A fossil mammal tooth smaller than a grain of rice does not announce itself loudly. It must be hard won from sediment and stone. Then, under a microscope, it reveals itself – no longer just a speck of blackness but a surface of cusps, ridges and worn edges.

It is a small object, easily missed. Yet five such teeth from northern Alaska, belonging to three newly discovered species of long-extinct rodent-like mammals, hold an unexpectedly large history: of polar environments, shifting continents, winter darkness – and of mammals moving through this world that was colder, stranger and more connected than we once imagined.

The fossil mammal teeth at the centre of my new study with US colleagues come from the Late Cretaceous Prince Creek Formation, around 73 million years ago. At that time, northern Alaska’s palaeolatitude was roughly 80-85°N (10-15° closer to the north pole than it is today).

Our discovery shows the Arctic was not simply a cold, lifeless edge of the Cretaceous world, but a place where mammals adapted, diversified, migrated and originated. It raises deeper questions about what it means for a species to be native to a place whose landscapes, climates and inhabitants are forever changing.

The Arctic is not an empty landscape now, and nor was it in the deep past. It was a distinctive and demanding biome, with months of winter darkness, freezing temperatures and strong seasonality.

Rivers crossed this landscape. Plants grew through the long light of summer. Dinosaurs lived and thrived there, and evidence suggests they reared their young in the Arctic. There was also a diversity of birds, fish and mammals: tenacious denizens of the polar dark.

The teeth we found

Finding miniscule fossil mammals is not the kind of palaeontology that begins with a spectacular skeleton weathering from a cliff. It begins with bulk sediment. Bags and buckets of it.

The material is dug out, washed, sieved through fine screens, dried, then meticulously sorted grain by grain beneath a microscope.

Mammalian palaeontology often depends on fragments: a tooth cusp, a root, a worn edge of enamel. It asks for patience more than drama. The field site may be remote, the landscape vast, but the discovery happens at the scale of a fingertip. It is an act of attention.

The five teeth described in our new study are multituberculates: a group of rodent-like mammals that lived alongside dinosaurs. They are the longest-lived mammal group to date, with a history spanning well over 100 million years. Our own species Homo sapiens, by comparison, is only a few hundred thousand years old.

But their success was not permanent. After surviving for longer than any other mammalian lineage, multituberculates eventually vanished, with their last representatives disappearing around the Eocene-Oligocene boundary, roughly 34 million years ago.

The teeth we found at Prince Creek represent three new species: Camurodon borealis, Kaniqsiqcosmodon polaris and Qayaqgruk peregrinus. The last of these sits especially close to the heart of the story.

Qayaqgruk peregrinus has close evolutionary affinities with a group of multituberculates discovered in Mongolia,, indicating that, many millions of years ago, these minute mammals moved between Asia and North America through a polar land corridor.

This was not the world of Pangaea, the single giant continent of earlier Earth history. By 73 million years ago, the continents were taking on more familiar shapes, but they had not yet settled into the map we know today.

North America was split by a shallow inland sea, and its far northwestern edge lay near northeastern Asia, creating a high-latitude corridor through which animals could move between continents.

The Arctic corridor itself was not unknown. Fossils of dinosaurs, birds and some mammals have shown that animals were migrating between Asia and North America during the Cretaceous.

But for multituberculates, the picture was much less clear. Their fossil record left open whether they crossed between the continents early and repeatedly, or only later in the Cretaceous. Qayaqgruk peregrinus helps close that gap.

Its name draws on the language of the Iñupiat, Alaska Native people, from the region where the fossils were found. Qayaq is a legendary Iñupiaq hero and wanderer whose journeys are told in The Epic of Qayaq.

Many of its Mongolian relatives carry the suffix -baatar, meaning hero in Mongolian, so the name also links its Alaskan discovery to its Asian evolutionary affinities.

Kaniqsiqcosmodon polaris is the oldest known member of the Microcosmodontidae family of multituberculates, suggesting this lineage, later known from North America, may have had a polar origin. Camurodon borealis represents the northernmost known occurrence of the North American family Cimolomyidae.

Five teeth are not a complete ecosystem, but they are enough to show these mammals were not occasional strays at the edge of their range. They belonged to the Arctic.

Surviving mass extinction

Survival in this environment was not a passive condition. Multituberculates survived through the end-Cretaceous mass extinction, 66 million years ago, when as much as 75% of all life on Earth went extinct, including the non-avian dinosaurs.

The adaptations that helped these ancient Arctic species live though deep winter cold, short dark days and seasonal scarcity may have given them advantages when the world’s ecosystems were devastated.

This is not to say the Arctic was a refuge in any simple sense. Evolution is rarely simple – it is full of bottlenecks and unexpected openings. But tiny mammals already adapted to seasonal scarcity may have carried some advantages into a disrupted world.

The Alaskan multituberculates did not belong to the Arctic because they had always been there, or because their lineages had never moved. They belonged because they were part of that ecosystem for a time: shaped by its dark, its cold, its flora, its fauna, its seasons and its routes of passage.

Fossils make this kind of belonging harder to define, but more interesting. They give the word indigenous, at least when we use it for species, a chronology: deep time shows that belonging is not always a matter of original presence, but of ecological participation across changing landscapes.

These five multituberculate teeth from the Prince Creek Formation are small enough to vanish in a pinch of sediment. Yet in their enamel are continents, seasons, darkness, ancestry and journeys.

They show that the ancient Arctic was not an evolutionary margin. It was a living biome, a passage between worlds, and part of the deep history of mammalian evolution.

Sarah Shelley does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

The most destructive wildfire season on record in Europe was in 2025, with more than one million hectares burned and tens of thousands of people displaced by fires across the continent.

For people in Ireland and Britain, the type of destructive wildfires that ravage southern Europe each summer can seem like a distant problem. But these fires are not confined to the dry Mediterranean landscapes of Spain, Portugal and Greece. In recent years, they have started to extend into regions more commonly associated with rain-soaked hills and bogs.

In 2026, this trend has continued with major wildfires breaking out across Scotland, Northern Ireland and Ireland.

As fires spread across the Highlands and Moray in Scotland this April, public warnings focused heavily on dry weather, campfires and accidental ignitions. In Northern Ireland, cautions were issued as firefighters battled several large gorse fires across the Mourne Mountains and other upland ares.

Similar warnings were issued nationally in Ireland over the Easter bank holiday weekend, when the public was urged to avoid lighting fires or bringing barbecues into the countryside. The threat of wildfires is only expected to ramp up this summer as temperatures rise further.

These are important messages. But focusing only on how fires start risks missing a slower and less visible transformation already unfolding across many upland landscapes. The real wildfire story in places like Ireland and Scotland is not just about climate or how fires start. It is also about how rural upland landscapes themselves are changing.

Changing farming styles

Recent research explores how decades of agricultural policy reform under the EU’s common agricultural policy, alongside falling farming populations and declining active land management, are reshaping vegetation patterns across Ireland’s uplands.

Historically, many upland landscapes were actively managed through livestock grazing, cutting and controlled patch burning. These practices helped maintain open landscapes and reduced the build-up of highly flammable vegetation.

But that balance has shifted. Reduced grazing pressure and changing land management practices are contributing to the expansion of highly flammable vegetation such as gorse, heather and purple moor grass.

While lower grazing pressure can bring biodiversity benefits and support natural regeneration, it can also increase the amount and proliferation of flammable vegetation across the landscape, known as fuel loads and fuel continuity. In practice, this means larger and more connected stretches of vegetation that allow fires to spread more rapidly and across greater distances.

This is especially concerning in upland areas where the average age of people working on farms is rising, and active land management is declining. Rural depopulation and labour shortages mean fewer people are available to manage what is known as commonages in Ireland and common grazing in Scotland. That means less maintenance of grazing systems and a reduction in the small, controlled vegetation burns that historically decreased wildfire risk by clearing vegetation and creating firebreaks. As one upland farmer in County Kerry recently described it to me: “It’s a bomb waiting to go off.”

Increasing flammability

Climate change is intensifying these risks. Hotter, drier conditions increase the likelihood that vegetation will dry out, increasing flammability. But climate alone does not explain why some landscapes burn more severely than others.

Wildfire risk is also shaped by what is growing on the land, how landscapes are managed, and whether fuel loads are reduced or allowed to accumulate over time. Experts responding to the recent Scottish fires also highlighted the role of vegetation build-up, prolonged dry conditions and changing land management in shaping fire behaviour, warning that historically wetter regions may face increasing wildfire risks in the future.

Similar patterns have already emerged across parts of southern Europe, where rural depopulation and land abandonment have contributed to increasingly severe wildfire regimes.

Recent research from Italy has shown abandoned land, declining grazing and reduced active land management have contributed to fuel accumulation, and to the build-up of dense, continuous vegetation – conditions associated with increasingly large and severe wildfires. While the climates and landscapes of Ireland and Scotland differ from the Mediterranean, similar long-term changes are beginning to emerge here.

---

Read more: The Pennine hills are full of holes – here’s how they’re helping fight climate change

---

This creates a difficult tension for policymakers and conservationists. Reduced grazing pressure and natural regeneration can support biodiversity recovery in upland systems. Yet these same changes may also increase wildfire risk where vegetation becomes dense, continuous and unmanaged. The challenge is therefore not choosing between farming or conservation, but finding ways to support landscapes that can sustain biodiversity, rural livelihoods and wildfire resilience together.

Wildfire risk in Ireland and Scotland can no longer be understood simply as a problem of careless ignitions or extreme weather. It runs much deeper than that. It is increasingly tied to long-term changes in how upland landscapes are farmed, governed and managed.

If future policy is serious about reducing wildfire risk, it must look beyond seasonal warnings and begin addressing the deeper forces reshaping our uplands.

Will Hayes is affiliated with the Department of Geography at Royal Holloway, University of London, and the Leverhulme Centre for Wildfires, Environment and Society. Funding for this research was provided by the Leverhulme Trust.