any first contact with modern humans would have been closer in time to these remains than any other archaic hominin remains in Asia to the east of India.
The late (ca. 100,000 years ago) East Asian hominins could be a new species entirely that is absent from the archaeological record elsewhere. They could be a late phase of H. erectus. They could be a modern humans suffering from diseases or disorders or with remains distorted by pressure or warping over the millennia or could have been inaccurately dated as older than they actually are due to some methodological problem. Or, they could be highly admixed modern human-archaic hominin hybrid individuals.
Given mountain evidence for a first wave of Out of Africa migration ca. 125,000 to 100,000 years ago, and an absence of any other modern human remains east of India that are anywhere near that old (despite the huge ecological impact of modern human settlement in many other places), I am strongly inclined to believe that either the dating or that the tentative classification of these remains as modern human is wrong. The circumstances of discovery in the case of the Llujiang remains were not a meticulous archaeologist's dream of care and optimal methodology, so some degree of skepticism about methodology issues is appropriate in that case. The Zhirhendong remains discovered in 2007 are more securely dated and more professionally evaluated, for example in this paper by Liu, et al in 2010 also discussed here), but I don't share their confidence that the highly fragmentary remains are really modern human and the possibility that a more modern fragment of human bones were somehow muddled in with older dirt, rock and fauna remains from previous archaic hominin occupants of the cave is not ruled out as definitively as one might hope for remains that are outliers by more than 60,000 years from all other modern human remains for many hundreds of miles around the site (many of those at least 40,000 years old themselves and hence not in locations where preservation of remains for thousands of years was a seriously problem).
These remains are also 60,000 years or more removed from any lithic tool sets found in the region which would be associated with modern humans or even with Neanderthals. Normally, lithic tool remains are far more abundant than teeth and jaw bones.
Modern human-like features arising in populations descended from H. erectus due to convergent evolution seems like at least as likely an explanation, and perhaps both modern human admixture seen in Melanesia and Australia and the directly sampled Denisova DNA involved this highly evolved (in a mostly convergent direction) version of H. erectus, rather than more primitive archaic species like the H. erectus type fossils.
The presence of two separate kinds of archaic admixture in Denisova DNA strengthens the case for possibility that there were hominin species for which we have no remains was present in Asia. The missing species could be the Denisova itself, or the Denisova remains could be from H. erectus or H. florensis, with the unidentified archaic species being one for which we also have no archaeological finds of remains.
The interpretive bias that I am guilty of and I admit this, is that the conjectures and narratives that I am piecing together are "minimalist". I have preferred explanations that most parsimoniously explain the data with a minimum of missing species or data points implied by that narrative. It could indeed be the case that the real story is much more complex. But, I am reluctant to consider possibilities that imply some very major new discoveries that have yet to be made merely because they are "possible" when there is no real positive evidence to support them.
Put another way, my Baysean prior is that the data that we have already, while not complete, is quite comprehensive, rather than being the mere tip of the iceberg with new unexpected discoveries lurking in every cave and every eroded away landslide.
I have no doubt that there are new discoveries to be made, but the evidence that we are missing an entire species or genera of archaic hominin that stretched from Indonesia to Siberia at some point in a range of tens of thousands of years seems relatively low in my mind after hundreds of thousands of years of H. erectus stability over a range almost as vast in Asia. A missing species is by far the more interesting possibility, but it also seems speculative and if we must speculate, assigning the possibility to a relict population of a known archaic hominin species that has been attested in Eurasia seems like the most likely possibility.
John Hawks' latest post on the subject does some analysis of the possible species affinity of the Denisova hominins:
The first reported third molar has length and breadth dimensions within the size range occupied by australopithecines and early Homo, both H. habilis and H. erectus. There are no distinctive morphological characters that would allow it to be assigned to any taxon. . . .
The similarity of Denisova and Neandertal genomes suggests that they emerged from a single population. This possibly was the early Middle Pleistocene population of Eurasia. Green and colleagues  derived Neandertals from a common ancestor with living Africans only 250,000-400,000 years ago. A model including the Denisova data, reported by Reich and colleagues  puts the emergence of a Neandertal-Denisova clade at between 190,000 and 520,000 years ago, and the divergence of the Neandertal and Denisova branches around 50,000-100,000 years later. This is later than the time of the last unequivocal H. erectus fossils, and more than a half million years after the Trinil individual -- type specimen of Homo erectus -- lived. Based on this chronology, the Denisova genome does not represent Homo erectus or any hominin population derived from the initial diversification of Homo.
A slower mutation rate might bring the most recent specimens of Homo erectus into the temporal range of the Denisovan population, by elevating the time of divergence of Denisovans, Neandertals and contemporary African peoples. Whole-genome sequencing of human parent-offspring pairs, as well as resequencing of more limited regions of the genome, has suggested a slower rate of substitutions , . This would be consistent with a much earlier diversification of Denisovan and Neandertal genomes, making it possible that they came from an early Middle Pleistocene adaptive radiation.
The Denisovans were too closely related to living humans to represent Homo erectus, as it is currently understood. Homo erectus occurred widely in Asia, including China and Java, and Africa during the span from 1.95 million to 750,000 years ago. In China and Java, fossils attributed to Homo erectus persisted until 200,000 years ago. There is no unequivocal fossil of Homo erectus after 200,000 years ago.
The mtDNA sequence of the Denisova genome is an outgroup to a clade including both humans and Neandertals, and appeared to branch from our ancestors roughly a million years ago. In addition to the finger bone, Reich and colleagues recovered the mtDNA of a second individual from Denisova Cave, represented by an isolated third molar . Reich and colleagues  showed that the mtDNA divergence between Denisova and the modern-Neandertal clade is deeper than expected given the nuclear genome genealogical divergence. They also showed that the nuclear genomes of Neandertals and Denisovans are somewhat closer than either is to the majority ancestors of living people. They discuss two possible explanations.
One scenario a mixture of the Denisovans with a more ancient Pleistocene population, followed by introgression of a more ancient mtDNA clade into the Denisovans. This would assert an ancient structured population preceding the origin of Denisovans, presumably from one of the Middle Pleistocene populations of Africa or Eurasia. A second scenario is incomplete lineage sorting, in which an earlier mtDNA divergence was captured by the Denisova and Neandertal populations at the time of their divergence and differentially lost from them. . . .
It has become possible to make some good estimates of demographic history using only a single diploid genome, using a technique developed by Li and Durbin. Meyer and colleagues applied this technique to the Denisova genome, finding that its genetic history contrasts with that of living human populations. Whereas African populations and living humans outside Africa share signs of population growth in the last 100,000 years, the Denisova genome seems to have come from a population that had long been small, and may have been shrinking.If the Denisova are too closely related to modern humans to be H. erectus, then H. erectus becomes a prime candidate for the 4% undetermined archaic admixture in the Denisova on top of the 17% Neanderthal admixture that may simply have reflected the geographic location on the farthest fringe of the Denisova range immediately adjacent to and overlapping with Neanderthal territory. In terms of relatedness, seeing the Denisova sharing a clade derived from H. Heidelbergensis with modern humans and Neanderthals, all of which diverge out of Africa or West Eurasia from H. erectus at a much later date than H. erectus, seems to be a pretty decent fit with the autosomal and mtDNA data.
But, what is going on from 200,000 years ago (the last unequivocal Homo erectus fossil) until 40,000 years ago (when modern humans are unquestionably dominant in all of SE Asia, East Asia, Melanesia and Australia and Denisova admixture in populations that are Denisova admixed today appears to have been complete)?
Did the Denisova entirely replace H. erectus over its entire range, only to be replaced in turn by modern humans? If so, what made this replacement possible and why did they otherwise leave such a slight footprint in the fossil and paleofauna/paleoflora record? Could it be that H. florensis, were not pygmy H. erectus and were instead a completely different Denisova species?
The Denisova Diaspora Model
Alternately, perhaps the Denisova were not a replacement for H. erectus. Instead, they may have been a somewhat more adaptable and resourceful secondary hominin species contemporaneous with H. erectus and Neanderthals that could not compete directly with either H. erectus (whom they did not replace over most of the H. erectus range) or Neanderthals, perhaps due to their smaller size (on the theory that H. florensis are type examples rather than pygmy examples of the species), for example.
Perhaps, instead, the Denisova sought habitats such as the Denisova cave and islands on the far side of the Wallace line in Indonesia that were desirable to them because they were adaptable enough to get there and live there, while H. erectus was not capable of living there and competing with them.
In this analysis, it was H. erectus and not modern humans that kept the Denisova from becoming numerous on the mainland Asian side of the Wallace line.
This would also imply that H. erectus had almost no admixture with humans anywhere, perhaps because, unlike the Denisova and Neanderthals, they were simply too fully speciated from modern humans to produce fertile hybrid offspring. The 4% unknown archaic admixture in the Denisova, which I have hypothesized is Homo erectus admixture into the Denisova, is suggestively about the same as the amount of Neanderthal admixture in modern human Eurasians, suggesting that perhaps this intermediate species may have been about as distinct from H. erectus as modern humans were from Neanderthals (something that is approximately true if they are a relict population derived from late pre-Neanderthal H. Heidelbergensis). The higher level of admixture in the Denisova cave with Neanderthals (17%) may have reflected not just proximity, although surely that was one factor, but also a considerably higher level of compatibility and similarity between the Denisova and Neanderthals than between Neanderthals and modern humans.
One can perhaps imagine the Denisova as a relict pre-Neanderthal population of Europe that was replaced by the Neanderthals entirely there where the Neanderthals had evolved to be better adapted to the local conditions, but persisted in distant environments of Asia that Neanderthals were less suited to whenever they could find territory not definitively claimed by well established H. erectus populations.
This Denisova diaspora model is a good fit for both the odd geographic distribution of the Denisova range and for its small and declining demographic history which is a poor fit for a model in which the Denisova replace H. erectus over a wide range prior to the arrival of modern humans.