March 24, 2010

Ancient mtDNA from Denisova Cave

Nicholas Wade writes in the NY Times:
But they say the genetic material extracted from the bone, an element called mitochondrial DNA, belonged to a distinct human lineage that migrated out of Africa at a different time from the two known archaic human species. Homo erectus, found in East Asia, left Africa two million years ago, and the ancestor of Neanderthals emigrated some 500,000 years ago. The number of differences found in the child’s DNA indicate that its ancestors left Africa about one million years ago, the researchers say. Their report is published online in the journal Nature.

...

The finger bone was found in a layer laid down on the cave floor between 48,000 and 30,000 years ago, according to radiocarbon dating. At that time, toward the end of the Pleistocene Ice Age, which ended 10,000 years ago, the climate was probably much colder. The people of the new lineage presumably wore clothes, Dr. Krause said, because chimpanzees and gorillas cannot withstand much cold, suggesting that fur alone is inadequate protection.
Nature doi:10.1038/nature08976

The complete mitochondrial DNA genome of an unknown hominin from southern Siberia

Johannes Krause et al.

Abstract

With the exception of Neanderthals, from which DNA sequences of numerous individuals have now been determined1, the number and genetic relationships of other hominin lineages are largely unknown. Here we report a complete mitochondrial (mt) DNA sequence retrieved from a bone excavated in 2008 in Denisova Cave in the Altai Mountains in southern Siberia. It represents a hitherto unknown type of hominin mtDNA that shares a common ancestor with anatomically modern human and Neanderthal mtDNAs about 1.0 million years ago. This indicates that it derives from a hominin migration out of Africa distinct from that of the ancestors of Neanderthals and of modern humans. The stratigraphy of the cave where the bone was found suggests that the Denisova hominin lived close in time and space with Neanderthals as well as with modern humans2, 3, 4.

Link

69 comments:

Andrew Oh-Willeke said...

Wow! What an amazing find.

The case that this is a population different that Homo Erectus really calls for more evidence, however.

While it is usually safe to make the assumption that mtDNA lineages are mostly a product of random genetic drift, this assumption seems particularly problematic in the case of a hominid population living in Siberia during an ice age at the far northern range of pre-Cro-Magnon hominid inhabitation. Mitochondria are among other things important regulators of our metabolism and metabolism would be subject to powerful selective pressures in this chilly environment. So normal mtDNA clock estimates may be suspect.

There is also no particularly good reason that to think that homo erectus left Africa in a single wave. Hominid remains thusfar indicate that Homo erectus remained morphologically quite static for millions of years, was present not far from land bridges to Eurasia for most of that time period, had a range that extended South to Southern Africa and East to Java, and lived a low population density hunter-gatherer existence. Even the earliest evidence for homo erectus using fire are only about 750,000 years ago. They weren't effective enough as hunters to produce mass extinctions either.

Given the lack of population density, great capacity to travel long distances compared to other primates, their limited dent in the food supply, and their limited package of possessions (that flowed from the crude stone and wood tool set associated with homo erectus), it is would hardly be surprising for new individuals to wander more or less continuously out of Africa for millions of years, without being held back by existing populations. Sparse population density hunter-gatherers can't hold territory.

Andrew Oh-Willeke said...

Following up on my comment before, evidence from <a href="http://dienekes.blogspot.com/2009/11/time-independent-evolutionary-mtdna.html>ancient and modern penguin mtDNA</a> suggests that mtDNA mutation rates are two to six times greater in cold areas that is normally assumed. This implies that a mutation that looks like it is 1,000,000 years ago could in fact come from an early Neanderthal lineage.

This would be consistent with the stratiography that show material goods from the Upper Paleolithic era accompanying the sample and little evidence of strata disturbance in the cave.

Our knowledge of ancient Neanderthal DNA is still limited by a geographically distinct small sample size, in a population that is know to have had at more than one subcontinental subpopulations and a subpopulation this far to the East would have been distinct again from the ones from which our other samples are drawn.

A high mutation rate for mtDNA and a hetrogenous out of African Neanderthal population with founders effects in subpopulations in Eurasia seems like a far more parsimonious explanation of the evidence than the five species at once theory suggested in the New York Times article.

If Neanderthal out of Africa is 500,000 years ago, and the common ancestor of known Neanderthal DNA and this sample is really 83,000 to 250,000 before that, rather than 500,000 years earlier, then the ideas that ancestors of both lineages could have been found in a pre-out of Africa Neanderthal population isn't very far fetched.

Maju said...

"The case that this is a population different that Homo Erectus"...

For me she looks like H. erectus. What else migrated out of Africa c. one million years ago?

"Following up on my comment before, evidence from ancient and modern penguin mtDNA suggests that mtDNA mutation rates are two to six times greater in cold areas that is normally assumed" [link corrected].

Actually the overall data would suggest if anything the opposite: a slightly higher mutation rate at warm zones. The pinguin data, along with many other, rather suggest that the molecular clock as most geneticists use it nowadays is wrong: it ticks too fast. It should be much slower.

However it can also be too irregular to be reliable at all.

n/a said...
This comment has been removed by the author.
n/a said...

I see no reason to assume "some new creature" "carried this mitochondrial genome out of Africa about a million years". Modern human mtDNA can say nothing about where our ancestors lived prior to its convergence, and this finding is very open to multiregional interpretations.

Maju said...

I'm double checking my data and what really means, IMO, is that the molecular clock (by usual standards) is wrong and everything is about doubly older.

H. sapiens and H. neanderthalensis must have began their divergence with the Acheulean migration, which is the last technocultural link between Eurasia and Africa before H. sapiens.

The Acheulean migration happened some 900,000 years ago (Lycett 2009). So the X-woman can perfectly belong to an older H. erectus migration c. 1.8 or two million years ago and the proportionality of the MC remains intact.

What don't remain intact are the assumptions about the speed of the MC, which should be half as fast (roughly) than the usual standards.

Mason said...

Here's the whole paper:

http://viewer.zoho.com/docs/p5bbbD

Maju said...

Thanks. :)

eurologist said...

Or, alternatively, this is another heidelbergensis derivative. I have argued before that (i) the beginning of heidelbergensis is somewhat ill-defined, and (ii) outside the extreme-cold adapted northwest Neanderthals, other heidelbergensis-derived humans would have survived in the East (and would be responsible for the archaeological record there that does not match erectus.

There really is very little reason to restrict heidelbergensis to after ~600,000 ago. In fact, several decades ago, when I first got interested in this topic, standard assumption was that heidelbergensis was at least 800,000 years old. Also, there is simply too much of a gap with ergaster if heidelbergensis' beginning is dated late, and the split between Neanderthals and modern humans gets too close to or even overlaps with heidelbergensis' beginnings (which makes no sense, at all).

Maju said...

I don't believe much that H. sapiens derives from H. heidelbergensis/antecessor, which is a species only well described in Europe. For me it derives from H. rhodesiensis, described in Africa and only there.

There's no visible archaeological flow between Eurasia and Africa after the Acheulean migration some 900,000 years ago until H. sapiens enters the scene. If H. heidelbergensis would have migrated in either direction a lot later than the Acheulean we should be able to detect some related technocultural flow. But both Mousterian and the African technologies of H. sapiens evolve separately from an Acheulean background.

Maybe I'm wrong but the evidence points to that and the only obstacle are "beliefs" such as the molecular clock (which is not evidence) or the belief that H. heidelbergensis and H. rhodesiensis are related (which is also not supported by any evidence but their parallel cranial development, a trait generic to all H. erectus senso lato - i.e. a clear evolutionary pressure).

eurologist said...

I don't believe much that H. sapiens derives from H. heidelbergensis/antecessor

Neither do I - perhaps you misunderstood me. I was merely pointing out the often-underestimated antiquity of heidelbergensis in Eurasia, and the possibility of both Neanderthals and (earlier) Eastern population (as this find) to be their derivatives - as other archaeological finds have always pointed to, IMO.

You can't on the one hand demand a slower molecular clock and put the Neanderthal-modern human separation back, and at the same time dismiss actual archaeological data on heidelbergensis.

That is, once you realize earlier times are more realistic, heidelbergensis becomes more important whichever way you look.

eurologist said...

I would tend to postulate something like this:

- heidelbergensis was in Europe from about 1,000,000 years ago to ~ 35,000 years ago, including all of its derivatives;
- Neanderthals separated from the modern human line roughly when Northwestern heidelbergensis completely separated from Africa, i.e., ~600,000 years ago; on the other hand,
- Eastern/ Southeastern heidelbergensis became isolated more shortly after arrival, and did not share most Neanderthal developments - thus putting the separation to modern humans closer to ~ 1,000,000 years
- some (but very limited) genetic flow likely happened between all three/four

Maju said...

It's very difficult, or rather impossible to clarify the matter. I tend to think in the line that the distinction between H. erectus, heidelbergensis, antecessor and rhodesiensis is very subtle and mostly linked to brain size (which tends to grow with time, maybe with some jumps more related to chronology than geography). Only when Neanderthal and Sapiens species take off, we can really see some overall cranial differences and these are markedly different in either species.

The only thing it could help would be to test the DNA (if possible) of the various specimens across the world, what could give a more clear picture.

Alternatively, I could accept a back-migration to Africa but supported by something (technological or cultural flows we can identify at least, as these are found more often than body remains).

Also, on a side note, I'm not aware of any "South Eastern" H. heidelbergensis or anything of the like. The only one towards the East would be the Dali skull which is from North China, right?

German Dziebel said...

It's interesting that their fig. 3 apparently represents a seriated dendrogram, so that the location of the clusters literally represents the degree of similarity between sequences. Denisova and Neandertal sequences are closer to Australian aboriginal sequences than to Bushmen and Pygmy sequences. I'm surprised that they didn't include the Mungo man sequence from 40K year old remains, which is more divergent than African sequences and which falls right between modern Australian and Neandertal sequences. This all potentially looks like Eurasian/Australian humans may contain a direct link to Eurasian hominids, with African humans out of the picture.

pconroy said...

German,

That's exactly what I thought too - more evidence that Mungo Man might actually not be alone - that there might be whole populations out there who lived into almost modern times, but who contained mtDNA indicating much earlier lineages...

pconroy said...
This comment has been removed by the author.
German Dziebel said...

"more evidence that Mungo Man might actually not be alone"

Yes. Typical out of Africa phylogenies are built off of sets that encompass lineages that have survived into the 21st century, not the lineages that went extinct. Population size was apparently much lower in East Eurasia, Australasia and the Americas than in Europe and Africa throughout the Pleistocene. Hence, allelic diversity is generally higher in Africa and (often) in Europe. The rate of lineage extinction outside of Africa and Europe must be higher. In order to really get a complete genetic picture of ancient human dispersals, ancient DNA coming from "eastern provinces" is essential.

Maju said...

Not sure if it may be considered a dendogram or not but fig. 3 shows no affinity whatsoever between Australian Aboriginals and Neanderthals. In that graph you can actually rotate any branch around the axis of its root and it means the same.

I understand you can do the same with a standard dendogram.

Neanderthals are not closer to any particular modern human group than to any other. And if you have any doubt, you can always check the data of Caramelli'03, specially the very visual fig.1.

"Maju,

Are you suggesting that Neanderthals emerged in Africa and migrated to Europe?"

Neanderthals? Nope. H. ergaster yes (leading to Neanderthals by a long sequence known to exist only in Europe).

I'm suggesting that there is no particular reason that we are related to Neanderthals closer than the African and European branchs of H. ergaster are, what archaeologically means Acheulean and c. 900,000 years.

German Dziebel said...

"Not sure if it may be considered a dendogram or not but fig. 3 shows no affinity whatsoever between Australian Aboriginals and Neanderthals. In that graph you can actually rotate any branch around the axis of its root and it means the same.

I understand you can do the same with a standard dendogram."

There are two types of dendRograms, with and without seriation. The position of outgroups (Neandertals and Denisova) in Fig. 3 may not be accidental. Check Mungo man research to see where its most divergent sequence together with the so-called "nuclear insert" (a fossilized mtDNA sequence found at higher frequencies outside of Africa) are positioned vis-a-vis Neandertal sequences.

Maju said...

The position in relation with any subset of another set is totally accidental.

All H. sapiens share a single mtDNA ancestor and this is well established. Similarly we know that all H. neanderthalenis sequenced to date also share a common distinct ancestor. The tree could perfectly have only three branches (2, one of which has 2). In fact it has only three branches, even if two are shown with diverse sub-branches.

Maju said...

And anyhow Mungo Man is not in that graph (though it is in the Caramelli paper and shows no closer relationship to Neanderthals whatsoever).

German Dziebel said...

"The position in relation with any subset of another set is totally accidental."

I would like to hear it from the authors of the papers, not from you. Once again, there're seriated dendrograms, and there're those that are composed only of hierarchical sets.

"And anyhow Mungo Man is not in that graph (though it is in the Caramelli paper and shows no closer relationship to Neanderthals whatsoever)."

The fact that it's not in the graph means nothing. I would like to check the original research on Mungo man published by Adcock et al. both in Australian Journal of Anthropology and in PNAS. But, for now, what I remember clearly is that the most divergent ancient Australian sequence was more divergent than the sequences collected from modern Bushmen and Pygmies. It means that if you'd included it into the Fig. 3 of this paper, it would have been on the same side as the Denisova and the Neandertal sequences. And geographically (that's where seriation comes in) Australians are closer to South Siberia than to Africa and most likely represent an offshoot of an originally East Eurasian population. Meanwhile, the fact that ancient modern humans showed to have DNA that's different from the DNA attested in extant modern humans weakens any claim that because Neandertal DNA is different from modern human DNA we're not closely related to Neandertals. Notably, the Neandertals' range has recently become more easterly.

I'm not suggesting anything definitive here. Only that this surprising find may be a turning point in our understanding of ancient human dispersals.

"All H. sapiens share a single mtDNA ancestor and this is well established."

See above. All EXTANT H. sapiens sequences have been shown "on paper" and "on computer screen" to possibly coalesce in Africa. Phylogeographically, however, there are mighty gaps between the "centroids" of the major continental mtDNA macrohaplogroups: East Africa for L, (South)east Asia for N, South or East Asia for M. Notably, again, N, which is the major haplogroup in Australia, is the furthest removed from Africa. This is consistent with the growing amount of ancient DNA (including the "nuclear insert") that seems to document lots of extinct hominin/human lineages in (Austral)asia (and, surprisingly, in America if we look at the high frequencies of the insert there).

Maju said...

If you know anything about mtDNA (or haploid Y-DNA in general) you know they are hierarchical sets. Strictly so.

"I would like to check the original research on Mungo man published by Adcock et al. both in Australian Journal of Anthropology and in PNAS".

It used to be available online. I remember reading it. Anyhow, the Adcock 2001 results have been heavily criticized, among other reasons because no independent analysis were made and post-mortem damage was more than likely. In general it's considered one of the most unreliable aDNA analysis ever made and the sequence can be safely discarded.

"All EXTANT H. sapiens"...

All extant and fossil H. sapiens, except very arguably Mungo Man, which is considered a mere error by everyone you can read.

The sequence is in any case, as unrelated to Australian Aboriginals as to any other human: in fact it's totally out of the known genetic record of H. sapiens and would constitute a distinct branch in the phylogenetic tree you are discussing so vehemently.

German Dziebel said...

"In general it's considered one of the most unreliable aDNA analysis ever made and the sequence can be safely discarded."

That must be why Caramelli 2003 decided to include it in their "very visual Fig. 1."

"Anyhow, the Adcock 2001 results have been heavily criticized..."

Of course, people are going to criticize heavily a finding that contradict their theory. Adcock et al., like most of us, understand that one fact is not enough. But since 2001 more and more inconvenient findings have surfaced, while makes us revisit Adcock et al. through a more favorable lens.

"The sequence is in any case, as unrelated to Australian Aboriginals as to any other human: in fact it's totally out of the known genetic record of H. sapiens and would constitute a distinct branch in the phylogenetic tree you are discussing so vehemently."

With all the uncertainties around mutation rate, lineage extinction rate, Pleistocene population size, species boundaries, etc., your categorical assertions sound like Soviet slogans.

"If you know anything about mtDNA (or haploid Y-DNA in general) you know they are hierarchical sets. Strictly so."

Now, you just have to study up on what seriated dendrograms are...

Maju said...

aDNA without independent testing... one of the first ever aDNA tests. Even the most reliable aDNA results have some burden of doubt on them... and they have to discard many sequences that don't produce reliable results...

It's a clear case of error. At least a very likely one, and as the LM3 is now again buried under Aborigin custody... it's extremely unlikely he'll be ever tested again (what would be the thing to do).

"But since 2001 more and more inconvenient findings have surfaced"...

Like?

German Dziebel said...

"aDNA without independent testing... one of the first ever aDNA tests."

Sure. See above.

"Even the most reliable aDNA results have some burden of doubt on them"

Even the most "reliable" modern DNA results have some burden of doubt on them. Unlike linguistics or kinship studies, the whole field is still very new.

"Like?"

I'm not going to bother. Do your homework. Okay, I'll mention one study that recently came out. It's called "The complete mitochondrial DNA genome of an unknown hominin from southern Siberia." Had we asked a paleoanthropologist in 2001 how likely is it to find a widely divergent DNA sequence in a 40-30,000 year old pinkie bone, he would answer "highly unlikely."

terryt said...

"This all potentially looks like Eurasian/Australian humans may contain a direct link to Eurasian hominids, with African humans out of the picture".

I wouldn't put African humans entirely out of the picture. A great deal of evidence shows that some humans with an African ancestry reached Australia. In fact their haplogroups are the only surviving ones in any modern humans, anywhere. Others may well have been 'drifted out'.

"Had we asked a paleoanthropologist in 2001 how likely is it to find a widely divergent DNA sequence in a 40-30,000 year old pinkie bone, he would answer 'highly unlikely'".

Especially from Siberia.

"I would like to check the original research on Mungo man published by Adcock et al. both in Australian Journal of Anthropology and in PNAS".

Here it is:

http://www.pnas.org/content/98/2/537.full

T. Kosmatka said...

Something just crystallized in my mind regarding the new Siberian MtDNA.

There are only three choices.

1.) It's from a new species.
2.) It represents a previously unknown deep clade division within Neanderthals.
3.) It represents a lost deep clade division within ancient anatomically modern humans.

Since the siberian MtDNA represents a deeper clade division than the division that separates humans and Neanderthals, then the following must be true:

A.) If the Siberian MtDNA belonged to a Neanderthal, then all human MtDNA would be a subset of Neanderthal MtDNA diversity.

B.) If the Siberian MtDNA belonged to a modern human, then all Neanderthal MtDNA would be a subset of modern human MtDNA diversity.

This Siberian MtDNA is the clear outgroup. Placing it within the range of diversity of either Neanderthals or humans would necessarily require that group to absorb the entire range of variation of the other.

One species or three.

Onur Dincer said...

Kosmatka, it can be any homo species (except maybe H. habilis) with the current data we have (just a pinky bone and ancient mtDNA of an individual). For instance, if the current molecular clock estimates are true, it can be a H. erectus specimen whose ancestors separated from the ancestors of modern humans and neanderthals, for instance, not 2 million years ago as some of the oldest remote-to-Africa H. erectus specimens might have been, but about 1 million years ago (compatible with the current data we have about H. erectus). But if Maju's clock estimates are true, it can even be a H. erectus specimen whose ancestors separated from the ancestors of modern humans and neanderthals about 2 million years ago. Of course, it can also be H. ergaster, H. heidelbergensis or any other homo species (maybe a completely new homo species we have never encountered until today).

eurologist said...
This comment has been removed by the author.
German Dziebel said...

"Especially from Siberia."

Exactly. Next to Neandertals and Upper Paleolithic technologies.

"Here it is:

http://www.pnas.org/content/98/2/537.full"

Thanks Terry. I have the other one somewhere in my files. It contained an expanded overview and phylogeny. I just need to find it over the weekend.

"A great deal of evidence shows that some humans with an African ancestry reached Australia. In fact their haplogroups are the only surviving ones in any modern humans, anywhere."

There are NO African haplogroups outside of Africa. For M and N to be considered African, they need to be found in Africa. However, only their derived haplotypes are found there. Also, L lineages need to be found in India, Siberia, Australia, etc. They haven't been found either in modern or ancient populations outside of Africa. These are facts. The rest are interpretations.

German Dziebel said...

Both Fig. 1 and Fig. 1 of Adcock et al. 2001 (see the link above provided by Terry) show that Mungo man mtDNA and the nuclear insert are intermediaries between modern human and Neandertal sequences.

Maju said...

No. Even Adcock says that it's within modern human variantion, a supposed direct descendandt from the MRCA (of humans).

I was checking sequences and there's a good deal of lineages that share two of the five mutations that separate KS8 from CRS. The other three are either private... or from a lineage I could not spot (check L0a3).

"There are NO African haplogroups outside of Africa".

By definition? I'm pretty sure that I see Africans (people born and raised in Africa and of African ancestry) going up and down my street every single day. But maybe even if their carriers have no passport or an unstable visa their DNA has changed nationality somehow.

Exactly as L3m (pre-M, M) and L3n (pre-N, N) seem to have changed "nationality" (or is it "continentality") when they crossed the Red Sea (or is it the Gaza Wall?)

But otherwise you are wrong and lying in this Dziebel.

German Dziebel said...

"No. Even Adcock says that it's within modern human variantion, a supposed direct descendandt from the MRCA (of humans)."

I know you don't read studies that you talk about. May be because you're illiterate. Adcock's Fig. 2 shows LM3 and the nuclear insert as forming one clade that's outside the clade encompassing "contemporary humans and most ancient Australians." The article is full of quotes to the same effect. In their Australian article, Adcock et al. added that on pure phylogeographic grounds the distribution of the nuclear insert (highest frequencies in the Americas, lowest in Africa) coupled with the location of the most divergent modern human mtDNA lineage in Australia contradicts the out of Africa model of human dispersals. See Adcock et al. Lake Mungo 3: A Response to recent critiques // Archaeology in Oceania 36 (2001).

"Exactly as L3m (pre-M, M) and L3n (pre-N, N) seem to have changed "nationality" (or is it "continentality") when they crossed the Red Sea (or is it the Gaza Wall?)"

You're still thinking about human prehistory in biblical terms.

"But otherwise you are wrong and lying in this Dziebel."

I've done lots of different things in my life, but when I talk to you, Luis, I feel myself something I've never been before. An exorcist.

terryt said...

"One species or three".

Probably one. It's just that mtDNA lines of two of the 'subspecies' have died out.

Onur Dincer said...

Probably one. It's just that mtDNA lines of two of the 'subspecies' have died out.

I concur. As long as they could produce fertile offspring (at least in principle), they should be considered as members of a single species (as its subspecies, judging by their clear genetic and morphological divergences).

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

judging by their clear genetic and morphological divergences

We may also add to this list the cultural divergences, at least between anatomically modern humans and neanderthals.

Onur Dincer said...
This comment has been removed by the author.
terryt said...

"E.g., Homo sapiens sapiens, Homo sapiens neanderthalensis and Homo sapiens altaianus respectively".

Totally agree. The presence of humans (of some sort) in the Altai as early as this ('between 48,000 and 30,000 years ago') indicates there is nothing impossible about modern humans living there too. And it's quite possible that the Homo sapiens altaianus haplogroup didn't itself emerge from Africa. It may have developed in a H. erectus population outside Africa. In which case the 'modern human' haplogroups may derive from one that had moved into Africa some time after H. sapiens neanderthalensis had in turn diverged from it.

"There is also no particularly good reason that to think that homo erectus left Africa in a single wave".

Or to think that there was no movement back into Africa at times. We know 'modern humans' have moved back in periodically, so why not earlier humans?

"I know you don't read studies that you talk about. May be because you're illiterate".

I gave up on Maju when he claimed that all modern human phenotype variation already existed within the small population of modern humans who emerged from Africa sometime between 100,000 and 50,00 years ago. And the difffering modern regional phenotypes (especially the East Asian one) didn't develop until the Neolithic, surely a time of expanding population numbers and so unlikely to be a period of selection or drift to fixation of particular phenotypes.

Maju said...

"... if anatomically modern humans, neanderthals and the lineage of the X-Woman are all subspecies of one and the same species"...

Who says that?

"I gave up on Maju when he claimed that all modern human phenotype variation already existed within the small population of modern humans who emerged from Africa sometime between 100,000 and 50,00 years ago".

I didn't say that because I don't normally discuss phenotypes. What is clear is that, at least pigmentation-wise, all the major alleles involved do exist among Africans and that the various Eurasian populations show marked founder effects of these alleles (as one would expect).

Vide Coops 2009, fig. 4.

So it's probable that you are more or less right in the words that you put on my mouth without proper reference.

terryt said...

"I didn't say that because I don't normally discuss phenotypes".

In this case you did say that. Within your comment you included the whole suite of typically East Asian phenotype: straight black hair, relatively little body hair and the eyefold, as well as the yellowish skin. And you definitely claimed their widespread presence in the East was in no way a result of Paleolithic founder effect. You claimed a major change in the region at the Neolithic, although I earlier forgot to add the bit where you claimed there had been no migration involved in the widespread appearance of that eastern phenotype from Indonesia in the south to the Russian Far East in the north. You insisted the change was in no way the result of any sort of expansion.

Maju said...

Quote and link please. You have the bad custom of misreading me.

Most of what you say I don't recognize. So if you're going to bitch about me, at least document your claims.

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

Terry, did you mean this part of one of Maju's comments in the "Major East-West divide in Indonesian Y chromosomes" thread?


However the limited fossil data does suggest a more "Australoid" (or just "archaic") morphology in SE Asia in general (but also in all East Asia) before Neolithic, so we have what may be two (largely) different issues: genetics and phenotype.

On the other hand, Australian fossils don't seem clearly "Australoid" either before a recent timeframe (see P. Brown's Australian page), the most archetypally AA skull is that of Roonka, which is about 7000 years old.

So I suspect it's about time we begin considering morphology not strictly attached to genetics or ancestry. At least I do try to keep some distance from morphology when analyzing all this data because otherwise I'd go not just "hyper-recentist" but even believer in recent arrival from Mars, as there are no clear precusors for the Mongoloid type (there are not even brachi/mesocephalic Paleolithic skulls almost anywhere!, except maybe Minatogawa, but Brown says he's not Mongoloid either).

So I'd either argue for nutrition, recent social selection for those phenotypes or wild epigenetics, rather than recent replacement from nowhere.

Onur Dincer said...

Who says that?

I. If anatomically modern humans, neanderthals and the lineage of the X-Woman could produce fertile offspring from sexual intercouse with each other, they should be considered members of a single species. And if each of the three separated and diverged from each other for hundreds of thousands of years, they should also be considered subspecies of that single species.

Maju said...

"If anatomically modern humans, neanderthals and the lineage of the X-Woman could produce fertile offspring from sexual intercouse with each other, they should be considered members of a single species".

But we don't know that. It's a big IF with nothing to support it.

"And if each of the three separated and diverged from each other for hundreds of thousands of years, they should also be considered subspecies of that single species".

All species began as "subspecies". I don't understand what's the point here.

...

Also I don't see how my quote above resembles what Terry said.

Onur Dincer said...

But we don't know that. It's a big IF with nothing to support it.


I know that it is a speculation. I was just evaluating the necessary implications of that speculation if it is confirmed one day.

All species began as "subspecies". I don't understand what's the point here.

Actually, it seems clear to me that these three Homo lineages diverged enough to be classified as different species or subspecies.

The real issue is the one I evaluated in my previous statement: whether these three diverged lineages always remained as subspecies or turned into different species in the course of time.

Also I don't see how my quote above resembles what Terry said.


It was the closest thing I was able to find in your comments to Terry's description.

Maju said...

"It was the closest thing I was able to find in your comments to Terry's description".

I know you have good intentions but better let him do his homework: he often misreads (twists, deforms) what I say in extreme ways, what makes discussion kind of difficult because you have to insist once and again in what exactly you meant (or even said word by word).

terryt said...

"Terry, did you mean this part of one of Maju's comments in the 'Major East-West divide in Indonesian Y chromosomes' thread?"

No. His main claims are in the 'Indonesian Y-haps mostly Paleolithic', or something similar. His comments there include:

"For me the data evidences that there is now enough variability to imagine every possible scenario".

So let's look at this hypothesis. One hundred thousand years ago all of Africa and Eurasia south of about 50 degrees N was occupied by humans, or closely related species. All modern humans are believed to descend from just the African subset of this population.

But it's extremely unlikely that the OoA contained a representative sample of the contemporary African genetic variation. Therefore modern humans descend from just a subset of a subset. Is it really credible to believe they were an extremely varied population? more varied even than are modern Africans?

The vast majority of Africans today have tightly curly hair. Even moderately curly hair is uncommon, and where it is found it is most likely a product of migration in from Eurasia. So either something very strange happened in Africa after the OoA, or the hypothesis of 'enough variability to imagine every possible scenario' is wrong.

Other related comments he made there:

"A lot of 'Mongoloids' have curly hair (Tibet, Amazonia, etc.) It's somewhat atypical but not a definitory trait".

To which I'd point out that you have to really look for it, and it is quite possibly a result of migration in from elesewhere, or a pre-Mongoloid survival.

Concerning the pronounced eyefold:

"Epicanthic fold. A trait that is not totally absent in other populations (Khoisan, North Europeans, etc.) nor universally present among East Asians, much less Native Americans".

Concerning yellowish skin:

"This trait must have coalesced (founder effect?) early in the human dispersal process because the known pigmentation traits of East Asians seem to arise then (ref- Coop'09)".

And:

"I say that such traits are flesh and were probably widespread before Neolithic. ... but the patterns of the derived East Asian MC1R haplotype do suggest that these traits are old, very old".

But the East Asian suite of characteristics cannot be early Upper Paleolithic in SE Asia. Straight hair is very rare in Australian Aborigines and Papuans, so the East Asian suite of characteristics must have arrived in SE Asia after these people had passed through. Even in Polynesians the suite is not strongly represented. In fact it's only marginally more pronounced in Indonesia, and probably immigrant there. So what are are we left with?

terryt said...

"But we don't know that. It's a big IF with nothing to support it".

It's often claimed that a single group of chimpanzees contain greater haplogroup diversity than does the whole of humanity. Therefore, obviously, just 100,000 years is insufficient to separate chimpanzee species. Just because the human/Neandertal/Denisova haplogroups separated less than one million years ago in no way automatically means they were unable to form fertile offspring. For example:

http://www.pnas.org/content/99/1/43.figures-only?cited-by=yes&legid=pnas;99/1/43

From the article: 'The dotted line indicates the chimpanzee-bonobo divergence time (1.8 million years) estimated with the Pt?1 haplogroup'. Chimpanzees and bonobos are quite capable of forming fertile offspring.

So if the three human species did not produce offspring it is unlikely to have been because of genetic incompatability, but a result of cultural factors.

Maju said...

Glad that you bother providing quotes, Terry. They are not what you said but well...

"Straight hair is very rare in Australian Aborigines and Papuans"

And pre-admixture Ainu (NE Asians). And Australia and Papua are not SEA. What we were discussing there and elsewhere (here at Dienekes' for example) was the existence (per Karafet) of a Paleolithic divide at Wallace Line.

Whatever the case, you are assuming too many things from mere phenotypes, which we know can vary wildly with minimal genetic/epigenetic variations.

And also you are wildly off topic.

...

"Just because the human/Neandertal/Denisova haplogroups separated less than one million years ago in no way automatically means they were unable to form fertile offspring".

It does not mean that, agreed. But we know now that there was no effective admixture between Neanderthal and H. sapiens, so it's reasonable to hypothesize that there was no effective admixture either between Neanderthals and Eurasian H. erectus.

Even if this hypothesis is wrong, from what we know (which admittedly is not much) we must still deal with Denisova as a different population in principle.

You can't just assume that the less likely option is right just because it might be right. It might also be wrong.

Onur Dincer said...

'The dotted line indicates the chimpanzee-bonobo divergence time (1.8 million years) estimated with the Pt?1 haplogroup'.

More recent research suggests they diverged much later, less than 1 million years ago:

http://mbe.oxfordjournals.org/cgi/content/abstract/22/2/297

Chimpanzees and bonobos are quite capable of forming fertile offspring.

That is why common chimpanzees and bonobos should be classified as subspecies of a single chimpanzee species.

If one day a fertile human-chimpanzee hybrid is produced, humans should also be included in the chimpanzee species as its subspecies (see Jared Diamond's "The Third Chimpanzee: The Evolution and Future of the Human Animal"). The same is true for the other hominina lineages. Of course, we can use homo rather than chimpanzee as the name of the species which would include all hominini (the tribe that comprises all hominina and pan). If gorillas can produce fertile offspring with humans and chimpanzees, they also should be included as a subsepcies of the same species with humans and chimpanzees.

Onur Dincer said...
This comment has been removed by the author.
Maju said...

"More recent research suggests they diverged much later, less than 1 million years ago"...

And even more recent research instead suggests an older date again, which could be more or less coincident with the one proposed by Stone in 2001.

Caswell 2008 proposes a divergence time of at least 1.3 million years ago but she also suggests that the formation of the Congo river c. 1.5-2 million years ago is a more likely reference, because it is this river what keeps chimpanzees and bonobos separated.

I discussed Caswell's most interesting paper here and is one of the reasons why I think that one of the most basic assumptions when calculating molecular clock hypothetical ages, the Homo-Pan divergence date, is radically wrong, being the real one at least 15% larger (the difference between 7 and 8 million years) or even as much as 50% larger (the difference between 5 and 10 million years).

"Chimpanzees and bonobos are quite capable of forming fertile offspring".

I missed this claim (from Terry, I imagine). Are they really? Do we know of any real case? It's the first time ever I read such claim and I have never heard of an hybrid Pan, much less a fertile one.

pconroy said...

Here a great post by "The Atavism" blog on this:
http://theatavism.blogspot.com/2010/03/does-forty-thousand-year-old-finger.html

Especially a nice diagram explaining how Denisova could be a Neanderthal:
http://img.photobucket.com/albums/v387/science_boy/tubetree_corrected-1.png

Maju said...

PConroy: what that article and possibly others seem to suggest is that we have to review all the archaeology only because of some MC estimate. And that's simply what MC fans have been suggesting all the time at other levels: when archaeology and MC estimates don't match (what happens way too often) it must be because the archaeological record is wrong or misinterpreted.

But the fact is that, with all its shortcomings, archaeology and C14 are still much more solid, more accurate or simply better than whatever the molecular clock esotherisms can offer.

The Denisova "problem" is easily solved by adjusting the MC in accordance with what we know from the archaeological record: Neanderthals diverging c. 1 mill. or 900,000 years ago (and not just 700-300 Ka) and the early colonization of Eurasia by H. erectus happening c. 2-1.8 Ma ago.

But, well, I know: I'm a heretic because I'm highly skeptic of the potential of the MC as we know it.

onix said...

hehe so asia has an archaic erectus contemporary not from china to showcase now, enriching the human lineage with at least one other species, that actually might just show there were not any species at all:) whats new? lol

terryt said...

"that actually might just show there were not any species at all"

I'm reasonably convinced that there has only been one 'species' of Homo since Homo erectus, and that all 'subspecies' since that time have been able to form fertile hybrids. What we see in the fossil record is little different from what we see today with the various human geographic varieties.

"I'm a heretic because I'm highly skeptic of the potential of the MC as we know it".

Regardless of the 'actual' dates the 'relative' dates are still relevant. Either way we have the Denisova fossil branching off the line leading to modern humans before the line leading to Neanderthals does. To me this suggests that modern humans are actually a product of a migration INTO Africa (from Asia) after both the Denisova population (?) and the Neanderthals had branched off. Then modern humans came back out.

Maju said...

"To me this suggests that modern humans are actually a product of a migration INTO Africa (from Asia) after both the Denisova population (?) and the Neanderthals had branched off. Then modern humans came back out".

It's a possibility and one that is very fashionable these days it seems. However there's no archaeological evidence whatsoever supporting that hypothesis, though admittedly there is also a blank of evidence in Africa to support the in situ evolution of our species from H. erectus.

As I see it, what supports an old divergence tree (of c. 1 million years) between our species and Neanderthals is (1) the Acheulean migration out of Africa, pretty well documented and with older known dates of c. 900,000 BP in South Iberia and (2) the very marked morphological differences between the two species.

Sincerely I fail to see how in just 200,000 or even less years the two species could have diverged so much. There is only one point of convergence: large brains but that is clearly a product of evolutionary pressure/trend on a path already taken by the whole genus Homo long before. Otherwise the two species are markedly different, always within the specifics of the genus as a whole. H. sapiens has been diverging for maybe as long and nowhere, not even in the most remote corner nor the most isolated population, it has been able to accumulate so many differences (not even a fraction).

And for me this finding of Denisova, which I strongly suspect bearing a pre-Acheulean H. erectus lineage, whatever the species, actually does nothing but add evidence in favor of this thesis.

But of course the definitive evidence can only come from archaeology, filling the blank we have in the evolution of our species between H. erectus/ergaster and H. rhodesiensis/sapiens, the same that Atapuerca has filled in the gap between the former and Neanderthals.

So far this has been elusive but Africa is huge and has not the same archaeological research density as does Europe. However the indirect evidence in form of industrial typologies, I understand, rather supports separate evolutionary processes in Europe (towards Neanderthal) and in Africa (towards H. sapiens).

Emotionally I'd might prefer that the roots of humankind would be at Atapuerca, just a few kilometers from where I live and where my roots are but the data does not seem to support that. And the Denisova hominin's phylogeny does not either.

terryt said...

"As I see it, what supports an old divergence tree (of c. 1 million years) between our species and Neanderthals is (1) the Acheulean migration out of Africa, pretty well documented and with older known dates of c. 900,000 BP in South Iberia and (2) the very marked morphological differences between the two species".

But that would simply place the Denisova fossil as even more ancient, and still the oldest mtDNA we have extracted. So that in turn would place the migration back into Africa some time before the Acheulean and after the original H. erectus emergence from Africa. Which actually makes sense, because the Acheulean didn't make it to Central, East or SE Asia although H. erectus was certainly present through most of that region.

Maju said...

"But that would simply place the Denisova fossil as even more ancient"...

I have already said what I see here: an H. erectus lineage from the first migration into Eurasia.

"So that in turn would place the migration back into Africa some time before the Acheulean"...

Why should there be any migration back into Africa? Where's the archaeology to support it?

"Which actually makes sense, because the Acheulean didn't make it to Central, East or SE Asia although H. erectus was certainly present through most of that region".

I was not clearly aware of that detail but actually it supports the idea of this lineage being H. erectus from the first OoA wave of all.

In brief, the sequence of OoA migrations is:

1. H. erectus with chopper industry c. 2 Ma ago -> Denisova

2. H. ergaster with Acheulean c. 1 Ma ago -> Neanderthal

3. H. sapiens (MSA, 130-70 Ka ago).

And there is absolutely no need for any backmigration to Africa, which is not justified archaeologically. What there is need is for a radical revision of the the molecular clock.

Vincent said...

I don't know how reliable this is, but a Denisovan specimen may have had a human father with Y-DNA R http://www.reddit.com/r/genetics/comments/2gt9d2/denisova_cave_neanderthal_was_ydna_haplogroup_r/

Vincent said...

Correction: it was a Neanderthal living in the same cave as the Denisova hominin, not a Denisovan proper

Maju said...

@Vincent: sounds intriguing but someone should double and even triple check that claim because it sounds most unlikely. And as it's said "exceptional claims require exceptional evidence".

The Denisova Cave Neanderthal toe was found in a layer deeper than the one of the "Denisovan" (H. heidelbergensis) fingers, so it must be older than 50 Ka BP, probably 60-70 Ka BP.

Assuming that the Y-DNA haplogroup R existed so early in time (what is quite dubious even for someone like me, who tends to favor old chronologies), the signature of possible H. sapiens admixture in this specimen is maybe not impossible but would be very weak in any case. I mean: it is known that admixed populations are pulled upwards in ML trees and that is indeed the case of the Altai Neanderthal in the autosomal tree approximation (but then again this could just mean that this population was more ancestrally diverged relative to known European Neanderthals, not necessarily implying admixture) but the displacement is very mild, what would allow at most a minimal H. sapiens (or H. heidelbergensis) admixture. That would mean that if the alleged finding is true the male lineage had been interbreeding with unmixed female Neanderthals for many generations. Not impossible but, sincerely, a highly unlikely scenario, and one that would push the Y-DNA R ancestor even further back in time.

So yeah: triple check it, please, because it looks a very weird scenario for many reasons.

terryt said...

"I don't know how reliable this is, but a Denisovan specimen may have had a human father with Y-DNA R"

Neanderthal or Denisovan, I see no reason why it should be impossible. Interbreeding almost certainly went in both directions.

Vincent said...

Right Maju, now it looks like we are in presence of contamination, as suggested by a commenter in that thread. But I still don't get why the guys from Max Planck institute uploaded a human Y chromosome labeling it as Altai Neanderthal.

terryt said...

"Assuming that the Y-DNA haplogroup R existed so early in time"

That occurred to me as soon as I'd posted my comment. It is impossible that it would be R although some sort of CF may be remotely possible.