Before farming reached Europe, the continent had been home for tens of thousands of years to people who hunted, fished and gathered — and who were, genetically, far from uniform. When ancient-DNA models talk about "hunter-gatherer ancestry" they mean a small family of distinct populations, labelled by the region their genomes were sampled from. This guide introduces the three that matter most for anyone with West Eurasian ancestry, and shows how to measure your share of each.
The three abbreviations#
WHG — Western Hunter-Gatherers. The Mesolithic population of western and central Europe from roughly 12,000 to 6,000 BC, best known from individuals such as Loschbour in Luxembourg and La Braña in Spain. Their genomes are strikingly homogeneous across the continent and quite distinct from everyone who came after. The source-population page is Western Hunter-Gatherers, 12,000–6,000 BC.
EHG — Eastern Hunter-Gatherers. The Mesolithic population of the forest-steppe and forest zone of eastern Europe — samples from Karelia and the Samara region are the classic ones. EHG carried a substantial share of ancestry related to the Upper Palaeolithic Siberian population known as Ancient North Eurasians, which sets them apart from WHG.
CHG — Caucasus Hunter-Gatherers. A Late Upper Palaeolithic and Mesolithic population from the southern Caucasus, first described from the Satsurblia and Kotias Klde genomes in Georgia (Jones et al., 2015). Genetically close to the early Neolithic populations of Iran, and important because CHG- or Iran-related ancestry forms roughly half of the Western Steppe Herder profile.
There are others — Scandinavian hunter-gatherers (a WHG–EHG mixture), the Baltic groups, the Iron Gates population of the Danube — but WHG, EHG and CHG are the sources most models are built from, and the ones a report will name.
How hunter-gatherer ancestry survived#
When Anatolian farmers spread across Europe from about 6500 BC, they largely replaced the WHG populations they met — but not completely. Ancient genomes from the Middle Neolithic show a resurgence of hunter-gatherer ancestry: farmer communities across Europe carried more WHG ancestry two thousand years after the transition than at its start, the result of continued mixing at the margins (Lipson et al., 2017; Mathieson et al., 2018). Modern Europeans therefore carry WHG ancestry along two routes: directly, and folded inside the already-mixed Neolithic farmer populations of later eras.
EHG ancestry reached most of Europe differently — as roughly half of the Yamnaya profile that spread with the steppe migrations after 3000 BC. The same applies to CHG. This matters for measurement, because a model that offers both "Eastern Hunter-Gatherer" and "Western Steppe Herder" as sources is offering two populations that share ancestry, and the weights will trade off against each other.
What a hunter-gatherer percentage measures#
As with any admixture weight, a hunter-gatherer share is a property of a model, not a fixed fact about a person. It depends on which hunter-gatherer population is used as the proxy, which other sources are offered alongside, which outgroups the model is tested against, and how much of your file survives the merge. In the earliest-era models, WHG is typically the smallest of the three classic components for most Europeans — commonly in the single digits to low teens — which makes it the one most in need of a z-score: a 7% weight with |Z| below 2 is not a measurable contribution, whatever the pie chart shows.
Regionally, published work places direct WHG-related ancestry highest around the Baltic and in parts of the Balkans and Iberia's north, and lowest in the Mediterranean south and in Sardinia. The figures are indicative, not lookups; the analysis exists to measure your file.
Testing it with qpAdm#
Formal modelling is the method the papers above used. In our qpAdm analysis the Hunter-Gatherer & Neolithic Farmer era is built precisely to resolve this structure: Western Hunter-Gatherers, Anatolian Neolithic farmers and Western Steppe Herders as the classic three-way model, with Eastern Hunter-Gatherers and other curated sources available where a file supports a finer split. The report prints every weight with its standard error and z-score, the model's p-value and the complete right set — and, in its Reading view, the full model record with each weight's 95% confidence interval and the nested-model table, downloadable as plain text — so a small hunter-gatherer weight can be judged rather than believed. How to read qpAdm results explains the reading; the buyer's guide explains the product. The file the analysis needs is the raw export you already have; the free Raw DNA File Check will tell you whether it is dense enough.
Testing it with Global25#
With a coordinate row, the same structure can be fitted in the free admixture calculator against curated early-era panels, or in the paid Global25 analysis, which prints every source offered to the calculator. The distance lens is also instructive here: a modern European row sits a long way from every Mesolithic individual in G25 space, because everyone alive today is a mixture the hunter-gatherers were not part of. Seeing Loschbour or La Braña at the top of a distance list is a sign the panel is wrong, not a discovery. qpAdm vs Global25 covers when each method is the right instrument.
Where the individuals come from#
Every WHG, EHG and CHG genome in the reference panel is a published individual from an excavated burial, curated in the Allen Ancient DNA Resource — see AADR explained. You can browse them on the free Ancient Sample Atlas: filter by date to the Mesolithic and the map shows where each sampled hunter-gatherer was found. And if you want to know whether you share actual stretches of genome with any of them, that is a different analysis — segment matching, identity by state rather than descent — explained in Ancient DNA matches explained.
A worked reading#
Suppose a report prints, for the earliest era:
p-value: 0.31
Anatolian_Neolithic_Farmer 0.58 SE 0.03 Z 19.2
Western_Steppe_Herder 0.34 SE 0.03 Z 11.4
Western_Hunter_Gatherer 0.08 SE 0.02 Z 3.6
The model is admissible, every standard error is tight, and the hunter-gatherer source — small as it is — sits more than three standard errors from zero. That 8% is a measurable contribution. Now imagine the same weights with SE 0.05 on the last line: |Z| falls to 1.6, and the honest reading becomes "a hunter-gatherer component is not resolvable in this file", with the two-source nested model the one to report. The weight did not change; what changed is whether it can be believed.
What the result is not#
A hunter-gatherer share is not a claim that Loschbour, or anyone in a Mesolithic cemetery, was related to you. It is a statement that your genome is compatible with drawing that share of ancestry from a population like theirs, measured against specific alternatives. Nor does it say anything about temperament, diet or appearance — the pigmentation genetics of WHG individuals is a real research finding about them, not an inheritance you can read off a percentage.
The companions to this guide are Steppe ancestry and Neolithic farmer ancestry; the terms are in the glossary.
References#
- Lazaridis, I. et al. (2014). Ancient human genomes suggest three ancestral populations for present-day Europeans. Nature, 513, 409–413.
- Jones, E. R. et al. (2015). Upper Palaeolithic genomes reveal deep roots of modern Eurasians. Nature Communications, 6, 8912.
- Fu, Q. et al. (2016). The genetic history of Ice Age Europe. Nature, 534, 200–205.
- Lipson, M. et al. (2017). Parallel palaeogenomic transects reveal complex genetic history of early European farmers. Nature, 551, 368–372.
- Mathieson, I. et al. (2018). The genomic history of southeastern Europe. Nature, 555, 197–203.
- Posth, C. et al. (2023). Palaeogenomics of Upper Palaeolithic to Neolithic European hunter-gatherers. Nature, 615, 117–126.
