Monday, July 4, 2016

Economic overhaul + population shift in Late Neolithic Iran


Courtesy of Arbuckle et al. at the Journal of Archaeological Science. Emphasis is mine:

Abstract: In this paper we address the timing of and mechanisms for the appearance of domestic cattle in the Eastern Fertile Crescent (EFC) region of SW Asia through the analysis of new and previously published species abundance and biometric data from 86 archaeofaunal assemblages. We find that Bos exploitation was a minor component of animal economies in the EFC in the late Pleistocene and early Holocene but increased dramatically in the sixth millennium BC. Moreover, biometric data indicate that small sized Bos, likely representing domesticates, appear suddenly in the region without any transitional forms in the early to mid sixth millennium BC. This suggests that domestic cattle were imported into the EFC, possibly associated with the spread of the Halaf archaeological culture, several millennia after they first appear in the neighboring northern Levant.

These findings more or less correlate with the results in the new Lazaridis et al. preprint:

During subsequent millennia, the early farmer populations of the Near East expanded in all directions and mixed, as we can only model populations of the Chalcolithic and subsequent Bronze Age as having ancestry from two or more sources. The Chalcolithic people of western Iran can be modelled as a mixture of the Neolithic people of western Iran, the Levant, and Caucasus Hunter Gatherers (CHG), consistent with their position in the PCA (Fig. 1b).

In other words, the small cows weren't just imported into the Eastern Fertile Crescent; they came with people who also made a major genetic impact on the region.

Here's my own PCA featuring the relevant Lazaridis et al. samples. Key: Caucasus_HG = Caucasus Hunter-Gatherer; Iran_ChL = Iran Chalcolithic; Iran_HG = Iran Hunter-Gatherer; Iran_N = Iran Neolithic; Levant_N = Levant Neolithic.


Thus, it would seem that after the early Neolithic farmers from Iran migrated to South Asia, they were largely replaced in their own homeland by Halaf pastoralists and/or related groups. Moreover, their descendents in South Asia, and especially South Central Asia, were then largely replaced by pastoralists from the Bronze Age Eurasian steppe (for instance, see here).

Obviously, this doesn't square too well with the idea of a Proto-Indo-European homeland in the Zagros Mountains of western Iran, does it?

See also...

Yamnaya =/= Eastern Hunter-Gatherers + Iran Chalcolithic

Zarathushtra and his steppe posse

51 comments:

  1. There is was also an Increase of Iran Neolithic admixture in Anatolia and the Levant, during the copper age era. Y-DNA J is found in the levant only after the increase of of Iran N admixture in the region, while/Iran ChL/Iran HG Y-DNA is largely similar (J is dominant)

    ReplyDelete
  2. But how do you know this was Iranian Neolithic admixture and not just CHG admixture?

    Remember, Laz et al. model CHG as mainly Iranian Neolithic, which is an anachronistic approach.

    ReplyDelete
  3. And how do you know that the Halaf Culture people didn't have this type of CHG/Iranian Neolithic-like, but not actually from Iran, admixture?

    ReplyDelete
  4. I think this nMonte by Ryukendo gives better idea about the origins of Iran Chalcolitic

    [1] "distance%=1.2362 / distance=0.012362"


    Iran_Chalcolithic
    "Iran_Late_Neolithic" 49.4
    "Anatolia_Neolithic" 29.55
    "Satsurblia" 15.3
    "LBK_EN" 3.2
    "Nganasan" 1.55
    "Afanasievo" 1

    The heavy presence of Anatolia Neolithic without Natufian/Levant_N leave little options.
    My bet it is somewhere between Taurus/SE Turkey (for CHG) and North Iraq. Modern Kurdistan basically. Obviously they will have some Iran Neolithic which means that population turnover was much higher in Chalcolitic Iran.

    So if this Iran_Chl were from Halaf culture this gives an interesting information about the origins of Halaf culture itself.

    ""Previously, the Syrian plains were not considered as the homeland of Halaf culture, and the Halafians were seen either as hill people who descended from the nearby mountains of southeastern Anatolia, or herdsmen from northern Iraq.[5] However, those views changed with the recent archaeology conducted since 1986 by Peter Akkermans, which have produced new insights and perspectives about the rise of Halaf culture.""

    ReplyDelete
  5. And this is the Alberto's nMonte

    Iran_Chalcolithic
    "Iran_Neolithic" 44.35
    "Anatolia_Neolithic" 38.8
    "Satsurblia" 14.85
    "Ami" 1.2
    "Eastern_HG" 0.8
    "Hungary_HG" 0
    "Loschbour" 0
    "Esan_Nigeria" 0
    "Esperstedt_MN" 0
    "Israel_Natufian" 0
    "Motala_HG" 0
    "Levant_Neolithic" 0
    distance=0.01166

    ReplyDelete
  6. @Aram

    South-Eastern Turkey and Northern Iraq are, of course, also the oldest sites of auroch domestication. I have not read the paper yet, so I don't know why the authors connect Iranian cattle specifically to Tell Halaf or, more generally, the Levant. Perhaps they have a geographically inclusive definition of the Levant (?)

    My guess is that there were people coming directly from South-Eastern Turkey whose movement was more related to technological advances rather than cattle. Cows probably would not have conferred much of an advantage, since mixed-ungulate pastoralism was well established in Northern Iran & Southern Turkmenistan since ~6,200 BC and in all likelihood much better adapted to the mountain- & dry belt biom. In fact, I think cows are seldom kept by actual pastoralists even in contemporary Iran - their use being mostly limited to pastoral farmers.

    ReplyDelete
  7. No, cattle herding increased suddenly in western Iran from around 6,000 BCE, and the cows that were part of this new economy were new to the region, and smaller than the old cows.

    So if there was no advantage in keeping these new and seemingly improved small cows, then why is their arrival so important in the archaeological record?

    ReplyDelete
  8. The amount of genetic turnover between Iran Neolithic & Chalcolithic is still to be determined, given that in the Lazaridis paper Iran_Neol feeds into CHG, but I suspect Ryu & Alberto's estimates aren't too off.

    When considering the significant flux seen in other regions of the Near East (Levant BA, Anatolia Chalcolithic, Armenia _EBA), it would seem that there was just a lot of miscegenation within the area, although we cannot exclude complete turn over in local contexts, also.

    Anyhow, whilst SW Iran (The Zagros) might be interesting for Elamite ethnogenesis, I would be curious to get more data from northern Iran and adjacent parts of southern Central Asia.

    ReplyDelete
  9. The 'old' cows most likely weren't domesticated. The domestication of aurochs is a very difficult process and the initial pool used for breeding must have been small, so there probably wasn't a lot of variation by 6,000 BC.

    The pastures in the valleys of the north-west could have been exploited, hence the increase in archeological finds. But it is highly questionable that this would have conferred a huge advantage over the sheep & goat agro-pastoralism that already existed in Northern Iran & Turkmenistan. If there was a migration out of the Levant into Iran following the Halaf period, it certainly didn't 'largely replace' the locals.

    ReplyDelete
  10. Have a look at where the Chalcolithic Iranians cluster compared to the Mesolithic North Iranian (Iran_HG) and Neolithic Western Iranians.

    Clearly there was a massive shift in population structure at the same time as when the new cows arrived.

    And since the Halaf people lived in the northern Levant, then it's likely that in using the Levant Neolithic (Levant_N) sample from Jordan to model the Chalcolithic Iranians, Lazaridis et al. probably underestimated the level of this genetic shift.

    ReplyDelete
  11. M. KB

    "But it is highly questionable that this would have conferred a huge advantage over the sheep & goat agro-pastoralism that already existed in Northern Iran & Turkmenistan."

    Did it get wetter around that time?

    From my reading cattle have the advantage when there is lush pasture, sheep and goats have the advantage when it's drier.

    ReplyDelete
  12. "From my reading cattle have the advantage when there is lush pasture, sheep and goats have the advantage when it's drier."

    I think the point here is that even if the cattle were less optimal for the environment, they nonetheless did come. And the reason is because the people who had domesticated cattle moved into the region.

    Cattle probably were not any better for the area. But the new people knew how to care for them. It was a part of their culture.

    ReplyDelete
  13. @Davidski,
    Omg , Omg…This is my turf.

    First off, paper is wrong. Not by much….It was not the middle of 6th M Bc but EARLY 6h M (or even very late 7th).
    Arbuckle et al. say this about cattle the same time as Kadowaki and Guliyev (http://onlinelibrary.wiley.com/doi/10.1002/oa.2534/abstract;jsessionid=948ECA9AD36CC15FBC853E1FFA72D1F7.f01t04%20Goats%20from%20Shulaveri-Shomu%20site%20of%20Goytepe..) and they say exactly the same thing about goats. Yes, it was the time of appearance of the Shulaveri-Shomu in southern Caucasus. They showed up with cattle, Goats.
    The Goats paper is precisely regarding goats in Arukhlo which is sort of the “capital” of the Shulaveri-Shomu. Which lead me previous here to state that !what the hell they manage to extract DNA from goats next to inhumations but not DNA for those Guys buried next to them!!! – I know, might take some time and that is why I still have hopes on Nagoya University and Kadowaki. They will prove me right.

    And two more things noticeable and Important about the surgence of the Strange Shulaveri People Hidden beyond the Sevan and the small reef mountains:

    First is hulled wheat and barley and most of all … Spelt. And spelt is sort of a trade mark of the Bell beakers. Spelt is found in Shulaveri, them in Merimde, then in chalcolithic Iberia then In Bell beaker Switzerland.
    Second was horses – See by 5800 BC they had small horses which were not at all food. All around them in levant and Anatolia it was oranger (sort of a horse but really small) and in north Africa the Donkey. So, no horse. – But horses apparently survived in the Kura basin where the Shulaveri lived and so they were actually the only ones to have horses in 5500 BC. – Horses only really show in Levant and Anatolia by 3300! That is 2000 years later for a region just a couple hundred kilometers from them! North Africa much later them that.

    However, there is outlier horse in Shiqmin (4.500bc) and another in Beesheva (4,200bc)… them the paradox of such strangely large donkeys in El Omari by 4000bc and loads of horses in newly formed settlements in Iberia chalcolithic…. And the icing on the cake… a Portuguese Breed (sorraia) that clusters together with horse przewalski thought of being the closest to Tarpans horses that exist…

    ReplyDelete
  14. Guys, need to bear in mind the time --- Is all about time!

    1 – Up until 6,000 BC - CHG people is Chockh culture, over sevan lake hidden (see hypsometic map Davidsky has just 2 post bellow. Zoom in and you will get it) where Kotias lived. So, up until 6,000 BC they, CHG were left alone.
    2- From around 6200BC to 5000BC the Shulaveri came. I always accepted had come from the steppe by Iran, or crossed over near black sea (where you later see Maykop) because is where it’s easy to do it. But that would imply that they would be a steppe (EHG?) mixing with CHG (Kotias)? Does it make else? – Alternatively it might be so, but the where living in there when the cattle people came ( as per this Davidski paper) and mixed. So My Shulaveri would be CHG+EHG+ Anatolia_N.~
    If instead no Anatolia_N them they would be CHG + Steppe + IRAN_N
    3 – Halaf was not long. From 5500 BC to 5000BC - not protected by that heaven north of Sevan, its called the Halaf-To-Ubaid-transitional period. So they were replaced? It was the time one saw the Hussuna-Samarra, north Kurdistan as already Ophidian people (Snake people) as seen in Western Iran.
    4 – By 4.900 bc to 4000BC --- Ubaid. It was havoc. IT all changed? Ashes and settlement abandonment, form western Iran to the base of the Caucasus mountains where the Shulaveri lived.
    5. 4500 BC, Who ever came was URUK ophidian (snake) and came from the south of Iraq, and was devastating and evolved and powerful.—Ended up being the Mesopotamia and Sumerians (?).


    So how does this all fits into DNA data?

    ReplyDelete

  15. Hummm, by this video posted today on youtube its not gonna be the Germans to prove me right up in Merimde but maybe this Scott Woodward, professor of microbiology at Brigham Young University, will be the one in Fayum. Like the hardest way. Fayum is everybody! but ok. Right time and they will get R1b (M269) having binge cattle parties in Fayum!

    https://www.youtube.com/watch?v=T3bhNkAunZM

    Sample, baby, sample!!!

    ReplyDelete
  16. @ Olympus Mons

    Great catch on the Bell Beaker connection.

    @ Davidski

    "Obviously, this doesn't square too well with the idea of a Proto-Indo-European homeland in the Zagros Mountains of western Iran, does it?"

    The notion of such a theory, to the extent that one presents a non-strawman version of it, would be that Zagros Mountains people have cultural and demic influences on Caucasian people who in turn influence steppe people who in turn expand as Indo-Europeans, recirculating back from the Steppe to South Asia and West Asia among other places.

    ReplyDelete
  17. In D-stats you provided of the form D(Chimp, Row)(Mbuti, Colmn) SC Asia fits better as (Iran_Chl, minor CHG or Iran_Neo)+Steppe+South India. Or as (CHG, EEF, no or little Iran_Neo)+Steppe+South India. The stats at least suggest their non_ASI and non-Steppe ancestor was note exactly like Iran_Neo but had significant Anatolia_N/Levant_N ancestry. Maybe this is true for some SC Asians and not others. But my point is I don't think SC Asia is as simple as Iran_Neo+Steppe+ASI.

    ReplyDelete
  18. You shouldn't be using South Indians as a reference because they're mixed.

    ReplyDelete
  19. Then should I take out East Asian colmns and use an East Asian as reference for ASI. That should work. Intra East Asian diversity won't make an effect.

    ReplyDelete
  20. You should maybe use Onge + East Asian. I'll try and upload an Onge column and row later.

    ReplyDelete
  21. This wasen't a population replacement as much as it was a slow admixture event.

    ReplyDelete
  22. It looks like a population replacement, along with the sudden introduction of a new economy.

    ReplyDelete
  23. How can it be replacement, when the Neolithic component is a significant portion of the later Chal pops.

    ReplyDelete
  24. The Neolithic component is a significant portion of the later Chal pops.

    It only looks that way because CHG is modeled as mostly Iran_N.

    EHG is almost as high in Iran_ChL as in CHG, so Iran_ChL has to be mostly CHG, with minor Iran_N and Levant_N.

    The only way that Iran_ChL can be largely Iran_N, is if the newcomers to Iran were mostly Levant_N with significant EHG. That doesn't make much sense.

    ReplyDelete
  25. This comment has been removed by the author.

    ReplyDelete
  26. So two separate migrations of Levant_N and EHG into Iran, as opposed to one of a mostly CHG population with significant Levant_N admixture?

    No chance.

    ReplyDelete
  27. The two needn't be related: ie seperate streams of admixture from levant and north, resp.. This a couple of different scenarios could account for what we're seeing, and on the basis of current evidence, there is no evidence to trump up one particular conclusion

    ReplyDelete
  28. There is evidence. It's laid out in the latest paper...

    The Chalcolithic people of western Iran can be modelled as a mixture of the Neolithic people of western Iran, the Levant, and Caucasus Hunter Gatherers (CHG), consistent with their position in the PCA (Fig. 1b).

    And it doesn't make any sense to assume that EHG somehow skipped over CHG and made it to Iran instead of CHG.

    ReplyDelete
  29. Iran_ChL is 63% CHG and 17% Iran_N in the paper. That's basically the results that I'm getting in my models.

    So it was an almost total replacement.

    ReplyDelete
  30. Fair enough
    Very interesting

    ReplyDelete
  31. This comment has been removed by the author.

    ReplyDelete
  32. The Halaf people, or whoever moved into Iran at this time, may have already been a mixture of what we'd term as CHG, Levant_N and Iran_N.

    So in theory this may have been a complete population replacement.

    We have to work with what we've got now, but we need to remember that genetic structure is usually clinal, rather than discrete, so it's rather unlikely that we're dealing here with a straight two-way mix of CHG and Levant_N. It's just common sense to throw in something like Iran_N.

    ReplyDelete
  33. This comment has been removed by the author.

    ReplyDelete
  34. Onge are now in the rows in the D-stats datasheets here...

    http://eurogenes.blogspot.com.au/2016/06/d-statsnmonte-open-thread-3.html

    They're too closely related to put them in the columns as well, but if anyone's interested...

    https://drive.google.com/file/d/0B9o3EYTdM8lQZEs3azBGMUxCdzQ/view?usp=sharing

    ReplyDelete
  35. Iran Chalcolitic Y DNA

    I1662 Iran Copper Age J2a-PF5008(xL581) calls
    I1674 Iran Copper Age G1a1b-GG372 calls

    G1 according one study has North West Iranian origin. It expanded from Iran to South Central Asia then to Central Asia. But not to Europe.
    This particular branch reached India. But G1 is rare in India.

    J2a-PF5008 is more complicated. It is present from Europe, South Caucasus, Altai/Central Asia Levant Gulf region. Altai Iron age samples were J2a2-PF5008 according to ChrisR
    It could be responsable for CHG growth.
    This case is negative for L581 the most common SNP today.

    ReplyDelete
  36. Y-DNA G1 is pretty much restricted to Modern day Iran and parts of central Asia. I doubt that Iran ChL was 63% CHG and only 17% Iranian Neolithic. Modern Iranians get much higher than 17%, when it comes to Iranian Neolithic admixture. Just like the Steppe EMBA fit, the Iran ChL fit could be just a good statistical fit.

    ReplyDelete
  37. Armenia ChL is around 30% Iran Neolithic in the table, but had no CHG. IMO there is now way in hell that Iran ChL was only 17% Iran Neolithic.

    http://i.imgur.com/KD76Iq4.png

    ReplyDelete
  38. Modern Iranians get much higher than 17%, when it comes to Iranian Neolithic admixture.

    So what? If modern Iranians are mostly of Iranian Chalcolithic origin they'll still get a lot of Iranian Neolithic admixture, but it won't actually be admixture from Neolithic Iran.

    ReplyDelete
  39. This comment has been removed by the author.

    ReplyDelete
  40. Armenia ChL is around 30% Iran Neolithic in the table, but had no CHG. IMO there is now way in hell that Iran ChL was only 17% Iran Neolithic.

    lol

    CHG is mostly Iranian Neolithic in that table.

    ReplyDelete
  41. I'm not sure I understand where EHG came into this, but if there were elements like this you could peel the banana... the proto-Halaf/Yarmin Tepe ceramic phase may have an origin in Southern Turkmenistan with a earthenware culture like Jeitun mixing with Kelteminar fishers (a pit comb ware folk) in the SE Caspian. (And if not these specific groups, similar groups and interactions between West Asia fishers and Iranians).

    Here's an excerpt from Jordan and Zvelebil:
    "More tentatively, it is possible that as a part of this general process of ceramic dispersal, the production of pottery also spreads
    via a more southerly route from Central Asia along the eastern
    shore of the Caspian south, into northwest Iran and northern Syria."

    Apparently, you can show this was happening in the North Caspian as well.

    ReplyDelete
    Replies
    1. Maybe I misunderstood. Did you mean CHG instead of EHG? I didn't see any EHG in the numbers above..?

      Delete
  42. It's not only that Onge takes away the Iran_N and ENA from South Asian populations, the strange thing is that Onge itself takes Iran_N:

    Andamanese_Onge
    "Ami" 38.2
    "Iran_Neolithic" 36.45
    "Papuan" 21
    "Israel_Natufian" 2.2
    "Bougainville" 2.15
    "Eastern_HG" 0
    "AG3-MA1" 0
    "Levant_Neolithic" 0


    But with this model, all West Eurasians (including, and specially, modern with Basal Eurasian) get clearly overfitted, as Ami and Papua do. While Munda, India_South and Ust-Ishim get clearly underfitted.

    So it's not the "West Eurasian" part of Iran_N that Onge wants, it must be something else. If Iran_N has no ASI, I don't know what it could be.

    ReplyDelete
  43. @bellbeakerblogger

    Early Neolithic farmers in Iran were direct descendants of local hunter-gatherers.

    But during the Late Neolithic/Chalcolithic they were largely replaced by a new population coming from the west.

    This new population was mostly CHG, and CHG is partly made up of EHG and/or something that is closely related to EHG.

    So there was no migration of EHG into Iran from Central Asia. In fact, after seeing the results in this paper I doubt there was anything in Central Asia before the Bronze Age that could be confused with EHG.

    @Alberto

    There's no need to call it ASI. It's just something Central Asian that made up a part of the ancestry of Iranian foragers, at least as far back as the Mesolithic, and maybe the Paleolithic.

    ReplyDelete
  44. Ok, so modelling Onge as 100% Ami shows a much higher distance, but all West Eurasians (with or without BEA, and India_South, Munda and ust-Ishim) get correctly fitted. The problem is that the East Asian populations get so overfitted that the distance become huge.

    So Onge is equally related to West Eurasians as East Asians are, but it's much less related to East Asians than themselves are. And this is where the algorithm has to compensate by indluding something that will reduce that affinity to East Asians. Hhhmm, finding the right columns for each model seems to be quite tricky and important...

    ReplyDelete
  45. Given that the two Iranian neolithic samples were R2a, there must have been some sort of genetic movement from Central Asia if we're to assume that Y-DNA R was from there.

    Anyhow, I agree that the Iranian neolithic group in Iran likely did encounter new comers from the west and mixed in with them (Which is what you call a replacement). But I also think that this specific group likely spread out from the Zagros and all the way to the Indus Valley, the western part merged in with the western new comers, while the eastern part likely mixed with the local folks that lived in South/Central Asia (Possibly creating the ASI). Fast forward a few thousand years and you have Indo-Iranian movements from Central Asia into both South and West Asia, this component added to the ASI likely made up the ANI.

    ReplyDelete
  46. @Andrew


    "@ Davidski

    "Obviously, this doesn't square too well with the idea of a Proto-Indo-European homeland in the Zagros Mountains of western Iran, does it?"


    The notion of such a theory, to the extent that one presents a non-strawman version of it, would be that Zagros Mountains people have cultural and demic influences on Caucasian people who in turn influence steppe people who in turn expand as Indo-Europeans, recirculating back from the Steppe to South Asia and West Asia among other places."



    Exactly, except that re-circulation was likely more limited than some may think. Actually, the PIE out of NW Iran is a highly consistent hypothesis, in light of multiple lines of evidence. PhD Giacomo Benedetti keeps a very nice blog dedicated to an Iranian homeland. http://new-indology.blogspot.com/

    ReplyDelete
  47. Davidski said...
    " Iran_ChL is 63% CHG and 17% Iran_N in the paper. That's basically the results that I'm getting in my models. So it was an almost total replacement."


    Except, CHG itself is highly Irn_N. Thus, that exchange doesn't make IRAN_ChL, or more importantly CHG, much unique from Iran_N. So as for being a "total replacement", that is meaningless. There was genes which originated in Iran and, given the proximity of both areas, likely reciprocated freely between the regions.

    ReplyDelete
  48. There was no migration from Iran to the European steppe.

    You're out of your fucking mind.

    ReplyDelete
  49. Interesting result here. It looks like there was a fairly substantial admixture of IRAN Chl and Canaanites in the Southern Levant (Sidon and Jordan)

    http://www.sciencedirect.com/science/article/pii/S0002929717302768?via%3Dihub

    ReplyDelete

Read the rules before posting.

Comments by people with the nick "Unknown" are no longer allowed.

See also...


New rules for comments

Banned commentators list