The diverse genetic origins of a Classical period Greek army

 

https://www.pnas.org/doi/10.1073/pnas.2205272119

Here, we analyze genome-wide data from 33 individuals associated with the Battles of Himera and from Himera’s civilian population, as well as 21 individuals from two nearby settlements associated with the indigenous Sicani culture of Sicily, to provide insight into the genetic ancestry of Sicily’s inhabitants in the first millennium BCE and to provide additional data points for evaluating the role of ancient conflict in population interactions in the ancient Mediterranean.

Using qpAdm, this group (Sicily_IA, new data) can be modeled as an admixture of four sources that distantly contributed to the genetic composition of Europeans (P = 0.179): 

Northwestern Anatolian Neolithic farmers (Turkey_Barcin_N; 76.4 ± 1.2%), WHGs (6.4 ± 1.0%), early farmers from Iran (Iran_GanjDareh_N; 6.3 ± 1.5%), and Early Bronze Age (EBA) Steppe herders associated with the Yamnaya cultural complex (Russia_Samara_EBA_Yamnaya; 10.9 ± 1.6%), which indicates an increase of Iranian-related admixture compared with the preceding LBA Sicily group, which can be best modeled without that component. One line of evidence for some local genetic continuity is the almost exclusive presence of Y-chromosomal haplogroup G-Z1903 and its derivates among the males.

Most Himerans associated with the battles can be found clustering on the PCA closely with individuals from the Greece_LBA, consistent with a major contribution of individuals of primarily Greek ancestry in the Himeran forces and substantial genetic continuity between the LBA period in Greece and fifth-century-BCE Greek colonies in Sicily.  Seven of the 16 soldiers of the 480 BCE battle (Sicily_Himera_480BCE_1) and all 5 of the soldiers of the 409 BCE battle (Sicily_Himera_409BCE) are part of this main genetic cluster. Using the qpWave/qpAdm framework, we can model each of the soldiers in these two groups as deriving their ancestry either 100% from a group related to Greece_LBA or from an admixture between a Sicilian LBA or IA source and an Aegean-related source 

One irritating thing about this unsupervised ADMIXTURE run is they label the Euro HG component as WHG and not EHG. This will obviously have funny results when trying to find non-local ancestry, especially from Northeastern Europe or the Steppes, like Baltic_IA showing up as 50% "WHG"

Looks like the non-outlier Himerian soldiers can be modelled as a mix of Aegean_BA + Sicily_LBA/IA, somewhat bolstering the idea that these Himerians were descendants of Greek colonists mixed with the local Sicilians.

Two individuals (I10943/W0396 and I10949/W0403; Sicily_Himera_480BCE_3) fall with modern northeastern European groups and eastern Baltic populations of the first millennium BCE and can be modeled using exclusively BA individuals from Lithuania as a proxy source (P = 0.129). 

 One low-coverage individual, I17870/W0336, falls intermediate between Sicily_Himera_480BCE_2 and Sicily_Himera_480BCE_3 on both PCA and with respect to the main ancestry clusters inferred from ADMIXTURE. 

Two (I10944/W0461 and I10947/W1774; Sicily_Himera_480BCE_4) fall with individuals from IA nomadic contexts in the Eurasian Steppe. In qpAdm, their ancestry is consistent, with around 85–89% deriving from IA Central steppe nomads and 11–15% from an Aegean-like source, an admixture that plausibly could have taken place among the genetically diverse populations of the Steppe. Their mitochondrial DNA (mtDNA) haplogroups suggest east Eurasian genetic roots: A6a, found so far only in modern-day China (70, 71), and N1a1a1a, restricted to Russia, Kazakhstan, and Mongolia. 

Finally, one outlier (I10951/W0653; Sicily_Himera_480BCE_5) falls with modern Caucasus populations and intermediate to ancient Steppe and Caucasus individuals on the PCA and carries the highest proportion of the CHG component in ADMIXTURE (Fig. 2). A single one-way model with a group closely related to Armenia_MBA as the source fit the data (P = 0.293). Similarly, the second low-coverage individual, I17872/W0428, falls closest to populations from the Caucasus on the PCA .

Now, for the outliers.. Won't be surprised if I10944 and I10947 were Sarmatians, they have the typical East Eurasian mtDNA plus slight East Asian DNA (10-15%) in amounts typical of Sarmatians and Cimmerians.

All the soldiers who fall outside the Aegean genetic cluster are interred in mass graves Nos. 1–4 (SI Appendix, Figs. S1–S3), whereas all individuals from mass graves Nos. 5–7 (SI Appendix, Figs. S4–S6) fall within this cluster, a statistically significant difference (P = 0.0028 by a χ2 test with one degree of freedom; Fig. 3). This result mirrors strontium isotope evidence showing more nonlocals interred in mass grave.  Mass graves Nos. 1–4 and Nos. 5–7 also are spatially segregated and differ in size, with mass graves Nos. 1–4 comprising significantly more interments. 

Furthermore, individuals in mass graves Nos. 5 and 6 include grave goods, unlike the other mass graves from 480 BCE (32). The fact that these individuals also fall within the Aegean genetic cluster suggests a link between Aegean ancestry and prestige, as perceived by the individuals responsible for burying the fallen soldiers.

Individuals with “foreign” ancestry, all of whom also are identified as nonlocal on the basis of isotopic evidence (Fig. 3 and Dataset S2), were interred in larger mass graves Nos. 1–4, and individuals with genetic affinities to other Greek populations were interred in the smaller mass graves Nos. 5–7

It looks like the local soldiers, who were of Aegean descent, were given more prestigious burials than the non-local ones. 

 

The Scythians were half Asian

 A few people have been criticizing me for my latest twitter post on Scythians (post anything on these people and you get a bunch of wind, I guess people feel very strongly about being related to them?)

Anyways, the criticism is that early Scythians were supposedly 100% Andronovo, had a Caucasoid/Europoid phenotype and lacked any distinctive East Asian ancestry or the associated phenotype. The people saying so make one important mistake: they conflate Andronovo with Scythian. While it's true that the latter originally came from the former, it's also true that they're widely separated in space and time with true Scythian cultures only emerging in the 1st millenium BCE. It must also be kept in mind the Andronovo horizon was massive, spanning most of Central Asia and reaching the boundaries of modern day Mongolia on the East. Given so, it's impossible for the populations in the horizon to be homogenous in ancestry. 

Scythians can certainly be Andronovo descendants, but not all Andronovo descendants are Scythian, nor did all even survive till the 1st millenium BCE (late Iron Age onwards). So we must ask ourselves where the Scythian homeland is, and what were the genetics of people found nearby in the Iron Age. 

Animal style art is a distinctive feature of Scythian and Scythian-like cultures (Cimmerians, Sarmatians). The earliest such art is from the Arzhan kurgans (Arzhan 0, Arzhan 1) in modern-day Russia, right next to Mongolia and Kazakasthan.

Here is where the earliest Scythian kurgan mounds are from, by the way, nestled well on the easternmost edges of the Eurasian steppes. The Arzhan 0 kurgan is in fact a princely tomb, with identical construction to Arzhan 0 and of course, filled with animal style decorative art, and dated to the 9th century BCE by accelerator mass spectrometry. Read more about it here

The Arzhan culture is succeeded by the Pazyrk culture, another early-Scythian one dated to the 6th-3rd centuries BCE. Regardless to say, these are some of the oldest sites where the material culture archaeologists consider "Scythian" arose, and they're deep inside Asia, right next to a bunch of Siberian people where the Andronovo descendants could've picked up a ton of EA ancestry, of course. 

Kuzmina discusses this in her magnum opus in a bit of detail, though keep in mind she lacked any genetic data and relied purely on archaeolinguistics and anthropology. I'll quote the relevant part.

He suggested that the impulse from Central Asia was of crucial importance at the end of the Bronze Age. This impulse brought to Scythia the arrows, daggers, knives, and horse-gear of the Karasuk culture. The reason in support of this hypothesis was offered by M. P. Gryaznov’s (1980) discovery of the Arzhan barrow in Tuva. There, early articles of horse equipment along with the objects of Scythian art have been discovered. It has been suggested that the complex dated to the 8th century BC which is a considerably early date. Scholars from St. Petersburg assume that the Scythian culture formed in Central Asia in the 8th or perhaps in the 9th century BC. They even admit the probability of the influence of Chinese art on the formation of the animal style. The Central-Asian hypothesis, which has gained wide acceptance, is represented in the works of N. A. Bokovenko (Moshkova 1992)

 Kuzmina and 20th century Russian archaeologists did not have the power of aDNA, but we do. So let's see how actual Scythian men and women from these earliest sites plot on a PCA. 

Huh, what a surprise! every single early Scythian sample clusters with modern day Bashkirs. [click the image so it magnifies in your window, otherwise it will be blurry).

The oldest Scythian samples are Zevakinko_IA and Aldy Bel_IA (this one is from the famous Arzhan kurgans, by the way). How about checking them out on G25?

Distance to: RUS_Tuva_Aldy_Bel_IA

0.04341559 Bashkir
0.04764879 Tatar_Siberian
0.07581902 Tatar_Siberian_Zabolotniye
0.08711644 Nogai
0.09365923 Uzbek
0.10087071 Yukagir_Forest
0.10372093 Hazara_Afghanistan
0.10404151 Tubalar
0.10465838 Karakalpak

As expected. 

Distance to: RUS_Zevakino_Chilikta_IA

0.04804137 Bashkir
0.05933492 Uzbek
0.06077576 Tatar_Siberian
0.07615525 Nogai
0.07823705 Tatar_Crimean_steppe
0.07973288 Hazara_Afghanistan
0.08347611 Uygur
0.08489468 Turkmen_Uzbekistan
0.08967604 Hazara
0.09028372 Turkmen

Barely surprising to anyone who's been following the latest studies on this. How about checking out one of the later descendant cultures, such as Pazyryk? 

Distance to: KAZ_Pazyryk_IA

0.04310245 Tatar_Siberian
0.05177868 Bashkir
0.07341034 Nogai
0.07795478 Tatar_Siberian_Zabolotniye
0.08562940 Uzbek
0.08595528 Tubalar

Here's how the Pazyrk Scythians depicted themselves. 

One last thing, how much East Asian do they have exactly?

Target: RUS_Zevakino_Chilikta_IA

Distance: 2.7621% / 0.02762117
45.6 MNG_Khovsgol_BA
39.2 RUS_Krasnoyarsk_MLBA
15.2 TKM_Gonur1_BA

Target: KAZ_Pazyryk_IA

Distance: 1.4174% / 0.01417422
53.2 MNG_Khovsgol_BA
39.4 RUS_Krasnoyarsk_MLBA
7.4 TKM_Gonur1_BA

Target: RUS_Tuva_Aldy_Bel_IA

Distance: 2.6900% / 0.02690034
50.4 MNG_Khovsgol_BA
47.4 RUS_Krasnoyarsk_MLBA
2.2 TKM_Gonur1_BA 

Yep, I'm hardly surprised. They were basically ancient hapas, tongue-in-cheek. How do Bashkirs look like? This model will probably be a poorer fit on them, as we need more proximal sources, but let us see. 

Target: Bashkir

Distance: 3.7975% / 0.03797549
47.8 RUS_Krasnoyarsk_MLBA
46.8 MNG_Khovsgol_BA
5.4 TKM_Gonur1_BA

Very similar. Some Turanian (BMAC) and the rest being Andronovo + Khovsgol (East Asian proxy). This was also the conclusion Ruscone et al. came to. 

Our findings shed new light onto the debate about the origins of the Scythian cultures. We do not find support for a western Pontic-Caspian steppe origin, which is, in fact, highly questioned by more recent historical/ archeological work (1, 2). The Kazakh Steppe origin hypothesis finds instead a better correspondence with our results

So that's settled. The Scythians were a nearly 50:50 product of Steppe Iranics and East Asians with some additional Turanian throw in.