January 28, 2015

~55 thousand year old modern human from Manot cave in Israel

It seems that this abstract came online one day early on my news feed and will probably appear in Nature tomorrow. I will update this entry when the paper properly appears. This is of course very important because it directly proves that modern humans appear in Eurasia before the Upper Paleolithic revolution, and disproves the theory that modern humans spread UP technologies with an expansion out of Africa.

We will have to wait until tomorrow to see exactly what they compared it against. The abstract contrasts it with "other early AMH" from the Levant, which I presume means the Skhul/Qafzeh specimens of ~50ka earlier than Manot. But, they also say that it is similar to UP Europeans and recent Africans, which suggests that they did not find any particular similarities to old African skulls of which there are many.

UPDATE: The paper is now online

UPDATE I: The authors write:
Manot 1 could also have been a direct descendant of early AMH populations (such as Skhul/Qafzeh), but the differences in morphology between Manot 1 and the majority of fossils from these sites render this possibility unlikely (Supplementary InformationC).However, it should be noted that within- and between-group morphological variations in these populations are extremely large18, rendering any conclusion based exclusively on morphology as tentative. Nevertheless, the absence of
otherAMHspecimens in the Levant between the Skhul/Qafzeh material (,120–90 kyr ago) and the later appearing Manot 1 (,55 kyr ago) does not support the hypothesis of continuous representation and local evolution of AMHs in the Levant. 
On the other hand, the considerably fluctuating climatic conditions during MIS 5 and 4 (favouring an alteration of differently adapted populations), the unequivocal presence of Neanderthals in the region in the time gap between early AMHs and the Manot population, and the continuing evolution of AMHs in Africa19 advocate for the most parsimonious explanation, which is that theManot people re-colonized the Levant from Africa, rather than evolved in situ.
I don't buy this explanation. The key phrase is "the majority of fossils". In Figure 3 is it is clear that Qafzeh 9 is completely modern. Moreover, UP Europeans like Cro-Magnon 3, Mladec 6/5 are close to AMH Levantine specimens such as Skhul-5 and Qafzeh-6. Indeed, the late Neandertals such as Shanidar and Amud are already moving towards "modern" humans.

African samples like Omo-2, LH18, and and Jebel Irhoud 1/2 don't look anything like modern humans. It is quite strange that the authors interpret this evidence as discontinuity between ~100ka humans from the Levant and replacement by a fresh Out-of-Africa wave, when in fact all the Qafzeh/Skhul remains (and a couple of Neandertals) look much more plausible relatives of Manot than any of the ancient African samples. Not single sample has been found in Africa from the mysterious hypothetical ancestral population of modern humans that supposedly colonized Eurasia ~60ka. As far as I can tell, this theory has nothing to support it, except, perhaps, (i) some misinterpretation of old genetic data based on the now discredited "fast" mutation rate, and (ii) the belief that behavioral modernity first arose in Africa and coincided with the spread of modern humans from that continent into Eurasia.



Nature advance online publication 28 January 2015. doi:10.1038/nature14134

Levantine cranium from Manot Cave (Israel) foreshadows the first European modern humans

Authors: Israel Hershkovitz, Ofer Marder, Avner Ayalon, Miryam Bar-Matthews, Gal Yasur, Elisabetta Boaretto, Valentina Caracuta, Bridget Alex, Amos Frumkin, Mae Goder-Goldberger, Philipp Gunz, Ralph L. Holloway, Bruce Latimer, Ron Lavi, Alan Matthews, Viviane Slon, Daniella Bar-Yosef Mayer, Francesco Berna, Guy Bar-Oz, Reuven Yeshurun, Hila May, Mark G. Hans, Gerhard W. Weber & Omry Barzilai

A key event in human evolution is the expansion of modern humans of African origin across Eurasia between 60 and 40 thousand years (kyr) before present (bp), replacing all other forms of hominins. Owing to the scarcity of human fossils from this period, these ancestors of all present-day non-African modern populations remain largely enigmatic. Here we describe a partial calvaria, recently discovered at Manot Cave (Western Galilee, Israel) and dated to 54.7 ± 5.5 kyr bp (arithmetic mean ± 2 standard deviations) by uranium–thorium dating, that sheds light on this crucial event. The overall shape and discrete morphological features of the Manot 1 calvaria demonstrate that this partial skull is unequivocally modern. It is similar in shape to recent African skulls as well as to European skulls from the Upper Palaeolithic period, but different from most other early anatomically modern humans in the Levant. This suggests that the Manot people could be closely related to the first modern humans who later successfully colonized Europe. Thus, the anatomical features used to support the ‘assimilation model’ in Europe might not have been inherited from European Neanderthals, but rather from earlier Levantine populations. Moreover, at present, Manot 1 is the only modern human specimen to provide evidence that during the Middle to Upper Palaeolithic interface, both modern humans and Neanderthals contemporaneously inhabited the southern Levant, close in time to the likely interbreeding event with Neanderthals.

Link

21 comments:

Davidski said...

Holy shit, I thought it was a genome.

eurologist said...

This could be ~5,000 - ~7,000 years earlier than previous UP dates in the Levant - which is significant. It still aligns with models that posit UP expansion out of the SE/ S / SW subcontinent in both a southern (-> Levant) and northern (-> W Russia to C Europe) and perhaps intermediate (Anatolia) migration. Just that with these new dates, the S route was 10ky to 15ky earlier than the the entry into Europe (e.g., y-DNA C, F, G vs. IJ?).

Also, it will be interesting to follow connections to original remaining SW Asian populations from ooA.


"the expansion of modern humans of African origin across Eurasia between 60 and 40 thousand years (kyr) before present"

The authors have come such a long way - why not state "125 to >50 thousand years ago," instead?

spagetiMeatball said...

Will they be able to extract DNA from the skull? Do you think this guy was related to basal eurasians?

Anonymous said...

Is it possible that this individual represents a population from further East? The age here, and its correlation with Genetic data seems to make me think that his immediate ancestry could have been from between Mesopotamia and the Iranian Plateau.

terryt said...

"It is quite strange that the authors interpret this evidence as discontinuity between ~100ka humans from the Levant and replacement by a fresh Out-of-Africa wave, when in fact all the Qafzeh/Skhul remains (and a couple of Neandertals) look much more plausible relatives of Manot than any of the ancient African samples".

I agree. In fact if you disregard the mutation rate applied to the diagram in the paper you linked to here:

http://dienekes.blogspot.co.nz/2015/01/y-chromosome-tree-bursts-to-leaf.html

we see a long tail in the C'D'E'F line before each of its subgroups expand and diversify. This hiatus easily fits a long period spent outside Africa confined to some small geographic region. The only trouble is under the currant molecular clock that presence is far too recent to fit such an ancient OoA. However we know of several ancient Y-DNAs that are also far to ancient to fit the generally accepted mutation rate: Ust'-Ishm K2* at 45k, Kostenki C1b* at 37k and MA-1 R* at 25k.

I have never been able to understand why there should be such a thing as a molecular clock for dating haploid replacement. Sure, the mutations are probably regular but replacement of parent haplotypes is basically random.

eurologist said...

"Is it possible that this individual represents a population from further East? The age here, and its correlation with Genetic data seems to make me think that his immediate ancestry could have been from between Mesopotamia and the Iranian Plateau."

aniasi,

That is the million dollar question: where did the "basal Eurasians" reside and survive in sufficient numbers during an exceptionally dry period, and outside the Neanderthal domain? Must have been some place between the Gulf and Afghanistan - but there are not many large hospitable areas there during cold and dry periods.

The follow-up question is, did the changing environment drive these people back into the Levant, or was it triggered by the UP people arriving from the subcontinent? Genetics from the recent Russian sample suggest an early admixture, favoring the latter - as does the fact that ~5,000 years later, the Levantine population is known to be fully UP.

Unknown said...

at this stage in the game, it seems irresponsible to make any conclusions about ancestry, migration, and dispersal based on morphology alone.

without knowing its DNA, this skull doesn't really have anything new to offer, even if is more similar to modern humans and exhibits african traits.

the presence of AMHs in the levant living in the caves along side neanderthals was in my freshman anthropological humans origin text in the early 1990s.

Václav Hrdonka said...

"Qafzeh 9 is modern ... Cro-Magnon 3, Mladec 6/5 are close to Skhul-5 and Qafzeh-6" - Is not it because the Qafzeh 9 was female? There was a bigger diference detween male and female skull. The male one was more robust. The modernization of the humankind meant reduction of testosteron and raise of oxytocin for men, so that people became more woman-like.

terryt said...

"That is the million dollar question"

Yes.

"The follow-up question is, did the changing environment drive these people back into the Levant, or was it triggered by the UP people arriving from the subcontinent?"

My guess is changing environment allowed the outside Africa population to expand widely. Don't forget that this population is not much older than the Ust'-Ishim Y=DNA K2a* at 45,000 years. This latter date indicates that the Y-DNA B/CT split must have occurred at around 90,000 years ago. I am becoming more and more certain that that split indicates the emergence from Africa.

In other words this individual must date from around the time of F's original diversification, not to any out of Africa.

Fiend of 9 worlds said...

Not too exciting, really. It's already arguable previous skulls in the area are close to 'modern' already. We really need some DNA combined with a skull with easy to identify morphology to make any really conclusive statement.

Anonymous said...

@Eurologist

Do we then have to look further east? I am almost inclined to look towards India. Are there any y clades or mt clades that may be typical of basal eurasians?

Also, you mentioned "UP people arriving from the subcontinent." Do you mean the Indian subcontinent, and if so, I would be interested in hearing more. From what I see, the oldest rooted Mtdna Clades are all found in South Asia, and it might be interesting to see if there were two distinct populations in the area that later converged in the middle east.

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

aniasi,

IMO all evidence indicates that Europeans and West and Southwest Asians (the latter, with old local "basal Eurasian" admixture) came from the subcontinent, perhaps via three (a southern, a central, and a northern) migrations - but many of them ultimately came from SE Asia, except for Neanderthal admixture in W Asia. Of course, all of these are ooA ~125-105 thousand years ago.

Male haplogroups F, G, H, IJ, and LT were likely residents of the western part of the subcontinent or resided even further East, in order of naming. Yet, I still believe that IJ entered Europe way before the last glacial maximum - in fact, not long after C.

eurologist said...

In the above, I meant further West (not East) than the subcontinent - even West of the upper Indus river.

Anonymous said...

Eurologist,

Interesting. It certainly does make sense, and I would be interested to see if any ancient DNA recovered from the subcontinent correlates with this. I also agree with your timeline.

By any chance, has basal Eurasian been found in any east african populations? I can't accept that they simply stopped at the levant, and M! & U6 distributions are the same age and a present in North Africa.

terryt said...

"Male haplogroups F, G, H, IJ, and LT were likely residents of the western part of the subcontinent or resided even further west"

I think the latter actually more likely. We know the members of the F had undergone a major expansion some time before 45,000 years ago because the derived form, K2a*, was present in Central Asia by 45,000 years ago. If F as a whole had begun its expansion 10,000 years before that, which is more than just possible, perhaps the specimen here was a basal member of that haplogroup.

I would guess outside the subcontinent, even as far west as Anatolia/Lake Van.

terryt said...

"I would guess outside the subcontinent, even as far west as Anatolia/Lake Van". I've changed my mind on further consideration. I think Dienekes' Persian Gulf hypothesis is worth considering in this matter. I don't think a good case can be made for a South Asian origin for the F haplogroup as a whole. G doesn't fit such an origin and nor does IJ. H may not either but is more problematic.

capra internetensis said...

@terryt

It's hard to make any argument for F as a whole when we have all this unclassified F* around. Lots of it in South and Southeast Asia. A good deal of F* has turned out to be basal H, but apparently one Vietnamese F* was confirmed F(xGHIJK). The centre of gravity of F could completely change when a proper study of F* is done - or not.

The timing of splits within F is interesting. GHIJK, HIJK, and IJK break up in very rapid succession. G is the first to break off, but doesn't expand until much later. Going by number of mutations accumulated, H and K2 are the earliest clades to expand, followed by IJ and LT, and much later by I, J, and G. In fact the latter expand at about the same time as R, well after NO, P, and H1, and probably later than O and Q too. (I wish I had an age estimate for K2b1.)

So the order of splits looks somewhat west to east, but that could be reversed depending on the nature of F*, and the order of expansions is rather east (or centre) to west.

terryt said...

"In fact the latter expand at about the same time as R, well after NO, P, and H1, and probably later than O and Q too".

The time of expansion can vary greatly from the time it first split from its relations though. And independently of them.

"It's hard to make any argument for F as a whole when we have all this unclassified F*"

F's place of origin will be considerably illuminated once the relationship between F1-P91, F2-M427, F3-M481 and GHIJK is sorted out. Some claim the first three are part of a single branch within the FGHIJK clade. If the three Fs do indicate three separate branches it would place F's origin firmly in South Asia. Or very close to it.

F(xGHIJK) as a single South Asian clade makes the situation more complex. Neither G nor IJ show any necessary connection with South Asia. Both could easily have originated around the Persian Gulf.

H may or may not have an origin in South Asia. The haplogroup now includes the old F3-P96 as H2. And H3 is now separated from the old H1-M69 as an eastern branch. H's big expansion has been H1a1d-M291 in South Asia. But that is just the most recent burst of a diversification that looks to have taken place as H1 established itself in South Asia after moving in. Was H's expansion through pre-existing F haplogroups or did they come in together? F3-M481 is similarly an eastern branch of the FGHIJK(xGHIJK) haplogroup. As K2 is of IJK.

LT is another haplogroup that may or may not be South Asian in origin. L contains two branches: a 'European' L2-L595 and a widespread L1-M22. Although the L1 haplogroup could be called 'South Asian' just L1a-M27 has actually been found in India. Another L1 branch (L1c-M357) has been found in Pakistan and Iran. The whole haplogroup, including L1b-M317, looks to fit easily with an expansion from the Persian Gulf region along the southern Iran/Pakistan and western South Asian coast.

T's position is also ambiguous. Interestingly one of the three branches (T1a3-L1255) has been found so far only in Kuwait, within the Persian Gulf. T1a2-L131 is widespread through Europe but almost certainly spread with agriculture from Anatolia.

As I said, the conclusion as to where F developed after splitting from C will only be conclusive once the relationship between F1-P91, F2-M427, F3-M481 and GHIJK is sorted out. That will tell us whether many haplogroups originated in South Asia or if South Asian haplogroups are primarily a product of eastward movement from the Persian Gulf.

vanda said...

In response to Aniasi, about a place for early humans to be away from Neanderthal and a secure place for climate: Wasn't the Persian Sea a locked in landmass before 5,500 BC, which the flooded when the Laurentide ice sheets melted and raised the sea levels about 2-300 feet covering Doggerland and southwest India destroying many coastal civilizations?

terryt said...

"Wasn't the Persian Sea a locked in landmass before 5,500 BC, which the flooded when the Laurentide ice sheets melted and raised the sea levels about 2-300 feet"

It is quite possible humans did live in a refuge around the Persian Gulf. Dienekes has written several posts on the subject. However we're talking here of a refuge around 50,000 years ago.