ERA 03
Britain Before Farming
13,000 - 3,200 years ago
From the first people returning to Britain as the ice retreated to the arrival of farming that would change the island for good, this era follows the return of human life to a landscape long empty of it.
This era covers the return of the first people to Britain as the ice retreated, the loss of Doggerland, and the rise of farming that would carry new ancestral lines across a transforming Europe and, at last, over the water into Britain itself.
In This Era
09
Britain’s Earliest People
Caves, Cousins and Colonisation
As the ice retreated, Britain remained joined to continental Europe by dry land, allowing people to move back into a country from which humans had been absent for many thousands of years. None of these earliest people can yet be placed on the particular ancestral paths traced back from Ursula, Xenia, Oisin or Wodan, although some would prove to belong to neighbouring branches of the same wider genetic families. What follows is therefore the story of the country itself: repeatedly occupied and recolonised by different peoples, and of the landscape into which those later family lines would eventually arrive.
The earliest genomes recovered from Britain reveal just how varied those arrivals were. Rather than a single population spreading steadily across the country, at least two genetically distinct groups reached Britain within little more than a thousand years of one another.
The first is represented by a woman who lived at Gough’s Cave in Somerset around 15,000 years ago. Her ancestry belonged to the older population associated with Magdalenian Europe, closely related to people living in Belgium and Spain at around the same time. Her people hunted animals including red deer, wild cattle and horse. The cave preserves a darker part of their world as well. Human bones were cut, broken and deliberately modified, and some skulls were shaped into cups, evidence interpreted as part of complex mortuary practices that included cannibalism.
Little more than a thousand years later, a very different population had reached Britain. At Kendrick’s Cave on Great Orme in North Wales, a man who lived around 13,500 years ago carried ancestry related to the Villabruna population then spreading through western Europe. His people’s diet, rich in marine and freshwater foods, and his burial at a site associated with a decorated horse jaw and pierced tooth ornaments rather than butchered animal remains, point to a genuinely different way of life from the one practised at Gough’s Cave.
For Ursula, the Kendrick’s Cave man carries an additional significance. His mitochondrial DNA belonged to U5a2, a sibling branch of the U5a1 lineage traced back from Ursula. A close maternal cousin within her wider genetic family had therefore already reached the far northwestern edge of Europe by the closing stages of the Ice Age.
By the beginning of the Mesolithic, evidence of human presence reaches much farther north. At Heaning Wood Bone Cave in Cumbria, a young female was buried between 9,290 and 8,925 cal BCE. Nicknamed Ossick Lass, she is the earliest known human burial from northern Britain. Her body appears to have been deliberately placed in the cave, accompanied by tiny pierced periwinkle shells, probably once arranged or worn as beads.
Site map of Britain's earliest arrivals. The locations of Gough's Cave (Somerset) (Bello et al. 2011; Bello et al. 2017) and Kendrick's Cave (Great Orme, North Wales) (Płonka & Kowalski 2017) and Heaning Wood Bone Cave (Cumbria) (Warburton et al. 2026), the three sites that preserve the clearest surviving evidence of the people who first reached Britain after the ice retreated, alongside a selection of the objects recovered from each.
Several centuries later, Gough’s Cave enters the picture again. Thousands of years after the Magdalenian woman had lived there, people from a different Mesolithic population were returning to the same limestone gorge. Among them was the man now known as Cheddar Man, buried between 8,607 and 7,982 cal BCE.
His Britain was greener than the one known to the first people at Gough’s Cave. The open landscapes of the Late Glacial had given way increasingly to woodland, rivers and wetlands, and Mesolithic communities lived by hunting, fishing and gathering wild foods. Gough’s Cave itself preserves evidence of people returning to the area at widely separated periods rather than one continuous settlement.
Cheddar Man stood around 166 cm tall, a little above the broader average for Mesolithic men in western Europe. Even so, study of his skeleton placed him towards the smaller end of the range for men living at around the same time.
He also brings two of the wider family stories together in a single individual. His Y chromosome belonged to I-M423/I2a, within the same broad paternal family as Wodan, while his mitochondrial DNA belonged to U5b1, a major sibling branch of Ursula’s U5a maternal family. Neither is the particular branch traced back from Ursula or Wodan, but both are collateral branches of those wider genetic families. Oisin’s and Xenia’s descendants had yet to leave any trace here; their stories were still unfolding elsewhere.
The prominence of caves in this early British evidence is partly a consequence of what survives. It does not mean that these communities spent their lives deep underground. Much human activity took place around cave entrances and across the surrounding landscape, while caves could also serve other purposes, including burial and the treatment of the dead. Their protected conditions preserved bone, antler and other fragile remains that have vanished from many open-air sites. Britain’s earliest inhabitants lived across a much wider landscape than the caves alone suggest; caves are simply among the places where some of their most intimate traces survived.
Britain was not returned to by one people, but rediscovered again and again, by those arriving on foot across land joined to the rest of the continent, ground that was already, slowly, disappearing beneath the sea.
10
Beneath the North Sea
Rising Seas and Isolation
By the end of the Ice Age, Britain was still physically joined to continental Europe. Beneath what is now the southern North Sea stretched Doggerland, a broad low-lying country of rivers, wetlands, lakes and coastlines connecting the two. It was not merely a route into Britain, but an inhabited landscape in its own right.
That habitation has left a real, if scattered, physical record. Since the mid-twentieth century, fishing trawlers and Dutch coastal dredging have recovered thousands of bone, antler and flint tools from the seabed and beaches, along with dozens of human remains, preserved by the oxygen-poor conditions of the submerged sediment. Among them are several mandibles, robust and large-toothed by modern standards, one of them hauled up by a fishing boat and later confirmed through radiocarbon dating to belong to the Mesolithic period. Isotope analysis of around thirty such finds shows that Doggerland's inhabitants relied overwhelmingly on freshwater food sources rather than the sea, a picture at odds with the common assumption that a people whose remains now wash up on beaches must once have lived by the coast. Alongside tools of animal bone and antler recovered from the same waters, at least two barbed hunting points were found to have been made not from animal bone but from the bones of the dead themselves, a discovery that points to a similarly complex and deliberate treatment of human remains as that already seen at Gough's Cave, arising independently in a very different landscape.
For thousands of years this connection survived, but the sea was steadily dismantling it. As the remaining ice sheets melted, sea levels rose and the lowest ground disappeared. The process took millennia rather than centuries and did not proceed at a smooth, predictable pace. At times the rise accelerated, then slowed again, repeatedly redrawing river mouths and coastlines and turning familiar ground first into marsh, then tidal water and finally open sea. Sea level in the region rose in two distinct pulses rather than smoothly: one peaking around 10,300 years ago and a second around 8,300 years ago, each driven by meltwater surging from the melting ice sheets of North America and Antarctica, at rates comparable to the fastest sea-level rise now projected for the next century. For the people living there, these were not simply changes on a map. Coastlines, estuaries and hunting grounds familiar to one generation could look very different to the next, and there is some evidence that diet itself shifted accordingly, growing steadily more dependent on aquatic food over time as the landscape grew wetter.
By around 9,000 years ago, much of the former plain had already been lost and Doggerland was becoming increasingly fragmented, although higher ground around what is now Dogger Bank survived for longer. Recent modelling suggests the losses came earlier and more sharply than once thought, with a marked phase of land loss around 10,500 to 10,000 years ago, a period that also shows a measurable rise in the intensity of human activity recorded at nearby sites in the Netherlands and northwest Germany, as though communities were concentrating onto the ground that remained. It was a landscape already undergoing profound change before its most dramatic shock arrived.
Evidence from the Seabed: Reconstructing Doggerland's Final Centuries. This reconstruction traces the North Sea basin through five real coastline stages, from the open plains of 16,000 years ago to the last fragments of Dogger Island around 7,000 years ago, built from EMODnet bathymetry and the sea-level modelling of Lambeck et al. (2014) and Hoebe et al. (2024). The retreating Last Glacial Maximum ice sheet (Clark et al. 2022; Hughes 2016) is shown at 16,000 years ago, with the ice front of 21,000 to 19,000 years ago marked for comparison. Around 8,150 years ago, in the span between the 9,000 and 8,000-year stages shown here, the Storegga Slide, one of the largest submarine landslides known, triggered a tsunami whose deposits are recorded at 60 sites around the North Sea coast (Walker et al. 2020) and, remarkably, within a buried river channel on Doggerland itself (Gaffney et al. 2020). The triangles mark real archaeological finds from this drowned North Sea landscape, from a Mesolithic barbed antler point trawled up in 1931 to flint scatters recovered from Brown Bank and the Eurogeul, distinguishing finds recovered in a documented context from stray, isolated discoveries (after Walker et al. 2022, Table 5.2).
Then, around 6,150 BCE, one of the largest submarine landslides of the Holocene occurred off western Norway. Roughly 3,200 cubic kilometres of sediment collapsed beneath the sea, generating the Storegga tsunami. The resulting waves reached as high as 20 metres around the Shetland Isles and were still several metres high when they struck parts of the northern English coast.
For many years Storegga was portrayed as the catastrophe that finally drowned Doggerland. The newer evidence is more complicated. Much of the landscape had already disappeared before the tsunami struck, and its effects varied sharply according to local topography. Some surviving areas appear to have endured the wave and remained above water afterwards. Doggerland was therefore not erased in a single terrible day; Storegga was a violent interruption within a much longer history of rising seas and shrinking land.
For people directly in its path, however, the consequences could still have been severe. At Howick on the Northumberland coast, Mesolithic people lived beside a resource-rich coastal landscape, drawing food from the coast, woodland and surrounding waters. The tsunami coincided with a real and significant decline in the region’s Mesolithic population. Modelling also suggests that the state of the tide may have been critical: if Storegga arrived near high tide, inundation could have reached the settlement itself, threatening both the people living there and the coastal resources on which they depended. Whether Howick itself suffered such a direct hit remains uncertain.
Even Storegga was not quite the end. Fragments of the old landscape survived afterwards before continued sea-level rise finally removed the remaining connection. Britain had become an island, but it had not become isolated.
When the next great population movement reached Britain, it would come by boat.
07
The Last Glacial Maximum
Isolation, endurance and survival
19,000 BCE
Roughly 20,000 years ago, the Earth was in the grip of the most severe phase of the last Ice Age. The Last Glacial Maximum, lasting from around 26,500 to 19,000 years ago, saw immense ice sheets spread across much of North America and northern Eurasia. In North America, the southern margin of the ice reached deep into the Midwest, as far as present-day Iowa; in Europe, it extended into Germany and Poland; while in Asia, extensive mountain glaciation covered parts of the Tibetan Plateau and the mountain ranges of Central Asia, from the Zagros to the Altai. So much water was locked into ice that sea levels fell by around 120-130 metres, exposing vast areas of land now beneath the sea, while average global temperatures were about 6°C colder than today.
Across much of northern Europe, whole landscapes disappeared beneath the ice, while elsewhere conditions deteriorated into cold, wind-scoured tundra and steppe. Human populations contracted dramatically into a handful of glacial refugia, warmer and more sheltered pockets farther south where enough plant and animal life persisted to sustain them. Most important were the Franco-Cantabrian region of northern Spain and southwestern France, the Mediterranean margins of Iberia, Italy and the Balkans. These were not simply temporary camps occupied for a few bitter winters, but regions in which communities survived across generations while enormous territories farther north became sparsely inhabited or were abandoned.
NOAA Climate.gov Data: NOAA SOS
Reconstructed Boundaries of Survival. Each region shown here represents a place where people withstood the most extreme climate conditions of the Last Glacial Maximum, reconstructed from real physical data rather than estimation. The four European zones are bounded by reconstructed coldest-month temperatures, adjusted for the lower sea level of the time, which exposed additional coastal land now submerged. Real mountain terrain shapes these boundaries directly: the Balkan refuge (gold), has an irregular outline that unlike the clean-edged zones further west, reflects genuine climatic exclusion by the mountains within it, not an artistic choice. The same temperature threshold applied to West Asia found no true gap between the Levant coast and the Zagros foothills, so the region (teal), is shown as the single continuous range the data actually supports, with digitised glacier extents excluding the genuinely ice-bound ground of the Zagros high peaks and Anatolia. Oisin's Siberian range (navy), meets no comparable temperature threshold: conditions there remained too severe throughout the period to define a boundary by warmth at all. It is defined instead by real vegetation data, with glaciated terrain and mountain systems excluded, marking where the ground stayed open and ice-free rather than where it was ever mild. Full sourcing for each dataset is set out in the Evidence section. The black markers depict real, dated evidence. Pavlov and Dolní Věstonice sit outside every shaded zone, marking Ursula's and Wodan's earliest confirmed ancestors, before any refuge existed. Mal'ta falls within the Last Glacial Maximum itself. Xenia has no ancient individual yet found for this period, so her marker shows the Druze communities of the Levant.
Even within this long cold spell, the climate itself did not simply freeze into place. Ice-core and marine-sediment records show a repeating rhythm of abrupt swings layered on top of the Ice Age’s slower deep-freeze: rapid warming episodes known as Dansgaard-Oeschger events alternating with sudden, sharper cold snaps, while Heinrich events saw enormous discharges of icebergs carry debris far out into the North Atlantic. These changes could unfold astonishingly quickly, sometimes over decades rather than centuries, repeatedly shifting vegetation, animal ranges and the limits of habitable ground.
Greenland ice-core and temperature records. NGRIP δ¹⁸O and reconstructed Greenland air temperatures provide the chronological and climatic framework for the regional comparisons, showing repeated abrupt shifts during the last glacial period and the transition into the Holocene.
How each of the four lines weathered that crisis differed enormously.
The refugia were not culturally uniform. In the west, the Franco-Cantabrian world developed the distinctive Solutrean tradition, renowned for its exceptionally finely worked stone projectile points, while farther east the Epigravettian developed from the earlier Gravettian world in Italy and the Balkans. As habitable ground expanded and contracted, people and the animals they depended upon could move beyond the safest parts of the refugia, carrying materials, techniques and ideas between communities that harsher conditions had pushed apart.
For Ursula, survival of the Ice Age can be seen most clearly in her wider maternal family. U5 was already present among the Gravettian people of Central Europe before the climate reached its maximum severity, and ancient DNA shows that this maternal lineage survived the enormous population upheavals that followed, becoming one of the characteristic lineages of later European hunter-gatherers.
A remarkable trace of this wider maternal family has recently emerged from a very different kind of evidence. At Covarón Cave in Asturias, within the Franco-Cantabrian region, ancient human DNA has been recovered not from a skeleton but directly from unpigmented cave-wall surfaces close to prehistoric rock art. Two samples yielded mitochondrial DNA belonging to U5a’b, within the same wider maternal family as Ursula. The DNA cannot yet be securely dated to the Last Glacial Maximum or attributed to the people who made the nearby art, so it is not evidence that Ursula’s own line occupied the cave. What it does show is that genetic traces of this ancient maternal family can survive on surfaces touched by prehistoric people even where no human bones remain.
Wodan’s own line appears to have weathered the same crisis farther east. His wider paternal family, haplogroup I, was already present at Pavlov before the cold reached its maximum. Its I2 branch in particular has been linked to an Ice Age refuge in the Balkans and eastern Europe, not far from the wider region in which Wodan’s lineage had already appeared. This raises the intriguing possibility that his paternal family survived the worst of the Ice Age by retreating southwards into the same broad southeastern European refuge. Of the four refugia, the Balkans was the most exacting: its jagged extent is shaped directly by the mountains within it.
Oisin’s world was different again. His deeper paternal family remained far to the east across Siberia and Central Asia, where extreme cold combined with very low rainfall to prevent the vast continental ice sheets seen across northern Europe. Instead, huge areas became part of the mammoth steppe, a cold, dry grassland supporting mammoth, reindeer, horse and other large herbivores, Oisin’s paternal line therefore endured the Last Glacial Maximum in a harsh but largely ice-free landscape. Coldest-month temperatures across this range still fell to between minus twenty-seven and minus forty degrees, far beyond anything survivable by the standard that defined the refugia of Europe. Real alpine glaciers did exist within it, in the Altai, the Sayan and the mountains close to Baikal itself, but the vast steppe between them remained open and passable. What sustained Oisin's people here was not shelter from the cold, which nowhere in this landscape existed, but a landscape that stayed walkable and huntable despite it.
Xenia’s wider maternal family lay farther south in West Asia, beyond the great European ice sheets but not beyond the effects of the Ice Age. Much of the interior became colder and drier, with open steppe replacing richer vegetation, while a single continuous stretch of more sheltered ground, running from the Levant coast to the Zagros foothills, continued to support people, animals and plants through the worst of the climate. Archaeological sites such as Palegawra Cave in the Zagros of northern Iraq show that hunter-gatherers were living in this region during the Last Glacial Maximum, using caves and rock shelters and exploiting the plants and animals available around them. Genetic evidence suggests that Xenia’s wider X2 maternal family began spreading more widely across Eurasia around or just after the Last Glacial Maximum, probably from the Near East. No Ice Age burial carrying her own X2c branch has yet been identified, but together the genetic and archaeological evidence gives us a much clearer picture of the West Asian world in which her maternal story continued.
CHELSA-TraCE21k regional temperature reconstructions. Modelled mean annual temperature and coldest-month minimum temperature are compared across Iberia, the Balkans, the Levant and Siberia from 21 ka BP to the present, placing the four regions on the same climatic scale.
-
![]()
Iberia: marine temperature and pollen records
Marine sediment and terrestrial pollen records show how sea-surface temperature, woodland and open vegetation changed through the glacial period and deglaciation. -
![]()
Balkans: pollen, speleothem and climate reconstructions
Records from Lake Ohrid, Sofular Cave and Tenaghi Philippon combine vegetation, hydroclimate and seasonal climate evidence across the southern Balkans and Black Sea margin. -
![]()
Levant: pollen and hydroclimate records
Dead Sea and Soreq Cave records trace changes in desert-steppe vegetation, Mediterranean woodland and regional hydroclimate through the last glacial period. -
![]()
Siberia: lake productivity and vegetation records
Lake Baikal and Lake Ochaul records combine aquatic productivity, plant-wax biomarkers and pollen to reconstruct the open, cold environments of southern Siberia.
Ursula and Wodan both likely moved south from their shared Gravettian world into the refuges of Europe. Oisin's line had no single refuge to retreat to; instead, people moved north and south across the open steppe, following game across a landscape that remained harsh wherever they went. Xenia's line barely moved at all, already settled within a broad zone that the Ice Age left largely undisturbed.
By around 19,000 years ago, the Last Glacial Maximum was beginning to loosen its grip. The ice would not disappear quickly or smoothly, and dramatic climatic reversals still lay ahead, but across Eurasia the geography of human possibility was starting to change. The long contraction was ending, and movement back into the abandoned landscapes of Europe was about to begin.
Four lines endured the same Ice Age in four different worlds. When the thaw finally came, each travelled their own path into a changing landscape.
08
When the Ice Retreated
From Refugia to Recolonisation
By around 19,000 years ago, the great ice sheets were beginning to retreat, but Europe did not simply return to the world that had existed before them. Tundra and steppe gradually gave way to grasslands, scrub and woodland, animals returned to territories they had abandoned, and people began spreading northwards again from the regions in which they had survived the coldest millennia. Ancient DNA shows that this was not simply the return of the same people to the same places. Europe’s population was being reshaped as different hunter-gatherer groups expanded, met and sometimes replaced one another.
The warming itself was anything but smooth. Around 14,700 years ago, temperatures rose abruptly at the beginning of the Bølling-Allerød warm period, transforming environments across Europe. Regional records show that the scale of this warming varied considerably from place to place, with changes of several degrees in some areas rather than a uniform rise across the continent. Then, around 12,900 years ago, the climate lurched sharply backwards. The Younger Dryas brought a return to much colder conditions, particularly across northern and western Europe, where summer temperatures in the north may have been around 6°C lower than today. The cold lasted for more than a thousand years before another rapid warming around 11,700 years ago finally marked the beginning of the Holocene.
Against this shifting background, Oisin’s paternal line appears in western Europe for the first time. At Villabruna, in the foothills of the Dolomites in northern Italy, a hunter-gatherer buried between 12,190–11,770 BCE carried R-L754/R1b1a, part of the direct paternal lineage traced here from Oisin. We might call him Orso. More than ten thousand years after Mal’ta Boy had carried the deeper R lineage in Siberia, that paternal line had now appeared in western Europe in the remains of a real individual.
Villabruna 1: skull to face. Left: cranium of Villabruna 1, the individual behind Orso, R1b1's earliest confirmed European carrier, found at Ripari Villabruna, Italy (Vercellotti, G., et al., 2008, used under CC BY-NC 4.0). Right: facial approximation. The preserved cranium of Villabruna 1, an adult male aged approximately 25 years and dating to around 14,000 years ago, provides the anatomical basis for this evidence-informed facial approximation. Cranial form and proportions were informed by the osteological study of Vercellotti et al. (2008), while pigmentation was informed by ancient-DNA evidence reported for Villabruna in Fu et al. (2016). The reconstruction incorporates the ancestral alleles at SLC24A5 (rs1426654) and SLC45A2 (rs16891982), two major loci associated with lighter skin pigmentation in present-day Europeans, together with the derived variant at the HERC2/OCA2 locus (rs12913832) strongly associated with blue eye colour. The resulting portrait is an AI-assisted, evidence-informed approximation rather than a definitive likeness; soft tissue, hair and other features that cannot be recovered directly from the cranium or genome remain interpretive.
Orso was part of a much broader population movement already under way. Villabruna-related ancestry was present in Italy by at least 17,000 years ago, and from around 14,000 years ago spread much more widely through central and western Europe. By the time Orso was buried, people carrying this ancestry were already becoming an important part of the post-Ice Age hunter-gatherer population across much of the continent.
Villabruna 1: at a glance. A summary of the archaeological, skeletal and genetic evidence for Villabruna 1, the individual behind Orso, R1b1a's earliest confirmed European carrier, found at Ripari Villabruna, Italy. The location map is built from real elevation data (SRTM15+, via OpenTopography) and Natural Earth boundary data. The burial, dental and skeletal illustrations are evidence-informed reconstructions based on published photographs (Vercellotti et al. 2008; Oxilia et al. 2015), not direct reproductions of those images. The painted stones panel is based on Aimar et al. (1994)'s original excavation description, cross-checked against photography from the Museo MUVAR, Sovramonte. Full citations for each finding appear within the relevant panel.
These people may also have looked rather different from many Europeans living there today. Among Villabruna-related hunter-gatherers, genetic variants associated with blue or light-coloured eyes were extremely common, while two of the major variants associated with lighter skin pigmentation were still almost entirely absent. This suggests that at least some of them combined darker skin with light eyes, one of several combinations found among Europe’s prehistoric hunter-gatherers.
Skin pigmentation continued to change over the thousands of years that followed. Lighter-pigmentation variants increased at different times in different European populations rather than spreading across the continent in a single wave. One important evolutionary pressure was the balance between protection from ultraviolet radiation and the need for enough UV exposure to make vitamin D at higher latitudes. Ancient DNA shows that one major light-skin variant began increasing in frequency from around 13,000 years ago.
Their physical build was changing too. Early Upper Palaeolithic Europeans had often been remarkably tall: average adult male stature has been estimated at around 174 cm, compared with roughly 164 cm among Mesolithic Europeans, a difference of about ten centimetres,. The greatest reduction appears to have taken place after the Last Glacial Maximum, with a smaller further decline into the Mesolithic. Ancient DNA suggests that at least part of this longer-term change had a genetic component rather than simply reflecting nutrition or living conditions.
Farther east, hunter-gatherer populations also carried ancestry connected to the Ancient North Eurasians of Siberia. As these eastern and western populations encountered one another, they mixed, adding another layer to the genetic reshaping of post-Ice-Age Europe. By the beginning of the Holocene, the continent was already very different from the one that had entered the Last Glacial Maximum.
At the far northwestern edge of this changing world lay a territory that was not yet an island. Britain was still connected to continental Europe by a broad expanse of dry land, and people were beginning to move back into landscapes from which humans had been absent for thousands of years.