Who Fought for the Greeks at Himera? Ancient DNA Reveals Mercenaries From the Balkans, Northern Europe, the Steppe and the Caucasus

Hero image: AI generated illustration of the battle of Himera in 480 BC, with Greek hoplites fighting beside foreign mercenaries, among them a steppe horse archer, a northern European warrior and a Balkan light infantryman, and the Carthaginian camp burning on the shore. Armour, clothing and weapons are approximations, and the scene is not a reconstruction of any documented moment. The figures do not portray the individuals whose genomes are discussed here.

Greek historians wrote the story of their wars as a story of Greeks. Herodotus listed the many peoples who fought for Carthage: Phoenicians, Libyans, Iberians, Ligurians, Sardinians and Corsicans. About the Greek side, the sources say much less. They mention citizens, allies and, in passing, hired soldiers.

At Himera, the dead of two battles have now spoken for themselves. Their teeth record where they grew up, and their genomes record where their families came from. Together, they show that the army that saved Greek Sicily in 480 BC was a far more international force than any ancient author admitted.

A Greek city between two worlds

Himera was founded around 648 BC on the north coast of Sicily by Greek settlers, among the indigenous Sicani and close to the Phoenician and Carthaginian towns of the west of the island. It was the westernmost Greek city on that coast, and in 480 BC it became the battlefield of a war between the Greek tyrants of Sicily and Carthage.

That year, the Carthaginian general Hamilcar landed with a large army to restore the deposed ruler of Himera. Gelon, tyrant of Syracuse, and Theron, tyrant of Akragas, came to the city's rescue and destroyed the Carthaginian force; Hamilcar died in the battle. Herodotus records a Sicilian tradition that it was fought on the same day as the Greek naval victory over Persia at Salamis, a coincidence modern historians do not accept.

Seventy years later, Hamilcar's grandson Hannibal Mago returned. A relief force from Syracuse arrived, fought and then withdrew, and in 409 BC Himera fell. The city was destroyed and never rebuilt.

The graves of two battles

From 2008, excavations ahead of a railway extension, directed by archaeologist Stefano Vassallo, uncovered more than 10,000 graves in the West Necropolis of Himera. Among them were mass graves of soldiers: seven attributed to the battle of 480 BC and one to the battle of 409 BC, holding 132 adult men in all. Several skeletons still carried the weapons that killed them, including a spear in the chest of one man and a spearhead in the abdomen of another.

In 2021, Katherine Reinberger, Laurie Reitsema, Britney Kyle and colleagues analysed strontium and oxygen isotopes in the teeth of 62 of these soldiers, which record where a person lived as a child. About two thirds of the 480 BC soldiers had grown up outside the area of Himera, against about a quarter of the 409 BC soldiers, and only 9% by the strontium measure alone. The authors concluded that the 480 BC army included many men from elsewhere, consistent with allies and mercenaries, while the 409 BC army was mostly local.

Seven Aegeans and nine outsiders

In 2022, Reitsema, Alissa Mittnik, Britney Kyle, David Caramelli, Ron Pinhasi, David Reich and colleagues published genomes of 54 individuals from Sicily in PNAS: 16 soldiers from the 480 BC battle, 5 from the 409 BC battle, 12 civilians from Himera, and 21 indigenous Sicilians from Polizzello and Monte Falcone, spanning the 8th to the 5th centuries BC.

The 480 BC soldiers split clearly. Seven belonged to a group related to Bronze Age Greece. The other nine did not. Two had a mix of central and eastern Mediterranean and central European ancestry and carried the Y-chromosome lineage E-V13, common in the Balkans. Two were closest to Bronze Age populations of the eastern Baltic. Two carried 85 to 89% ancestry related to Iron Age nomads of the central Eurasian steppe, with the East Asian maternal lineages A6a and N1a1a1a. Two were related to the Caucasus, and one low coverage genome fell between groups.

The isotopes matched the genomes exactly. All nine outsiders had grown up outside the area of Himera, while five of the seven Aegean soldiers were local. The graves matched too. All nine outsiders lay in mass graves 1 to 4, where the dead averaged about 30 years of age; all seven Aegean soldiers lay in mass graves 5 to 7, where they averaged 45 and where some graves contained grave goods. The 409 BC soldiers, by contrast, all belonged to the Aegean group, and all five had strontium values consistent with a local childhood.

The authors interpret the outsiders as mercenaries, paid soldiers recruited from far beyond Sicily. The ancient sources support the idea: Diodorus reports that Gelon of Syracuse employed many foreign mercenaries, and that about 7,000 of them were granted citizenship in 466 BC. In 409 BC, by contrast, Himera fought its final battle largely with its own citizens.

Our test: placing every soldier

Eighteen of the soldiers are in the public Global25 files: 14 from the 480 BC battle and 4 from the 409 BC battle, with coverage between 50 and 74% (one 409 BC genome has only 16%). The two low coverage outsiders are missing. We checked scale compatibility: the average of the seven Aegean soldiers lies only 4 units from the published average for the same group (distances multiplied by 1,000), and our average of 19 Iron Age Sicani from Polizzello lies 2.5 units from the published Sicani average. A single genome of this quality lies about 25 to 30 units from the average of its own population, which sets the scale for what counts as close.

We first placed the soldiers among 24 populations of the Bronze and Iron Ages, from Iberia and Carthage to the Baltic, the steppe and the Caucasus.

The picture is immediate. The seven Aegean soldiers of 480 BC and the four soldiers of 409 BC sit together beside Mycenaean and Classical Greece. The two E-V13 men sit among the Iron Age populations of the northern Balkans. The two north-eastern men sit at the other end of the map, beside Bronze Age Poland and Lithuania. The two steppe men sit beside the Early Sarmatians of the southern Urals, and the Caucasus man beside Early Iron Age Georgia.

Soldier by soldier

For each soldier, we then searched all Bronze and Iron Age populations in the dataset for the closest match, and added the Y-chromosome lineage reported in the 2022 dataset.

The Aegean soldiers are closest to Mycenaean Greece (22 to 34 units) and to Archaic and Classical Greece. Their paternal lineages are those of the ancient eastern Mediterranean: J2a, G2a, T1a and L2, with one R1b lineage from a rarer branch than the one common in western Europe. Their closest present-day populations are in southern Italy, Sicily and Greece, including the Greeks of the Mani peninsula, but at 38 to 62 units, far less close than to the Bronze Age Greeks.

The two E-V13 men are closest to the Early Iron Age people of Kalakaca and Gomolava, in Serbia (25 and 29 units), and to other Iron Age groups of the Danube and the western Balkans. Their Y-chromosome lineage, E-V13, is associated with Bronze and Iron Age populations of the Carpathian basin and the Balkans. They probably came from the northern Balkans or the Danube region, more than 900 km from Himera, perhaps through the Greek cities of the Adriatic or the northern Aegean.

The two north-eastern men are the most surprising. They are closest to the Bronze Age Trzciniec culture of Poland (34 to 36 units) and to Iron Age individuals with a north-eastern profile from the Scythian world of present-day Ukraine. One carried R1a-Y35, a branch of R1a-Z280 common today in eastern Europe; the other carried I2a. Their nearest present-day populations are Lithuanians and Russians of the Pskov region. They came from somewhere in the forest and forest-steppe zone of north-eastern Europe, about 2,000 km away. The Greek colonies of the northern Black Sea, such as Olbia, traded deep into that region and are a plausible route.

The two steppe men are closest to Iron Age steppe nomads: the Early Sarmatians of the southern Urals, and Sarmatian and Scythian groups of the Volga and Don. These were the Iranian-speaking horsemen whom the Greeks called Scythians, and further east Sauromatians. Their paternal lineages, R1a-Z2124, a branch of R1a-Z93 typical of these nomads, and N, together with their East Asian maternal lineages, fit that origin. No present-day population lies close to them; the nearest, the Tatars of the Volga, are more than 60 units away.

The Caucasus man is closest to the Early Iron Age people of Georgia and to later groups of the North Caucasus (30 to 35 units), and his nearest present-day relatives are mountain Georgians of Tusheti and Khevsureti. Unusually, he carried a lineage of R1a, more common on the steppe than in the Caucasus.

A six-source model gives the same answer in percentages. The Aegean soldiers come out at 90 to 100% Aegean or Sicani ancestry. The north-eastern men come out at 93 to 95% north-eastern European, the steppe men at 91 to 92% steppe and the Caucasus man at 92% Caucasus. The 2022 study modelled the steppe men as 11 to 15% Aegean; our test gives 8 to 9%, inside the range that our spike-in tests can return for a true value of 15%. Global25 cannot say whether this small share is real or a by-product of the model. The two E-V13 men come out as a mix of northern Balkan and Mediterranean ancestry, but this split is unreliable, as explained below.

The soldiers of 409 BC

The four soldiers of 409 BC with Global25 data all fall in the Aegean cluster, among the 480 BC Greeks. Their closest ancient matches are Mycenaean and Classical Greece and Iron Age Thrace. Their Y-chromosome lineages are J2a, T1a and R1b-L2, a branch of R1b-U152 common today in Italy. One 409 BC soldier carried the same T1a2b lineage as one of the Aegean soldiers of 480 BC. Nothing in their genomes points beyond the Greek and Sicilian world.

Three traps in the data

Trap 1: Greek or Sicilian? Global25 cannot tell

The six-source model splits several soldiers between Aegean and Sicani ancestry, and in two 409 BC soldiers it puts most of the weight on the Sicani. It is tempting to read this as Greek settlers marrying into the local population, and the 2022 study did find signs of such mixing among Himera's civilians. But the Aegean and Sicani sources are only 50 units apart in Global25, about the noise of a single genome.

Our spike-in tests show the problem. A soldier with no Sicani ancestry at all can come back with up to 47% Sicani, and a true 50% can come back anywhere between 8 and 85%. The outsiders, on the other hand, are read reliably: a foreign-born soldier with no Mediterranean ancestry comes back at no more than 7 to 10%. Global25 can tell a Greek from a Sarmatian. It cannot tell a Greek from a Sicilian.

Trap 2: the northern Balkans overlap with Greece

Iron Age populations of the northern Balkans lie only about 70 units from Mycenaean Greece. In the six-source model, a Greek genome with no Balkan ancestry can return up to 27% foreign ancestry, mostly in the Balkan block, and a Balkan genome can return a large Mediterranean share. For the two E-V13 men, the reliable evidence is their distance to Iron Age Serbia and their Y-chromosome lineage, not the exact percentages.

Trap 3: labels and lineages

The Global25 files label the two steppe men "Sarmatian". Strictly, the Sarmatians appear in written sources only later; in 480 BC, the steppe north of the Black Sea belonged to the Scythians, with the Sauromatians further east. The genomes say steppe nomad, not a specific tribe. The Y-chromosome data need similar care: the reproduction of the 2022 dataset that we used assigns R1a-Z2124 and N to the two steppe men in one order, and another public summary assigns them in the opposite order, so we do not attribute them to individuals. And present-day populations closest to a soldier are not his descendants: they are simply the modern groups whose average lies nearest, after 2,500 years of further migration.

What this means

The Greek victory at Himera was won by an army that was only partly Greek. Alongside citizens of the Sicilian Greek cities fought young men from the Danube and the western Balkans, from the forests of north-eastern Europe, from the steppe beyond the Black Sea and from the Caucasus. They had grown up far away, they died young, and they were buried apart from the older local Greeks. Seventy years later, when Himera fought its last battle, the men who died were its own.

Ancient historians rarely named the hired soldiers who fought in Greek armies. Their genomes, and the chemistry of their teeth, now give them back at least a place of origin.

Global25 coordinates used in this article

The six model sources and the 18 soldiers with Global25 data. Labels give the battle, the genetic group and the sample number. They can be pasted directly into the Vahaduo tool.

Source_Greece_Mycenaean,0.116782,0.162180,-0.008183,-0.060950,0.022835,-0.026411,-0.001058,-0.003323,0.005502,0.043609,0.007827,0.012664,-0.014628,0.000702,-0.020643,-0.007611,0.014603,0.001723,0.009666,-0.004115,-0.006813,0.004390,-0.004498,0.000374,-0.002048
Source_Sicily_Sicani_Polizzello,0.125864,0.169861,0.015382,-0.056627,0.047701,-0.025849,-0.003525,-0.001348,0.026889,0.063121,0.002675,0.011824,-0.023950,-0.004498,-0.022387,-0.009623,0.007006,0.001667,0.005974,-0.010485,-0.006705,0.003117,-0.004255,-0.007807,-0.000189
Source_Serbia_EIA_Kalakaca_Gomolava,0.129290,0.152628,0.027840,-0.011134,0.033327,-0.007710,0.001216,0.000814,0.008193,0.024645,0.001471,0.006083,-0.014289,-0.000283,-0.010323,-0.003783,0.005522,0.001110,0.007719,-0.003252,-0.007083,0.005106,0.000739,0.005621,-0.001684
Source_Poland_MBA_Trzciniec,0.122170,0.119155,0.079258,0.090978,0.038623,0.032398,0.005405,0.009500,-0.004193,-0.030282,0.001299,-0.009891,0.014891,0.007133,0.003212,0.008950,0.007801,-0.003294,0.001089,0.006294,0.000249,-0.002926,0.005094,-0.005663,-0.001617
Source_Russia_EarlySarmatian_SouthernUrals,0.110978,0.042398,0.036864,0.070414,-0.021234,0.026704,0.000294,0.000692,-0.024032,-0.036356,-0.005805,-0.002511,0.000074,-0.017272,0.026432,0.015049,-0.016331,0.003420,-0.002106,-0.001970,-0.013476,-0.001391,-0.002095,0.005211,0.001826
Source_Georgia_EIA_Treli,0.108587,0.131816,-0.052797,-0.028101,-0.039269,-0.000893,0.006627,-0.006046,-0.059394,-0.024347,-0.001624,0.008033,-0.018285,0.001762,0.006352,-0.015195,0.000704,-0.005473,-0.003645,0.007629,0.007212,0.001805,-0.000419,-0.006121,0.003113
Himera_480BC_Aegean_I10945,0.114961,0.154360,-0.014708,-0.054264,0.017849,-0.027889,-0.002350,-0.002308,0.003886,0.041368,0.001137,0.016635,-0.021853,0.003165,-0.017236,-0.019358,0.005998,-0.000633,0.015712,-0.014757,-0.021836,0.013107,0.001232,0.008555,-0.001557
Himera_480BC_Aegean_I10948,0.118376,0.165531,-0.002263,-0.059432,0.030159,-0.022311,-0.003760,-0.000231,0.013499,0.049204,0.006333,0.010940,-0.025124,-0.005230,-0.018865,-0.003845,0.004824,0.001394,0.011564,-0.006628,-0.002995,0.002844,0.006162,0.008194,-0.001676
Himera_480BC_Aegean_I10952,0.114961,0.160454,0.004903,-0.060078,0.027390,-0.020638,-0.007050,-0.002538,0.008181,0.045012,0.006171,0.007493,-0.011298,0.003716,-0.017915,-0.013789,-0.005867,-0.001140,0.006411,-0.012631,0.000250,0.006925,-0.004314,0.006266,-0.005628
Himera_480BC_Aegean_I7217,0.121791,0.157407,-0.009428,-0.063954,0.032621,-0.027331,0.003055,-0.002538,0.004704,0.047928,0.007145,0.010341,-0.016501,-0.009083,-0.018458,-0.012729,0.007041,0.002787,0.005405,-0.022136,-0.004367,0.007296,-0.001602,0.012893,0.003473
Himera_480BC_Aegean_I7218,0.117238,0.161469,-0.001131,-0.062662,0.025235,-0.031794,0.004465,-0.000462,0.012271,0.051755,0.006983,0.008842,-0.012785,-0.008945,-0.021987,0.006099,0.033639,0.000127,0.009176,-0.007253,-0.011729,-0.003215,-0.005669,0.008555,-0.010777
Himera_480BC_Aegean_I7219,0.125205,0.173656,-0.001508,-0.068153,0.024312,-0.022032,0.001175,0.001846,0.004500,0.043554,0.001786,0.010790,-0.024975,-0.010046,-0.027008,-0.012066,0.018645,-0.002027,0.013198,0.000375,-0.013975,-0.000618,0.002835,-0.004338,-0.001916
Himera_480BC_Aegean_I7221,0.117238,0.157407,-0.012822,-0.063631,0.018157,-0.025379,-0.005170,-0.004846,-0.004500,0.035901,0.003085,0.004946,-0.016055,-0.000550,-0.018051,-0.008221,0.012386,-0.000887,0.002514,-0.001376,-0.011105,0.006554,-0.003081,0.004217,-0.005987
Himera_480BC_NorthBalkan_I10946,0.127482,0.147252,0.031678,-0.005814,0.036314,-0.007530,-0.006580,-0.004846,0.009408,0.031345,0.011042,0.004646,-0.013825,0.004542,-0.014522,-0.011138,-0.005607,-0.003421,0.010056,-0.010630,-0.011729,0.013849,-0.003081,0.001566,0.008023
Himera_480BC_NorthBalkan_I10950,0.125205,0.143190,0.027153,-0.010013,0.040007,-0.009482,0.003055,0.002769,0.006545,0.028429,0.000162,0.009142,-0.014123,0.007019,-0.021580,-0.010872,0.016820,-0.000633,0.009679,0.000875,-0.008235,0.006801,-0.003328,-0.000482,-0.001557
Himera_480BC_NEEurope_I10943,0.126344,0.122879,0.077687,0.080104,0.030467,0.028447,0.010810,0.014538,-0.015953,-0.040456,0.001137,-0.006145,0.019475,0.014450,-0.009636,0.014320,0.025946,-0.005954,-0.007542,0.011255,0.000749,-0.006677,0.005916,-0.012050,-0.003353
Himera_480BC_NEEurope_I10949,0.124067,0.112724,0.078441,0.075259,0.038469,0.031515,0.012926,0.007154,-0.001636,-0.024784,-0.000162,-0.010041,0.011447,0.022845,-0.010586,0.007690,0.002347,0.002407,0.004022,0.011380,-0.000873,-0.010758,0.006532,-0.015665,0.005508
Himera_480BC_Steppe_I10944,0.109270,0.043668,0.042238,0.057817,-0.019696,0.011435,0.001175,0.000692,-0.026384,-0.029704,-0.003735,0.001798,-0.000743,-0.002752,0.014929,-0.001193,-0.021253,0.003421,0.001006,-0.002876,-0.015348,0.002720,0.001972,0.001566,0.000958
Himera_480BC_Steppe_I10947,0.108132,0.042652,0.040352,0.052003,-0.025543,0.024821,-0.007520,-0.002308,-0.024747,-0.024784,-0.006983,-0.001349,-0.003122,-0.021744,0.015336,0.017237,0.010040,-0.004181,0.010684,0.003001,-0.015847,-0.002844,0.003204,0.007109,0.008263
Himera_480BC_Caucasus_I10951,0.112685,0.121864,-0.045254,-0.019057,-0.036930,0.007251,0.002350,-0.007384,-0.062380,-0.036812,0.000162,0.011989,-0.028097,0.005643,0.015065,-0.012463,0.007823,-0.004054,-0.019609,0.012631,0.003244,-0.000742,0.001725,-0.012532,-0.006107
Himera_409BC_I7223,0.117238,0.158423,0.012822,-0.043282,0.037545,-0.022032,-0.001175,0.000000,0.017180,0.048839,0.001299,0.006894,-0.020961,-0.003165,-0.008279,-0.019491,-0.003390,0.004687,0.010559,-0.010130,-0.006489,0.003833,-0.005546,0.002651,0.002515
Himera_409BC_I7224,0.126344,0.163500,0.003017,-0.044251,0.025543,-0.021196,-0.006580,-0.007615,0.019225,0.045923,-0.000162,0.010191,-0.021110,-0.004542,-0.009908,0.000663,0.025816,-0.007348,-0.000503,-0.003502,-0.008111,0.001113,0.006409,0.009399,-0.007065
Himera_409BC_I7225,0.134311,0.161469,0.005280,-0.061047,0.041546,-0.019243,-0.001410,-0.004384,0.003068,0.047928,0.005684,0.012889,-0.029881,-0.000688,-0.020629,-0.020286,0.002086,-0.002787,0.004777,-0.013506,0.004367,0.002349,-0.002835,0.002048,0.006826
Himera_409BC_I17866,0.120652,0.165531,-0.005280,-0.068799,0.013233,-0.022590,-0.004700,-0.003461,0.014726,0.044466,0.001137,0.016635,-0.029583,0.002615,-0.008415,-0.006364,0.004433,0.012922,0.006913,0.004127,0.000624,0.013478,0.002835,0.004820,-0.008143

References

1. Ancient DNA Reitsema LJ, Mittnik A, Kyle B, Catalano G, Fabbri PF, Kazmi ACS, et al., Caramelli D, Pinhasi R, Reich D (2022). The diverse genetic origins of a Classical period Greek army. PNAS 119(41): e2205272119. doi.org/10.1073/pnas.2205272119
2. Isotopes Reinberger KL, Reitsema LJ, Kyle B, Vassallo S, Kamenov G, Krigbaum J (2021). Isotopic evidence for geographic heterogeneity in Ancient Greek military forces. PLOS ONE 16(5): e0248803. doi.org/10.1371/journal.pone.0248803
3. Press release University of Tubingen and University of Georgia (2022). The diverse genetic origins of a Classical period Greek army. idw-online.de
4. Ancient source Herodotus, Histories 7.165 to 167, on the Carthaginian army and the battle of Himera.
5. Ancient source Diodorus Siculus, Library of History 11.20 to 26, 11.49, 11.72 and 13.59 to 62, on the battles of 480 and 409 BC and Gelon's mercenaries.
6. Data Global25 PCA coordinates by Davidski, and the Moriopoulos 2026 collection of Global25 coordinates.