Thursday, April 6, 2017

On mobility in the Eastern Mediterranean during the Bronze and Iron Ages


At Scientific Reports Meiri et al. present and analyze an updated dataset of ancient cattle and pig DNA from the Eastern Mediterranean. At the moment, ancient pig DNA is actually one of the best resources for studying human population movements in the region during the tumultuous Bronze and Iron Ages.

However, this is likely to change later this year or next year, with the publication of high density ancient human genome-wide DNA data for the Minoans, Mycenaeans, Philistines and other main players in the Bronze and Iron Age Eastern Mediterranean.

In any case, interestingly, pig mitochondrial (mtDNA) haplogroup Y2 is found on the Pontic Steppe during the Neolithic-Chalcolithic (7000-3500 BCE). It then appears during the Early Middle Bronze Age (3500-1550 BCE) in Greece and Anatolia. I do wonder if these pigs migrated south with the speakers of Proto-Greek and Proto-Anatolian?

Abstract: The Late Bronze of the Eastern Mediterranean (1550–1150 BCE) was a period of strong commercial relations and great prosperity, which ended in collapse and migration of groups to the Levant. Here we aim at studying the translocation of cattle and pigs during this period. We sequenced the first ancient mitochondrial and Y chromosome DNA of cattle from Greece and Israel and compared the results with morphometric analysis of the metacarpal in cattle. We also increased previous ancient pig DNA datasets from Israel and extracted the first mitochondrial DNA for samples from Greece. We found that pigs underwent a complex translocation history, with links between Anatolia with southeastern Europe in the Bronze Age, and movement from southeastern Europe to the Levant in the Iron I (ca. 1150–950 BCE). Our genetic data did not indicate movement of cattle between the Aegean region and the southern Levant. We detected the earliest evidence for crossbreeding between taurine and zebu cattle in the Iron IIA (ca. 900 BCE). In light of archaeological and historical evidence on Egyptian imperial domination in the region in the Late Bronze Age, we suggest that Egypt attempted to expand dry farming in the region in a period of severe droughts.

...

Haplotype Y2 is considered to have a Near Eastern origin [27, 28]. However, the existence of pig haplotype Y2 in our Greek samples during the Early Helladic II (one radiocarbon determination – 2875–2581 cal BCE) (Fig. 3) together with the findings of Mesolithic wild boar remains in Romania and northeast Italy [33, 35] challenge this conventional wisdom. The absence of haplotype Y2 from Anatolia in the Neolithic (despite a large sample size, n = 38 [28]) on one hand, and its presence in Romania during this period on the other [33] suggest a west-to-east translocation, from Greece to Anatolia no later than the Early Bronze Age.


Meiri et al., Eastern Mediterranean Mobility in the Bronze and Early Iron Ages: Inferences from Ancient DNA of Pigs and Cattle, Scientific Reports 7, Article number: 701 (2017) doi:10.1038/s41598-017-00701-y

51 comments:

  1. Well , as a friend tells me , Indian Zebus were not anti-Semitic ! .

    ReplyDelete
  2. "pig mitochondrial (mtDNA) haplogroup Y2 is found on the Pontic Steppe during the Neolithic-Chalcolithic (7000-3500 BCE). It then appears during the Early Middle Bronze Age (3500-1550 BCE) in Greece and Anatolia."

    This long distance migration, in spite of the fact that the study notes that pigs are not suitable for long distance herding, provides another insight. The Bronze Age migrants to Greece and Anatolia who brought the pigs with them must have migrated from the Pontic Steppe to Greece and Anatolia by ships traversing the Black Sea, instead of by land.

    "Haplotype Y2 is considered to have a Near Eastern origin [27, 28]. However, the existence of pig haplotype Y2 in our Greek samples during the Early Helladic II (one radiocarbon determination – 2875–2581 cal BCE) (Fig. 3) together with the findings of Mesolithic wild boar remains in Romania and northeast Italy [33, 35] challenge this conventional wisdom."

    The date is material. This is a few hundred years earlier than strongly suggestive archaeological evidence of Indo-European migration to Greece and Anatolia, a century or two before 2000 BCE. But, is also quite a bit later than some of the estimates of when that migration took place made by linguists, when tend to assume an Indo-European migration to Anatolia that is much earlier than other Indo-European migrations (often before 3500 BCE).

    "The absence of haplotype Y2 from Anatolia in the Neolithic (despite a large sample size, n = 3828)"

    A loss of formatting from the original in the quote above creates a misleading impression. Where it says "despite a large sample size, n=3828", the source text actually says "n=38" and has a superscript end note reference in a different color which is "28." So, the real sample size is 38 and not 3828 which would be unprecedented.

    ReplyDelete
  3. A loss of formatting from the original in the quote above creates a misleading impression. Where it says "despite a large sample size, n=3828", the source text actually says "n=38" and has a superscript end note reference in a different color which is "28." So, the real sample size is 38 and not 3828 which would be unprecedented.

    Fixed it.

    ReplyDelete
  4. Pigs aren't steppe animals
    They're mostly from forested zone of Europe (eg GAC).

    ReplyDelete
  5. Catacomb and Khvalynsk sites on the steppe often show pig remains.

    ReplyDelete
  6. The paper, which has amongst its author Israel Finkelstein, one of the archaeologists who disproved all what the Bible said about Old Jewish History, and is at the base of the book of Shlomo Sand, The Invention of the Jewish people, says clearly that there was a migration of the Sea Peoples (amongst them Philishtim) from Aegean countries to Palestine. This is the premise to say that some Jewish haplotype (I am saying European and not Middle Eastern ones) may have entered the Jewish pool then. It is a possibility that I have always taken into account. We'll see in each case when it is true. So far the Jewish haplotype (only one) of the R-M269* has found a SNP in common with a Bulgarian, so something is breaking in the 19 SNPs upstream this "Jewish cluster". We'll see next if it entered from Sea Peoples or more recently from Khazars or others.

    ReplyDelete
  7. Khvalynsk> Sredny Stog> Ezero> Anatolia. I think with thousands of years inbetween the pigs surely could have walked a few kilometres per year.

    ReplyDelete
  8. RK,

    An additional point of interest; I've found that PCoA on Fst allows for excellent "basal" models.

    Being able to do something along these lines has long been an abiding interest for me, and I've always found my own attempts with D-stats and PCA-based nMonte, and the efforts of others with qpAdm, to be inadequate.

    In these fits, only four streams of West Eurasian ancestry were allowed: AG3-MA1, Iran_Neolithic, WHG, and Levant_Neolithic, along with a bunch of (supposedly) "unadmixed" ENA and African reference populations.

    South Asia:

    Dravidian_South

    35.60% Iran_Neolithic + 13.20% AG3-MA1 + 4.95% WHG + 2.05% Luhya_Kenya + 0.95% Levant_Neolithic

    23.05% Ami + 11.75% Onge + 8.45% Papuan

    Distance=14.1444

    GujaratiD

    46.15% Iran_Neolithic + 15.60% AG3-MA1 + 5.70% WHG + 2.40% Levant_Neolithic + 0.55% Luhya_Kenya

    16.20% Ami + 7.60% Onge + 5.80% Papuan

    Distance=12.8

    GujaratiC

    45.90% Iran_Neolithic + 16.65% AG3-MA1 + 6.40% WHG + 3.70% Levant_Neolithic

    14.80% Ami + 7.15% Onge + 5.40% Papuan

    Distance=11.7104

    GujaratiB

    44.35% Iran_Neolithic + 17.60% AG3-MA1 + 7.50% WHG + 6.65% Levant_Neolithic

    13.65% Ami + 6.65% Onge + 4.00% Papuan

    Distance=10.7969

    GujaratiA

    45.55% Iran_Neolithic + 19.65% AG3-MA1 + 8.10% WHG + 7.65% Levant_Neolithic

    11.30% Ami + 5.10% Onge + 2.65% Papuan

    Distance=10.0363

    The ENA levels are identical to what I get with my ASI simulations (using the PCA data); 40%-45% in non-Brahmin/non-tribal South Indians, and around 20% in GujaratiA.

    Also, the patterns here are rather nicely in-sync with David's Basal K7 ADMIXTURE test.

    ReplyDelete
  9. Continuing…

    South Central Asia:

    Pashtun (“Pathans”)

    46.70% Iran_Neolithic + 17.75% AG3-MA1 + 11.90% Levant_Neolithic + 8.00% WHG

    10.20% Ami + 3.65% Onge + 1.80% Papuan

    Distance=9.3252

    Kalash

    53.65% Iran_Neolithic + 20.25% AG3-MA1 + 8.20% WHG + 5.35% Levant_Neolithic

    9.15% Ami + 1.70% Onge + 1.70% Papuan

    Distance=9.2269

    Tajik_Ishkashim

    40.70% Iran_Neolithic + 20.85% AG3-MA1 + 13.70% Levant_Neolithic + 10.60% WHG

    10.25% Ami + 2.85% Onge + 1.05% Papuan

    Distance=8.9494

    Tajik_Shugnan

    39.15% Iran_Neolithic + 21.85% AG3-MA1 + 16.25% Levant_Neolithic + 12.40% WHG

    7.95% Ami + 1.80% Onge + 0.60% Papuan

    Distance=9.1692

    Tajik_Rushan

    39.60% Iran_Neolithic + 19.85% AG3-MA1 + 17.25% Levant_Neolithic + 12.70% WHG

    9.05% Ami + 1.20% Onge + 0.35% Papuan

    Distance=9.2977

    West Asia:

    Brahui

    53.30% Iran_Neolithic + 14.60% Levant_Neolithic + 13.75% AG3-MA1 + 5.90% WHG

    7.50% Ami + 2.65% Onge + 1.60% Papuan

    0.70% Gambian

    Distance=8.9841

    Makrani

    53.80% Iran_Neolithic + 16.20% Levant_Neolithic + 11.90% AG3-MA1 + 6.10% WHG

    6.00% Ami + 2.35% Onge + 1.30% Papuan

    2.35% Gambian

    Distance=8.6231

    Iranian_Bandari

    46.70% Iran_Neolithic + 22.70% Levant_Neolithic + 11.10% AG3-MA1 + 5.70% WHG

    5.15% Ami + 2.20% Onge + 1.25% Papuan

    5.20% Gambian

    Distance=8.3521

    Iranian_Persian

    44.00% Iran_Neolithic + 31.65% Levant_Neolithic + 10.80% AG3-MA1 + 6.10% WHG

    5.60% Ami + 1.50% Onge + 0.35% Papuan

    Distance=9.127

    Iranian_Mazandarani

    50.40% Iran_Neolithic + 28.90% Levant_Neolithic + 10.40% AG3-MA1 + 4.65% WHG

    4.55% Ami + 0.85% Onge + 0.25% Papuan

    Distance=9.7075

    Europe:

    Finnish

    36.65% WHG + 27.25% Levant_Neolithic + 18.70% AG3-MA1 + 9.00% Iran_Neolithic

    7.30% Ami + 1.10% Onge

    Distance=10.7748

    Russian_Kargopol

    34.30% WHG + 26.30% Levant_Neolithic + 18.90% AG3-MA1 + 11.75% Iran_Neolithic

    7.65% Ami + 1.10% Onge

    Distance=10.722

    English

    38.15% Levant_Neolithic + 32.65% WHG + 15.55% AG3-MA1 + 10.80% Iran_Neolithic

    2.35% Ami + 0.50% Onge

    Distance=11.3052

    All in all, I’d say that these are solid results.

    In addition, here is something that's very interesting. I tried to model the Pashtuns with contemporary, geographically close reference populations. This is what I find, using the PCoA.

    “Pathans”

    20.70% Tajik_Ishkahim + 18.85% Tajik_Shugnan
    37.85% Brahui
    22.60% Punjabi_Lahore

    Distance=0.6355

    Now, compare this to what I find with the Global_10 data (scaling was applied).

    “Pathans”

    22.1% Tajik_Shugnan + 17.2% Tajik_Ishkashim
    37.6% Brahui
    23.1% Punjabi_Lahore

    Distance=0.0619

    Virtually identical! I was mildly shocked when I saw this.

    This sort of thing really bolsters my confidence in the use of Fst-based data.

    ReplyDelete
  10. @xyyman

    Both Natufians and the ancient Egyptians sampled to date are clearly West Eurasians, not Sub-Saharan Africans and not East Africans.

    It's your choice if you refuse to accept this, but I will delete all of your posts that contradict this fact.

    ReplyDelete
  11. I think a few words should be said about the Zebu admixture in the Iron Age sample from Israel, found in the above study.

    The authors speculate, justifiably, that the mating of Zebu males with Taurine females could have happened during the rule of the Egyptians in the Levant.

    This is possible but we know from osteological and material evidence that the presence of Zebu in both the Near East & Egypt dates to much earlier. In Iran and the Caucasus it dates to as early as 3,000 BC. The Zebu was already present in Anatolia during the Hittite rule.

    The authors of the paper also argue that the Taurine cattle spread around Eurasia with the movement of people from the fertile crescent. Applying this theory, it can very well be said that the movement of Zebu into Iran and subsequently into the Caucasus and & Levant was through the agency of human movements into these regions with their ultimate origin in South Asia.

    We know from archaeology that there were Indus colonies in Mesopotamia, in Central Asia and on the SE Arabian coast. We also now have the evidence of mtdna M subclades (with their origin in South Asia), in early bronze age North Caucasus, as well as in Mesopotamia and now recently in Egypt (mtdna M). Interestingly enough, the recent paper on WE mtdna expansions in South Asia (Silva et al), argued that there was evidence of expansion of some mtDNA M subclades (including a subcade of M5 i.e. M5a2a4) during the bronze age from South Asia into Iran. Could the New Kingdom Egyptian M5 be related to this ?

    We know that the Mitanni kings from Syria had marital relations with the kingdom of Egypt. This could have been one of the ways in which the Indian specific mtdna M5 as also the spread of Zebu into Egypt happened.

    ReplyDelete
  12. @Jaydeep

    Interesting about the M5a2a4. The biggest surprise of the Egyptian aDNA was the super-high M5 frequency. I wonder what will turn up in the Y DNA.

    I wonder about Indian Ocean connections as well. There's the maritime trade between Mesopotamian and Harappan civilizations, the appearance of Sahel crops in the South Indian Neolithic, possibly really early plantains in West Africa.

    ReplyDelete
  13. @ Sein, that matching result you found between Globe10 and this PCoA is pretty interesting. Is that a general phenomenon of other populations? I know that I would be surprised if it replicated in Europe as Globe10 didn't seem to me to have enough information to recreate fine scale European genotype PCA (which was replicated by PCoA on European Fst).

    @ Davidski, btw, on your Fst table, I guess you calculated the same way as Laz 2016 - "Estimation of FST coefficients - We estimated FST in smartpca with default parameters, inbreed: YES, and fstonly: YES."? I cross referenced the scores and they're pretty identical (or at like 0.001 / 0.002 closeness, so shouldn't have a major effect).
    I still have some concerns about whether the ancients have effects from damage, so I was wondering, do you know if the Fst scores can be run over just transversion SNPs to limit any effects of damage further? (Or you may have already done that!). I guess from your prior stuff that might SNPs down to 60,000 or something but that might be sufficient for structure.

    ReplyDelete
  14. Btw, with the Fst PCoA, I thought this might be an interesting example to some, showing why I am concerned about the effect on this of any potential slight amplification of extra distances at fine scales between ancients and moderns:

    So let's take the European subset of matrix, and add in the ancestors who would appear to be most proximately relevant to the main streams of ancestry in Europeans (I chose Central_MN, Iberia_Chalcolithic, Iberia_MN, Yamnaya_Kalmykia, Yamnaya_Samara, Andronovo, Armenia_MLBA, Bell_Beaker_Germany, Corded_Ware_Germany, Hungary_BA, Srubnaya, Unetice): https://pastebin.com/PciV3uVr

    Now run PCoA on this: http://i.imgur.com/1MYvUeT.png

    There are similarities to what we'd expect for a European PCA, but we see that the whole Europe_MN->Steppe axis is displaced from the Europeans.
    This holds true for axis 3, which doesn't seem to resemble anything much. This is what we'd expect if all the ancients had systematically raised Fst to recent people, with correct Fst to each other.

    Now let's subtract a standard effect of 0.003 from Fst between ancients and moderns (while leaving their Fst to each other identical): https://pastebin.com/Y606rFC1

    (In the grand scheme of things, 0.003 is a tiny Fst, but on the European fine structure scale, it's relatively large).

    Run PCoA again: http://i.imgur.com/6RKImtI.png

    In axis 1 and 2, the moderns fit cleanly on the Europe_MN->Steppe axis. Moreover in axis 3, the positioning of ancient LNBA is very as expected - Bell Beaker with Scots and Orcadians, Unetice with Icelandic, Hungary_BA closest to present day Central Europeans. (If Baltic_BA were on, I think it would likely sit with the North Slavic-Baltic region of the graph).
    This continues into axis 5.

    This is maybe even clearer if we remove the Basques, who tend to be axis dominant likely due to higher unique drift - http://i.imgur.com/aIO3ZwB.png

    Changes are a bit more minimal in neighbour joining: http://i.imgur.com/ARgYV5f.png ... but you still get a very nice embedding of Hungary_BA in Central Europe with the subtraction on ancient to modern Fst: http://i.imgur.com/MhWRTJB.png

    Ideally we could know if this subtraction is a correct procedure, and have it emerge from running the Fst calculation rather than this manual adjustment.

    ReplyDelete
  15. "We know that the Mitanni kings from Syria had marital relations with the kingdom of Egypt. This could have been one of the ways in which the Indian specific mtdna M5 as also the spread of Zebu into Egypt happened."

    Can I be forgiven for having a dirty thought leap into my head upon reading this paragraph? ;)

    ReplyDelete
  16. Matt,

    To be honest, this was my first attempt to use contemporary + geographically adjacent reference populations.

    But, now that you mention it, I would like to compare and contrast how nMonte performs with PCA and PCoA data, when the references are recent + neighboring populations.

    I'll definitely look into it.

    ReplyDelete
  17. @ Sein, cool, don't worry if you have more interesting models to look into, only if you have time for comparisons.

    Few structural graphics based on the Fst scores (after a -0.003 adjustment for ancients):

    World Structure: http://i.imgur.com/bIiwRD8.png

    Structure within cluster 1 (when I split populations into 3 clusters based on their PCoA scores): http://i.imgur.com/M6qA2r7.png

    PCoA Axis 1 vs 2: moderns + ancients -http://i.imgur.com/76itnzh.png ; moderns only - http://i.imgur.com/lCrZ6QC.png

    I like the last moderns only graphic - although it's obviously potentially historically misleading, it shows some cool distinctions that show up with this swathes of populations (esp re: distinct Iranian and other South Asian clines).

    (to be honest, the Fst adjustment doesn't matter much at this scale, it's on the finer scale that it comes more into play)

    ReplyDelete
  18. Matt,

    Ah, you're right; lots of interesting patterns on those moderns-only PCoA axes.

    Looking at something like that, it's easy to imagine that this sort of data will be excellent for analyzing fine-scale regional structure.

    Frankly, I'm glad that we can now count this in our general arsenal of analyses; you came up with a very interesting and effective approach, one which definitely deserves further development (it'll be interesting to see if your concerns with regard to aDNA damage can be addressed).

    ReplyDelete
  19. one of my pet theories is
    - LBK brought their pigs with them
    - Ertobolle adopted LBK pigs then mixed them with north Euro boar (for whatever reason, cold resistance maybe)
    - the now Euro pigs spread east with funnelbeaker etc replacing the original species
    - eventually they spread into the mid-east via Hyksos or whoever bringing a new strain of swine flu the locals had no resistance to

    hence the anti-pork thing

    ReplyDelete
  20. Jaydeep

    "This could have been one of the ways in which the Indian specific mtdna M5 as also the spread of Zebu into Egypt happened."

    Seems plausible to me that particular resources developed where they were first found and then spread outwards with a geographical crossroads like the mid-east ending up being the first to get the whole collection.

    ReplyDelete
  21. @Ryukendo K

    https://drive.google.com/file/d/0B9o3EYTdM8lQUHhHa1NKZDlTV0U/view?usp=sharing

    https://drive.google.com/file/d/0B9o3EYTdM8lQeS1Fc3JBTTJVbk0/view?usp=sharing

    Haven't checked these, so there might be some problem samples. If there are, just take them out of the Fst_matrix file and run the PCoA with Past.

    ReplyDelete
  22. @Grey

    " eventually they spread into the mid-east via Hyksos or whoever bringing a new strain of swine flu the locals had no resistance to

    hence the anti-pork thing"

    More likely Trichinosis.

    ReplyDelete
  23. @Grey

    "- Ertobolle adopted LBK pigs then mixed them with north Euro boar (for whatever reason, cold resistance maybe)"

    People at that time probably had no real sense of breeding animals or plants for a purpose. They simply bred what was available, and at some point made a selection for the next generation. Either by killing the weak, or by picking out the strong to reproduce, or just by statistical survival/reproduction levels (without conscience selection at all).

    So hybridization with local boars could have just happened on accident, but the hybrid offspring were more fit to local conditions, so they had a reproductive advantage.

    ReplyDelete
  24. Jaydeep and others ,

    Remember that at Megiddo, there were maryannu with Indo-Iranian names in the Amarna letters (14th century BC).

    ReplyDelete
  25. @Matt

    Transversions only...

    https://drive.google.com/file/d/0B9o3EYTdM8lQdzRLeEMwblZ1bVU/view?usp=sharing

    https://drive.google.com/file/d/0B9o3EYTdM8lQZk82TTlZSkNCX2M/view?usp=sharing

    ReplyDelete
  26. @ Davidski - thanks for running those. However, they don't seem to show any reduced differentiation from moderns (if anything differentiation seems pretty similar to the allsites, but generally very slightly raised, globally). So if there is a systematic "ancients higher Fst with moderns", we at least know it looks like we can't deal with it through transversions site restrictions.

    @ Ryu and Sein

    Upthread I mentioned using a global shift of -0.003 Fst on ancients, to create a European+ancients PCoA where moderns fit on a steppe->MN cline and specific linkages showed in higher dimensions with regional LNBA populations.

    Doing that with world data to produce this matrix (https://pastebin.com/SuBUmjvk) then running a PCoA on that (https://pastebin.com/6xg37jMN - restricted to 25 dimensions).

    With nMonte I started to get the following fits - https://pastebin.com/VqjwyWjR.

    You can contrast these with the fits for populations on a PCoA I ran with about the same number of axes, and the same calc populations *without* the ancient adjustment before running PCoA here - https://pastebin.com/H1vaw8MF where you have England, Lithuania, Sardinian picking up approx >4% Yukaghir Forest that isn't real from any other method. On other runs I didn't upload, this was a systematic problem and actually this was an even bigger for Iranians, Finns, Palestinians, Sicilians with even more than 4% proportions.

    Now the big immediate shift for the fits on the adjusted data is that these fits do *not* contain any excess of modern ENA / African related ancestries, and the fit % is also much closer, and the fits are generally as logical as they were before. They look better.

    Among Europeans, only Finns pick up significant traces of Siberian ancestry (at a level that works out aroun 7% ENA) and Palestinians and Persians get expected levels of African and ENA.

    So on the positive, this gives me a bit more confidence that there may be a slight issue here with ancients having slightly raised Fst to moderns. Not enough to be an issue the way Lazaridis etc were using them, but causes a problem with what we're doing in the PCoA+nMonte, where to offset the differentiation, these admixed / Asian / African moderns can get drafted in if allowed to, in excess of the real contribution.

    On the negative, it does mean that using the Fst this way seems really sensitive to even slight differences in Fst at the level of fine structure. I've offset this for the ancients with the -0.003 adjustment, but likely in some instances this slightly overshoots. It also means to get the best fits at very fine scale it may be necessary to go to 4 or 5 decimal places.

    On the actual fits themselves, there are some interesting patterns. In addition to the Bell Beaker / Unetice signature in Western Europe, there are a few of unusual things:

    1) Basques as a fit of almost purely steppe and local Iberian farmers with no additional Bell Beaker signature, 2) Eastern Europe gets a *lot* more Hungary_BA, Corded Ware and Steppe_LNBA, looking like a synthesis of these with no influence from BB or Unetice, 3) total absence of Armenia_MLBA in "Pathan" even when allowed to take it, in favour of a Steppe+Iran_Neo synthesis, in contrast to Persian.

    Of course, I wouldn't want to read too much into these, as my -0.003 is a bit arbitrary.

    ReplyDelete
  27. @ Ryu, Naxi/Burmese/Han NChina/Tu does seem to form some pole of East Asian variation under certain sets of Fst (taking those Dave has put up for you ITT):

    For a tight set of populations lensed around Han: http://i.imgur.com/IXqZsds.png

    For a wider set of populations: http://i.imgur.com/8DazZac.png

    In both these PCoA you have a North-South East Asian cline (depending on the set, termining at Murut Austronesian and Yakut or in Buryats and She), then you have a dimension separating what seems to be ASI (and to a lesser extent general West Eurasian) influence. These swap places for 1 and 2 in distinction depending on the populations we run.

    Then the Axis 3 seems to be separating Naxi and NW East Asia from both Siberians and Austronesians / SE Asians and seems fairly general. This separates what appear to be Western and Northern Tibeto-Burman populations from otherwise similar combinations of the broad Siberian+SE Asian tendencies.

    (Other axes add more separation further down the pike - e.g. in the general Japanese get their own axis of separation, likely linked to the Jomon - but lots of these lack a clear geographical interpretation, and PCoA in higher dimensions sometimes overfits populations on one dimension to balance another, capturing distance so useful for nMonte, but compounding problems with geographic interpretation.)

    ReplyDelete
  28. Karl_K

    "People at that time probably had no real sense of breeding animals or plants for a purpose."

    yes, good point

    #

    Jingus Jendal said...

    "More likely Trichinosis."

    more likely but less fun

    ReplyDelete
  29. Matt,

    Interesting stuff, I'll definitely try some analyses with the adjusted ancients Fst-based PCoA.

    RK,

    Your suggestions here are of considerable interest, and I find the Kusunda-HG idea to be very intriguing.

    ReplyDelete
  30. Karl,

    "People at that time probably had no real sense of breeding animals or plants for a purpose."

    Aha. Which means 'abscence of proof' is 'proof of abscence'. Better try Purgatori.

    "They simply bred what was available, and at some point made a selection for the next generation. Either by killing the weak, or by picking out the strong to reproduce, or just by statistical survival/reproduction levels (without conscience selection at all)".

    Is that a concept found in Dante, or a concept you just made up, rather?

    "So hybridization with local boars could have just happened on accident, but the hybrid offspring were more fit to local conditions, so they had a reproductive advantage."

    1. Does this mean that the Purgatory have a beginning AND an end?
    2. Compared with todays average European, would the brain-size of the Cro-Magnons and the Neanerthals be smaller or bigger?

    ReplyDelete
  31. Rob,

    "Pigs aren't steppe animals
    They're mostly from forested zone of Europe (eg GAC)."

    Correct. Besides, both thrives the best among green hills with braod-leaved threes, nuts, roots and mushrooms. Besides whatever spillover their neighbouring two-legs care to leave from their households.

    Don't forget that also PWC/TBK, such as Ajvide, kept quite a few pigs. Probably goats, too.

    Compared to dogs and aurochs domesticating of (wild) reindeer, deer, pigs and goats is quite easy. Especially the latter ones, that seem to have a natural curiousity/affinity to humans.

    As the LP population of Eurasia learned to know their hunting-grounds and main foodsources very well, they would soon learn everything worth knowing about each and every specie of their megafauna. The use of certain animals as icons - as signs of dignification/purpose/position - is apparently part of several paleolithic family-sites.

    Since trapping was the most usual hunting-method in Paleolithic Eurasia we may suspect that the Mesolithic pioneers of arctic Eurasaia had inherrited their way of life - including the botanic and zoological knowledge needed to survive at the nortthern ties of Eurasia, as well as the southern.

    From the domesticated species of today we should suspect that the proto-dog (Wolf) and the first holsteinian cattles (Aurochs) should be the most challenging to domesticate. Together with the Taipan, from whihc the first large, warm-blooded horses seems to have been bred.

    The domestication of food-plants are definitly a trait from the Paleolithic Europe, as well. 19.000 years old seeds found at archaeological sites proves Oats was used in Italy and of Grapes around Geneva well before 8.500 years ago - when "agriculture" as "ONE packet" is s-u-p-p-o-s-e-d to have start spreading - "Out-of-Anatolia".

    In the 1950-ies, when Gordon B. Childe formulated this suggestion the first time, it was an important conceptualization and thus an academic achivement - at it's time. Today we have basis of facts - from a variety of new discoveries and (even) sciences, making it possible to run all kinds of fact-stats to complete Wildes basic 'concept' - today misconstrued as a "teory" and threated as a "consentual fact".

    With discoveries of domectication dated back to the Late Paleolithic - like 35.000+ yrs old mortars, ropes and dog-dna - it seems clear that Wilde's packaging and conceptulaization ('Out-of-Mesolithic-Anatolia') needs some major adjustments.

    Like describing the flexed and ochred skeletons from the Natufian Pit-Graves, 9.000+ BP, with the flexed ochre-grves in NW Europe, 9.500+ BP - to descibe a common "pioneer-period", where y-dna G spread in the Mediterranean part of Europe, while y-dna I populated the Atlantic facade - from Gibraltar to North Cape. The concept of agriculture seem to have been a part of both lines economical strategies - based on goats and sheep - and/or dogs and pigs. Besides the obvious harvesting from nature, by plucking, cutting, catching, trapping, hooking, hitting or shooting.

    The later spread of domesticated aurochs and horse was obviously not an option before the Holocene climate reached its 'optimum' - ans the old tundra became green heaths and meadows - from Iberia to Siberia.

    The spread of domesticated cattles and horses can obviously be connected to the distribution of some later y-dna-lines - such as R1a/b. Perhaps also Q and T.

    ReplyDelete
  32. @Gioiello
    "amongst its author Israel Finkelstein, one of the archaeologists who disproved all what the Bible said about Old Jewish History

    Finkelstein did nothing of the sort, and I'm sure he'd disagree with your claim. The Tanakh is twenty-four different books written (and rewritten) in the Levant over a thousand year period. Some books are poetry, some law, some are histories and some are mortality parables. You can say parts are improbable, other parts allegory and other parts are fables, but to say "disproved" is stupid.
    That's like saying the Homeric Epics are "disproved", or the Puranas are "disproved". It's ignorant and ahistorical.

    ReplyDelete
  33. @ Spanked
    "@Gioiello
    "amongst its author Israel Finkelstein, one of the archaeologists who disproved all what the Bible said about Old Jewish History

    Finkelstein did nothing of the sort, and I'm sure he'd disagree with your claim. The Tanakh is twenty-four different books written (and rewritten) in the Levant over a thousand year period. Some books are poetry, some law, some are histories and some are mortality parables. You can say parts are improbable, other parts allegory and other parts are fables, but to say "disproved" is stupid.
    That's like saying the Homeric Epics are "disproved", or the Puranas are "disproved". It's ignorant and ahistorical".
    Ask Jews if they consider the Bible what you say, or if they think that
    1) Abraham came from Ur
    2) Jews were slaves in Egypt
    3) Moses brought them to the Promised land (which would be the land where they stayed before as they have done to-day with the Aliyah)
    4) there were kings (Saul, David, Salomon)
    5) this kingdom was separated in two: Israel and Judah
    etc etc etc
    For what I know, Finkelstein "disproves" all that by an archaeological point of view.
    I, amongst many others, am saying that European Jews (Ashkenazim and Sephardim) have a little to do by a genetic point of view with those Jews.
    About who is "ignorant" and "ahistorical" very likely we have different points of view.

    ReplyDelete
  34. My understanding is that in a region of limited grazing resources, ie the levant and more arid border regions, pigs represented a poor choice as a domesticate. They cannot be utilized for milk, or wool, and their hides are less versatile than cattle or goats. Moreover they tended to compete with humans for food resources, ie nuts and fungi, whereas goats, sheep, and cattle were grazers of plant matter that were otherwise unexploited by humans. So the cultural aversion to pigs arose from a cost/benefit analysis in a semi-marginal environment. Thusly in wetter, more densely vegetative habitats pigs were more likely to be utilized as domesticates.

    ReplyDelete
  35. RK and Matt,

    I've been trying to use a set of references that can work with every contemporary West Asian, Central Asian, and South Asian population, using the adjusted ancients Fst-based PCoA.

    So far, this setup works great for all the populations I've tested.

    Iranian_Persian:

    61.25% Iran_Chalcolithic
    14.40% Sarmatian_Pokrovka
    13.15% Srubnaya
    7.05% Jordan_EBA
    1.05% CHG

    3.10% Thai

    distance=2.0571

    "Pathan"

    30.50% Iran_Chalcolithic
    30.25% Sarmatian_Pokrovka
    15.05% Iran_Neolithic
    7.95% Srubnaya
    5.45% MA1

    10.80% Thai

    distance=5.3796

    Tajik_Ishkashim

    29.05 Srubnaya
    19.90% Sarmatian_Pokrovka
    19.70% Iran_Chalcolithic
    15.10% Iran_Neolithic
    6.35% AG3-MA1

    6.30% Thai + 2.80% Ulchi + 0.80% Mongola

    distance=3.5645

    Tajik_Rushan

    29.75% Srubnaya
    27.90% Sarmatian_Pokrovka
    24.65% Iran_Chalcolithic
    9.10% Iran_Neolithic
    2.40% AG3-MA1

    3.30% Thai + 1.75% Ulchi + 1.15% Mongola

    distance=2.7231

    Makes sense.

    ReplyDelete
  36. @batman

    "1. Does this mean that the Purgatory have a beginning AND an end?
    2. Compared with todays average European, would the brain-size of the Cro-Magnons and the Neanerthals be smaller or bigger?"

    I have extreme difficulty following your leaps of topic and logic.

    It is well known that humans did not have any firm grasp of the science behind genetics and selection until quite recently. Do you deny that? Before the 1800s, there was no concept of genes, or that individual traits could be separated from each other by breeding. It was mostly assumed that traits blended together, simply because nobody before Mendel wrote down what happens when you carefully cross individuals for several generations.

    Of course natural/artificial selection always occurs, nonetheless.

    So, what is your point?

    ReplyDelete
    Replies
    1. Stallions Fighting competitions found practices by the Vikings. Bulls Fighting competitions practiced by Celts especially in Ireland and found today within the Balkans and Portugal. Herens Cows fighting in the Alps.Surely this points towards some kind of selectioñ by ancients....

      Delete
  37. @Karl Well apparentry the Scythians gelded(Neutered)some stallions.This surely point at some kind of selectioñ.

    ReplyDelete
  38. @Ric

    No doubt. I said from the start that people selected the best for the next generation, and limited breeding by unsuitable animals. This is obviously how multiple dog breeds emerged.

    All I meant was that breeding was in general for an entire phenotype, but not for individual traits (such as the cold tolerance that Grey suggested when I first commented).

    People have been selecting for 'the best' plant or animal for a long time. They just were not selecting for 'cold tolerance' as a particular seperable trait.

    ReplyDelete
  39. @Karl yes I agree with you there. Nature usually did the biggest selection.

    ReplyDelete
  40. This is ine of a few reasons why there was a major jump in agricultural productivity in the early 20th century. People finally understood that genes and the associated traits could be broken up and inherited as distinct units, instead of as a general blending.

    So. To be clear. People have selected plants and animals for many thousands of years. But, only as an organism, not for individual genes or traits that could be passed into other breeds on an individual basis.

    ReplyDelete
  41. @Gioiello
    Disputing a founding myth is hardly "disproving the bible", and Finkelstein ignores an awful lot of archaeological evidence for many of his other claims related to Samaria and Kingdom of Judah. The man hardly has the support of mainstream archaeology. His trick is getting grants and fame by playing the iconoclast.
    Between this and your Italian chauvinism, I see you have an attraction to nutty revisionist claims that appeal to your prejudices.

    ReplyDelete
  42. @ Spanked

    Before you said that Finkelstein wouldn't agree with me, now you say that archaeologists (and you) don't agree with him. I am here with my name, surname, place of origin and residence, Y and mt. Show me your Jewish Y and mt and I'll say you if you come from "Old Jews" or from some other people. For free. What FTDNA and people like you don't do.

    ReplyDelete
  43. So the Y2-pig spread from the Crimea.

    ReplyDelete
  44. Karl,

    "So, what is your point?"

    That you have no idea about the first, selective breeding and cross-breeding started. Which makes your assumptions - in this case - worthless. However well you've studied Mendels peafarm.

    Selective breeding - on a continual basis to reach specific, genetic results - had to be at work for quite a few genrations before the domesticated dog-races appears. No later than the Late Paleolithic.

    Well before agriculture grew large and wide, during the Holocene, there had been a phase of selective breeding of what became major food-plants. As the neolithic grew into every nook and corner of Eurasia there was a continous tradition at work, to select and stimulate the food-plants that gave the better yields - on various grounds and in fluctating climates.

    The view you maintain regarding the skills, insights and knowledge of the mesolithic pioneers and their neolithic descendants - claiming "they knew nothing" about fenotypes/genotypes and selective breeding - is nothing but a gross, unsubstantiated speculation.

    ReplyDelete
    Replies
    1. @batman They probably did not use the sofisticated words we use today to describe their basically similar methods. Heheheeeh..

      Delete
  45. @Sein, those seem pretty logical and easier to cross compare than a big soup of populations.

    Thought you might want to try some fits with the other matrix that Davidski posted up, for Ryukendo, with a slightly different set of SCA / SEA populations -

    Fst Matrix (adjusted ancient) - https://pastebin.com/wP3ae9jF

    PCoA (adjusted ancients 25 dimensions) - https://pastebin.com/K8nD55Gm (full PCoA - https://pastebin.com/H86550Ta)

    Example fits:
    South Central Asia and admixed SE Asia (using Hakkipikki for ASI proxy) - https://pastebin.com/mSRxwh21

    Europe -https://pastebin.com/Qf7iXLJR (Couple images: http://i.imgur.com/5OtzPrx.png and http://i.imgur.com/Fapl0dD.png).

    ReplyDelete
  46. @batman

    "The view you maintain regarding the skills, insights and knowledge of the mesolithic pioneers and their neolithic descendants - claiming "they knew nothing" about fenotypes/genotypes and selective breeding"

    You can think whatever you want, but the facts remain, and it's not about words and terms, it is about understanding what is actually happening, and having both the reason and ability to conduct careful breeding for many years.

    Mesolithic people certainly did not have that luxury, and neither did Neolithic people, or Bronze Age people, or 99.999% of 18th century Europeans.

    This has nothing to do with being smart.

    As an older example. Neolithic Europeans, Neolithic East Asians, agricultural South Asians and Americans, etc. They had very sophisticated societies, but didn't manage to invent the wheel and axle after many thousands of years of high cultural advancement.

    The fact is this. Before the 19th century, nobidy on Earth actually understood genetics or breeding animals. And still they didn't REALLY understand it until the mid 20th century.

    You do not have to understand selection, at all, to domesticate an animal or a plant. All you have to do is select.

    I assume you believe that honeybees are so smart and appreciate beauty so much that they decided to select flowers to be beautiful to the human eye?



    ReplyDelete
  47. So Y1-Pigs = Hatti
    Y2-Pigs = Proto-Anatolians/Hittites
    Arm1T-Pigs =Kura Araxes

    ReplyDelete
  48. Matt,

    Ah, the funny thing is that I ran a few models off of that PCoA, yesterday.

    Looking at Eastern Iranian peoples:

    Tajik_Pomiri

    43.85% Sarmatian_Scythian + 11.85% Steppe_EMBA
    32.40% Iran_Chalcolithic
    6.05% Mongola
    4.25% Cambodian
    1.60% AG3-MA1

    distance=4.9832

    Pashtun_Afghanistan

    46.80% Iran_Chalcolithic
    29.60% Sarmatian_Scythian + 6.60% Steppe_EMBA
    6.95% Mongola
    6.30% Cambodian
    3.75% AG3-MA1

    distance=5.8026

    Pashtun

    42.30% Iran_Chalcolithic
    20.30% Sarmatian_Scythian + 11.80% Steppe_EMBA
    13.55% Cambodian
    5.60% Iran_Neolithic
    5.10% AG3-MA1
    1.35% Mongola

    distance=6.9227

    Very sensible.

    Now that you've adjusted the ancients' distances to moderns, I'll definitely retry these models. I think it does much to clean things up.

    Also, I like those models; very interesting output, and some consistent patterns.

    Although, I would express caution with regard to the use of modern South Indians, since they are 55%-65% ANI/West Eurasian (probably 60% West Eurasian, looking at nMonte with PCA, David's "Basal K7" ADMIXTURE test, looking at some basal models I've tried for South Indians using the Fst PCoA, etc), so they are as much proxies for ANI as they are for ASI.

    Paniya are better, since they seem to be 60% ASI, but unfortunately we still have that 40% ANI (Iranian_Neolithic-related and ANE-related).

    A Mesolithic South Indian genome would be amazing.

    ReplyDelete
  49. It surely would have been nice to see if the Y2-Pigs reached Central Europe and when ? This could have provided a few more nails for some coffins.

    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