What does modern genetics actually tell us about Bhumihar ancestry? Discussions online often reduce the subject to a single question — how much “Steppe DNA” do Bhumihars have? — but the real genetic history is considerably more interesting.
Bhumihar populations, like other populations of northern India, descend from several ancient ancestry streams that mixed over thousands of years. These include ancestry related to early South Asian hunter-gatherers, populations associated with the Indus or Indus-periphery genetic cline, and Bronze Age pastoralists from the Eurasian Steppe.
Importantly, Bhumihars were actually included in the population modelling used in one of the most influential ancient-DNA studies of South Asia. That gives us something far more useful than speculation based on surnames, caste histories or consumer DNA calculators.
The quick answer: what does Bhumihar DNA look like?
The available peer-reviewed evidence places Bhumihars firmly within the wider genetic landscape of northern South Asia. Their ancestry is not derived from one ancient population. Instead, it reflects several prehistoric ancestry streams that were already mixing long before modern caste identities took their present form.
In a specific analysis published in the supplementary material of Narasimhan et al., the estimated Central Steppe MLBA-related proportion was about 0.28 for Bhumihar_Bihar and 0.25 for Bhumihar_UP. [2]
That immediately requires an important qualification. A population estimate of 28% does not mean that every person who identifies as Bhumihar possesses precisely 28% Steppe ancestry. Population-genetic estimates depend on the people sampled, the ancient reference populations selected, and the statistical model used.
What does the genetics of a caste actually mean?
Caste is primarily a historical and social identity. It is not a biological species, subspecies or race.
Geneticists can nevertheless detect differences between many South Asian communities because marriage within particular groups became common over long periods of time. This practice of marrying predominantly within a population is called endogamy.
If a community remains relatively endogamous for many generations, frequencies of genetic variants can gradually diverge from those in neighbouring communities. Researchers can then identify statistical clusters or differences in average ancestry proportions.
This distinction is especially important when discussing South Asian population genetics because scientific findings are sometimes pulled into debates about caste identity that genetics itself cannot resolve.
The three major ancient ancestry streams
Older discussions of Indian genetics often divided ancestry into ANI, or Ancestral North Indian, and ASI, or Ancestral South Indian. Ancient DNA has since allowed researchers to reconstruct much more of what lay behind those statistical categories.
For a simplified explanation of the ancestry of many present-day northern South Asian populations, three ancient sources are particularly useful:
| Ancient ancestry | What it represents | Importance in South Asia |
|---|---|---|
| AASI-related | Deep ancestry connected to hunter-gatherer populations that lived in South Asia before later West Eurasian-related population movements. | Found in varying proportions throughout the Indian subcontinent. |
| Indus / Iranian-related | Ancestry related to populations connected with the Indus-region genetic cline, which themselves carried ancestry related to ancient populations east of the Zagros. | Became a major ancestry source for later South Asian populations. |
| Steppe MLBA-related | Ancestry associated with Middle-to-Late Bronze Age pastoralists of the Eurasian Steppe. | Entered South Asia later and contributed particularly strongly to many northern populations. |
What is AASI ancestry?
AASI stands for Ancient Ancestral South Indian. Despite the name, it should not be understood as meaning simply “modern South Indian”.
AASI is a population-genetics concept representing ancestry related to ancient hunter-gatherers who lived in South Asia before the formation of many later populations.
People throughout the subcontinent carry AASI-related ancestry. Northern Indians possess it, southern Indians possess it, and many tribal and caste populations possess it in differing proportions.
Researchers have historically had to model this ancestry indirectly because an ideal, completely unmixed ancient AASI reference genome has not been available in the way that ancient genomes are available for several Eurasian populations.
This is one reason percentages produced by different ancestry models should not be treated as absolute measurements.
Iranian-related ancestry does not mean recent Iranian ancestry
One of the easiest aspects of Indian population genetics to misunderstand is the expression “Iranian-related ancestry”.
It does not mean that a large fraction of Bhumihar or Indian ancestry derives from modern Iranians.
Rather, ancient DNA indicates that populations associated with the Indus region possessed ancestry related to very old populations connected to the eastern Iranian plateau, alongside ancestry deeply rooted in South Asia.
The ancient genome analysed from the Indus Valley Civilisation-associated site of Rakhigarhi was particularly important. Shinde and colleagues reported that the individual lacked the Steppe pastoralist ancestry that became widespread in parts of South Asia later. [3]
The study also argued against a simple model in which agriculture in the Indus region resulted merely from a mass migration of western Iranian farmers.
ANI and ASI: the older model of Indian ancestry
A landmark 2009 study led by David Reich showed that much of the genetic variation among Indian populations could be modelled using two hypothetical ancestral populations known as Ancestral North Indians (ANI) and Ancestral South Indians (ASI). [4]
ANI showed greater genetic affinity to West Eurasian populations. ASI represented ancestry more deeply rooted within South Asia.
It is important to remember that ANI and ASI were never excavated tribes or known historical communities. They were statistical constructs inferred from the DNA of living populations.
Ancient DNA later revealed a more detailed picture.
ANI can broadly be understood as having formed from an Indus-related population that subsequently received substantial ancestry from Steppe pastoralists. ASI formed largely through mixture between Indus-related ancestry and populations carrying considerably more AASI-related ancestry. [1]
ShaluSharma.com has previously discussed the older ANI/ASI framework in its article Races and People of India . Ancient-DNA research now allows that older framework to be understood in much finer detail.
What exactly is Steppe DNA?
The Eurasian Steppe is the enormous grassland zone stretching across much of Eurasia.
During the Bronze Age, mobile pastoralist societies expanded across these regions. For South Asian population history, ancestry related to Middle and Late Bronze Age Steppe populations is particularly important.
These populations are often discussed in connection with archaeological horizons such as Sintashta and Andronovo.
Steppe populations themselves were not genetically “pure”. They were products of earlier mixtures involving several ancient Eurasian populations. Their descendants subsequently moved over very large distances.
Ancient DNA indicates that Steppe-related ancestry entered South Asia through Central Asia during the second millennium BCE, after the mature urban phase of the Indus Valley Civilisation. [1]
The incoming ancestry did not simply replace South Asian populations. Instead, Steppe-descended groups mixed with people already living in the subcontinent.
Is Steppe DNA the same thing as European DNA?
No.
Bronze Age Steppe populations contributed ancestry to several later populations, including populations in Europe, Central Asia and South Asia.
As a result, some modern Europeans and South Asians share ancestry that ultimately traces back to related prehistoric Steppe populations. That does not make a modern Indian carrying Steppe ancestry genetically “European”.
The relevant migrations occurred thousands of years before modern European and South Asian national identities existed.
How much Steppe ancestry do Bhumihars have?
This is where Bhumihar population genetics becomes particularly interesting.
Narasimhan and more than one hundred collaborators published The Formation of Human Populations in South and Central Asia in Science in 2019. [1]
Their work combined ancient genomes from South and Central Asia with genome-wide information from modern South Asian populations.
Bhumihar populations from Bihar and Uttar Pradesh were explicitly included in the modelling.
| Population in study | Central Steppe MLBA-related ancestry | Steppe / Indus-related ratio | Modern Indian Cline Z-score |
|---|---|---|---|
| Bhumihar_Bihar | 28% | 0.56 | -7.0 |
| Bhumihar_UP | 25% | 0.48 | -5.0 |
These values come from Table S86 of the study's supplementary material. [2]
The Bihar Bhumihar population was particularly notable because it showed one of the strongest deviations towards a greater amount of Steppe pastoralist-related ancestry relative to Iranian/Indus-related ancestry than predicted by the researchers' general Modern Indian Cline model.
In the same table, groups with similarly negative Z-scores included several Brahmin populations and Sikh Jatts. [2]
How should these ancestry percentages be interpreted?
Ancient ancestry percentages are statistical estimates rather than direct measurements in the everyday sense.
Researchers frequently use techniques such as qpAdm to test whether the ancestry of a target population can be modelled as a mixture of selected ancient reference populations.
The reference populations are proxies. They represent populations genetically similar to ancestral sources rather than necessarily being the exact population from which every later individual descended.
Change the reference populations, the individuals included or the underlying statistical model and the numerical estimate can shift.
There is also normal variation between individuals. Two unrelated Bhumihar individuals need not have identical ancestry proportions.
Recent family history, regional origin and marriage outside the community can create further variation.
Why do online ancestry calculators produce different numbers?
Hobbyist ancestry calculators and commercial DNA services frequently use different reference populations from those used in academic papers.
One calculator might use categories labelled “Steppe”, “Sintashta”, “Zagros”, “AASI” and “Indus Valley”. Another might combine some of those components.
These percentages cannot automatically be compared with the 25–28% estimates in Narasimhan et al.
Are Bhumihars genetically close to Brahmin populations?
There is evidence of genetic affinity between the Bhumihar populations analysed and some Brahmin populations, although this must be interpreted carefully.
One relatively early Bhumihar-specific study was published by Richa Ashma and V. K. Kashyap in 2003. The researchers examined 15 autosomal microsatellite markers among four caste populations in Bihar: Brahmin, Bhumihar, Rajput and Kayasth. [6]
Their analysis found substantial genetic similarity between these populations, with the Bhumihar sample clustering relatively closely with the sampled Bihar Brahmins.
That finding is interesting, but it should not be overstated.
Fifteen microsatellite markers contain far less information than today's genome-wide datasets containing hundreds of thousands or millions of genetic variants. The study also predates the ancient-DNA revolution.
The 2019 ancient-DNA analysis provides another interesting observation. Several Brahmin populations and the Bhumihar populations appeared among groups with relatively elevated Steppe-to-Indus-related ancestry ratios. [2]
Does this prove that Bhumihars are Brahmins?
No.
A genetic test cannot determine ritual status or settle historical disputes about caste classification.
Two communities may be genetically similar because they separated from a common ancestral population, because they exchanged genes before strict endogamy developed, or because both received similar proportions of much older ancestry.
Questions surrounding Bhumihar social history therefore require evidence from history, anthropology, genealogy and social practice as well as genetics.
Endogamy and the formation of modern Indian genetic structure
Ancient Indian populations were not always as genetically separated as many later caste communities became.
A major study by Moorjani and colleagues found evidence for widespread mixture between populations related to ANI and ASI approximately 1,900 to 4,200 years ago. After that period, gene flow between many communities declined substantially. [5]
This transition towards stronger endogamy helps explain a striking feature of present-day South Asian genetics: neighbouring communities can display detectable differences despite living in the same geographical region for centuries.
Basu and colleagues similarly reconstructed substantial population structure within India and identified the long-term influence of endogamy. [10]
Founder effects
Endogamy can create what geneticists call a founder effect.
Imagine a comparatively small founding population in which an uncommon genetic variant is present. If the group expands while most marriages remain within that population, the variant can become much more frequent than it is in surrounding communities.
Nakatsuka and colleagues analysed more than 260 South Asian groups and identified dozens of populations with founder effects stronger than those observed in Finns and Ashkenazi Jews, two populations already well known in medical genetics for founder-associated diseases. [7]
However, it would be inappropriate to make specific claims about Bhumihar disease risks without sufficiently large Bhumihar-specific genomic and clinical studies.
Bhumihar Y-DNA, R1a and R1a-Z93
Another popular question is whether Bhumihars carry the Y-chromosome haplogroup R1a.
The first thing to understand is that the Y chromosome represents only one line in a person's family tree: the uninterrupted paternal line from father to son.
It therefore provides different information from autosomal DNA.
R1a is a widespread Eurasian Y-chromosome haplogroup with multiple branches. Of particular importance to South Asian population history is R1a-Z93, a branch widespread across parts of Central and South Asia.
Ancient-DNA research has identified R1a-Z93 lineages among Bronze Age Steppe populations associated with the migrations relevant to later Indo-Iranian population history.
A genetic chronology of the Indian subcontinent by Silva and colleagues found that the distribution and estimated expansion dates of R1a-Z93-related lineages were consistent with significant Bronze Age male-mediated movements into South Asia. [8]
Does R1a mean a person is 100% Steppe?
Absolutely not.
A Y chromosome represents one paternal lineage. Autosomal DNA represents ancestry inherited from ancestors throughout the family tree.
A man can carry an R1a-Z93 Y chromosome while the majority of his total ancestry comes from non-Steppe ancestral populations.
Likewise, another person can possess considerable autosomal Steppe ancestry without carrying R1a at all.
Published Bhumihar-specific Y-chromosome datasets remain much more limited than would be necessary to define a single reliable “Bhumihar Y-DNA profile”. Bhumihars should not be assumed to possess one universal paternal haplogroup.
Was Steppe migration into South Asia male-biased?
This is one of the more fascinating findings to emerge from modern South Asian genetics.
Several analyses have found evidence that Steppe-related ancestry entered South Asian populations disproportionately through men.
Narasimhan and colleagues observed that Steppe ancestry in modern South Asians was enriched on the Y chromosome relative to autosomal ancestry, supporting a model of male-biased migration. [1]
Silva and colleagues similarly identified heavily sex-biased dispersals in the genetic chronology of South Asia. [8]
This provides a plausible explanation for how a paternal lineage associated with Steppe populations could become relatively frequent even where Steppe-related ancestry represents only a minority of the total genome.
What about Bhumihar mitochondrial DNA?
Mitochondrial DNA, usually shortened to mtDNA, is inherited predominantly through the maternal line.
Just as a Y chromosome tells us about one paternal line, mitochondrial DNA tells us about one maternal line.
South Asian mtDNA diversity contains many lineages with very deep roots in the subcontinent. This differs from the clearer evidence for later migration found among some Y-chromosome lineages.
The contrast between paternal and maternal ancestry is one of the pieces of evidence supporting sex-biased migration during parts of South Asian prehistory.
However, there is no single “Bhumihar mitochondrial haplogroup”. Large populations contain many maternal and paternal lineages.
Do Bhumihars have Indus Valley Civilisation ancestry?
In the broad population-genetic sense, the answer is yes: substantial Bhumihar ancestry can be traced to populations genetically related to the Indus-region ancestry cline.
The Indus Valley Civilisation flourished across large areas of present-day Pakistan and north-western India during the third millennium BCE.
Ancient-DNA research indicates that populations connected with this civilisation carried mixtures of ancestry ultimately related to ancient Iranian-related populations and deeply rooted South Asian populations. [3]
Later, some populations carrying this ancestry mixed with Steppe-derived groups. Others mixed more extensively with populations containing greater proportions of AASI-related ancestry.
The resulting combinations contributed heavily to the formation of later South Asian populations.
Were Bhumihars directly descended from Steppe pastoralists?
Only partly.
Describing Bhumihars simply as descendants of Steppe people would ignore most of their ancestry.
Even using the relatively high estimates in the 2019 model, around three-quarters of the ancestry of the sampled Bhumihar populations was not modelled as Central Steppe MLBA-related.
The larger picture is therefore one of mixture.
Bhumihar population history includes ancestry rooted deeply in South Asia, ancestry associated with the Indus-region genetic cline, and later Steppe-related ancestry, followed by further demographic history in northern India.
Did Steppe ancestry arrive with Indo-European languages?
Genetic data cannot tell us the language spoken by a particular ancient individual. DNA does not contain vocabulary or grammar.
Researchers can, however, compare the timing and direction of population movements reconstructed from ancient DNA with archaeological and linguistic evidence.
Narasimhan et al. argued that the combined genetic, archaeological and linguistic evidence supports the movement of Indo-Iranian languages into South Asia in association with Bronze Age populations carrying Steppe ancestry. [1]
The important distinction is that genes do not themselves equal languages. Population movement can help spread a language, but language adoption can also occur without complete population replacement.
Why older Indian DNA studies sometimes reached different conclusions
Anyone researching Indian genetics will eventually encounter older studies whose conclusions appear to differ from modern ancient-DNA research.
For example, Sahoo and colleagues published an influential Y-chromosome study in 2006 arguing that a large recent paternal contribution from Central Asia was not necessary to explain much of Indian Y-chromosome diversity. [9]
That should not be read simply as one set of scientists being “wrong”.
Population genetics changed dramatically when researchers became able to obtain usable DNA from ancient skeletons.
Earlier researchers were usually trying to reconstruct ancient migration using DNA from people living thousands of years later.
Ancient DNA allowed researchers to examine people who lived before, during and after prehistoric population movements. Improvements in Y-chromosome resolution also allowed broad haplogroups such as R1a to be broken into increasingly precise sub-branches.
This is why modern ancient-DNA studies generally provide a much more detailed reconstruction of Bronze Age population movements.
Can genetics prove Bhumihar caste identity?
No.
A genome does not contain a label saying Bhumihar, Brahmin, Rajput, Kayasth or any other caste.
Caste identities developed through combinations of social history, marriage networks, occupation, political organisation, ritual practices, geography and cultural transmission.
Long periods of endogamy can subsequently create genetic structure correlated with some of those identities.
That distinction matters.
Is there a Bhumihar gene?
No.
There is no single mutation that makes a person Bhumihar.
When population geneticists distinguish between communities, they are examining statistical differences across large numbers of variants throughout the genome. Most of those variants are also found in numerous other populations.
The differences concern frequencies and combinations rather than unique caste-specific genes.
Can a DNA company identify somebody as Bhumihar?
Ordinary commercial ancestry tests are unlikely to provide a dependable “Bhumihar percentage”.
With sufficiently large reference datasets, researchers can sometimes distinguish closely related South Asian communities statistically because endogamy has created detectable population structure.
But this is very different from reliably looking at an unknown genome and assigning social identity to the person.
Close neighbouring populations share the overwhelming majority of their genetic ancestry.
The broader genetic history of Bihar and eastern Uttar Pradesh
Bhumihar history is also inseparable from the history of the Gangetic plain, particularly Bihar and eastern Uttar Pradesh.
This region has experienced population movement for thousands of years and later became one of the most important political and cultural centres of ancient India.
The population history reconstructed by genetics therefore predates the emergence of Bihar, Uttar Pradesh or modern caste categories by millennia.
Readers interested in the wider culture and history of the region may also enjoy ShaluSharma.com's guides to the Bhojpuri language , Maithili language of Bihar and Nepal and Golghar and the history of Patna .
What makes Bhumihar genetics interesting?
The significance of the Bhumihar data is not that Bhumihars possess one mysterious or unique ancestry.
Rather, the population provides a particularly useful example of how the complex demographic history of northern India can still be detected in modern genomes.
One ancestry stream reaches deep into the prehistoric hunter-gatherer populations of South Asia.
Another derives from populations carrying ancestry related to ancient groups east of the Zagros that became an important part of the Indus civilisation's genetic landscape.
Another arrived through Central Asia with descendants of Bronze Age Steppe pastoralists.
These populations mixed. Their descendants continued moving and mixing across South Asia. Much later, marriage became increasingly restricted within particular communities, helping to create the genetic structure detectable today.
Bhumihar genetics therefore makes little sense if reduced to a debate about whether the population is “Steppe”, “Indus”, “AASI” or “ANI”.
Frequently asked questions about Bhumihar DNA
How much Steppe DNA do Bhumihars have?
In Table S86 of the supplementary material accompanying Narasimhan et al. (2019), the Bhumihar_Bihar population was estimated at approximately 28% Central Steppe MLBA-related ancestry, while Bhumihar_UP was estimated at approximately 25%. These figures apply to the populations and statistical model used in the study rather than every individual Bhumihar.
Do Bhumihars have high Steppe ancestry?
Relative to many South Asian populations included in the same analysis, the sampled Bhumihar groups displayed a substantial Steppe-related signal. Bhumihar_Bihar was also one of the strongest negative outliers in the study's Steppe-to-Indus-related ancestry analysis.
Are Bhumihars genetically Brahmins?
Some genetic studies show similarities between sampled Bhumihar and Brahmin populations, but genetics cannot determine Brahmin social or ritual identity. A 2003 microsatellite study from Bihar found the sampled Bhumihar group genetically close to sampled Bihar Brahmins.
Do Bhumihars have AASI ancestry?
Yes. Like other North Indian populations, Bhumihars carry ancestry ultimately related to ancient South Asian hunter-gatherer populations. AASI-related ancestry is found throughout the Indian subcontinent in differing proportions.
Do Bhumihars have Indus Valley ancestry?
Population-genetic models indicate substantial ancestry related to populations belonging to the Indus-region genetic cline. This should not be interpreted as proof that any particular modern family descends directly from residents of one specific Harappan city.
Do Bhumihars carry R1a?
R1a-Z93 is an important paternal lineage in the wider genetic history of northern South Asia and Bronze Age Steppe migration. However, there is no single universal Bhumihar Y-chromosome haplogroup, and detailed Bhumihar-specific haplogroup frequency data remain limited.
Does R1a prove Steppe ancestry?
R1a-Z93 can provide evidence about one paternal lineage, but a Y chromosome represents only one branch of a person's ancestry. It cannot be used by itself to calculate someone's total autosomal Steppe ancestry.
Are Bhumihars genetically European?
No. Bhumihars are a South Asian population. Some prehistoric ancestry is shared with ancient Steppe populations that also contributed ancestry to parts of Europe, but those migrations occurred thousands of years before modern European identities developed.
Is there a Bhumihar gene?
No. No single genetic variant defines Bhumihar identity. Population-genetic differences arise from statistical patterns across many genetic variants, ancestry proportions, founder effects and long-term endogamy.
Can a DNA test tell whether someone is Bhumihar?
Not reliably from an ordinary consumer DNA test. Large research datasets can sometimes distinguish population-level patterns, but caste identity cannot be determined from one unique DNA marker.
Scientific references and further reading
The ancestry estimates and historical discussion in this article are based primarily on peer-reviewed population-genetics and ancient-DNA research rather than commercial DNA tests or social-media ancestry calculators.
Read the full paper at PubMed Central
View the Science supplementary material
View the paper
View the paper
View the paper
View on PubMed
View the paper
Read at PubMed Central
View the paper
View the paper