A new piece in the puzzle of Antarctica’s land-mammal fossil record

When we think about Antarctic fossils, we usually imagine the spectacular marine record. In contrast, evidence of terrestrial mammals was mostly limited to the allochthonous isolated teeth and bones from Seymour (Marambio) Island. 

A new study in Polar Biology (Mansilla-Vera, H., de Valais, S., Goin, F., Gelfo, J.N. First mammal tracks from the Eocene of Antarctica. Polar Biol 49, 101 (2026). https://doi.org/10.1007/s00300-026-03539-1) describes a small mammalian footprint from the Fossil Hill Formation on King George (25 de Mayo) Island, in the South Shetland Islands. The site lies within Antarctic Specially Protected Area 125a, a locality already known for its exceptionally well-preserved Eocene fossils, including plant remains, feathers, bird footprints, and invertebrate trace fossils. The Fossil Hill Formation has been assigned an early Eocene age, and the footprint-bearing horizon is estimated to be approximately 55.5 million years old.

The specimen, catalogued as T-373, consists of a partial set of a footprint and handprint preserved as natural casts in fine-grained volcanic sediment. The footprint is only about two centimetres long and wide, but it contains important information. It shows a digitigrade posture, a compact sole or metapodial pad, and at least three digital impressions arranged in an anteriorly convex arch. From its preserved area, we estimated a minimum body mass of approximately 1.3 kg for the animal that produced it.

What kind of mammal made the track?

This is where the evidence becomes particularly interesting and also where caution is essential. The track is not sufficiently well preserved to assign it confidently to a particular ichnotaxon or species. Several types of carnivorous mammal footprints have superficially similar characteristics, but T-373 differs from established ichnogenera in important details. Its combination of proportions, pad morphology, digit arrangement, posture, and estimated body size nevertheless provides useful clues.

Probably the track was generated by a small sparassodont metatherian. Sparassodonts were an extinct group of South American predatory mammals that occupied carnivorous niches during much of the Cenozoic. Their fossil record is particularly important for understanding the evolution of mammalian predation in South America.

Why consider a sparassodont if no sparassodont skeleton has been found on King George (25 de Mayo) Island?

The answer lies in combining several independent lines of evidence.

First, the estimated size of the trackmaker—at least about 1.3 kg—is substantially larger than the smallest mammals known from the Eocene Antarctic fossil record, while remaining consistent with the size of small hathliacynid sparassodonts. Second, the morphology of the footprint is compatible with a small, digitigrade mammal with a broadly mustelid-like body plan. Third, sparassodonts were already present in South America during the Paleocene and Eocene, while land connections between South America and Antarctica provided potential routes for terrestrial dispersal.

None of these observations, individually, proves that a sparassodont produced T-373. Together, however, they make this interpretation the most plausible among the available alternatives.

A missing predator in the Antarctic fossil record

The importance of the footprint extends beyond identifying its possible producer.

The land-mammal fossil record in Antarctica is remarkably sparse. Most known Paleogene mammalian remains come from the La Meseta and Submeseta formations of Seymour (Marambio) Island and consist primarily of isolated teeth, with fewer skeletal remains. Until now, the Antarctic record of terrestrial mammals has therefore been strongly biased both geographically and anatomically.

King George (25 de Mayo) Island tells a different story. The Fossil Hill Formation preserves a terrestrial ecosystem outside Seymour (Marambio) Island, including abundant plant remains and avian traces. This track adds a mammalian component to this record and, importantly, represents the first direct ichnological evidence of a land mammal in the Eocene of Antarctica.

Footprints are particularly valuable in this context because they record an animal at the place where it walked. Unlike isolated bones or teeth transported after death, a track can provide direct evidence that an animal occupied a particular terrestrial environment. In the case of this footprint, it therefore represents the first autochthonous evidence of a mammal from the Antarctic Eocene, rather than remains transported into a marine or estuarine setting.

Reconstructing a more complex Antarctic food web

The discovery adds a new piece to the Antarctic mammalian history puzzle. 

The Austral continent was very different 55 million years ago. Despite its high-latitude position, warmer climatic conditions and forested landscapes supported terrestrial ecosystems that included mammals, birds, plants, and other organisms. The known mammalian community included small metatherians and other mammals occupying insectivorous, frugivorous, folivorous, and granivorous niches, together with larger herbivorous and browsing forms such as xenarthrans, and some South American native ungulates (SANU) astrapotheres, and litopterns.

Predation was already recognized as an important component of these ecosystems. Birds, including large phorusrhacids—the famous “terror birds”—have been proposed as major terrestrial predators and were present in these ecosystems. The new footprint adds another possibility: a small mammalian carnivore moving through the same landscape.

A single footprint cannot reconstruct an entire ecosystem. But until we find continental Antarctic outcrops with a non-transported mammalian fossil assemblage, this indirect evidence is crucial. It reveals the presence of a group that might otherwise remain invisible in ancient communities.

A fossil bone tells us that an animal existed. A footprint can tell us that an animal was moving across a particular substrate, in a particular environment, at a particular moment in geological time. When direct evidence from fossil teeth and bones is scarce—as they are in Antarctica—these traces can provide one of the lost pieces of the ancient biodiversity puzzle.

A glimpse of Antarctica before isolation 

The Eocene was a transitional period in Antarctica's history. Terrestrial ecosystems persisted while the continent was still connected to other southern landmasses, but this configuration eventually changed. The final Gondwana breakup led to Antarctica's progressive isolation. Following the Eocene–Oligocene transition, the establishment of the Antarctic Circumpolar Current, progressive cooling, and the development of a major Antarctic ice sheet profoundly altered the environmental conditions that had previously supported diverse terrestrial ecosystems, ultimately contributing to the extinction of land mammals from Antarctica.

The footprint is only a few centimetres across. Its scientific significance is considerably larger. It provides evidence for a terrestrial mammal on King George (25 de Mayo) Island, extends the known geographic record of Antarctic Paleogene mammals, and suggests that a small mammalian predator may have been part of the Eocene Antarctic food web.

Although there are now no autochthonous land mammals—or even terrestrial vertebrates—in Antarctica, its scarce and incomplete fossil record reminds us that profound climatic changes can reshape the conditions that make life possible. At a time of unprecedented anthropogenic impacts on the environment and Earth’s climate, palaeontology offers us a unique perspective. By reconstructing the puzzle of ancient life from isolated pieces, we can learn from the past and gain valuable insights into the possible trajectories of our future.  

Read the paper here