Castlepook’s Sixth Interstadial

Dateline

“31,393 BCE” is a modern chronological label, not a year that can be recovered from an Irish record. Nobody in Ireland then used a calendar that can be converted into a numbered BCE year, and neither archaeology nor palaeoclimatology can identify an event, a person or a season with that precision. The label corresponds only approximately to 33,300 years before present, with “present” conventionally fixed at 1950. The responsible scale is therefore a broad span of calibrated millennia, not a single year.

That span falls within the Late Pleistocene, during Marine Isotope Stage 3 (MIS 3), a long and unstable part of the last Ice Age. On the Greenland ice-core chronology, the approximate date lies close to the end of Greenland Interstadial 6, a relatively milder interval dated broadly to c. 33,740–33,360 years before AD 2000, before a return to colder Greenland Stadial 6 conditions. This is a useful North Atlantic comparison, not an annually exact Irish weather record: the chronological uncertainties alone are far larger than a single modern year.

A Landscape in Transition

Ireland was not yet at the fullest extent of the last Irish ice sheet. The better-dated maximum of that ice cover came later, broadly between c. 27,000 and 23,000 calibrated years ago, when ice covered the island and much of the surrounding continental shelf. Around the early thirty-fourth millennium years ago, conditions were nevertheless cold, variable and increasingly severe. Ice masses, periglacial ground and open vegetation made the island a very different place from later wooded, settled Ireland.

Lower global sea level exposed areas now beneath the Atlantic approaches and Irish Sea. But it would be careless to turn that broad fact into a confident map of a dry route into Ireland at this particular moment. Ice loading, changing sea levels, the shape of the seabed and the configuration of local glaciers all affected connections between Ireland, Britain and continental Europe. Later palaeogeographical modelling shows how complicated and regional those changes were; it does not supply a simple, permanent “land bridge” for 31,393 BCE.

The central historical development was thus environmental. Shifts in temperature, snowfall, sea ice, growing glaciers and exposed ground altered where plants could grow, where grazing animals could feed, and whether a small travelling group could safely move through the north-western edge of Europe. In such a world, mobility and survival depended on seasons and habitats rather than on fixed frontiers.

Derryvree’s Buried Wetland

The clearest Irish environmental evidence relevant to this general period comes not from a settlement, but from sediment sealed beneath a drumlin at Derryvree near Maguiresbridge, County Fermanagh. Roadworks in 1969 exposed freshwater silts and fine sands between two glacial tills. Their organic content preserved pollen, plant macrofossils, mosses and insects: an unusually precious glimpse of a landscape later overridden or enclosed by ice-derived deposits.

The original radiocarbon determination, 30,500 +1,170/−1,030 radiocarbon years BP, has subsequently been expressed as a broad calibrated range of roughly 31.7–36.9 thousand calendar years BP. That range encompasses the notional date in question but emphatically does not date the Derryvree wetland to one year, or even to one human lifetime. It records an interval within MIS 3 in which liquid freshwater, sediment and vegetation existed in south Fermanagh.

The fossils point to open tundra or tundra-steppe conditions, with sparse dry-ground vegetation and wet hollows rich in mosses and sedges. Finds included cold-adapted plants such as mountain-avens, Arctic willow, dwarf birch and bogbean, alongside aquatic or wetland species. Insect evidence suggests a muskeg-like setting: peaty, waterlogged ground with pools, plant debris and mossy margins. This was not a featureless white wilderness. Even in a harsh climate, sheltered basins could hold standing water and a varied small-scale patchwork of vegetation.

Yet Derryvree should not be mistaken for a national survey. It represents one locality in what is now Northern Ireland, at one broad phase of a fluctuating glacial cycle. Pollen can travel, plant remains favour places where they are preserved, and a buried basin cannot tell us exactly how every coast, mountain or plain looked. The secure conclusion is modest but important: parts of Ireland could sustain cold-open vegetation and freshwater habitats during middle Midlandian time.

Animals of the Cork Caves

Castlepook Cave, near Doneraile in north County Cork, offers a complementary record of animal life. Its old excavations and surviving museum collections have produced one of Ireland’s richest Late Pleistocene vertebrate assemblages. Reassessment and direct dating have shown that the cave accumulated material from more than one period, an essential warning against treating its fauna as the contents of a single moment. The site nonetheless demonstrates that southern Ireland was repeatedly reached by, or supported, a striking cold-stage community before the Last Glacial Maximum.

Among the animals recorded are reindeer, woolly mammoth, giant deer, brown bear, wolf, Arctic fox, stoat, mountain hare, collared lemming, Norwegian lemming and spotted hyaena. The combination evokes open country able to support grazing herbivores, with predators and scavengers exploiting carcasses and prey. Reindeer are especially apt symbols of mobility in such an environment, but their bones do not prove that a herd crossed a particular valley in 31,393 BCE. Equally, mammoth and hyaena remains show that these animals reached the region during MIS 3, not that all species stood together around the cave entrance at once.

Caves complicate the evidence. They can be dens, winter refuges, traps, drainage routes and repositories for bones dragged in by carnivores. Water, frost action, collapse, digging and early excavation can disturb original positions. Castlepook’s value lies precisely in its potential to preserve a long environmental archive; its weakness is that the archive has to be disentangled specimen by specimen.

The Reindeer Femur Question

One Castlepook reindeer femur has made this broad period unusually significant for Irish prehistory. It has been publicly reported as roughly 33,000 years old and as bearing a deep linear mark interpreted by zooarchaeologist Ruth Carden as a stone-tool cut or chopping mark made while the bone was fresh. County Cork heritage material presents the find as possible evidence of Palaeolithic hunters processing reindeer in the area around 31,000 BCE.

If that interpretation is confirmed, it would be an exceptionally early indication of human action in Ireland: not a legend of first settlement, but physical evidence that somebody used a sharp implement on a reindeer carcass. It would push the Irish human story far earlier than the securely dated terminal-Pleistocene butchered brown-bear patella from Alice and Gwendoline Cave, County Clare, dated to c. 12,810–12,590 and 12,810–12,685 calibrated BP.

The distinction between evidence and interpretation is crucial. The radiocarbon result dates the animal’s death, not necessarily the instant when the mark was made. A single marked bone cannot establish where an animal was killed, whether a person removed meat, hide, tendons or marrow, whether the bone entered the cave after butchery, or how long any visitors stayed. Nor has Ireland produced a securely associated hearth, stone-tool assemblage, dwelling surface, human burial or repeatedly used camp of comparable age. The most responsible description is therefore a possible, potentially transformative trace of Upper Palaeolithic activity, not proof of permanent occupation.

Society Beyond Recovery

No political affairs can responsibly be described for Ireland at this date. There is no evidence for rulers, territories, diplomacy, warfare, law, tribute or organised institutions. There is likewise no Irish evidence from this horizon for trade networks, named communities, language, religion, ritual practice, art, formal learning or burial custom. The absence of such evidence does not prove that no people ever entered the island during MIS 3; it means that none of these subjects can be reconstructed as Irish history for this particular period.

If people did reach Ireland, comparison with Upper Palaeolithic groups elsewhere in north-west Europe suggests that they would have needed extensive practical knowledge: warm clothing, fire, cutting tools, animal behaviour, route-finding and cooperation. But that is a comparative inference, not something demonstrated by an Irish tool kit. Castlepook has not yet yielded the equipment, camp debris or repeated stratified activity needed to turn that plausible human capability into a description of Irish everyday life.

Approaching Erasure

What happened after this broad interval helps explain why the record is so thin. Further climatic deterioration and ice-sheet expansion reshaped Ireland. Glacial erosion removed, redeposited or buried older surfaces, while later sea-level change submerged former coastal plains and travel corridors. Such processes make the absence of Palaeolithic camps less surprising, but erosion cannot itself prove an unseen population. It is a reason for caution, not a substitute for archaeological evidence.

For 31,393 BCE, then, Ireland’s recoverable history is principally an environmental one. Wetland sediments in Fermanagh show cold but living ground; the Cork caves preserve evidence of a diverse Ice Age fauna; and a reindeer femur at Castlepook raises a serious, still limited question about an early human encounter. The real significance of this notional year is not a recorded event. It is a threshold in which changing climate, animals and fragmentary bones reveal how much of Ireland’s deepest past remains to be discovered.

Primary Sources

Tags: Castlepook Cave, County Cork, Ireland Year by Year, Late Pleistocene, Marine Isotope Stage 3, Radiocarbon Dating