Bell Beakers and Mycenaeans from Yamnaya; Corded Ware from the forest steppe

eba-yamnaya-ancestry-hungary

I have recently written about the spread of Pre-Yamnaya or Yamnaya ancestry and Corded Ware-related ancestry throughout Eurasia, using exclusively analyses published by professional geneticists, and filling in the gaps and contradictory data with the most reasonable interpretations. I did so consciously, to avoid any suspicion that I was interspersing my own data or cherry picking results.

Now I’m finished recapitulating the known public data, and the only way forward is the assessment of these populations using the available datasets and free tools.

Understanding the complexities of qpAdm is fairly difficult without a proper genetic and statistical background, which I won’t pretend to have, so its tweaking to get strictly correct results would require an unending game of trial and error. I have sadly little time for this, even taking my tendency to procrastination into account… so I have used a simple model akin to those published before – in particular, the outgroup selection by Ning, Wang et al. (2019), who seem to be part of the only group interested in distinguishing Yamnaya-related from Corded Ware-related ancestry, probably the most relevant question discussed today in population genomics regarding the Proto-Indo-European and Proto-Uralic homelands.

eneolithic-steppe-best-fits
Supplementary Table 13. P values of rank=2 and admixture proportions in modelling Steppe ancestry populations as a three-way admixture of Eneolithic steppe Anatolian_Neolithic and WHG using 14 outgroups.
Left populations: Test, Eneolithic_steppe, Anatolian_Neolithic, WHG.
Right populations: Mbuti.DG, Ust_Ishim.DG, Kostenki14, MA1, Han.DG, Papuan.DG, Onge.DG, Villabruna, Vestonice16, ElMiron, Ethiopia_4500BP.SG, Karitiana.DG, Natufian, Iran_Ganj_Dareh_Neolithic.

I have used for all analyses below a merged dataset including the curated one of the Reich Lab, the latest on Central and South Asia by Narasimhan, Patterson et al. (2019), on Iberia by Olalde et al. (2019), and on the East Baltic by Saag et al. (2019), as well as datasets including samples from Wang et al. (2019) and Lamnidis et al. (2018). I used (and intend to use) the same merged dataset in all cases, despite its huge size, to avoid adding one more uncontrolled variable to the analyses, so that all results obtained can be compared.

I try to prepare in advance a bunch of relevant files with left pops and right pops for each model:

  1. It seems a priori more reasonable to use geographically and chronologically closer proxy populations (say, Trypillia or GAC for Steppe-related peoples) than hypothetic combinations of ancestral ones (viz. Anatolian farmer, WHG, and EHG).
  2. This also means using subgroups closer to the most likely source population, such as (Don-Volga interfluve) Yamnaya_Kalmykia rather than (Middle Volga) Yamnaya_Samara for the western expansion of late Repin/early Yamnaya, or the early Germany_Corded_Ware.SG or Czech_Corded Ware for the group closest to the Proto-Corded Ware population (see below), likely neighbouring the Upper Vistula region.
  3. I usually test two source populations for different targets, which seems like a much more efficient way of using computer resources, whenever I know what I want to test, since I need my PC back for its normal use; whenever I don’t know exactly what to test, I use three-way admixture models and look for subsets to try and improve the results.

I have probably left out some more complex models by individualizing the most relevant groups, but for the time being this would have to do. Also, no other formal stats have been used in any case, which is an evident shortcoming, ruling out an interpretation drawn directly and only from the results below.

Full qpAdm results for each batch of samples are presented in a Google Spreadsheet, with each tab (bottom of the page) showing a different combination of sources, usually in order of formally ‘best’ (first to the left) to ‘worst’ (last to the right) fits, although the order is difficult to select in highly heterogeneous target groups, as will be readily visible.

maykop-trypillia-intrusion-steppes
Disintegration, migration, and imports of the Azov–Black Sea region. First migration event (solid arrows): Gordineşti–Maikop expansion (groups: I – Bursuchensk; II – Zhyvotylivka; III – Vovchans’k; IV – Crimean; V – Lower Don; VI – pre-Kuban). Second migration event (hollow arrows): Repin expansion. After Rassamakin (1999), Demchenko (2016).

Corded Ware origins

The latest publications on the Yampil barrow complex have not improved much our understanding of the complexity of Corded Ware origins from an archaeological point of view, involving multiple cultural (hence likely population) influences. This bit is from Ivanova et al., Baltic-Pontic Studies (2015) 20:1, and most hypotheses of the paper remain unanswered (except maybe for the relevance of the Złota group):

In the light of the above outline therefore one should argue that the ‘architecture of barrows’ associated in the ‘Yampil landscape’ of the Middle Dniester Area with the Eneolithic (specifically, mainly with the TC), precedes the development of a similar phenomenon that can be observed from 2900/2800 BC in the Upper Dniester Area and drainage basin of the Upper Vistula, associated with the CWC [Goslar et al. 2015; Włodarczak 2006; 2007; 2008; Jarosz, Włodarczak 2007]. The most consuming research question therefore is whether ritual customs making use of Eneolithic (Tripolye) ‘barrow architecture’ could have penetrated northwards along the Dniester route, where GAC communities functioned. One could also ask what role the rituals played among the autochthons [Kośko 2000; Włodarczak 2008; 2014: 335; Ivanova, Toshchev 2015b].

This issue has already been discussed with a resulting tentative systemic taxonomy in the studies of Włodarczak, arguing for the Złota culture (ZC) in the Vistula region as an illustration of one of the (Małopolska) reception centres of civilization inspirations from the oldest Pontic ‘barrow culture’ circle associated with the Eneolithic and Early Bronze Age [Włodarczak 2008]. Notably, it is in the ZC that one can notice a set of cultural traits (catacomb grave construction, burial details, forms and decoration of vessels) analogous to those shared by the north-western Black Sea Coast groups of the forest-steppe Eneolithic (chiefly Zhyvotilovka-Volchansk) and the Late Tripolye circle (chiefly Usatovo-Gordinești-Horodiștea-Kasperovtsy).

gac-trypillia-usatovo-corded-ware
Globular Amphorae culture „exodus” to the Danube Delta: a – Globular Amphorae culture; b – GAC (1), Gorodsk (2), Vykhvatintsy (3) and Usatovo (4) groups of Trypillia culture; c – Coţofeni culture; d – northern border of the late phase of Baden culture;red arrows – direction of Globular Amphora culture expansion; blue arrow – direction of „reflux” of Globular Amphora culture (apud Włodarczak, 2008, with changes).

Taking into account that I6561 might be wrongly dated, we cannot include the Corded Ware-like sample of the end-5th millennium BC in the analysis of Corded Ware origins. That uncertainty in the chronology of the appearance of “Steppe ancestry” in Proto-Corded Ware peoples complicates the selection of any potential source population from the CHG cline.

Nevertheless, the lack of hg. R1a-M417 and sizeable Pre-Yamnaya-related ancestry in the sampled Pontic forest-steppe Eneolithic populations (represented exclusively by two samples from Dereivka ca. 3600-3400 BC) would leave open the interesting possibility that a similar ancestry got to the forest-steppe region between modern Poland and Ukraine during the known complex population movements of the Late Eneolithic.

It is known that Corded Ware-derived groups and Steppe Maykop show bad fits for Pre-Yamnaya/Yamnaya ancestry, and also that Steppe Maykop is a potential source of “Steppe-related ancestry” within the Eneolithic CHG mating network of the Pontic-Caspian steppes and forest-steppes. Testing Corded Ware for recent Trypillia and Maykop influences, proper of Late Trypillia and Late Maykop groups in the North Pontic area (such as Zhyvotylivka–Vovchans’k and Gordineşti) side by side with potential Pre-Yamnaya and Yamnaya sources makes thus sense:

Now, the main obvious difference between Khvalynsk-Yamnaya and Corded Ware is the long-lasting, pervasive Y-chromosome bottlenecks under R1b lineages in the former, compared to the haplogroup variability and late bottleneck under R1a-M417 in the latter, which speaks in favour – on top of everything else – of a different community of sub-Neolithic hunter-gatherers including hg. R1a-M417 hijacking the expansion of Steppe_Maykop-related ancestry around the Volhynian-Podolian Upland.

Akin to how Yamnaya patrilineal descendants hijacked regional EEF (±CWC) ancestry components mainly through exogamy, dragging them into the different expanding Bell Beaker groups (see below), but kept their Indo-European languages, these hunter-gatherers that admixed with peoples of “Steppe ancestry” were the most likely vector of expansion of Uralic languages in Eastern Europe.

corded-ware-from-trypillia-maykop
PCA of ancient Eurasian samples. Marked likely Proto-Corded Ware samples and potential origin of its PCA cluster based on qpAdm results. See full PCA and more related files.

Baltic Corded Ware

One of the most interesting aspects of the results above is the surprising heterogeneity of the different regional groups, which is also reflected in the Y-DNA variability of early Corded Ware samples.

Seeing how Baltic CWC groups, especially the early Latvia_LN sample, show particularly bad fits with the models above, it seems necessary to test how this population might have come to be. My first impression in 2017 was that they could represent early Corded Ware groups admixed with Yamnaya settlers through their interactions along the Dnieper-Dniester corridor.

However, I recently predicted that the most likely admixture leading to their ancestry and PCA cluster would involve a Corded Ware-like group and a group related to sub-Neolithic cultures of eastern Europe, whose best proxy to date are EHG-like Khvalynsk samples (i.e. excluding the outlier with Pre-Yamnaya ancestry, I0434):

corded-ware-pca-sub-neolithic-europe
Detail of the PCA of the Corded Ware expansion. See full PCA and more related files.

Late Corded Ware + Yamnaya vanguard

Relevant are also the mixtures of Corded Ware from Esperstedt, and particularly those of the sample I0104, which I have repeated many times in this blog I suspected to be influenced by vanguard Yamnaya settlers:

The infeasible models of CWC + Yamnaya_Kalmykia ± Hungary_Baden (see below for Bell Beakers) and the potential cluster formed with other samples from the Baltic suggest that it could represent a more complex set of mixtures with sub-Neolithic populations. On the other hand, its location in Germany, late date (ca. 2500 BC or later), and position in the PCA, together with the good fits obtained for Germany_Beaker as a source, suggest that the increase in Steppe-related ancestry + EEF makes it impossible for the model (as I set it) to directly include Yamnaya_Kalmykia, despite this excess Steppe-related ancestry actually coming from Yamnaya vanguard groups.

I think it is very likely that the future publication of EEF-admixed Yamnaya_Hungary samples (or maybe even Yamnaya vanguard samples) will improve the fits of this model.

These results confirm at least the need to distrust the common interpretation of mixtures including late Corded Ware samples from Esperstedt (giving rise to the “up to 75% Yamnaya ancestry of CWC” in the 2015 papers) as representative of the Corded Ware culture as a whole, and to keep always in mind that an admixture of European BA groups including Corded Ware Esperstedt as a source also includes East BBC-like ancestry, unless proven otherwise.

yamnaya-vanguard-corded-ware-chalcolithic-early
Yamnaya vanguard groups in Corded Ware territory before the expansion of Bell Beakers (ca. 2500 BC). See full map.

Bell Beaker expansion

A hotly (re)debated topic in the past 6 months or so, and for all the wrong reasons, is the origin of the Bell Beaker folk. Archaeology, linguistics, and different Y-chromosome bottlenecks clearly indicate that Bell Beakers were at the origin of the North-West Indo-European expansion in Europe, while the survival of Corded Ware-related groups in north-eastern Europe is clearly related to the expansion of Uralic languages.

NOTE. For the interesting case of Proto-Indo-Iranians expanding with Corded Ware-like ancestry, see more on the formation of Sintashta-Potapovka-Filatovka from East Uralic-speaking Abashevo and Pre-Proto-Indo-Iranian-speaking Poltavka herders. See also more on R1a in Indo-Iranians and on the social complexity of Sintashta.

Nevertheless, every single discarded theory out there seems to keep coming back to life from time to time, and a new wave of interest in “Bell Beaker from the Single Grave culture” somehow got revived in the process, too, because this obsession – unlike the “Bell Beakers from Iberia Chalcolithic” – is apparently acceptable in certain circles, for some reason.

We know that Iberian Beakers, British Beakers, or Sicilian EBA – representing the most likely closest source population of speakers of Proto-Galaico-Lusitanian, Pre-Celtic Indo-European, and Proto-Elymian, respectively – have already been successfully tested for a direct origin among Western European Beakers in Olalde et al. (2018), Olalde et al. (2019), and Fernandes et al. (2019).

This success in ascertaining a closer Beaker source is probably due to the physical isolation of the specific groups (related to Germany_Beaker, Netherlands_Beaker, and NE_Mediterranean_Beaker samples, respectively) after their migration into regions dominated by peoples without Steppe-related ancestry. Furthermore, Celtic-speaking populations expanding with Urnfield south of the Pyrenees also show a good fit with a source close to France_Beaker.

So I decided to test sampled Bell Beaker populations, to see if it could shed light to the most likely source population of individual Beaker groups and the direction of migration within Central Europe, i.e. roughly eastwards or westwards. As it was to be expected for closely related populations (see the relevant discussion here), an attempt to offer a simplistic analysis of direction based on formal stats does not make any sense, because most of the alternative hypotheses cannot be rejected:

Not only because of the similar values obtained, but because it is absurd to take p-values as a measure of anything, especially when most of these conflicting groups with slightly ‘better’ or ‘worse’ p-values represent multiple different mixtures of the type (Yamnaya + EEF) + (Corded Ware + EEF ± Yamnaya), impossible to distinguish without selecting proper, direct ancestral populations…

A further example of how explosive the Bell Beaker expansion was into different territories, and of their extensive local admixture, is shown by the unsuccessful attempt by Olalde et al. (2018) to obtain an origin of the EEF source for all Beaker groups (excluding Iberian Beakers):

bell-beaker-local-population-iberia
Investigating the genetic makeup of Beaker-complex-associated individuals. Testing different populations as a source for the Neolithic ancestry component in Beaker-complex-associated individuals. The table shows P values (* indicates values > 0.05) for the fit of the model: ‘Steppe_EBA + Neolithic/Copper Age’ source population.
burials-yamnaya-hungary
Map of attested Yamnaya pit-grave burials in the Hungarian plains; superimposed in shades of blue are common areas covered by floods before the extensive controls imposed in the 19th century; in orange, cumulative thickness of sand, unfavourable loamy sand layer. Marked are settlements/findings of Boleráz (ca. 3500 BC on), Baden (until ca. 2800 BC), Kostolac (precise dates unknown), and Yamna kurgans (from ca. 3100/3000 BC on).

Now, there is a simpler way to understand what kind of Steppe-related ancestry is proper of Bell Beakers. I tested two simple models for some Beaker groups: Yamnaya + Hungary Baden vs. Corded Ware + GAC Poland. After all, the Bell Beaker folk should prefer a source more closely related to either Yamnaya Hungary or Central European Corded Ware:

Interestingly, models including Yamnaya + Baden show good fits for the most important groups related to North-West Indo-Europeans, including Bell Beakers from Germany, the Netherlands, Italy, and Poland, representing the most likely closest source populations of speakers of Pre-Proto-Celtic, Pre-Proto-Germanic, Proto-Italo-Venetic, and Pre-Proto-Balto-Slavic, respectively.

The admixed Yamnaya samples from Hungary that will hopefully be published soon by the Jena Lab will most likely further improve these fits, especially in combination with intermediate Chalcolithic populations of the Middle and Upper Danube and its tributaries, to a point where there will be an absolute chronological and geographical genomic trail from the fully Yamnaya-like Yamnaya settlers from Hungary to all North-West Indo-European-speaking groups of the Early Bronze Age.

The only difference between groups will be the gradual admixture events of their source Beaker group with local populations on their expansion paths, including peoples of mainly EEF, CWC+EEF, or CWC+EEF+Yamnaya related ancestry. There is ample evidence beyond ancestry models to support this, in particular continued Y-DNA bottlenecks under typical Yamnaya paternal lineages, mainly represented by R1b-L51 subclades.

east-bell-beaker-group-expansion
Distribution of the Bell Beaker East Group, with its regional provinces, as of c. 2400 cal BC (after Heyd et al. 2004, modified). See full maps.

European Early Bronze Age

European EBA groups that might show conflicting results due to multiple admixture events with Corded Ware-related populations are the Únětice culture and the Nordic Late Neolithic.

The results for Únětice groups seem to be in line with what is expected of a Central European EBA population derived from Bell Beakers admixed with surrounding poulations of East Bell Beaker and/or late (Epi-)Corded Ware descent.

Potential models of mixture for Nordic Late Neolithic samples – despite the bad fits due to the lack of direct ancestral CWC and BBC groups from Denmark – seem to be impossible to justify as derived exclusively from Single Grave or (even less) from Battle Axe peoples, supporting immigration waves of Bell Beakers from the south and further admixture events with local groups through maritime domination.

PCA of ancient European samples. Marked are Bronze Age clusters. See full PCAs.

Balkans Bronze Age

The potential origin of the typical Corded Ware Steppe-related ancestry in the social upheaval and population movements of the Dnieper-Dniester forest-steppe corridor during the 4th millennium BC raises the question: how much do Balkan Bronze Age groups owe their ancestry to a population different than the spread of Pre-Yamnaya-like Suvorovo-Novodanilovka chieftains? Furthermore, which Bronze Age groups seem to be more likely derived exclusively from Pre-Yamnaya groups, and which are more likely to be derived from a mixture of Yamnaya and Pre-Yamnaya? Do the formal stats obtained correspond to the expected results for each group?

Since the expansion of hg. I2a-L699 (TMRCA ca. 5500 BC) need not be associated with Yamnaya, some of these values – together with the assessment of each individual archaeological culture – may question their origin in a Yamnaya-related expansion rather than in a Khvalynsk-related one.

NOTE. These are the last ones I was able to test yesterday, and I have not thought these models through, so feel free to propose other source and target groups. In particular, complex movements through the North Pontic area during the Late Eneolithic would suggest that there might have been different Steppe-ancestry-related vs. EEF-related interactions in the north-west and west Pontic area before and during the expansion of Yamnaya.

Mycenaeans

One of the key Indo-European populations that should be derived from Yamnaya to confirm the Steppe hypothesis, together with North-West Indo-Europeans, are Proto-Greeks, who will in turn improve our understanding of the preceding Palaeo-Balkan community. Unfortunately, we only have Mycenaean samples from the Aegean, with slight contributions of Steppe-related ancestry.

Still, analyses with potential source populations for this Steppe ancestry show that the Yamnaya outlier from Bulgaria is a good fit:

The comparison of all results makes it quite evident the why of the good fits from (Srubnaya-related) Bulgaria_MLBA I2163 or of Sintashta_MLBA relative to the only a priori reasonable Yamnaya and Catacomb sources: it is not about some hypothetical shared ancestor in Graeco-Aryan-speaking East Yamnaya– or even Catacomb-Poltavka-related groups, because all available Yamnaya-related peoples are almost indistinguishable from each other (at least with the sampling available today). These results reflect a sizeable contribution of similar EEF-related populations from around the Carpathians in both Steppe-related groups: Corded Ware and Yamnaya settlers from the Balkans.

mycenaeans-minyan-ware-greece-minoan
Cultural groups in and around the Balkans during the Early Bronze Age. See full maps.

qpAdm magic

In hobby ancestry magic, as in magic in general, it is not about getting dubious results out of thin air: misdirection is the key. A magician needs to draw the audience attention to ‘remarkable’ ancestry percentages coupled with ‘great’ (?) p-values that purportedly “prove” what the audience expects to see, distracting everyone from the true interesting aspects, like statistical design, the data used (and its shortcomings), other opposing models, a comparison of values, a proper interpretation…you name it.

I reckon – based on the examples above – that the following problems lie at the core of bad uses of qpAdm:

  1. In the formal aspect, the poor understanding of what p-values and other formal stats obtained actually mean, and – more importantly – what they don’t mean. The simplistic trend to accept results of a few analyses at face value is necessarily wrong, in so far as there is often no proper reasoning of what is being assessed and how, and there is never a previous opinion about what could be expected if the alternative hypotheses were true.
  2. In the interpretation aspect, the poor judgement of accompanying any results with simplistic, superficial, irrelevant, and often plainly wrong archaeological or linguistic data selected a posteriori; the inclusion of some racial or sociopolitical overtones in the mixture to set a propitious mood in the target audience; and a sort of ritualistic theatrics with the main theme of ‘winning’, that is best completed with ad hominems.

If you get rid of all this, the most reasonable interpretation of the output of a model proposed and tested should be similar to Nick Patterson’s words in his explanation of qpWave and qpAdm use:

Here we see that, at least in this analysis there are reasonable models with CordedWareNeolithic is a mix of either WHG or LBKNeolithic and YamnayaEBA. (…) The point of this note is not to give a serious phylogenetic analysis but the results here certainly support a major Steppe contribution to the Corded Ware population, which is entirely concordant with the archaeology [?].

Very far, as you can see, from the childish “Eureka! I proved the source!”-kind of thinking common among hobbyists.

The Mycenaean case is an illustrative example: if the Yamnaya outlier from Bulgaria were not available, and if one were not careful when designing and assessing those mixture models, the interpretation would range from erroneous (viz. a Graeco-Aryan substrate, as I initially thought) to impossible (say, inventing migration waves of Sintashta or Srubnaya peoples into Crete). The models presented above show that a contribution of Yamnaya to Mycenaeans couldn’t be rejected, and this alone should have been enough to accept Yamnaya as the most likely source population of “Steppe ancestry” in Proto-Greeks, pending intermediate samples from the Balkans. In other words, one could actually find that ‘the best’ p-values for source populations of Mycenaeans is a combination of modern Poles + Turks, despite the impracticality of such a model…

I haven’t been able to reproduce results which supposedly showed that Corded Ware is more likely to be derived from (Pre-)Yamnaya than other source population, or that Corded Ware is better suited as the ancestral population of Bell Beakers. The analyses above show values in line with what has been published in recent scientific papers, and what should be expected based on linguistics and archaeology. So I’ll go out on a limb here and say that it’s only through a careful selection of outgroups and samples tested, and of as few compared models as possible, that you could eventually get this kind of results and interpretation, if at all.

Whether that kind of special care for outgroups and samples is about (a) an acceptable fine-tuning of the analyses, (b) a simplistic selection dragged from the first papers published and applied indiscriminately to all models, or (c) cherry picking analyses until results fit the expected outcome, is a question that will become mostly irrelevant when future publications continue to support an origin of the expansion of ancient Indo-European languages in Khvalynsk- and Yamnaya-related migrations.

Feel free to suggest (reasonable) modifications to correct some of these models in the comments. Also, be sure to check out other values such as proportions, SD or SNPs of the different results that I might have not taken into account when assessing ‘good’ or ‘bad’ fits.

Related

On the Ukraine Eneolithic outlier I6561 from Alexandria

sredni-stog-eneolithic-late

Over the past week or so, since the publication of new Corded Ware samples in Narasimhan, Patterson et al. (2019) and after finding out that the R1a-M417 star-like phylogeny may have started ca. 3000 BC, I have been ruminating the relevance of contradictory data about the Ukraine_Eneolithic_o sample from Alexandria, its potential wrong radiocarbon date, and its implications for the Indo-European question.

How many other similar ‘controversial’ samples are there which we haven’t even considered? And what mechanisms are in place to control that the case of Hajji_Firuz_CA I2327 is not repeated?

Ukraine Eneolithic outlier I6561

It was not the first time that I (or many others) have alternatively questioned its subclade or its date, but the contradictory data seem to keep piling up. We can still explain all these discrepancies by assuming that the radiocarbon date is correct – seeing how it is a direct and newly reported lab analysis – because it is an isolated individual from a poorly sampled region, so he may actually be the first one to show features proper of later Corded Ware-related samples.

ukraine-eneolithic-from-caucasus
PCA of ancient Eurasian samples. An interpretation of the evolution of the Pontic-Caspian steppe populations in the Eneolithic. See full PCA.

The individual seems to be especially relevant for the Indo-European and Uralic homeland question. The last one to mention this sample in a publication was Anthony (2019), who considered it in common with two other Eneolithic samples from Dereivka to show how Anatolian farmer-related ancestry first appeared in the recently opened CHG mating network of the Pontic-Caspian steppes and forest-steppes during the Middle Eneolithic, after the expansion of Khvalynsk:

The currently oldest sample with Anatolian Farmer ancestry in the steppes in an individual at Aleksandriya, a Sredni Stog cemetery on the Donets in eastern Ukraine. Sredni Stog has often been discussed as a possible Yamnaya ancestor in Ukraine (Anthony 2007: 239- 254). The single published grave is dated about 4000 BC (4045–3974 calBC/ 5215±20 BP/ PSUAMS-2832) and shows 20% Anatolian Farmer ancestry and 80% Khvalynsk-type steppe ancestry (CHG&EHG). His Y-chromosome haplogroup was R1a-Z93, similar to the later Sintashta culture and to South Asian Indo-Aryans, and he is the earliest known sample to show the genetic adaptation to lactase persistence (I3910-T). Another pre-Yamnaya grave with Anatolian Farmer ancestry was analyzed from the Dnieper valley at Dereivka, dated 3600-3400 BC (grave 73, 3634–3377 calBC/ 4725±25 BP/ UCIAMS-186349). She also had 20% Anatolian Farmer ancestry, but she showed less CHG than Aleksandriya and more Dereivka-1 ancestry, not surprising for a Dnieper valley sample, but also showing that the old fifth-millennium-type EHG/WHG Dnieper ancestry survived into the fourth millennium BC in the Dnieper valley (Mathieson et al. 2018).

The main problem is that this sample has more than one inconsistent, anachronistic data compared to its reported precise radiocarbon date ca. 4045–3974 calBCE (5215±20BP, PSUAMS-2832). I summarized them on Twitter:

  • First known R1a-M417 sample, with subclade R1a-Y26 (Y2-), with formation date and TMRCA ca. 2750 BC (CI 95% ca. 3750–1950 BC), and proper of much later Steppe_MLBA bottlenecks. The closest available sample would be the Poltavka outlier of hg. R1a-Z94 (ca. 2700 BC), from a mixed cemetery that could belong to a later (likely Abashevo) layer; the closest related subclade is probably found in sample I12450 of Butkara_IA (ca. 800 BC).
  • NOTE. The formation date of upper clade R1a-Z93 is estimated ca. 3000 BC, with a CI 95% ca. 3550–2550 BC, suggesting that the actual TMRCA range for the subclade has most likely a lower maximum formation date than estimated with the available samples under Y3.

  • Ancestry and PCA cluster like Steppe_MLBA (see PCA below), different from neighbouring Sredni Stog samples of the roughly coetaneous Dereivka site (ca. 3600-3400 BC), and from a later Yamnaya sample from Dereivka (ca. 2800 BC), even more shifted toward WHG-related ancestry.
  • Allele for lactase persistence (I3910-T), found only much later among Bell Beakers, and still later in Sintashta and Steppe_MLBA samples. This suggests a strong selection in northern Europe and South Asia stemming from steppe-related (and not forest-steppe-related) peoples, postdating the age of massive Indo-European migrations.
  • Hajji Firuz Chalcolithic outlier

    My impression is that the Hajji_Firuz Chalcolithic outlier, initially dated ca. 5900-5500 BC, had much less reason to be questioned than this sample, since Pre-Yamnaya ancestry was (and apparently is still) believed by members of the Reich Lab to have come from south of the Caucasus, and to have arrived around that time or earlier to the North Caspian steppe, i.e. before the 5th millennium BC.

    The formation date of its initially reported haplogroup, R1b-Z2103, is ca. 4100 BC (CI 95% 4800-3500 BC), which seems also roughly compatible with that date and site – at least as compatible as R1a-Y3(xY2) is for ca. 4000 BC -, so it could have been interpreted as a migrant from the South Caspian region, potentially related to Proto-Anatolians, especially before the description of the Caucasus genetic barrier in Wang et al (2018). For some reason, though, the Hajji_Firuz sample was questioned, but this one didn’t even merited an interrogation mark.

    There was already a similar situation with two samples (RISE568 and RISE569) initially reported as belonging to Czech Corded Ware groups, that turned out to be Early Slavs ca. 3,000 years younger, in turn more closely related to Bell Beaker-derived cultures of Central-East Europe. It seems little has changed since that case.

    All in all, my guess is that genomic data of I6561 would have been a priori more compatible with a later period, during the expansion of East Corded Ware groups: at least Middle Dnieper culture, potentially Multi-Cordoned Ware culture, but most likely a Srubnaya-related one, given the most likely SNP mutation and TMRCA date, and the haplogroup variability found in the few samples available from that culture.

    ukraine-eneolithic-from-srubna
    PCA of ancient Eurasian samples. Marked I6561 sample within the cluster formed by Srubnaya samples. See full PCA.

    Compatibility checks

    I tried to start a thread on the possibility that the radiocarbon date was wrong, and IF it were, how likely it would be that formal stats could actually show this, or how could we automatically prevent ancestry magic fiascos.

    In other words: if this guy were a Srubnaya-related individual actually dated e.g. ca. 1700 BC, and someone would try to ‘prove’ – based on the current open source tools alone – that he was the ancestor of expanding peoples of the 4th and 3rd millennium BC (i.e. Balkan outliers, Yamnaya, Corded Ware, you name it), could these results be formally challenged?

    I was hoping for some original brainstorming where people would propose crazy, essentially impossible to understand statistical models, say plotting dozens of well-studied mutations of different geographically related ancient samples with their reported dates, to visually highlight samples that don’t exactly fit with such a feature-based time series analysis; I mean, the kind of theoretical models I wouldn’t even be able to follow after the first two tweets or so. I didn’t receive an answer like that, but still:

    I have nothing to add to these answers, because I agree that all contradictory data are circumstancial.

    The current absolute lack of this kind of validity checks for ancestry models is disappointing, though, and leaves the so-called outliers in a dangerous limbo between “potentially very interesting samples” and “potentially wrongly dated samples”. Radiocarbon date is thus – together with compatibility of population source in terms of archaeological cultures and their potential relationship – a necessary variable to take into account in any statistical design: an error in one of these variables means a catastrophic error in the whole model.

    Formal stats

    For example, in these qpAdm models, I assumed Srubnaya, Ukraine_Eneolithic_outlier, and Bulgaria_MLBA samples were roughly coetaneous and potentially related to the Srubnaya-SabatinovkaNoua cultural horizon, hence stemming from a source close to:

    1. Abashevo-like individuals (whose best proxy to date should be Poltavka_outlier I0432) potentially admixed with Poltavka-like herders; or
    2. Potapovka-like individuals potentially admixed with Catacomb-like peoples (whose best proxy until recently were probably Yamnaya_Kalmykia*).

    *To avoid adding more potential errors by merging different datasets, I have used only proxy samples available in the Reich Lab’s curated dataset of published ancient DNA.

    srubnaya-noua-sabatinovka-mlba
    Srubnaya and Noua-Sabatinovka cultural horizon during the MLBA. See full maps.

    Apart from the lack of more models for comparison (I’m not going to dedicate more time to this), the results can’t be interpreted without a proper sampling and context, either, because (1) Poltavka_o may actually be from a much later group closely related to Srubnaya; (2) Bulgaria_MLBA is only one sample; and (3) there are only two samples from Potapovka; so the models here presented are basically useless, as many similar models that have been tested looking just for a formal “best fit”.

    So feel free to chime in and contribute with ideas as to how to detect in the future whether a sample is ancestral to or derived from others. I will post here informative answers from Twitter, too, if there are any. I don’t think a discussion about the potentially wrong date in this specific sample is very useful, because this seems impossible to prove or disprove at this point. Just what tools or data would you use to at least try and assess whether samples are compatible with its reported date or not – preferably in some kind of automated sieve that takes dozens or hundreds of samples into account.

    On the bright side, there is so much more than formal stats to arrive to relevant inferences about prehistoric populations, their movements and languages. That’s why I6561 didn’t matter for the conclusion by Anthony (2019) that it was the R1b-rich Eneolithic Don-Volga-Caucasus region the most likely Indo-Anatolian and Late Proto-Indo-European homeland, due to the creation of a wide Eneolithic mating network with extended exogamy practices, where Y-chromosome bottlenecks seem to be one of the main genomic data to take into account from the Neolithic to the Middle Bronze Age.

    And that is the same reason why it doesn’t matter that much for the Proto-Indo-European or Uralic question for me, either.

    Related

Uralic speakers formed clines of Corded Ware ancestry with WHG:ANE populations

steppe-forest-tundra-biomes-uralic

The preprint by Jeong et al. (2018) has been published: The genetic history of admixture across inner Eurasia Nature Ecol. Evol. (2019).

Interesting excerpts, referring mainly to Uralic peoples (emphasis mine):

A model-based clustering analysis using ADMIXTURE shows a similar pattern (Fig. 2b and Supplementary Fig. 3). Overall, the proportions of ancestry components associated with Eastern or Western Eurasians are well correlated with longitude in inner Eurasians (Fig. 3). Notable outliers include known historical migrants such as Kalmyks, Nogais and Dungans. The Uralic- and Yeniseian-speaking populations, as well as Russians from multiple locations, derive most of their Eastern Eurasian ancestry from a component most enriched in Nganasans, while Turkic/Mongolic speakers have this component together with another component most enriched in populations from the Russian Far East, such as Ulchi and Nivkh (Supplementary Fig. 3). Turkic/Mongolic speakers comprising the bottom-most cline have a distinct Western Eurasian ancestry profile: they have a high proportion of a component most enriched in Mesolithic Caucasus hunter-gatherers and Neolithic Iranians and frequently harbour another component enriched in present-day South Asians (Supplementary Fig. 4). Based on the PCA and ADMIXTURE results, we heuristically assigned inner Eurasians to three clines: the ‘forest-tundra’ cline includes Russians and all Uralic and Yeniseian speakers; the ‘steppe-forest’ cline includes Turkic- and Mongolic-speaking populations from the Volga and Altai–Sayan regions and Southern Siberia; and the ‘southern steppe’ cline includes the rest of the populations.

eurasian-clines-uralic-altaic
The first two PCs summarizing the genetic structure within 2,077 Eurasian individuals. The two PCs generally mirror geography. PC1 separates western and eastern Eurasian populations, with many inner Eurasians in the middle. PC2 separates eastern Eurasians along the northsouth cline and also separates Europeans from West Asians. Ancient individuals (color-filled shapes), including two Botai individuals, are projected onto PCs calculated from present-day individuals.

For the forest-tundra populations, the Nganasan + Srubnaya model is adequate only for the two Volga region populations, Udmurts and Besermyans (Fig. 5 and Supplementary Table 8).

For the other populations west of the Urals, six from the northeastern corner of Europe are modelled with additional Mesolithic Western European hunter-gatherer (WHG) contribution (8.2–11.4%; Supplementary Table 8), while the rest need both WHG and early Neolithic European farmers (LBK_EN; Supplementary Table 2). Nganasan-related ancestry substantially contributes to their gene pools and cannot be removed from the model without a significant decrease in the model fit (4.1–29.0% contribution; χ2 P ≤ 1.68 × 10−5; Supplementary Table 8).

west-urals-finno-ugrians-qpadm
Supplementary Table 8. QpAdm-based admixture modeling of the forest-tundra cline populations. For the 13 populations west of the Urals, we present a four-way admixture model, Nganasan+Srubnaya+WHG+LBK_EN, or its minimal adequate subset. Modified from the article, to include colors for cultures, and underlined best models for Corded Ware ancestry among Uralians.

NOTE. It doesn’t seem like Hungarians can be easily modelled with Nganasan ancestry, though…

For the 4 populations east of the Urals (Enets, Selkups, Kets and Mansi), for which the above models are not adequate, Nganasan + Srubnaya + AG3 provides a good fit (χ2 P ≥ 0.018; Fig. 5 and Supplementary Table 8). Using early Bronze Age populations from the Baikal Lake region (‘Baikal_EBA’; Supplementary Table 2) as a reference instead of Nganasan, the two-way model of Baikal_EBA + Srubnaya provides a reasonable fit (χ2 P ≥ 0.016; Supplementary Table 8) and the three-way model of Baikal_EBA + Srubnaya + AG3 is adequate but with negative AG3 contribution for Enets and Mansi (χ2 P ≥ 0.460; Supplementary Table 8).

east-urals-ugric-samoyedic-qpadm
Supplementary Table 8. QpAdm-based admixture modeling of the forest-tundra cline populations. For the four populations east of the Urals, we present three admixture models: Baikal_EBA+Srubnaya, Baikal_EBA+Srubnaya+AG3 and Nganasan+Srubnaya+AG3. For each model, we present qpAdm p-value, admixture coefficient estimates and associated 5 cM jackknife standard errors (estimate ± SE). Modified from the article, to include colors for cultures, and underlined best models for Corded Ware ancestry among Uralians.

Bronze/Iron Age populations from Southern Siberia also show a similar ancestry composition with high ANE affinity (Supplementary Table 9). The additional ANE contribution beyond the Nganasan + Srubnaya model suggests a legacy from ANE-ancestry-rich clines before the Late Bronze Age.

bronze-age-iron-age-karasuk-mezhovska-tagar-qpadm
Supplementary Table 9. QpAdm-based admixture modeling of Bronze and Iron Age populations of southern Siberia. For ancieint individuals associated with Karasuk and Tagar cultures, Nganasan+Srubnaya model is insufficient. For all five groups, adding AG3 as the third ancestry or substituting Nganasan with Baikal_EBA with higher ANE affinity provides an adequate model. For each model, we present qpAdm p-value, admixture coefficient estimates and associated 5 cM jackknife standard errors (estimate ± SE). Models with p-value ≥ 0.05 are highlighted in bold face. Modified from the article, to include colors for cultures, and underlined best models for Corded Ware ancestry among Uralians.

Lara M. Cassidy comments the results of the study in A steppe in the right direction (you can read it here):

Even among the earliest available inner Eurasian genomes, east–west connectivity is evident. These, too, form a longitudinal cline, characterized by the easterly increase of a distinct ancestry, labelled Ancient North Eurasian (ANE), lowest in western European hunter-gatherers (WHG) and highest in Palaeolithic Siberians from the Baikal region. Flow-through from this ANE cline is seen in steppe populations until at least the Bronze Age, including the world’s earliest known horse herders — the Botai. However, this is eroded over time by migration from west and east, following agricultural adoption on the continental peripheries (Fig. 1b,c).

Strikingly, Jeong et al. model the modern upper steppe cline as a simple two-way mixture between western Late Bronze Age herders and Northeast Asians (Fig. 1c), with no detectable residue from the older ANE cline. They propose modern steppe peoples were established mainly through migrations post-dating the Bronze Age, a sequence for which has been recently outlined using ancient genomes. In contrast, they confirm a substantial ANE legacy in modern Siberians of the northernmost cline, a pattern mirrored in excesses of WHG ancestry west of the Urals (Fig. 1b). This marks the inhospitable biome as a reservoir for older lineages, an indication that longstanding barriers to latitudinal movement may indeed be at work, reducing the penetrance of gene flows further south along the steppe.

eurasian-clines-uralic-turkic-mongol-altaic
The genomic formation of inner Eurasians. b–d, Depiction of the three main clines of ancestry identified among Inner Eurasians. Sources of admixture for each cline are represented using proxy ancient populations, both sampled and hypothesised, based on the study’s modelling results. The major eastern and western ancestries used to model each cline are shown in bold; the peripheral admixtures that gave rise to these are also shown. Additional contributions to subsections of each cline are marked with dashed lines. b, The northernmost cline, illustrating the legacy of WHG and ANE-related populations. c,d, The upper (c) and lower (d) steppe clines are shown, both of which have substantial eastern contributions related to modern Tungusic speakers. The authors propose these populations are themselves the result of an admixture between groups related to the Nganasan, whose ancestors potentially occupied a wider range, and hunter-gatherers (HGs) from the Amur River Basin. While the upper steppe cline in c can be described as a mixture between this eastern ancestry and western steppe herders, the current model for the southern steppe cline as shown in d is not adequate and is likely confounded by interactions with diverse bordering ancestries. Credit: Ecoregions 2017, Resolve https://ecoregions2017.appspot.com/

Given the findings as reported in the paper, I think it should be much easier to describe different subclines in the “northernmost cline” than in the much more recent “Turkic/Mongolic cline”, which is nevertheless subdivided in this paper in two clines. As an example, there are at least two obvious clines with “Nganasan-related meta-populations” among Uralians, which converge in a common Steppe MLBA (i.e. Corded Ware) ancestry – one with Palaeo-Laplandic peoples, and another one with different Palaeo-Siberian populations:

siberian-clines-uralic-altaic
PCA of ancient and modern Eurasian samples. Ancient Palaeo-Laplandic, Palaeosiberian, and Altai clines drawn, with modern populations labelled. See a version with higher resolution.

The inclusion of certain Eurasian groups (or lack thereof) in the PCA doesn’t help to distinguish these subclines visually, and I guess the tiny “Naganasan-related” ancestral components found in some western populations (e.g. the famous ~5% among Estonians) probably don’t lend themselves easily to further subdivisions. Notice, nevertheless, the different components of the Eastern Eurasian source populations among Finno-Ugrians:

uralic-admixture-qpadm
Characterization of the Western and Eastern Eurasian source ancestries in inner Eurasian populations. [Modified from the paper, includes only Uralic populations]. a, Admixture f3 values are compared for different Eastern Eurasian (Mixe, Nganasan and Ulchi; green) and Western Eurasian references (Srubnaya and Chalcolithic Iranians (Iran_ChL); red). For each target group, darker shades mark more negative f3 values. b, Weights of donor populations in two sources characterizing the main admixture signal (date 1 and PC1) in the GLOBETROTTER analysis. We merged 167 donor populations into 12 groups (top right). Target populations were split into five groups (from top to bottom): Aleuts; the forest-tundra cline populations; the steppe-forest cline populations; the southern steppe cline populations; and ‘others’.

Also remarkable is the lack of comparison of Uralic populations with other neighbouring ones, since the described Uralic-like ancestry of Russians was already known, and is most likely due to the recent acculturation of Uralic-speaking peoples in the cradle of Russians, right before their eastward expansions.

west-eurasian-east-eurasian-ancestry
Supplementary Fig. 4. ADMIXTURE results qualitatively support PCA-based grouping of inner Eurasians into three clines. (A) Most southern steppe cline populations derive a higher proportion of their total Western Eurasian ancestry from a source related to Caucasus, Iran and South Asian populations. (B) Turkic- and Mongolic-speaking populations tend to derive their Eastern Eurasian ancestry more from the Devil’s Gate related one than from Nganasan-related one, while the opposite is true for Uralic- and Yeiseian-speakers. To estimate overall western Eurasian ancestry proportion, we sum up four components in our ADMIXTURE results (K=14), which are the dominant components in Neolithic Anatolians (“Anatolia_N”), Mesolithic western European hunter-gatherers (“WHG”), early Holocene Caucasus hunter-gatherers (“CHG”) and Mala from southern India, respectively. The “West / South Asian ancestry” is a fraction of it, calculated by summing up the last two components. To estimate overall Eastern Eurasian ancestry proportion, we sum up six components, most prevalent in Surui, Chipewyan, Itelmen, Nganasan, Atayal and early Neolithic Russian Far East individuals (“Devil’s Gate”). Eurasians into three clines. (A) Most southern steppe cline populations derive a higher proportion of their total Western Eurasian ancestry from a source related to Caucasus, Iran and South Asian populations. (B) Turkic- and Mongolic-speaking populations tend to derive their Eastern Eurasian ancestry more from the Devil’s Gate related one than from Nganasan-related one, while the opposite is true for Uralic- and Yeiseian-speakers. To estimate overall western Eurasian ancestry proportion, we sum up four components in our ADMIXTURE results (K=14), which are the dominant components in Neolithic Anatolians (“Anatolia_N”), Mesolithic western European hunter-gatherers (“WHG”), early Holocene Caucasus hunter-gatherers (“CHG”) and Mala from southern India, respectively. The “West / South Asian ancestry” is a fraction of it, calculated by summing up the last two components. To estimate overall Eastern Eurasian ancestry proportion, we sum up six components, most prevalent in Surui, Chipewyan, Itelmen, Nganasan, Atayal and early Neolithic Russian Far East individuals (“Devil’s Gate”).

A comparison of Estonians and Finns with Balts, Scandinavians, and Eastern Europeans would have been more informative for the division of the different so-called “Nganasan-like meta-populations”, and to ascertain which one of these ancestral peoples along the ancient WHG:ANE cline could actually be connected (if at all) to the Cis-Urals.

Because, after all, based on linguistics and archaeology, geneticists are not supposed to be looking for populations from the North Asian Arctic region, for “Siberian ancestry”, or for haplogroup N1c – despite previous works by their peers – , but for the Bronze Age Volga-Kama region…

Related