Thylacosmilus

Thylacosmilus atrox was a large South American sparassodont of the Late Miocene and Early Pliocene. Its upper canines grew continuously, with roots extending backward above the eye sockets; long bony flanges on the lower jaw protected the teeth when the mouth closed. The resemblance to Smilodon evolved independently and conceals important differences in teeth, skull mechanics and ancestry. The genus is included in the ancient mammal catalogue.

Thylacosmilus consumed animal tissue, but its exact ecological role remains disputed. One model treats it as an ambush predator that held prey with powerful forelimbs. Another connects its tooth and jaw anatomy with opening carcasses and reaching soft organs. No associated prey or stomach contents have settled the question.

Quick facts

Discovery, holotype and the name

Elmer Riggs of the Field Museum described Thylacosmilus in 1933 from fossils collected at Corral Quemado in Catamarca Province, Argentina. Holotype FMNH P14531 preserves a well-known skull and parts of the postcranial skeleton. Riggs published a fuller anatomical account the following year.

The name is commonly rendered as “pouched knife”, while atrox means fierce or formidable. Riggs initially named T. lentis as well, but later authors usually treat it as the same species. Earlier Argentine names, especially Achlysictis lelongi, complicate priority; Thylacosmilus atrox remains the name in widest use.

The type beds were once called Pliocene. More recent correlation places much of the material in the Late Miocene, with the youngest records reaching the Early Pliocene. A single date near three million years ago does not describe the full history of the genus.

Not a cat and not a living marsupial

Thylacosmilus was not a cat. It belonged to Sparassodonta, an endemic South American radiation of predatory metatherians. Sparassodonts are closer to marsupials than placental mammals, but their branch separated from the lineages of living opossums, carnivorous marsupials and kangaroos.

Sabre-like canines arose independently in several mammals. In placental cats, Smilodon and the earlier Machairodus show a separate route to enlarged canines. Thylacosmilus reached a similar head silhouette through a different dental system and metatherian anatomy.

Open-rooted canines and mandibular flanges

The upper canines had open bases and continued to grow. Their roots curved backward above the orbits and occupied much of the skull roof. In cross-section the teeth were triangular rather than thin, blade-like knives; enamel was very thin and did not cover every surface.

The tips diverged slightly. Long bony flanges on the lower jaw lay beside the upper canines with the mouth closed. They may have reduced the chance of lateral damage. A soft-tissue sheath around the teeth is not fossilised and remains one possible reconstruction, not a preserved structure.

The upper incisors were absent, lower canines reduced, and the postcanine teeth did not form the same clean shearing pair seen in many sabre-toothed cats. Blunt wear on their crowns differs from the strongly cutting wear expected in a conventional hypercarnivore. That combination prompted questions about the classic model.

What biomechanical models can test

Comparative modelling in 2013 found that jaw-closing muscles alone produced a relatively weak bite when the mouth was opened widely. The skull resisted loads applied as the head was driven by strong neck muscles. One possible motion therefore used the neck to guide the canines into tissue rather than relying on a powerful bite with the jaw nearly closed.

A low modelled bite force does not mean the animal could not kill. Very long canines required a large gape, a position in which jaw muscles lose mechanical advantage. Strong neck muscles and stable forelimbs could compensate for that limitation.

At the same time, the slightly divergent canines and loss of the upper incisors make a direct copy of a Smilodon-style bite uncertain. Models test how bones respond to specified loads; they do not reveal which movement a living animal used most often.

Body, mass and movement

The postcranial skeleton indicates a robust terrestrial animal with especially strong forelimbs. The elbow and hand were suited to stable support and forceful movement, rather than long, fast pursuit. The hind limbs were comparatively less powerful.

Mass estimates range from about 80 kilograms to above 100 kilograms, with some calculations higher still. Results vary according to whether a humerus, femur or skull measurement is used and whether the comparison sample consists of cats or other metatherians. It is more useful to picture a jaguar-sized animal with a heavier front end than to assign one exact weight.

Strong forelimbs could restrain prey or hold a carcass during feeding. The skeleton cannot tell whether those movements were performed on a living animal, so mechanical capacity is not observed behaviour.

Predator, carcass opener or both?

The traditional hypothesis describes an ambush hunter: a short rush, a forelimb hold and a neck-driven bite into soft tissue. Powerful shoulders, a wide gape and resistance to vertical loading fit that sequence.

In 2020 Christine Janis and colleagues proposed an alternative. The lack of upper incisors would make scraping meat from bone difficult; blunt postcanine wear did not look like that of a typical meat slicer; and canine shape might suit opening carcasses. They suggested feeding on soft organs and a possible scavenging specialisation.

The study did not prove that Thylacosmilus never hunted. An unusual predator could have killed some prey and also opened carcasses, using its teeth differently from a cat. The most accurate summary is a specialised consumer of animal tissue whose proportions of hunting, scavenging and carcass processing remain unknown.

Environment and disappearance

Thylacosmilus lived in open woodland, scrub and plains of South America. Notoungulates, litopterns, large armadillo-like mammals and terrestrial birds formed part of the wider fauna. Changes in Miocene herbivore and predator communities altered the available ecological opportunities.

Its disappearance has sometimes been linked to northern cats arriving after the formation of the Isthmus of Panama. Long, secure coexistence with Smilodon has not been shown: Thylacosmilus disappeared before large sabre-toothed cats became widespread in South America. Climate and faunal changes provide a more complex background than direct replacement by one competitor. Similarity to large omnivores such as Entelodon is superficial; the two belonged to different branches and lived on different continents.

What the skeleton does not preserve

Fossils directly confirm the skull, open canine roots, mandibular flanges and powerful postcranial skeleton. Hair is likely, but colour and length are unknown. External ears, lips and soft tissues around the protruding teeth are not preserved.

Stripes, spots and a cat-like muzzle are artistic choices. A careful reconstruction retains the narrower skull, absence of a cat-like tooth row and heavy forequarters instead of copying a feline face onto the animal.