Steppe ancestry in South Asia

The peoples who brought Indo-European languages to South Asia — the Indo-Aryans of Vedic tradition and their Iranian cousins — were not aolithic population. They were pastoralists from the Pontic-Caspian Steppe, carryi carried a distinctive genetic signature that ancient DNA has now traced from the grasslands of modern Ukraine to the mountains of the Hindu Kush and the plains of the Ganges.^[1]

The migration

The story begins with the Yamnaya culture, a semi-nomadic pastoral people who expanded across the Pontic-Caspian Steppe beginning around 3300 BCE. Their ancestry — a mixture of Eastern Hunter-Gatherers (EHG) from the forests of northern Eurasia and Caucasus Hunter-Gatherers from the south — became one of the most consequential genetic signatures in human history.^[1]^[7]

From the Yamnaya descended the Sintashta culture (c. 2100–1800 BCE), which appeared on the southern Ural Steppe and produced the earliest known spoke-wheeled chariots. The Sintashta people are widely identified with the early Indo-Iranians: their burial rites — horse sacrifice, fire rituals, weapons — closely match descriptions in the later Vedic and Avestan texts. Every Sintashta male tested to date has carried the Y-chromosome haplogroup R1a-Z93, the lineage that would become the signature paternal marker of the Indo-Iranian world.^[2]^[7]

The Andronovo cultural horizon, which grew out of Sintashta and spread across the Central Asian steppe from roughly 2000 to 900 BCE, carried this ancestry further south and east. One stream moved onto the Iranian plateau, bringing Iranic languages. Another crossed the mountains into the Indian subcontinent, carrying Indo-Aryan languages and the R1a-Z93 lineage into the Swat Valley and beyond.^[1]^[2]

Ancient DNA evidence

The landmark study by Narasimhan et al. (2019), published in Science, sequenced 523 ancient genomes from across Eurasia and established the genetic framework for South Asian prehistory. The findings revealed three ancestry clines that succeeded each other in time:^[1]

Before 2000 BCE, the Indus Periphery Cline — people with varying proportions of Iranian farmer-related and Ancient Ancestral South Indian (AASI) ancestry — formed the primary population of the Indus Valley Civilization. This cline had no detectable Steppe ancestry.^[1]

After 2000 BCE, the Steppe Cline appeared, represented by individuals from the Swat Valley time transect (c. 1400 BCE onward). These people were a mixture of Indus Periphery-related populations and Central_Steppe_MLBA groups carrying approximately 41% Steppe-derived ancestry. The Steppe ancestry in South Asia has the same genetic profile as that found in Bronze Age Eastern Europe, tracking a movement of people that affected both regions.^[1]

The Modern Indian Cline, which emerged after the mixing of these populations, is represented today in diverse South Asian groups. Modern South Asians can be modelled as a mixture of three source populations: Iranian farmer-related ancestry, AASI, and Central_Steppe_MLBA — the Steppe pastoralists who arrived in the second millennium BCE.^[1]

The study estimated that up to 30% of the ancestry of some modern South Asian groups derives from these Steppe pastoralists. The proportion is highest in the northwest and declines toward the south, creating a gradient that mirrors the direction of the original migration.^[1]

The paternal lineage: R1a-Z93

The Y-chromosome haplogroup R1a tells the story of the migration in paternal terms. The parent clade R1a-M417 diversified into two great branches roughly 5,800 years ago: R1a-Z282, which stayed in Europe and rose to dominate the eastern part of the continent, and R1a-Z93, which turned east and south to become the signature lineage of the Indo-Iranian world.^[2]

The R1a-Z93 branch further subdivided into subclades that preserve the split between the two arms of the Indo-Iranian world. The Z2123/Z2125 sub-branches are found at high frequency among Pashtuns, Kyrgyz, and the descendants of the ancient Scythians and Sarmatians. The L657 sub-branch is the largest clade in India. When a Pashtun and a Bengali Brahmin both test R1a-Z93, they are reading the same Bronze Age chapter from two different pages.^[2]^[7]

Among modern populations, R1a-Z93 reaches its highest frequencies in the northwest of the subcontinent. Pashtun groups such as the Yusufzai carry it at approximately 80%, while North Indian Brahmins reach around 67%. The frequency declines southward, falling to roughly 12% among Tamil speakers — a pattern consistent with a migration that entered from the northwest and diluted as it spread.^[2]^[6]

The Steppe ancestry, however, is disproportionately concentrated in the paternal line relative to the autosomal genome. Among Pashtuns, the Y-chromosome R1a-Z93 frequency (~80%) is roughly double the genome-wide Steppe ancestry (~38%). This pattern — a Y-chromosome that outruns the genome — is consistent with a male-biased Bronze Age expansion, where Steppe men mixed with local women as they spread.^[6]

Pashtuns and the Steppe signal

The Pashtuns carry among the highest proportions of Steppe ancestry in South Asia, making them a key population for understanding the genetic legacy of the Indo-Iranian migrations.^[3]^[4]

Y-chromosome studies of Afghan populations have consistently found high frequencies of R1a among Pashtuns. Haber et al. (2012) reported R1a1a-M17 at 51% in a Pashtun sample from the Afghan Hindu Kush. Di Cristofaro et al. (2013) found 56% across two Pashtun samples from Baghlan and Kunduz. These figures are among the highest recorded for any South Asian population.^[3]^[4]

Autosomal analyses confirm the picture. The HGDP Pashtun samples, analysed through the HarappaWorld admixture calculator, show an "NE Euro" component of approximately 11% — a value that tracks the Steppe pastoralist signal. When combined with the "Caucasian" component (approximately 20%), which represents a related but distinct West Eurasian ancestry, the total West Eurasian contribution to the Pashtun genome is substantial. The remaining proportions — "Baloch" (approximately 34%, representing Neolithic farmer ancestry from West Asia) and "S Indian" (approximately 19%, representing the indigenous South Asian substrate) — complete the picture of a population shaped by multiple waves of migration.^[5]

The gradient is telling. Populations closer to the Indian heartland — such as the UP Muslims and Pathans of Uttar Pradesh — show higher S Indian and Baloch proportions and lower Caucasian and NE Euro proportions than Afghan Pashtuns. The trend is consistent with the direction of the Steppe migration: higher Steppe ancestry in the northwest, declining toward the south and east.^[5]

The Steppe signal, however, is only one layer of Afghan paternal ancestry. The older layer — haplogroup L and its sub-branches — arrived with Neolithic farmers from West Asia thousands of years before the Steppe pastoralists. Together, R1a-Z93 and L-M357 account for the majority of Afghan paternal lineages.^[5]

The Caucasus barrier

The Caucasus Mountains have long been recognised as a genetic barrier between Europe and West Asia. North of the mountains, populations such as the Lezgins, Chechens, and Adyghe show elevated levels of Steppe-related ancestry — the "NE Euro" component in admixture analyses. South of the mountains, Georgians and Abkhazians carry less Steppe ancestry and more of a distinct Caucasus-related component.^[5]

This pattern is consistent with the mountains impeding the spread of Steppe ancestry southward. The north Caucasus populations, exposed to migrations from the steppe, absorbed more of the Steppe genetic signal. The south Caucasus populations, shielded by the mountain barrier, preserved a more ancient genetic profile. The observation does not mean that Caucasus populations are the source of the Steppe component — rather, they are recipients of it, reflecting the broader east-west cline of Steppe ancestry across Eurasia.^[5]