Quickies
Quite worth mentioning:
Wu Liu et al., The earliest unequivocally modern humans in southern China. Nature 2015. Pay per view → LINK [doi:10.1038/nature15696]
AbstractThe hominin record from southern Asia for the early Late Pleistocene epoch is scarce. Well-dated and well-preserved fossils older than ~45,000 years that can be unequivocally attributed to Homo sapiens are lacking1, 2, 3, 4. Here we present evidence from the newly excavated Fuyan Cave in Daoxian (southern China). This site has provided 47 human teeth dated to more than 80,000 years old, and with an inferred maximum age of 120,000 years. The morphological and metric assessment of this sample supports its unequivocal assignment to H. sapiens. The Daoxian sample is more derived than any other anatomically modern humans, resembling middle-to-late Late Pleistocene specimens and even contemporary humans. Our study shows that fully modern morphologies were present in southern China 30,000–70,000 years earlier than in the Levant and Europe. Our data fill a chronological and geographical gap that is relevant for understanding when H. sapiens first appeared in southern Asia. The Daoxian teeth also support the hypothesis that during the same period, southern China was inhabited by more derived populations than central and northern China. This evidence is important for the study of dispersal routes of modern humans. Finally, our results are relevant to exploring the reasons for the relatively late entry of H. sapiens into Europe. Some studies have investigated how the competition with H. sapiens may have caused Neanderthals’ extinction (see ref. 8 and references therein). Notably, although fully modern humans were already present in southern China at least as early as ~80,000 years ago, there is no evidence that they entered Europe before ~45,000 years ago. This could indicate that H. neanderthalensis was indeed an additional ecological barrier for modern humans, who could only enter Europe when the demise of Neanderthals had already started.
When asked in private correspondence earlier today what did I think of this, I replied that María Martinón (second listed author) is a top expert in tooth morphology and that, if she says they are unmistakably H. sapiens, I have to believe it.
I also replied a bit more extensively that this should be no surprise, that evidence in favor of a c. 100 Ka BP migration of H. sapiens into South and Southeast Asia has been piling up for some time already. Some of the most important pieces of evidence are the Zhirendong jaw (also from Southern China, dated to c. 100 Ka BP) and the African-like Katoati toolkits (NW India, dated to c. 96 Ka BP). These dates are roughly coincident with the end of the Abbassia Pluvial (c. 125-90 Ka BP), which is in turn coincident with the period of evidence for earliest H. sapiens presence in Arabia and Palestine.
In other words, our ancestors crossed into Arabia and Palestine (and maybe other less well documented nearby regions of West Asia) around 125 millennia ago (with a second wave c. 90 Ka ago). The Neanderthal admixture episode probably happened soon after. Then they moved to South and SE Asia, quite possibly pressed by growingly arid conditions in Arabia, and this second migration took place around 100 millennia ago (earlier is not yet supported but can't be fully discarded).
All this has major implications for molecular clock calibration, of course: mtDNA L3 should be c. 125 Ka old and M some 100 Ka old, similarly Y-DNA CF should be around 100 Ka old as well. This is the kind of stuff that makes genetics-oriented people skeptic but the molecular clock is a mere educated hunch, while the archaeological data is serious evidence that cannot be ignored.
ReplyDelete"All this has major implications for molecular clock calibration, of course: mtDNA L3 should be c. 125 Ka old and M some 100 Ka old, similarly Y-DNA CF should be around 100 Ka old as well. This is the kind of stuff that makes genetics-oriented people skeptic but the molecular clock is a mere educated hunch, while the archaeological data is serious evidence that cannot be ignored."
I don't think so...
1) There is no archaeological evidence for continuity between early Eurasian H. sapiens and those who took part in the later (from ~60 kya) expansion, like Ust'-Ishim, who had Neanderthal admixture from 50-60 kya.
2) Most importantly, we have several early UP genomes from across Eurasia (Oase, Tianyuan, Ust'-Ishim) that confirm conventional Y-DNA/mtDNA TMRCA estimates.
Are you advocating for mass extinction of all the H. sapiens that existed in Asia and a second undocumented migration just to justify your broken "molecular clock"? It sounds as extremist and far-fetched that I have no words, really.
DeleteUst-Ishim carried Y-DNA NO* (or something like that), what is long derived from CF, there is no way his lineage could have evolved in a mere 10-20 Ka period. It is all kinds of wrong, even within the usual molecular clock parameters. As for Tianyuan's mtDNA B is derived from N, which I understand that must have expanded much later than M, probably around the Toba episode, of which N/R benefitted most likely, as did their main "partner" Y-DNA K2 (and maybe C too).
Also the route out of Africa was effectively closed in that period, not just by very arid climatic conditions but also by the migration of Neanderthals to as far south as Yemen.
In other words, I think you are confusing the post-Toba expansion of Y-DNA K2 and mtDNA R with the out of Africa migration, which is much older.
DeleteUst'-Ishim belongs to K2, and is close to the root. The mutation rate derived from his genome confirms conventional Y-DNA age estimates. The mtDNA mutation rate derived from branch shortening of his R* lineage confirms conventional mtDNA age estimates.
DeleteThey actually used the supposedly "broken" molecular clock to date his mtDNA lineage, and it was dated to 49 kya (so actually a slight overestimate since UI lived 45 kya). You can attempt to misrepresent the evidence, but the evidence is clear.
Send me copy of the original paper because I read all kind of stuff on UI and never got my free blogger copy. After arduous search I found this graph at a Russian (?) blog, which (from memory) is the original tree. It is clear that UI hangs from the NO or otherwise K2a (could stil be pre-NO) node, although it's exact relation with N is not clarified. So we are talking of NO* or K2a* or pre-NO*, because NO = K2a is one of the two major basal branches of K2.
DeleteThe "near the root" position does not correspond therefore so much to UI himself but to the NO =K2a node from which it hangs.
It is not as underived as you claim. Ruler on hand UI is 15% downstream of K2. A rustic extrapolation would make K2 some 60 Ka old, with huge error margins probably. So I rather feel that UI vindicates Karafet and my own post-Toba proposed chronology for K2.
The intriguing part is how NO* got to West Siberia so early. Was it carried as a minor relative lineage by the much more clear westward flow of P1 via North India and later lost by drift? Or did it actually arrive via the Northern route from East Asia? Considering that he is similarly close to West Eurasians as to East Asians and that by that time we should expect to have already a notable differentiation between both macro-populations, I'd say he's a very early Taiga migrant in westward direction and the product of East-West remixing in such inhospitable latitudes.
Even 60 kya sounds too old. Considering how close UI is to the root, I think their K(xLT) estimate should be rather solid. In the Supplementary Info, they dated K(xLT) to 50 kya with a 95% interval of 47-55 kya, CF to 71 kya (62-80 kya) and CT to 72 kya (63-81 kya).
Delete"The Ust’-Ishim Y-chromosome carries no additional mutations belonging to any of the sub-haplogroups of K(xLT); however, there are 6 additional mutations that are not observed in the 23 present-day humans to which we compare."
I wouldn't call him NO* just because he may share very few mutations with the NO node (although they make no mention of that?), he may be pre-NO however.
Also, his mtDNA lineage has just one mutation downstream of R according to the paper. You should find that significant, at least.
NO* or pre-NO*. It is clear in the tree that he does not hang from the K2 node but from the branch shared with haplogroup O (no N samples present), so he does share at least some mutation(s) with these. Lacking the paper, I cannot be more precise.
DeleteRe. UI's mtDNA, R0 is also just one coding region mutation downstream of R, as are many other basal branches (R1, R2'JT, R6, R7, R11'B6, B, R30, R31 and P). So this is the signature of a star-like explosion (synonym of human rapid expansion, although more modest than that of M or H, and overlapping almost that of N). In my dedicated page, I argue for the following rough chronology:
1. Arrival to South Asia (M node): c. 93-83 Ka BP
2. First expansion (many M sublineages): c. 85-75 Ka BP
3. Slowing down of the expansion (Toba) and N node: c. 77-67 Ka BP
4. Reactivation of the expansion and clear arrival to Australasia (R node): c. 69-59 Ka BP
5. Expansion to less hospitable areas (NE Asia, the Neanderlands) (many R sublineages): c. 61-51 Ka BP
Obviously UI belongs to step #5, and his chronology is not too distant from it, more so if the 45 Ka BP date is uncalibrated C14, what means maybe 5-10 Ka more in "real time".
The new Siberian paper explains the details on UI's paternal lineage. He shares just one mutation with NO. So he's pre-NO, not NO*, which may not even have existed yet.
Delete"Ancient Siberian Ust’-Ishim is thought to belong to K2 Y-DNA clade. We observed that all 16 K2 – specific SNPs were present in Ust’-Ishim. However, Ust’-Ishim also shares one out of 47 derived alleles specific to NO branch (hg19 coordinate ChrY 7690182; Supp. Fig. 12). Therefore, Ust’-Ishim’s haplogroup most likely belongs to NO clade that diverged shortly after NO-QR split (45 kya; 95% CI: 42-48 kya), but before N-O split (38 kya; 95% CI: 35-41)."
But of course, when it's just one mutation and he's so close to the K2 root, he could just be called K2* so as not to mislead people into thinking that his Y-DNA has a close evolutionary relationship with NO.
DeleteOne mutation downstream K2 is the one shared with NO, but then there are other mutations donwstream in the UI specific line that nobody can ignore. And they do add up in the chronology.
DeleteI was very shocked by the claim of "NO-QR split" (K2 node) being "45 kya; 95% CI: 42-48 kya". That's totally absurd, the minimal 100% CI is some time before 45 Ka and it is very curious that in fig. 4a, UI, unlike other calibration points is displaced towards the present by almost 10% against the precision of his dating, argued above by yourself. 45 Ka is a terminus ante quem, there's 100%, thanks to UI that the K2 node is significantly older than 45 Ka BP. Whether it is 50 or 60 or whatever is I guess debatable but necessarily some time older than 45 Ka BP.
They are cheating again.
They are using the mutation rate of 0.82 x10^-9 mutations/site/year which comes from modern pedigree studies.
DeleteAny given node or sample will randomly deviate from the expected number of mutations by some amount - 5% of samples will fall completely outside of the 95% confidence interval, of course. So although we are pretty sure that the true date of the NO node really is more than 45 kya (radiocarbon also has a confidence interval, nevermind uncertainty of the calibration curve) that doesn't matter, with a CI of 42-48 kya half the time the true date will be older than 45 kya. So this actually makes sense.
From the paper:
"We also calculated that Ust’-Ishim’s Y-DNA haplogroup had 352 and 335 fewer private SNPs relative to what was expected for a modern-day individual (based on QR and NO clade individuals respectively). This provides two independent estimates of 41,100 years (95% CI: 37,100-45,000 years) and 39,200 years (95% CI: 35,600-42,900) for Ust’-Ishim’s age."
Ust'-Ishim's age from radiocarbon was 46,880–43,210 cal BP, so the high end of the CI for the mutation rate they are using overlaps the radiocarbon CI - barely.
Using just the fossil aDNA data you get a slower mutation rate which makes the ages about 10% older. Personally, I prefer the slower rate, but it could be wrong. It could be that Ust'-Ishim man and Anzick-1 happen to randomly have not quite as many mutations as expected.
Well, it's apparent that you can't use a single rate, when this one contradicts the facts. You have to use diverse rates that do fit the facts. You can't posit a theoretical model that clashes with reality. Age(K2) is necessarily much older than age(UI), so its average just cannot be more recent - and you don't fix that with the CI, because it's still age(K2) and not age(UI).
DeleteThe question is not to save and rescue a fixed mutation rate, whichever it is, but to produce models that make good matches with reality, with observed facts. Otherwise is like the obsession with circular orbits that dampened astronomy until Kepler. Sometimes just a minor adjustment, like Kepler's circle → ellipse "reform" solves everything.
Like Mandelbrot said: clouds are not spheres, mountains are not cones, coasts are not circles and bark is not smooth, nor does lighting travel in straight line.
K2 can't be that much older than UI. UI has 6 private mutations (1/47 shared with NO, apparently); this means he can't be very removed from the K2 ancestor. Y-DNA mutates every few hundred years. Even if you favor a slower mutation rate for whatever reason, he is very few mutations removed from the root, so it doesn't add up to much. We're talking thousands of years, rather than tens of thousands. 50-55 kya is not too crazy, but 60 kya is basically impossible.
DeleteOf course K2 is (slightly) older than UI, the authors realize this as well. He just happens to be so close to the root that the confidence intervals include the time before he lived.
Must be reasonably older: just take any of those trees, measure the distance between UI and present, then the distance between UI and the K2 node, apportion and add to 45 Ka, that's roughly how old K2 is.
DeleteUI is not the way you describe it something extremely close to the root of K2, it is clearly drifted downstream of it. So a 50-60 Ka age for K2 is reasonable when you calibrate using UI.
... "the authors realize this as well."
Not the one you quoted, quite apparently. The confidence interval does not fix that slip 48 Ka for K2 is still too young. The problem is that your reference author does not want to assume the implications of UI's actual age for the overall tree.
And it's not just "slightly".
''So a 50-60 Ka age for K2 is reasonable when you calibrate using UI.''
DeleteThus refuting your own claim of K2 expanding soon after Toba...
I'd personally estimate K2 to be 55kya based on the first attested human presence in New Guinea Highlands, which thought to be of about 50kya. For this estimation to be correct, we have to assume that the C1b2 migration hadn't reached the Highlands so early, due to its coastal current distribution. If we stick to the idea that K2b1 made the first move into the Highlands, then we can easily come up with a 55kya approximation for K2.
Hold on, I have theorized (not strictly "claimed") that the expansion of mtDNA N could have been soon after Toba, i.e. c. 70 Ka BP. The expansion of mtDNA R (and therefore of Y-DNA K2) seems a bit more recent, maybe c. 60 Ka BP. My lowest limit would be c. 55 Ka BP because in that time we already see presumably R-carrying (R0?) people with UP tech in West Asia and soon after (49-47 Ka BP) in Europe and Altai.
DeleteI'm not going to get right now dampened into Y-DNA MC calculations, partly because I'm with a bad cold, and partly because it needs plenty of time and good data, but I must emphasize that one of the problems is that there is no alignment with a "present" of the drift of the various branches, as we can see for example in the original UI Y-DNA tree, so effectively the MC ticks are partly random. And this randomness can make present day lineages to have "ticked" at totally different rates: one branch being 50% longer than another in many cases.
As for Papua, I can only presume that the colonization of the highlands took place some time after the colonization of the lowlands (those peoples arrived from the sea after all) so the date you mention is at best a terminus ante quem. But anyhow my educated guess or estimate is that the arrival of M27 could be around 65 Ka BP and that of Q c. 55 Ka BP. But remember that it is just an educated guess, just like any other estimate. Y-DNA K2b1 should have arrived with mtDNA Q but I haven't studied the case of C1b2, so I no idea if it should be considered an "allied lineage" (they traveled together) or an older one (traveled with Melanesian M27?)
Of course the SNP counting method, being so recent, still has to be perfected, because we can't as of yet be certain about the exact patterns of the mutations and whether they may be influenced by other factors, Just like why Y-STRs estimates were proven to be ineffective, but in truth, that's more because people, the scientists included, misunderstood them and failed to see that different markers behave differently and are more likely to mutate as the repeats increase, making some haplogroups more diverse because they'd be modal for a higher value at a certain marker. Because of that, the community wasn't able to establish adequate mutation rates. All of this thus resulted in a distorsion of reality especially when the number of markers is small, which is what we usually encounter in most studies with their miserbable reduced sets. So yeah, eventhough I wouldn't expect SNPs to be as complex as STRs, I'm still prepared to further changes as time passes. For the moment, we have an overwiew of the ancientness of haplogroups which helps us locating in time the various expansions.
DeleteAnd Q, descendant of M29'Q, did not arrive in Melanesia, it rather coalesced there as shown by the presence of M29 only in island Melanesia with only a marginal presence in the very east of ''mainland'' Papua New Guinea which could represent a back-migration from nearby New Britain. As M29 is one codR mutation away from M, it appears to be extremely ancient and thus can't be linked with K2'expansion, although Y-DNA M could have later expanded with a subset of Q. M29, as well as M27 and M28, probably enterred New Guinea from northern Wallacea and moved quickly along the coast till they've reached Island Melanesia, and the Y-DNA counterpart of such migration is C*. Don't forget that the mountainous belt of New Guinea would have act as a somewhat strong barrier that prevented people enterring into Australia and that could explain why many uniparental lineages observed in Melanesia aren't found in Australia. Migrations that enterred New Guinea from southern Wallecea didn't encounter these geographical restrictions and were thus able to reach Australia. As for the initial australian migration represented by Y-DNA C1b2b (C4) and mostly mtDNA N, it probably never landed in New Guinea but crossed directly from Nusa Tenggara as shown by otherwise austalian mtDNA O's presence in Flores.
Sorry, I obviously meant P, which is derived from R, not Q.
DeleteI think Ust'-Ishim coming from the east via the northern route is well worth considering.
ReplyDeleteThe original paper is now up on academia.edu. Very often authors upload their papers to academia.edu or researchgate.net shortly after they are published and so they can be accessed even while still paywalled at the original journal site. You just have to google for them.
http://www.academia.edu/8920635/Genome_sequence_of_a_45_000-year-old_modern_human_from_western_Siberia
But the meat is in the supplementary info which is free as usual:
http://www.nature.com/nature/journal/v514/n7523/full/nature13810.html#supplementary-information
The ~45 000 ybp date of U-I is calibrated and actually has a rather small confidence interval. The date of the K2 and R nodes based on it of course have quite large confidence intervals and are based on the assumption of relatively constant mutation rate, but such as they are they fit very well with expansion of Y hg K and mt hg R at about the beginning of the Upper Paleolithic 50 000 years ago.
It is of course possible that mutation rates have changed over time in the modern human lineage, or that they are quite irregular in general, in which case the pre-UI dates could simply be wrong as you think. So far ancient DNA has fit into the usual scheme but we need more high-quality samples, especially very old ones, to be sure of it.
Your view has the advantage of putting the spread of CDEF and L3 lineages in the Eurasian population well before Toba, in which case the disjunct distributions of D and E and M and N can be explained by all the linking forms in India getting buried under volcanic ash. The usual dating puts the TMRCA of BCDEF very roughly 100 000 years ago (which suits an early Out of Africa) and the split of CDEF, which was followed shortly after by the break-up of CF and DE, in the same time frame as the Toba eruption (both have a confidence interval of many thousand years but they very much overlap). In that case CDEF could be a post-Toba expansion from some refuge, which would not explain the distribution very well, or it could be an expansion that happened to start shortly before Toba, which would be rather a big coincidence, and would have trouble explaining the diversity of L3 in Africa.
Thanks for the links, Capra.
DeleteI don't mean that my tentative and rough chronology is the last word. We probably won't have anything like that until enough calibration points are confirmed. My own calibration points were somewhat arguable: "assuming map 2 to correspond to Jwalapuram (since c. 80 Ka BP) and map 5 to the earliest Aurignacoid cultures (Emirian, since c. 55 Ka BP or maybe a bit earlier)". Nowadays I'd assume an older date for map 2, probably 100 or 96 Ka BP, based on the mentioned oldest H. sapiens data in India and China, because people had to be in both regions already in those dates, so that date is clearly post-M. That would produce longer "ticks", and more so if I take UI as the other end reference (although I still think that 50 or 55 Ka BP are more correct, because we have to account for the time between the R+1 tick and the R+2 one, to which UI had not yet reached).
We can get a rough estimate (and I believe my mtDNA map sequence are a good tool for such exercises) but it will never be anything precise, after all if each tick is of c. 10 Ka, for instance, there's no reason why a particular tick could not have happened in 3, 5, 15 or 25 Ka of interval: it's a lottery. This stuff can be modeled, from repeated modeling, average tendencies can be deduced but in the end every new step is dependent on the previous ones and that is Chaos, because it cannot be predicted, just as no forecast issued today for the weather of December 25 is anything but a random guess.
So in the end we need hard facts: ancient DNA with a reliable associated date, the more the merrier. Once we get those multiple calibration points, we can get a bit more accurate but we can also perfectly find that the clock seems to tick faster in some periods and slower in others for no obvious reason.
What you say about Toba's ash can be indeed a very convenient explanation for the disjunction between the diversity of South and East Asia. After all it was India the most affected area, what may well have wiped out many lineages. However other explanations are possible too: mtDNA N and Y-DNA C and D may just not have found their niches to thrive in South Asia because they arrived slightly later or whatever other reason. But I agree that they are less plausible, so Toba should be accounted for and IMO it seems to fit reasonably well with another hard to explain phenomenon: the expansion of some SE Asian lineages, like mtDNA R (bringing along some N) and Y-DNA K2, which I believe are contemporary and UI is a good example of their very ancient "partnership", even if he seems to be in the wrong geography (a bit too far Northwest for what I'd expect).
Naah, if any Y-DNA lineage had to be to be the counterpart for mtDNA M, it's definitely D. But in all likelhood, it is surely not F because no F is ancient enough compared to D or even C to have expanded let's say 15ky earlier than N. But if you don't believe one bit of the estimates even in a relative way, then what de we make of the fact that mtDNA M peaks in Japan, Tibet and the Andamans, although for a population as isolated as the Andaman Islanders, it's easy for an already predominant line to achieve fixation !!! Also keep in mind that aside from M1(which is still originallly far eastern), there are no M lines in the west, which again perfectly goes in line with Y-DNA D's distribution in comparison to F or C.
DeleteIt is defintely India that is fooling you, but mtDNA M's basal diversity in India is more of an eastern/central thing and even there it is not higher than what is observed in southeast Asia. You would probably ask about the lack of basal N lines in India, but that's because you associate R to Y-DNA K2 rather than to HIJK as whole, a connection that would explain both the highest diversity of R in India as well as the predominance or R derived clades in west Eurasia. mtDNA N would be more linked to Y-DNA C and F(xHIJK), and for all we know, the indian F* clades could form a monophyletic clade with the rest of HIJK. That also goes in line with the fact that mtDNA R is very ancient, only a bit less than greater N.
Y-DNA D on the other hand is lacking in India but if we were to admit that it came first, it becomes easy to at least consider a possible extinction after so many migration maves. Don't forget about autosomal ASI being very close to Onge but not wholly so. But the Andaman Islands were peopled from Burma most likely and we are thus forced to accept a very ancient population layer involving India and southeast Asia. Later early migrants in India mixed with these Andaman related people, forming the ASI component in the process. Also, there's no way that the Andamans were peopled recently and I still think they are the vestige of some really ancient layer that predates even the initial colonization of Sundaland and Sahul, which invlolved probably Y-DNA C and mtDNA N. Also, the fact that a landbridge connecting Burma to the Islands has not been proven yet shows that it's very ancient. SAying thet peopling occured via India is as dumb as saying the Isles were peopled recently, although there was a second migration, contributing mainly to the genetics of the now almost extinct Great Andamanese, clearly distinct from the older and smaller Onge/Jarawa/Sentinelese block.
My impression is that not just D but CF in general are associated with mtDNA M. Assuming that only D was related to it would be unable to explain the reality of South Asia (no D) and make extremely difficult to explain that of East Asia and Australasia. The most important Y-DNA lineage in the first phase of the Eurasian colonization (the one most tightly associated with mtDNA M) was in any case F or rather CF. To my eyes, D seems to be rather a minor lineage that traveled along and that had only limited success, although of course it is possible that it was more common in the distant past. A similar situation on the matrilineal side of things is that of mtDNA N but this haplogroup got luckier in the second, post-Toba, phase (but associated to CF-derived Y-DNA lineages, not D).
Delete"because no F is ancient enough"
Based on what?
See: http://forwhattheywereweare.blogspot.com/p/y-dna-ages.html
CF seems to be comparably old to E and F is only a bit more recent than CF. That 1000 GP dataset is clearly insufficient to characterize D and C properly but for E, F and CF it should work perfectly (another issue is calibration, as usual).
"You would probably ask about the lack of basal N lines in India, but that's because you associate R to Y-DNA K2 rather than to HIJK as whole"...
It doesn't matter: mtDNA R is not that numerically important in South Asia and it is clear that the basal diversity of M in the subcontinent is much larger than that of R. As an additional reason, HIJK (or F(xG)) shows no sign of ever having been to SE Asia before the K2 stage, while mtDNA N quite apparently was instead (R seems centered rather towards South Asia but its "mother" N not, and is quite clearly SE Asian by origin - the pre-N precursor was either too small to leave a legacy or its signal was wiped out by catastrophes or genocides).
I think that the most correct interpretation is to understand the "first Asians" as mostly Y-DNA CF and mtDNA M, with a few carrying Y-DNA pre-D and mtDNA pre-N but having no success (at least none that we can detect) before reaching to SE Asia. Actually it's very likely that the same happened to Y-DNA C, whose presence in South Asia is most likely product of backflow from the East, even if very ancient.
One can interpret the F+M early "homogeneity" of South Asia as product of Toba but another quite possible explanation is much simpler: drift. Dominant lineages tend to displace smaller ones, unless these get lucky and make a founder effect. These founder effects seem to have happened in the Eastern fan-like dispersal only.
"Don't forget about autosomal ASI being very close to Onge"...
DeleteI am of the opinion that such thing is a myth, not well founded. Andamanese do have some South Asian admixture (visible in mtDNA only) but overall they must be considered an Eastern population, along mainstream East Asians, Australian Aborigines, Papuans, Orang Asli, Ainu and Filipino Negritos (each of them is a distinctive derivative of that Eastern fan, with limited or no relation to the others, nor to South Asia, afterwards). I do think that ASI does exist and is roughly representative of the genetic landscape of pre-Neolithic South Asia but I strongly disagree with the use of Onge as proxy (you could use Dai, Chinese or Papuans and would be nearly the same result).
Andamans were peopled probably in the wider context of East Asian colonization but not too early. You may have noticed that in my maps there is no dot in the Andamans and that is because their matrilineages appear to have a more recent age than the period considered, i.e. the Andamans were probably only settled after 50 or even 40 Ka BP. I'd have to recheck the data for the Andamanese lineages to produce a more precise estimate but an older version of the same analysis put them in early UP chronology. That they aren't anymore in it is surely caused by the discovery of more basal mutations in the meantime (each time I used the most up-to-date phylogeny).
"... we are thus forced to accept a very ancient population layer involving India and southeast Asia".
Well that fits roughly with my understanding about greater mobility in pre-UP times between South and SE Asia than afterwards. But that's way too generic. It's probable that there was first a major flow eastward right after the M node (i.e. c. 100 Ka BP or soon afterwards) and later a major flow westward right after the N node (coincident with the R node and possibly 70-60 Ka BP). There may have been some other minor flows along the coast but I'd say that stuff like the Andamanese genesis should be closely related to the second episode (the details are necessarily obscure in any case).
''Assuming that only D was related to it would be unable to explain the reality of South Asia (no D) and make extremely difficult to explain that of East Asia and Australasia.''
DeleteThat's exactly why I told you not to be fooled by India and because D could have been wiped out by the numerous later migrations and local expansions that shaped the genetic structure of the subcontinent, in chronological order those involving C1, F-H and K. This one actually probably coalesced somewhere near Gujurat and even returned to India many thousands years later in the form of K2b2 (P) and K2a (mainly O, but also N and ''X''), adding itself to the already common LT. Ah, and let's not forget J2 of course. So it becomes easy to conceive that a relatively small-sized population bearing Y-DNA D and mtDNA M would have been dwelling in eastern India and SEA until the arrival of these other distinct populations. After some time of absorbing the original female gene pool and disriminating the male one, we would obtain the present picture. After all, you yourself support the same scenario for North Africa in order to explain the discrepancy between uniparental lines, Y-DNA E-M81 and mtDNA R (HV, JT, U).
As for Australasia, it's clear that there were two main waves. Or else how would you explain Australia and only Australia's high basal diversity of N(xR), actually the highest in the world. So yeah, we have C1b2 first with N, with C1b2b (previous C4) or rather pre-C1b2b, moving quickly and leaving no trace in Sundaland/Wallacea, and C1b2a(previous C2) staying behind and absorbing previous and/or incoming elements. This is why the C2 bearing places are very different from Australia mtDNA wise, having abosrbed much M and also R lines, while Australia's quick move only allowed for a minimal contribution from M lines, which would have happened in Sundaland (or further North in Burma/Thailand) while picking up Y-DNA D's female counterparts. So yeah, D, being the first and thus likley less advanced technologically, was easily segragated.
And when I said no F is ancient enough, it was in the sense of if HIJK is associated with, not N, but derived R, no F line would have separated early enough given that HIJK itself is very early in F's phylogeny. So, it would be difficult to associate Y-DNA F with mtDNA M, given the following (let me use the notion of equality to keep things simple) :
- G, or pre-G = N1 : G is very early in F's phylogeny and it appears to have been alone in west Asia for some time. There's no M there, instead it appears to be linked to this Basal N line, which also has South asian representatives. If even this so early of an F split appears to be linked to N, then which lines would be with older M???
- C1b2 = N, C1a2(C-V20) = N : I already explained these associations to you and I bet you're already familiar with the first one. If these very early clades were with N, how would F with its latter expansion would be with 10ky older M?
How on earth would mtDNA N be present with the earliest CF movements but not M? HIJK can't be with M if not only earlier pre-G was with N, but more earlier C as well. Plus, the fact that we see no M in the west in contrast to early C-F also argues in favor of thisCF/N link. It looks too clear and if we were to postulate that M was wiped out in all of these geographically distant areas, it would be too many coincidences to appear realistic. So really, it's only India that is still making people believe that F = N...
As for the Andamans, I didn't support the use of Onge as a proxy for ASI, I only said that ASI is part ''Onge'', rather than the reverse. So Andamans don't have South asian admixture (unless we're talking about the Great Andmanese and paternal south asian lines but here we're dealing about the older Onge/Jarawan/Sentinelese block), it's actually South Asians that have inherited some of their genes from an andamanese-like population that was once wandering in SA+SEA and that's exactly what I was trying to show you with mtDNA M in India. Plus M31, and M32 are too rare in the mainland to have been drifted to these levels. They may have been more common in extremely remote times in Burma but not anymore. And why would you consider this to be SA admixture rather than SEA given that the islands were settled from Burma, looks like you are tricked by their modern belonging to India. But have you seen how these guys look like? I guess you did but how come you guys can't accept their extreme ancientness? They're not even remotely close to either Indians or Burmans. As for M32, it is extremely ancient and even if the Onge belong solely to M32a, it's because of bottlenecking of course. So M32 could have been basal when it came to the Andamans, with M32a coalescing there and eventually taking over. M31a1 is apparently more related to the Mainland, but it could still be very ancient or even be assigned to the second migration. In admixture runs, Onge appear as a three-way mixture between South Asians, East Asians, and Oceanians, but I reckon this is not actual admixture but rather reflects the role that this andaman-like mtDNA M-bearing population played in the formation of these three populations.
DeleteIn the very last line of my first post, I said:
Delete''So really, it's only India that is still making people believe that F = N...''
I obviously meant ''So really, it's only India that is still making people believe that F = N...'' and not N...
What's up with me, I meant ''...F = M...'' and not N
DeleteI think Y-DNA G was a private (F*) lineage that just happened to get lucky in the West Eurasian colonization. Much of the same can be said of IJ, although it is further downstream and their role seems bigger. I don't think there was any mtDNA N in West Eurasia before the P1 wave brought it from SE Asia, along with the more obvious R. The only clearly "western" (South Asian) matrilineage involved was M1.
DeleteRe. Andaman Islands, both mtDNA lineages are shared with India (from my wiki-notes):
→ M31 (in Andaman Islands, Bengal, Munda, Orissa, Saharia)
→ M32'56
→→ M32 (M32a in Andaman Islands, M32c in Madagascar)
→→ M56 (in South Asia)
Y-DNA D* seems of course most related to SE Asia.
Re. Australasia: I'd also think of two (or maybe even more) waves.
"C1b2 = N, C1a2(C-V20) = N"
Obviously mtDNA N is much more common than Y-DNA C1, so nope.
"How on earth would mtDNA N be present with the earliest CF movements but not M?"
That's not at all what I'm saying. When CF (and the minor lineage D) first expanded, it did so with mtDNA M. However there were obviously some other minor L3-derived lineages, at the very least pre-N (if there were others, they vanished). At some point this lineage had a major lucky strike and, associated with Y-DNA C and K2 expanded almost worldwide. That lucky strike could well be associated with the Toba event.
N was in West Eurasia before P1 or would you argue that basal N1, clearly associated with early western Y-DNA F lines such as G, is SE asian? P1's migration almost certainly happened by following the southern edge of the Himalayas, thus they ended up in Central Asia. mtDNA N1's distribution is definitely not like that, as only highly derived N1a has some presence in Central Asia/Southern Siberia. The rest are overwhelmingly western and they're certainly linked with G/H2 (previous F3), two very early splits within F.
DeleteFor the Andamans, who cares about the relationship with India? That's exactly one of the reasons why I have speculated about a very ancient population roaming in this entire region. But what is certain is that they (the first Andamanese) came via Burma and not India, that's too far. And it was Y-DNA D with mtDNA M. D in India was wiped out entirely, which is not surprising like I already said quite a few times now, while it survived in the Andamans due to strong isolation. Ah, and aren't you aware of the two alleged DE cases in India? I've checked their haplotypes before and they're the same, which indicates that they're not misclassified, unless their YAP+ state is a parallel mutation but the chance for the YAP mutation to happen twice is extremely reduced, voire practically impossible. Furthermore, they don't look (in terms of Y-STRs) like any D or E I know, so they're most likely some yet unclassified D clade that may or may not be directly related to the andamanese one.
mtDNA N(xR) is the one I'm talking about when dealing with Y-DNA C/C1. And its not very common either in Europe, and we know mtDNA survies better than Y-DNA so its not surpising that mtDNA X or whatever would be more sucessful than C-V20. But in any case, C-V20, a clade as old as K itself(!), is exclusively european and didn't move with mtDNA M. And C1b2 does ''equal'' N, Australia nails the coffin here. I know my paragraphs are sometimes overwhelming but I reckon they're worth reading...
I actually meant how come we seemingly have mtDNA N(xR) in every place where very early CFs are found (Europe-C1a2, Australia-C1b2a, West Asia-G), but no M, apart from Australia but according to me it's because before ending up there, they (C1b2 males) have absorbed some M lines somewhere in Southeast Asia (probably north of Sundaland). But the problem is that you argue that CF and D were part of the same initial move, while I suggest they weren't,... For K2, it expanded after C, that's for sure, and it expanded with some R clades, of which the expansion is distinct from that of earlier N(xR).
Naively, using the latest (16th) PhyloTree build, coding region mutations from N to Western lineage:
DeleteN1: 3 (incl. I)
N2: 4 (incl. W)
X: 4
R0: 2
R2'JT: 2
U: 4
So it seems apparent that the oldest Western haplogroups (should add M1 but would not change anything) are R0 (exclusively West Eurasian) and R2'JT (shared West Eurasian and South Asian). Luckily the latest builds have not changed anything substantial in this aspect and my reconstruction still stands. U, X and N2 seem to have expanded later (maybe much later, Gravettian era looks like a good time-frame) and N1 seems to sit between these two.
So I see no reason to think that Western N subclades were all the time sitting in the Persian Gulf but rather that they arrived in the same general migration Westward which brought R subclades to the West. Actually N looks totally SE Asian, so they should in any case have migrated from there to here somehow, even if they left only a weak legacy in India (N2a, W and N5, which is sister of N1). But actually they are not that important in West Eurasia either: the three R subclades clearly dominate the landscape everywhere.
"For the Andamans, who cares about the relationship with India?"
Well, first of all in order to be precise, second to underline that there are complex South-SE Asia relations in those early times, third to balance the SE Asianness of Y-DNA D*.
"And it was Y-DNA D with mtDNA M."
While y-D is clearly only associated to some mt-M subclades, making a simple pairing of ALL mt-M with y-D makes no sense whatsoever.
"D in India was wiped out entirely, which is not surprising"...
It was probably a minor lineage. That's the most likely reason of its extinction in South Asia: drift. Toba may have helped but the case is that D was almost certainly from the beginning the young cousin that the CF bros brought along, much as pre-N was relative to big sis M. Smaller lineages are more likely to vanish... unless they stop being small via founder effect. It's plausible that there were some other lineages in the migration that were not so lucky, we just can't get our hands on them.
"Ah, and aren't you aware of the two alleged DE cases in India?"
Vaguely but AFAIK they were not tested for the marker of E, so everybody imagined they'd be E1b migrants from Arabia or something like that.
"Furthermore, they don't look (in terms of Y-STRs) like any D or E I know, so they're most likely some yet unclassified D clade that may or may not be directly related to the andamanese one".
That's interesting. I'll take your word on that.
Anyhow the Andamanese D* hang directly from Daic D* → Hong Shi 2008, fig. 3.
Re. y-C and mt-N, there are some other instances of clear pairing, notably Australia. But there's always also M subclades implicated. You can't disentangle N and M, except maybe in West Eurasia and even here there is M1. And even less you can disentangle N(xR) and R, except in Australia. It's a mess, although it can be understood as such "mess", where lineages exist in real, complex populations and not populations in simple lineages that never mix nor change allegiance. Some lineages are bigger since early on and hence more resilient, some are smaller and therefore more fragile... but a few of these smaller lineages get lucky and become big and resilient. The histories of matrilineages and patrilineages can sometimes be paired (are clearly parallel) but obviously people is not all the time marrying their cousins, so these allegiances appear and disappear.
So then, tell we which mtDNA line would you associate with C-V20. Due to the fact that mtDNA mutates slowly, we can't really say how much time passed between the few codR mutations that separate each line from the N node. We can have somewhat of an overview but it's not perfect and the fact that, just like C-V20, some lines were bottlenecked doesn't help. Most haplogroup R lines expanded really fast and were very successful just like HIJK, but this wasn't necessarily the case for basal N lines, thus they're more likely to carry longer tails. Also, JT has 4 codR mutations from N and HV has much more. Plus the PhyloTree didn't receive any update since 2 years...
DeleteAbout the Anadmans again, daic D* is the oldest split within D, which is in turn possibly the earliest split of all non-african clades. Again one of the reasons for which I firmly believe in this older eastern NRY D- and mt M-bearing layer. Also, mtDNA M32, like I was saying, would represent the first migration to the Islands, as it is only one codR mutation away from M (split between M32 and M32 was really fast, just like, for example, Y-DNA K2 clades separated almost immediately from each other).
Don't forget to answer my initial question and also how come no M lines survive in the west, again in perfect harmony with Y-DNA D. And don't bring on M1 as it's not western eurasian but rather african and, above all, it is not basal and has clear ties with the far east.
"... tell we which mtDNA line would you associate with C-V20".
DeleteThat's the kind of question I would never make. Why? Because it is clearly a minor lineage in spite of its fluke finding in the La Braña bros. This oddball Western C should obviously be considered a minor lineage (along with some other oddballs like H2) within a migration best described by other much larger ones, notably P1, IJ, G and T. The main mtDNA lineages associated with them are R0, JT, U, N1, N2, X and M1.
You see the difference, right? I don't imagine pairs of lineages but complex and realistic populations with arrays of various Y-DNA and mtDNA lineages. Naturally these populations would have subpopulations, in which some of the implicated lineages would become more or less successful, or even not be present at all. One of these subpopulations could be more pioneering, another maybe more coy (later roles could be reversed maybe), one could expand more dynamically in certain region, a second one in another region, a third one maybe got stuck in a valley with limited expansion options, etc. It's people and not lineages who "write history", even if it is unwritten paleohistory, only recorded in stones, bones and aminoacids.
"... how come no M lines survive in the west"...
And M1? M1 is a fully fledged Western lineage and seems quite approximately associated with Y-DNA T in their Western expansion.
But you do have a point because it seems quite apparent that the girls were somehow keeping the group together, allowing for some notable patrilineages to join the population but much more rarely matrilineages. A simple explanation would be matrilocality (alias uxorilocality), in which the father's role is adopted by the maternal uncle (what would make cultural transmission pretty much associated to matrilineages and not Y-DNA). I can't be sure this was the case but it's a possible scenario. Let's not forget that these peoples were the ones soon after engraving and sculpting all those famous "venus" icons, which must have got some sort of maternal symbolism. A key part of that female-centered cultural transmission would have been the needle and the art of sewing multi-layered clothes, almost necessary to conquer the cold north.
"Most haplogroup R lines expanded really fast and were very successful just like HIJK, but this wasn't necessarily the case for basal N lines, thus they're more likely to carry longer tails".
DeleteThere are four N subclades that seem to be as old as R, i.e. they do have very short stems. They make 1/3 of all basal N sublineages. It's true that R sublineages do show a stronger dynamism but some of it was clearly also present in N already, after all N and R stars are of about the same size, regardless of what their derived lineages did afterwards. The N node is still one of the most clear signals of expansion in the human mitogenome, after those of M and H (really huge stars) and about the same size as that of R.
In a simplified reconstructive narrative, Sykes style, N sent 16 (or 17 if we accept the proposed N14) daughters around the world. Some went West, following the trail their ancestors had taken to the East along the great river, some went North following the coasts and some went south to the new lands that the great crocodile had lead some people to. A few remained in the homeland of their mother.
Naturally some were more successful than the others, and obviously R and her own "daughters" were the most successful ones. They didn't only head West, as their mother did but many headed back to the rising sun, to their grandmother's homeland, where they succeeded (R is much more common than N(xR) in SE Asia) and joined their aunts in their various routes north and south.
Obviously this is even more easily explained if N and R had approximately the same homeland in fact and R was the oldest of all N sublineages (even if we can't confirm that from the mutation rate, just a hunch).
...
"Also, mtDNA M32, like I was saying, would represent the first migration to the Islands, as it is only one codR mutation away from M"
Problem is that M32 is not the Andamanese line, it is M32a. M32c is only found in Madagascar AFAIK but it is not particularly likely that the Austronesian migrants who carried it picked it in the Andamans (not impossible but as likely as any other origin, be it in Indonesia or even in India). And M32 is not one but five mutations downstream of M. That's why I did not put any red dot on the Andamans in my latest version of the reconstructed Eurasian expansion based on mtDNA. Instead I assume a geographically ill defined M32'56 in the India-Indonesia arch (maybe in coastal Myanmar, which has a very large M diversity) and then subsequent migrations in various directions, only one of them being the Andamans:
→ M56 to India in step M+2 (~75 Ka BP)
→ M32a to Andaman in step M+5 (~45 Ka BP)
→ M32c to maybe Indonesia in step M+12 (or surely earlier, because this is the Malagasy lineage in fact)
"And don't bring on M1 as it's not western eurasian but rather african"...
Actually it is West Eurasian and there is no doubt that it expanded from Arabia, probably in association with Y-DNA T primarily. The West Eurasian expansion had very direct African offshoots, most notable of which are the African branches of M1 (not the whole lineage), the African branch of X (X1, primarily Egyptian) and U6 (primarily NW African). These African offshoots were quite probably related to the early LSA (African version of the UP). Similarly I think that not just Y-DNA T is related to this African extension of the West Eurasian colonization but also Y-DNA J1, which is highly diverse in NE (but not NW) Africa.
Erratum: "They make 1/3 of all basal N sublineages", should be 1/4.
DeleteThis comment has been removed by a blog administrator.
ReplyDelete^ Banned long ago and the main reason I had to begin pre-moderating comments.
DeleteHow does one put this delicately : id take any revolutionary conclusions from China with a healthy dose of skepticism. They've been trying to "prove" AMHs have been in the Far East for a much longer time. Not to say they're wrong , but the dating was based on indirectly Dating the nearby stalagmites; so id like to see independent verification of the stratigraphic context of the finds.
ReplyDeleteFinally it's a bare minimum they aDNA type the teeth- that'll help pass the whiff test
Actually the traditional Chinese bias used to be that they were descendants, at least partly, of local H. erectus, a fully completely "race" or "subspecies" of Homo. I don't think that such issues are a problem anymore and I believe that disqualifying findings that have gone through ample peer review and subsequent publications only on anti-Chinese bias is not reasonable and rather seems a burning nail attitude of defense of one's biases instead (such as the clearly pseudo-scientific molecular-clock-o-logic blind faith, based almost only on scholasticism, the worst enemy of science). In any case, there is also the Indian data, as well as other findings from South China (Zhirendong seems the best dated early H. sapiens in Asia, East of Palestine, and he/she was ~100 Ka old).
Delete"Finally it's a bare minimum they aDNA type the teeth- that'll help pass the whiff test"
I keep reading that type of stuff without considering that usually it is not likely to produce any usable data (too old, too hot climate). For example recently in relation to H. naledi (I was flippant!) So far the only methodology that seems to work for such problematic contexts seems to be the sampling of the petrous bone (what implies a skull at the very least) and is totally new, published only in the last months, and is not without challenges either.
In any case it'd be not "the bare minimum" but a second phase of the research. Not everyone is at the avant-guard levels of the Max Plank Institute, actually almost everyone is not and has difficulties of all types to get near their referential but hard to imitate standards.
This comment has been removed by the author.
DeleteThis is southern China, not Siberia, recoverable aDNA is wildly unlikely. They couldn't even recover enough carbon to get a radiocarbon date on the teeth. Miracles should not be the bare minimum.
DeleteThey did get a radiocarbon date for one of the animal bones of 43 000 years, but that was a minimum because it was at the radiocarbon background of the lab, i.e. the real date may have been infinite.
New material is always great. I accept that the teeth are from AMHs and very old - although I am not confident of the dates at this point. My guess would be that Svante Pääbo could at least coax some mtDNA out of a sample given enough time. While that wouldn't help with the dating, it would shed some light on which path the ancestors of these people took to get to China. Looking forward to it.
ReplyDeleteJoy
The find is paradigm changing if the date is correct. But, there is very good reason to be skeptical despite any really obvious methodological flaws. Separate and apart from the arguments from big picture that has emerged from ancient DNA, there is the complete absence of an hominin archaeological record (H. Erectus, H. Sapiens, or otherwise) at about 80kya, to corroborate a human presence then. A date of 60kya or even 70kya, in contrast, would be just on the cusp of the modern human archaeological record in SE Asia/E Asia, and that discrepancy between the indirectly determined date and the actual age could arise in all sorts of subtle ways.
ReplyDeleteIt is almost always the case that relics of tools, fires, garbage piles, mud encrusted footprint fossils, intentional marking on cave walls, etc. greatly outnumber the quantity of human remains found at the same age, so finding human remains tens of thousands of years before any of the earliest modern human type archaeological relics known in a region from Burma to the Pacific Coast of China from the Arctic to the Antarctic is quite problematic.
As Maju notes, a complete extinction of first wave modern humans can be pretty problematic too. What would do it? Not Toba way to the east in China. And, it is hard to imagine a population too thin to be sustainable making it so far east of Africa. Settlements would have been established all along the route and even if lots of them were fragile, it is hard to see why they all die off.
Without some corroboration, it is very hard to accept this as authentic. Extraordinary claims require extraordinary proof.
Systematic skepticism is not a good idea. Reasoned, healthy skepticism needs good reasons (and I see none). I haven't yet got a copy of the paper but for what I've been told the site is unquestionably dated on unquestionable geological grounds: the 80 Ka BP is a minimum ante quem date and the 120 Ka BP is a maximum post quem one. So the average date is rather 100 Ka BP but researchers are being "conservative" in their phrasing.
DeleteYou won't find in East Asia any kind of tools that can be considered "unmistakably modern" before a very late "mode 4" that arrived via Siberia. That's a characteristic of East Asian "modernity": a persistent "mode 3" flake industry, with almost no specific traits, that offers almost no clues. There's also no rock art. So following your (knee-jerk) logic, Andrew, there were no H. sapiens in North China until 30 Ka BP, and in SE Asia until almost the Neolithic. That obviously makes no sense whatsoever.
However you are denying the most clear evidence of all: human remains, which are piling up. I presume that in the case of South Asia you'll do the opposite: there are unmistakably "modern" (African-like) toolkits, yet no skulls nor bones, so from your knee-jerk "skepticism" you will argue, I presume, that without bones those toolkits could have made by Denisovans or Neanderthals, right?
Whatever it takes to keep worshiping the pseudo-scientific molecular-clock waste of time.
There are three different pre-80 Ka modern human findings in South China, contemporary African-like tools in India in two different sites. Clear precursors in Arabia and Palestine. An ideal climatic window. What else can you need?
There's also people who similarly claim that the world is 4 Ka years old and even that the Earth is flat. I'd rather not be compared with them, will you?
I have very little doubt that modern humans reached India significantly before Toba, where archaeological remains are abundant and there is a clear departure from Neanderthal associated tool kits as you indicate. But, the pre-Toba archaeology stops in India and never picks up again. There is also a significant ill understood gap in the Homo Erectus record of tools and remains which stops about 100kya with nothing replacing it for tens of thousands of years. The fact that modern humans left such clear remains in tropical and subtropical India, makes the absence of comparable relics in SE Asia and Southern East Asia all the more striking.
ReplyDeleteThere are two older and questionable indications from South China of human remains pre-Toba (i.e. 75kya), and now this third instance and a lot less questionable, but not beyond question either. Those three sets of human remains are the entire hominin archaeological record from 100kya to 70kya from Burma to the Pacific. How can any early modern human population have left so few traces?
I certainly don't doubt that these are very old hominin remains and that they are very likely Homo Sapiens. Odds on, the dating of the material they dated was also right. If there was a dating issue, I'd suspect it to be some circumstance that caused the teeth to end up physically in the vicinity of older materials for some reason, of which many are possible. I don't think that the scientists on this find are being dishonest, but thinking like a Baysean, there is good reason to suspect that the straightforward and naive interpretation is not the most likely to be correct, just as you'd immediately doubt the accuracy of an indirect date of jewelry with a classical Greek era design found in Peru in a 2000 BCE dated stratum.
But, reassessments of initial dating not done with the remains themselves are not all that unprecedented. Lots of European Neanderthal and Cro-Magnon sites have been reassessed in that way and so have a couple of the leading sites in Indonesia. Those reassessments, in each case, were in the direction suggested by the overall paradigm.
Post-Toba, there is a set of remains ca. 65kya in mainland SE Asia, and of course, lots of stuff in and around Australia by the time you get to 50kya (with some Australian remains arguably 60-65kya).
Like you, I'm not terribly impressed by molecular clock estimates, although I do put much more weight in phylogeny determinations which can sometimes be calibrated with ancient DNA, and I do put much more weight on linkage disequilibrium methods where the dating is much more reliable.
Even the Denisovan and Neanderthal remains found in the Denisovan cave had a lot of archaeological relics other than remains associated with them, and lots of other Nenderthal traces nearby. These seem to be just teeth. If there's more corroboration, I'm happy to hear it and reassess. But, it seems pretty spare to me.
There's no gap that I know of in the Indian occupation. Anyhow, how do you envision people migrating to India and not to SE Asia, considering that:
Delete1. There is no barrier, no deserts, no impassable mountain barriers, no oceans!
2. The basal genetic pool of Eurasians is similarly split between the two regions (what long ago inspired the coastal migration theory, which, right or wrong, attempted to explain a very fast expansion not just in India but also in SE Asia).
3. If anything we see a backflow from SE Asia (probably after Toba but not necessarily right after it) and not any second immigration from Africa or the expansion of a reserve from West Asia.
Basically I think that the people who are "skeptic" is being just molecular-clock dogmatic, fanatics in denial of actual evidence, just like a creationist before the skull of a T-rex.
I also suspect that the 80,000-120,000 years ago estimate is really 100,000 years ago +/- 20,000 years with 20,000 years representing one standard deviation. Thus, a 60,000-70,000 year age for the remains might be 1.5-2 standard deviations from the mean estimate value, which is not that remarkable a difference from the mean estimate.
ReplyDelete