The American “Cheetah” Was a Puma Cousin That Probably Ate Fish in the Arctic

For decades, the American “cheetah” had a wonderfully tidy job in the story of Ice Age North America.

It was the sprinter. Its long legs and slender body seemed built to chase pronghorn across open country, much as African cheetahs chase antelope today. The resemblance even helped inspire an explanation for why modern pronghorn can outrun every predator that currently pursues them: perhaps they are still carrying the speed needed to escape a cat that vanished thousands of years ago.

Then three supposed puma fossils from the Yukon started producing strange genetic results.

The bones were roughly 31,000 years old and came from Bluefish Caves and Upper Quartz Creek, far north of the American “cheetah's” previously recognized range. When researchers compared their DNA with Miracinonyx trumani material from Wyoming, the mystery resolved itself.

The Yukon bones were not pumas. They belonged to Miracinonyx.

That identification extended the cat's recognized range by more than 20 degrees of latitude into the Arctic. The chemistry preserved in the bones then produced a second surprise: the northern animals had apparently been feeding from an aquatic food web, probably by eating anadromous fish.

A cheetah-shaped puma cousin

A 2026 Current Biology study used nuclear paleogenomes to place Miracinonyx on the cat family tree. Its sister lineage is the modern puma, Puma concolor, not the African cheetah.

The study estimates that the two lineages separated about 2.6 million years ago. Miracinonyx was not itself a puma, but an extinct relative that developed some remarkably cheetah-like anatomy.

That relationship had been emerging for years. In 2005, mitochondrial DNA from a Wyoming fossil also grouped Miracinonyx with pumas. Mitochondrial DNA represents only one inherited genetic line, however, and the mitochondrial history of a cat species can differ from the history recorded across its nuclear genome. The new nuclear data provide much stronger confirmation.

This is convergent evolution: similar pressures producing similar anatomical solutions in animals on different evolutionary branches. Miracinonyx and the African cheetah both acquired bodies suited to rapid movement through open country, but they arrived there separately.

Even the resemblance has limits. A 2023 analysis of its elbow found a structure intermediate between those of a puma and a cheetah. Living cheetahs have forelimbs highly specialized for efficient high-speed pursuit. Miracinonyx appears to have retained greater ability to grapple with prey.

The result is an animal without a clean modern counterpart. It was not simply a long-legged puma, and it was not an American copy of an African cheetah.

Finding an aquatic food web in a bone

Establishing that Miracinonyx reached the Yukon raised another question: what was it eating there?

Part of the answer was stored in its bones.

Within a food web, animal tissues generally become more enriched in the heavier nitrogen isotope, nitrogen-15, with each trophic step. Herbivores tend to be enriched relative to the plants they eat, and predators tend to be enriched relative to their prey.

There is an important complication. The nitrogen values at the base of one ecosystem can differ from those in another. A high value cannot automatically be read as proof that an animal occupied a particular trophic level or ate a particular species.

The researchers therefore combined bulk isotope measurements with measurements of nitrogen in individual amino acids. This compound-specific technique helps separate an animal's position in a food chain from variation in the underlying environmental baseline.

Miracinonyx specimens from Natural Trap Cave in Wyoming had isotope signatures consistent with terrestrial carnivores eating herbivores. The Yukon fossils were different. Their chemistry indicated an unusually elevated trophic position and pointed toward an aquatic food web.

The researchers concluded that the most plausible explanation was consumption of aquatic resources, probably including anadromous fish. These fish spend much of their lives feeding at sea before migrating into freshwater rivers, carrying the chemical signature of a marine food web inland.

Fish were locally available. Bluefish Caves, which produced one of the Miracinonyx specimens, also contains Pleistocene fish fossils.

What the evidence cannot reveal is how the cats obtained those meals. They might have caught fish in shallow water, taken them during seasonal migrations or scavenged carcasses along riverbanks. Isotopes record food that became part of an animal's body. They do not record the behavior that put it there.

The sample is small, too. Three Yukon animals cannot establish how every northern Miracinonyx behaved. The similar signal in specimens from more than one locality makes a recurring dietary pattern plausible, but it cannot show how widespread that pattern was across the species.

The problem with a very persuasive nickname

None of this means the earlier reconstruction was entirely wrong. Miracinonyx had adaptations associated with running, and the Wyoming isotope evidence fits a predator of terrestrial herbivores. It may have hunted pronghorn, and it may have contributed to the evolutionary pressures behind their extraordinary speed.

The new study neither proves nor disproves that idea. Instead, it reveals how incomplete the familiar picture was.

A cheetah-shaped fossil became a cheetah-like hunter in scientific reconstructions and paleoart. Once that analogy took hold, it encouraged a view of Miracinonyx centered on open-country pursuit and specialized terrestrial prey.

The Yukon fossils complicate that view. They place the species in Arctic Beringia, a region that included open, grass-rich steppe-tundra, but they also reveal a diet sharply different from the one reconstructed for the Wyoming population.

That combination matters. The Arctic discovery is not surprising because Miracinonyx abandoned open country. It is surprising because the same species appears to have occupied an enormous geographic range while drawing food from very different ecosystems.

There are limits to what six fossil animals can tell us. We still do not know the true running speed of Miracinonyx, how often it climbed, how it killed prey or why it disappeared near the end of the Pleistocene. Genetic changes involving circadian regulation may hint at life under extreme Arctic light cycles, but that proposed connection remains unproven.

What the bones establish is already enough to retire the simplest reconstruction.

Miracinonyx was a puma cousin with the outline of a cheetah and a range stretching into the Arctic. In at least one northern population, its bones preserve chemical evidence consistent with an aquatic diet, probably including fish. The animal did not fail to live up to its nickname. The nickname failed to contain the animal.