November 27, 2011

Ancient mtDNA from Cardial culture (early Neolithic Iberia)

Molecular Ecology DOI: 10.1111/j.1365-294X.2011.05361.x

Ancient DNA from an Early Neolithic Iberian population supports a pioneer colonization by first farmers

C. GAMBA et al.

The Neolithic transition has been widely debated particularly regarding the extent to which this revolution implied a demographic expansion from the Near East. We attempted to shed some light on this process in northeastern Iberia by combining ancient DNA (aDNA) data from Early Neolithic settlers and published DNA data from Middle Neolithic and modern samples from the same region. We successfully extracted and amplified mitochondrial DNA from 13 human specimens, found at three archaeological sites dated back to the Cardial culture in the Early Neolithic (Can Sadurní and Chaves) and to the Late Early Neolithic (Sant Pau del Camp). We found that haplogroups with a low frequency in modern populations—N* and X1—are found at higher frequencies in our Early Neolithic population (∼31%). Genetic differentiation between Early and Middle Neolithic populations was significant (FST∼0.13, P less than 10−5), suggesting that genetic drift played an important role at this time. To improve our understanding of the Neolithic demographic processes, we used a Bayesian coalescence-based simulation approach to identify the most likely of three demographic scenarios that might explain the genetic data. The three scenarios were chosen to reflect archaeological knowledge and previous genetic studies using similar inferential approaches. We found that models that ignore population structure, as previously used in aDNA studies, are unlikely to explain the data. Our results are compatible with a pioneer colonization of northeastern Iberia at the Early Neolithic characterized by the arrival of small genetically distinctive groups, showing cultural and genetic connections with the Near East.

Link

36 comments:

Maju said...

I was in wait of learning what exactly they have found with this study and finally Jean Manco has put up the results in her site and they are: 3H, 2N*, 1U5, 1K and 1X1.

The only haplogroup really exotic (but more in a North African than in West Asian sense) is X1. I would also consider K "Neolithic" probably but N* is more like "what the heck" than clearly immigrant and more than half the haplogroups from these sites are presumably pre-Neolithic: H and U5. The gene pool looks strikingly modern overall in any case.

It may be of your interest that I have just found another paper (published in Spanish language only, from 2009 but unknown) on very early Neolithic DNA from Navarre, which is also strikingly modern: 3H*, 2H3, 1U, 1K, 1I and 1 HV.

In this case the only likely "Neolithic" arrival are K (I never know well what to think of K but its expansion is young enough to be Neolithic and is also first detected in Syria) and maybe HV(xH) (detected in the Paleolithic of Italy). However if this HV would be HV0 (and hence likely V), the exoticism is less clear.

In any case the mtDNA pools in Iberia (Neolithic but also Epipaleolithic) look very modern, with the occasional oddity but not something that makes you think of any massive replacement either with the Neolithic arrival nor later on.

Dienekes said...

Only U5 is likely to be pre-Neolithic from that bunch.

Maju said...

Sorry Dienekes but mtDNA H has been reported in Epipaleolithic Portugal (and separately in late Paleolithic Morocco) and at least one of the Portuguese sequences is unmistakably H1b (no CRS and no other possible error). H1 is pre-Neolithic at least in SW Europe (if you think it's "recent" then insert it into some Epipaleolithic wave that caters to your imagination but facts are facts).

Anyhow, I listed the haplogroups slightly wrongly above. They are:

Early (c. 5400 BCE) Catalan Neolithic (Can Sadurní): 3N* (two share the same haplotype however), 1H, 1K, 1U5, 1X1

Early (c. 5200 BCE) High Aragonese Neolithic (Chaves): 2H, 1K

Late (c. 4000 BCE) Catalan Neolithc (Sant Pau): 1N*, 1H20, 1K

The previous list only included the Early Neolithic sites (together). I find the High Aragonese site (Chaves) more similar to the nearby Navarrese site of Paternabidea, mentioned above, than to the Catalan sites.

Both are quite modern-looking, unlike the Catalan sites, suggesting maybe a backflow from the Atlantic (I'm also considering the modernity of the Portuguese samples of Chandler 2006). I wonder if Megalithism (which is quite narrowly associated with Neolithic in many areas) played a greater role than usually acknowledged in spreading modern West European genetics, at least mtDNA ones.

Dienekes said...

Sorry Dienekes but mtDNA H has been reported in Epipaleolithic Portugal

Your link doesn't work.
And, you're wrong. It was announced in some Iberian conference and never published (since 2005).

Either they screwed up and didn't submit, or they submitted and their work was shot down by the reviewers. In any case, there is no evidence for haplogroup H in pre-Neolithic Iberia (or anywhere else in Europe for that matter).

Maju said...

Correct(ed) link: http://dl.dropbox.com/u/21063754/dna2005.pdf (from Zilhao's personal page).

The sequences are shown (in unorthodox but understandable way) in fig. 6 (the reduced median network). I worked with them and checked all possible solutions with the most up to date PhyloTree build. At least one of those nodes is 100% sure H1b. All the others reported as H are also possible (and even likely) H sequences but we can't confirm without coding region tests.

Unknown said...

Well as I see it the meso/paleolithic mtDNA is looking like U in the north (U2?, U4? and/or U5?) and H and N in the south. So far we have yet to find a U with CRS in aDNA and we find lots of CRS which was more probably H anyhow. So lets forget the idea that the CRS could be U, it is increasingly unlikely. The earliest European samples are N and most probably H in Italy (Paglicci). 10,000 year old nearby Morocco is also dominated by H. These facts also fit with the south being H and N.

H was the dominant mtDNA of meso/paleolithic southern Europe in my opinion.

http://www.buildinghistory.org/distantpast/ancientdna.shtml

The proposed signature neolithic mtDNA of the near East, J, has appeared only once in 5500 Germany. Which puts it on par with the single sample of far eastern C5 found. So J was not part of the early mesolithic in Western Europe. It may have arrived with the much later megalithic culture but never dominated Europe as population replacement requires.

T2 which perhaps might be a better indicator (dominates neolithic Syria), is also almost completely absent.

I dont know what to make of the Ozbal paper which I have not previously noticed. But the many ?s that Jean shows are troubling. Also this seems to have 4 members of one family. And H3 is virtually absent from modern Turkey. I am inclined to ignore it as contamination or one Western European lady living in the Halaf culture.

X1 is a subclade of N so some of the N could potentially by X1 I suppose. X1 being Druze and North African. But in any case if this is seen as a signature of the arrival of the neolithic from the near east that also was a genetic megaflop as it is almost unknown in Western Europe today.

K I am not sure of. It appears in hunter gather Sweden also. I suppose it could be neolithic.
But again, it never dominates.

If the neolithic genocide occured the folk from the east did not bring their women with them.

Unknown said...

http://www.bris.ac.uk/archanth/staff/zilhao/#Mesolithic-Neolithic_Transition

working link.

Unknown said...

I just used this neat haplogroup predictor tool on some of the ancient DNA samples listed on Jean's page.

http://nnhgtool.nationalgeographic.com/classify/

Paleolithic
-----------
Gravetian(Russian)Sunghir S2/S3= H

Solutrean (Spain) [NE-NM82.2] = H

Magdalenian (Spain) 1P1 = H

Mesolithic
----
Portugal K13 = H

Whereas it is possible to criticise these tools the most likely candidates for these samples and others is H. Occams Razor.

Dienekes said...

Well as I see it the meso/paleolithic mtDNA is looking like U in the north (U2?, U4? and/or U5?) and H and N in the south.

That's a tall claim, since there is no published pre-Neolithic mtDNA from southern Europe, except the Paglicci sequences, but that cannot be considered "south" since it was during the Ice Age when the ancestors of both North and South Europeans cannot have lived in northern Europe.

In any case, I have no hopes that Maju will not stop parading the 2005 Iberian conference paper as evidence of Paleolithic continuity. That study is not even cited by Gamba et al., which ought to give you a clue about its quality.

Dienekes said...

I just used this neat haplogroup predictor tool on some of the ancient DNA samples listed on Jean's page.

You should go back and read the Warning! in Jean's page. And, also Sungir is listed as HV? or U?, and Solutrean NE-NM82.2] as R0?, your haplogroup H "predictions" notwithstanding.

Maju said...

Sunghir are "unmistakable" H17'21. At least as much as HVS-I can tell us. I do not know where Jean got her labels for the double Sunghir haplotype but I do know that she is at least as reluctant as you are, D., of accepting a pre-Neolithic H. Like you, she is trapped in a labyrinth of misleading molecular clock hunches and the belief of the "Zilhao school" of total demic replacement, which is contested by most other prehistorians I know of (and even the data of the only incursion Zilhao himself made into ancient population genetics, linked above).

Obviously identifying mtDNA H of all haplogroups by means of just HVS-I is nigh on impossible (many H sublineages and notably H1 are "CRS" by haplotype). It seems also real that there was more commonality of U5 and U4 in Paleolithic times but it is almost impossible that all the R* and R-CRS HVS-I sequences are all U*, as was the case with two Swabian CRS haplotypes.

The greatest expansion of unmistakable U5 and U4 clades is actually found in the Epipaleolithic only, being previously quite rare: 1/2 U5 in Andalusian Solutrean and 1/24 U4 in the Oranian of Taforalt. There are two other cases of U but are not U5 nor U4 but U* and U2, being probably as much unrelated to the argument as N* could be.

It is only in the Epipaleolithic when the U5 and U4 ancient samples swell, essentially taking over all the northern European plains... but remaining at relative low levels in Iberia: 1/9 U5b1c2 and 1/9 U4a3.

What I see with all this mtDNA data so far is continuity in Iberia (at least in Portugal) and much larger population changes in Central and Northern Europe, which do not seem to stop with Neolithic anyhow. Much of Southern Europe anyhow remains ill-sampled.

Reference in sequential maps (last updated on May 9th).

jackson_montgomery_devoni said...

So would you say it is true that within Europe mtDNA haplogroups U4 and U5 are the only pre-Neolithic or Mesolithic origin haplogroups found among modern Europeans?

Dienekes said...

I do not know where Jean got her labels for the double Sunghir haplotype but I do know that she is at least as reluctant as you are, D., of accepting a pre-Neolithic H. Like you, she is trapped in a labyrinth of misleading molecular clock hunches and the belief of the "Zilhao school" of total demic replacement, which is contested by most other prehistorians I know of (and even the data of the only incursion Zilhao himself made into ancient population genetics, linked above).

No, she is trapped in the school of "I won't say there was haplogroup H in pre-Neolithic Europe when no evidence for it has appeared, and it is absent in the vast majority (if not all) pre-Neolithic individuals whose haplogroup has been accurately determined".

You may want to believe in Iberian exceptionalism, i.e., that Iberia has been an island of genetic continuity since the Paleolithic, even though every other sample point in Europe points to population discontinuity even since the Neolithic (with the exception of Sardinia), but your blind faith in an unpublished and uncited Iberian conference paper from 2005 is not sufficient evidence.

So would you say it is true that within Europe mtDNA haplogroups U4 and U5 are the only pre-Neolithic or Mesolithic origin haplogroups found among modern Europeans?

Kostenki was U2. There is also U4 in Mesolithic samples from Central/Northern Europe, as well as non-U from the Pitted Ware culture, although those are from a time period long after the introduction of agriculture into the continent.

jackson_montgomery_devoni said...

Yes the non-U types present among the Pitted Ware people is probably due to admixture with farmers I should think.

Maju said...

Jean does report as "H" the Chandler 2005 Epipaleolithic and Neolithic Portuguese, as does with the Moroccans from Taforalt - and she does because they report it that way.

In the case of the lineage of Sunghir, it was not reported as any haplogroup in the source: just the HVS-I sequence: 16129A. However if you check that HVS-I sequence it can only be H17'27.

Well, I must correct: according to the last build of PhyloTree it can also be H1j (but no other possible modern lineage). While this news may cast doubt on it being H17'21 specifically, they reinforce the idea of H1 (or at least H) being extremely old in Europe.

Of course all these discussions and ways to nowhere could be avoided if academic researchers were not such cheapskates and would test for the coding region as well... but we have to work with what is given to us, and that says that H existed in Europe probably since Gravettian (or older) times.

"Kostenki was U2".

But U2 (or U*) are as related to U5 or U4 as R0a or HV4 are to H. (Let ourselves not be self-deceived by nomenclature issues: U is sister of R0, the former pre-HV, in the phylogenetic hierarchy, not of H, which is two nodes downstream, exactly as U2 is).

"There is also U4 in Mesolithic samples from Central/Northern Europe"...

Yes but that is Epipaleolithic ("Mesolithic") which is almost as recent as Neolithic (in some cases, those of the Pitted Ware retrograde Neolithic peoples, it is even more recent in fact). The only true Paleolithic U4 is from Taforalt (Morocco) and is not that old either (12,000 years ago).

If you ask me (that you won't, rhetorical only) what conclusions can we reach from this data, I'd say that it's best to remain most cautious but that both sisters U and R0 (and probably JT as well, albeit only located at one instance) have been in Europe since very early on and that we can detect some key descendants as follows: H since Gravettian (Russia, later Portugal as well), U5 since Solutrean (Spain, later in Northern Europe), U4 since the late Upper Paleolithic (Morocco, later in Northern Europe). These are ante quam (latest possible) dates.

It is always much more difficult to discern H in any case when using HVS-I but even with that major issue it has been found in Paleolithic Europe at least twice.

Dienekes said...

It is always much more difficult to discern H in any case when using HVS-I but even with that major issue it has been found in Paleolithic Europe at least twice.

No, it hasn't.
You don't understand basic phylogenetics. 16129 occurs in at least three different branches of the present-day R0 subtree alone, as well as in numerous other places of the tree. It's called hypervariable for a reason.

Maju said...

Not in solitary, Dienekes. Alone with no other HVS-I changes is found only in H (the two sublineages mentioned above).

Also H1b in Portugal has a different HVS-1 sequence.

You can keep in the "no" stubborn stand but there is nothing saying that such findings are wrong or that most of the other R-CRS or R* lineages are not H (what is clear is that they are not U5 nor U4 in any case). You either appeal to the failed "molecular clock" to claim alleged "impossibility" or you accept that it is at least possible and we should demand further research.

I'm for the second option.

Dienekes said...

Not in solitary, Dienekes. Alone with no other HVS-I changes is found only in H (the two sublineages mentioned above).

The fact that it occurs multiple times throughout the tree means that it does not define any particular lineage. Even you ought to be able to understand that.

The inference 16129 => H is unsound, and you are free to continue deluding yourself that it is proof that haplogroup H existed in Gravettian Europe.

jeanlohizun said...

Hey Dienekes how is this for a proof that mt-DNA H was in Europe in pre-Neolithic times?

http://www.nature.com/ejhg/journal/v8/n9/pdf/5200514a.pdf

Other workers have designated mtDNA haplogroups (clusters
of related mtDNA types) based on HVR I sequences4 or on
restriction site polymorphisms of the entire mitochondrial
genome,31,32 and used these to classify European mtDNAs.
Since the substitutions we observed in HVR I have been
shown to be associated with specific restriction patterns,7,31
we can assign the prehistoric sequences to haplogroups with
a good degree of confidence. The Mezzocorona sequence,
with substitutions at 16126 and 16294, falls into haplogroup
T, whereas the Borgo Nuovo and Villabruna sequences fall
into haplogroup H, and the ‘ice man’ sequence belongs to
haplogroup K.
Thus, each neolithic sequence falls into a
different haplogroup, further testifying to the high level of
mtDNA diversity in the Alps at the beginning of the neolithic
period.


The Villabruna is dated to 13831–14267 ybp according to Table-2.

jackson_montgomery_devoni said...

I would like to point out that some haplogroup H has also been found in early Neolithic Syria and possibly also in Neolithic Turkey. In contrast so far no U5 has been found anywhere in the Neolithic Near East.

GailT said...

U5b2a1a is frequently CRS, and with an estimated age of 16,000 years, and given the preponderance of U5 among European hunter-gathers, any CRS could well be U5b2a1a.

There is limited evidence, if any, for H in Paleolithic Europe. Until we get more ancient DNA, I don't see any point in speculating about whether HV or JT were there. New results might show that it was, but what I don't understand is the ideological commitment that some people have to wanting to believe it was there in spite of the lack of evidence.

If R0 and JT were present along with U in ice age refugia, it seems like we should have seen more R0, HV and JT among hunter-gathers. Occam's razor would suggest they weren't there, but it's also premature to reach that conclusion. I'm not invested in one answer or the other, I'm just anxious to see more ancient DNA so that finally have something to argue about.

Dienekes said...

>> Hey Dienekes how is this for a proof that mt-DNA H was in Europe in pre-Neolithic times?


Weak

Maju said...

"U5b2a1a is frequently CRS, and with an estimated age of 16,000 years, and given the preponderance of U5 among European hunter-gathers, any CRS could well be U5b2a1a".

It is more precisely the case of U5b2a1a2 (three back-mutations in CRS direction!), although I guess private branches in other lineages may also produce that. Good finding!

Still it'd be very odd that all the CRS would be that lineage precisely. I mean: it is not that dominant even within U5 and probably never was; it'd be a very curious and unlikely anomaly. So far the two cases where other markers was checked with CRS sequences ended up being R0(xH) [tested the AluI marker, unsure how credible it is] and U* [coding region test: 100% certain].

And then what about the other (often reported as H) R* lineages, which maybe fit H-whatever for one or two of the reported HVS-I? The hard fact is that we do not know and that they could all hypothetically end up being R-other lineages very rare or even extinct today. I wish somebody did the coding region tests to clarify the matter once for all.

But we have at least two distinct HVS-I sequences from (Epi-)Paleolithic peoples widely separated in space and time that must be H: one is H1b and the other H1j or H17'27. We also have, as Jackson notices, an Early Neolithic H5 (but not H1) in Syria by the same logic (HVS-I) - others were reported maybe but the HVS-I are actually not clarifying enough.

"If R0 and JT were present along with U in ice age refugia, it seems like we should have seen more R0, HV and JT among hunter-gathers".

Agreed... in theory. I have not spotted what seems to be JT* but in Nerja's Solutrean and then not again until Neolithic (as J and T separatedly). RO* (or HV*) was reported once prior to Neolithic (Italy) and I am very certain that there are two definite cases of H (Russia and Portugal) in Paleolithic Europe. But there is still "loads" of R* (with or without CRS) which by modern likelihoods should be H1 in >95% of cases (but that have shown not to be in 100% of tested cases, that is: two, hence the controversy).

On May 9th, I listed:

Early/Middle UP (n=9): 1 U5, 1 U2, 1 H (same sequence), 1 R0(xH)-CRS, 2 R*-CRS (one is dubious), 1 JT.

Late UP (n=6, Europe only): 2 U*-CRS (same sequence), 2 R*-CRS and 2 R* (other). [In addition 24 people from Morocco were: 1 U4, 13 R*-CRS and 10 R* (other)].

Epipaleolithic (n=16, Europe only): 6 U5, 3 U4, 2 R*-CRS, 3 R* (other), 2 L3* (N?, one probably L3d2 in fact, suggesting some African flow into Iberia prior to Neolithic). [In addition 11 Neolithic Syrians of same period were: 3 K, 1 H5, 5 R* (not CRS) and 2 L3* (N?).

This gives a lot of European R* (often reported as "H") to be explained: 25% in the early/middle UP, 80% in the late UP (plus 95% in Morocco) and more than 30% in the Epipaleolithic (when the Northern European U5/U4 numbers swell) (plus 45% in contemporary Neolithic Syria).

The swelling of U5/U4 in the Epipaleolithic is intriguing but even more is the swelling of R* (H?) in the late UP, don't you think?

Dienekes said...

But we have at least two distinct HVS-I sequences from (Epi-)Paleolithic peoples widely separated in space and time that must be H: one is H1b and the other H1j or H17'27.

They must not. Any further posts of misinformation will be deleted.

Maju said...

It is not 'misinformation': it is facts, facts that you have dismissed but not discussed.

Censorship will be adequately denounced in any case: this blog's free discussion quality has dropped a lot in the last years.

Dienekes said...

It is not facts, it is your ignorant misinterpretation of homoplasy as signifying something it does not.

Unknown said...

"but what I don't understand is the ideological commitment that some people have to wanting to believe it was there in spite of the lack of evidence"

LOL. That is the crux of the matter. Everyone thinks the other guy is the blind ideologue who cannot see the "TRUTH". Each of us beleives we are completely unbiased.

Evidence =
(1) Multiple authors identifying multiple Southern European samples as most probably H.

(2) Lots of H in 10,000 year old nearby Morocco. If it was in Morocco it stands to reason that it would most likely also be in Iberia from the long term population flows.

http://www.buildinghistory.org/distantpast/nafricaadna.shtml

(3) Sister clade V is also found in 10,000 year old Morocco. No way is that U5. If V is in Morocco why would the more prolific H be absent?

If it looks like a duck, quacks like a duck and swims like a duck. Most probably, it's a duck.

Maju said...

It might be homoplasy of course but that is just a possibility, a quite remote one in my opinion, and not a fact in any case.

We still have to account for the 80% of R* (with and without CRS) in Late UP Europe (and Morocco). Just claiming "U" (specially when not all U is the same, not even closely related in fact) is no satisfactory answer: there is a U5+U4 blooming in the Epipaleolithic of Central and Eastern Europe indeed but that is only one piece of the puzzle.

jackson_montgomery_devoni said...

So far it looks like any presence of mtDNA haplogroup H in pre-Neolithic Europe is rather weak.

Dienekes said...

Maju wrote:

"But we have at least two distinct HVS-I sequences from (Epi-)Paleolithic peoples widely separated in space and time that must be H"

and then he wrote:

"It might be homoplasy of course but that is just a possibility, a quite remote one in my opinion, and not a fact in any case."

If it "must be" H, then it "might not be" homoplasy.

Actually, it _is_ homoplasy, since it occurs in multiple spots on the tree. Even yourself can't decide on which branch to put it on _because of homoplasy_. And since you don't know which branch to put it on, and there is no magic tendency of H mtDNA to mutate in that spot, your overconfidence that it is H is unwarranted.

Samequeen said...

U5 is old, yes, and I think that K is also. I think U5 and K wandered in the same hunting bands.

Maju said...

If it "must be" H, then it "might not be" homoplasy.

Let me rephrase: it MUST be H, unless it is homoplasy, a very remote possibility, an extremely conjectural hypothesis you launch to reject the natural conclusion.

Exceptional claims demand exceptional evidence, Dienekes and there is just NO EVIDENCE that we are before not one but two cases of homoplasy.

Of course, in the end each one will choose to favor this or the other possibility. That is our option until definite evidence arrives but I am not saying nonsense: just pointing out to very striking data.

"Actually, it _is_ homoplasy, since it occurs in multiple spots on the tree. Even yourself can't decide on which branch to put it on _because of homoplasy_".

Alone it occurs only in two branches of H. In the previous build it was just one (or I and the search feature committed an error).

But even if that would be homoplasy, it is not the case for the Portuguese sequences at all, where the lineage is defined by not one but three markers at sites 16189, 16356 and 16362. This can't be homeoplasy.

Whatever the case, we are stuck in this difference of opinions, and we won't be able to advance until proper testing, including coding region markers, becomes more common.

Dienekes said...

Let me rephrase: it MUST be H, unless it is homoplasy, a very remote possibility, an extremely conjectural hypothesis you launch to reject the natural conclusion.

Lol, a remote possibility that has happened 3 times in the R0 subtree alone. Extremely conjectural indeed.

You don't know what homoplasy is. Go read up on it before posting here again.

Alone it occurs only in two branches of H. In the previous build it was just one (or I and the search feature committed an error).


"Alone" is meaningless. The point is that the site in question has mutated independently multiple times all over the tree. Since the CRS in HVSI does not indicate anyo particular haplogroup, and the mutation in question has occurred multipled times independently, its occurrence does not prove any particular haplogroup.

In fact, you've proven my case by noting that the mutation has popped up in a new branch of the tree between two builds.

it is not the case for the Portuguese sequences at all, where the lineage is defined by not one but three markers at sites 16189, 16356 and 16362.

There are no Portuguese sequences, there's a 2005 unpublished Iberian conference paper that isn't cited by anyone.

Whatever the case, we are stuck in this difference of opinions, and we won't be able to advance until proper testing, including coding region markers, becomes more common.

This isn't a difference of opinions, you are arguing for pre-Neolithic H in Europe on the basis of flawed data and logic.

Andrew Oh-Willeke said...

"Lots of H in 10,000 year old nearby Morocco."

Annie Mouse is quite on point here. There is H and V in pre-Neolithic North Africa, probably Near Eastern sourced, and if it is there, attributing H to Epipaleolithic repopulation of Europe along a Southern European coastal route following the LGM from the Near East is a plausible possibility.

IIRC, the North African remains are in association with fish remains, a defining characteristic of Southern European coastal Epipaleolithic sites as well, again suggesting that they are part of the same population movement.

Indeed, I think we slight ourselves by trying to reduce food production complexes into merely hunter-gatherer and food producer.

A big game hunter like Neanderthals, is not an omnivore who gathers berries-nuts-hunts small game-hunts big game-eats insects-collects shellfish-goes fishing (like modern human hunter gathers). And, hunting and gathering rooted in predominantly fishing from rich fisheries produces a more sedentary and materially affluent society than a nomadic terrestrial hunting and gathering society. Fishing from rich fisheries is really intermediate between hunting and gathering and farming/herding. Likewise, horticulture and herding are very different lifestyles and herding itself comes in variants (nomadic, sedentary, transhumance) that have different cultural implications.

It isn't implausible even, to think that reliable sedentary fishing societies (along with better weather) may have been important in making farming into a viable means of survival in the transition period during which the domestication of crops process was underway.

This slight variant on a source for H exclusively from early Neolithic populations really isn't deeply at odds with any data anyway. Pre-LGM Upper Paleolithic remains, and interior European Upper Paleolithic remains may be all U4 and U5 (perhaps with trace frequencies of other hgs in the contact period where some bride swapping may have taken place).

Of course, if the Near East was rich in mtDNA H during the Epipaleolithic from ca. 13000 BCE to 6000 BCE, it probably remained similar in genetic makeup when it started becoming a source of emigrating farmers and herders instead of fishermen, so there is probably no way to know, from the genetics alone, is mtDNA H in a population has a Near Eastern Epipaleolithic source or a Near Eastern early Neolithic source. Perhaps population U4/U5 replacement/dilution in Europe was first driven by a better fishhook, not wheat or sheep.

Conversely, even if pre-farming and herding populations of Europe were rich in mtDNA haplogroup H in the late Epipaleolithic, that doesn't mean that farmers couldn't have replaced the pre-existing population, because the genetics alone wouldn't tell us.

While Dienkes is right in not trusting unpublished papers, sooner or later we will have published research on ancient DNA which we know from conference papers is out there to be tested, and we will know.

Waggg said...

Anyway, in this specific study, it doesn't support very well the theory of all the mtDNA hgs H coming from west Asia with neolithic populations, IMO.

Here we have a mtDNA H20a (*) haplotype which is not found in Europe (I've read that less than 1% of the Iberian population is H20 but I couldn't find any source to confirm it and anyway it wouldn't change much) but only in the Caucasus (mostly) and in Asia minor and the near East area (Especially in Georgians and the west Caucasus having also the highest diversity there) it goes well with G2a I guess and with the neolithic process as well) while in other north Spanish/Provençal Cardial-derived aDNA, it was dominated by Y-DNA G2a (but there were also some likely European y-DNA hgs) and it also had several mtDNA H3, an haplogroup which is more frequent in south-western Europe and is absent in west Asia and Caucasus (and almost absent of south-eastern Europe).
It doesn't seem quite right to associate originally H3 with this specific neolithic (mostly, so far) G2a population in which we found so clear (but rare) tracks of modern haplogroups' subclade so specific to west Asia/Caucasus (like H20a).

It seems unlikely that these H20a and H3 were originally from the same set of population at this time, haplogroups with respective directions of expansion (obviously not much left tracks by H20a in Europe) seemingly going in the opposite direction.


(*) http://1.bp.blogspot.com/-keGzWiR3ivs/TtKpmfhw7XI/AAAAAAAAArY/FxIeW3aZd08/s1600/CatalanNeolithicmtDNA.png

and

http://www.snpedia.com/index.php/Haplogroup_H_%28mtDNA%29


PS : More generally, if H1 was neolithic in west Eurasia, i'm not sure we would find it in the Guanches' aDNA (the autochthonous population of the Canaries islands) or even probably this much in E1b Tuaregs from the depth of Sahel, IMHO.

Alice C. Linsley said...

Haplogroup X has been studied from the angle of the Ainu marriage and ascendency structure and linked to the Nilotic Horites.

http://jandyongenesis.blogspot.com/2012/01/kindling-of-ancient-memory.html