Thrinaxodon: a small cynodont in an Early Triassic burrow

Thrinaxodon liorhinus was a small cynodont that lived in the Early Triassic, soon after the end-Permian mass extinction. Fossils from South Africa and Antarctica preserve skulls, partial skeletons and rare animals inside sediment-filled burrows. They show an animal with a mixture of older therapsid features and traits that later became more specialised among mammals. Thrinaxodon was neither a dinosaur nor a mammal, and the fossils do not establish that it was a direct ancestor of mammals.

Quick facts

Name and type specimen

Harry Govier Seeley named Thrinaxodon and its species T. liorhinus in 1894. The genus name combines Greek elements commonly interpreted as “three-pronged” and “tooth”; the species name refers to the form of the nose. The type, NHMUK PV R 511, is held by London's Natural History Museum. It preserves more than a skull, giving researchers some postcranial anatomy as well as the teeth and jaws.

Older literature assigned several names to fossils now generally treated as Thrinaxodon, including Ictidopsis elegans, I. formosa, Thrinaxodon putterilli, Notictosaurus luckhoffi and Micrictodon marionae. Reassessment of overlapping anatomy reduced this list. A proposed synonymy with the poorly known Nythosaurus larvatus has remained less secure. The accepted species is T. liorhinus.

Fossils in South Africa and Antarctica

Many specimens come from the Lystrosaurus Assemblage Zone of the Beaufort Group in South Africa's Karoo Basin. The deposits formed after the end-Permian extinction, about 252 million years ago. Lystrosaurus occurs in the same broad fossil record, but shared layers do not prove that the two animals interacted regularly.

South African collections include dozens of skulls of different sizes and partial or nearly complete skeletons. This sample makes it possible to compare growth stages, tooth replacement and variation between individuals rather than extrapolating from one specimen.

In 1977, Edwin Colbert and James Kitching described 14 specimens from the lower Fremouw Formation in the Transantarctic Mountains. During the Early Triassic, Antarctica and South Africa were joined within Gondwana. The southern fossils therefore document a broad high-latitude range across connected land, not a crossing of today's ocean or polar ice.

Skull, teeth and jaw movement

The skull was large relative to the body, with an expanded region behind the eye and a prominent cheek arch. The dentary, or tooth-bearing bone of the lower jaw, had become enlarged. Several smaller bones at the rear were reduced but still participated in the jaw joint. The fully mammalian joint between the dentary and the skull had not yet evolved.

The teeth differed along the jaws: incisors at the front, large canines behind them and multi-cusped postcanine teeth farther back. The canines could seize prey, while the postcanines could puncture and cut. Their contact was less precise than in many later cynodonts and mammals, so Thrinaxodon should not be described as having a fully modern chewing system.

Muscle attachment surfaces and digital models allow researchers to infer separate components of the jaw-closing apparatus, broadly comparable with the temporal, masseter and pterygoid muscles of mammals. These are reconstructions from bone. The muscles themselves are not fossilised.

Secondary palate and breathing

Projections of the upper jaw and palate formed a substantial bony partition between the nasal passage and the mouth. A secondary palate can permit breathing while food is held or processed in the mouth. In Thrinaxodon, however, the bony partition was not completely closed along the midline.

A study of 68 skulls spanning different growth stages found a persistent median gap. Soft tissue may have covered some of it, but that tissue is not directly preserved. It is therefore more accurate to call the palate well developed but incompletely ossified than to present it as identical to the closed palate of a modern mammal.

The vertebral column and ribs also show a division between trunk regions. This has been compared with the conditions that could support a muscular diaphragm. A diaphragm is plausible, but no soft muscle survives and the arrangement of the skeleton cannot reveal its exact shape.

Limbs, posture and size

The forelimbs were more widely set than the hind limbs, which could be placed closer beneath the body. Its posture was not exactly that of a sprawling lizard or a modern mammal. Joint surfaces indicate a range of movement, so a single static illustration cannot represent every stance or gait.

Adult skulls can reach roughly 9–10 centimetres in length. Whole-body estimates of about 40–50 centimetres use proportions from the preserved skeleton and related cynodonts; they are not measurements of an unbroken specimen from nose to tail. No trackway has securely established a top speed. The short, strong limbs are consistent with terrestrial movement, while the burrows provide better evidence of sheltering behaviour than limb proportions alone.

Bone sections show rapidly deposited fibrolamellar tissue in young animals and more slowly formed parallel-fibred tissue nearer the outer surfaces of larger bones. This pattern supports rapid growth early in life followed by slowing. Clear annual pauses are weak or absent in many samples, but that does not by itself establish a constant body temperature or a mammal-like metabolism.

Skulls from about 30 to 96 millimetres change in the proportions of the snout, temporal region, eye socket and palate. Tooth replacement also continued during growth. Some names once applied to separate forms may therefore have reflected age-related differences rather than distinct species.

Burrow BP/1/5905

A sandstone cast of burrow BP/1/5905 contains an articulated Thrinaxodon skeleton. The shape of the tunnel and the animal's position show that a cynodont occupied the burrow about 251 million years ago. A reconstruction of the skeleton in cross-section found that its body and limbs could fit in the narrow passage.

The fossil directly records presence inside a burrow, not who made it. Thrinaxodon may have excavated the shelter or reused a tunnel made by another animal. Similarity to other cynodont burrows makes active digging a reasonable interpretation, but does not prove the origin of every individual tunnel.

Burrows could have buffered daily temperature swings, drought and predators in the unstable ecosystems that followed the extinction. Those are possible benefits, not all demonstrated functions of this particular structure. An individual cast does not show that every Thrinaxodon spent most of its life underground, hibernated or lived in a stable pair.

The Thrinaxodon and Broomistega association

Another burrow cast, BP/1/5558, contains two nearly complete skeletons identified by synchrotron scanning: Thrinaxodon specimen BP/1/7199 and a young temnospondyl amphibian, Broomistega putterilli BP/1/7200. They lie close together and were buried by several pulses of sandy sediment.

The amphibian had healing rib fractures, and its anatomy was not well suited to digging. Researchers proposed that the injured Broomistega entered a burrow already occupied by Thrinaxodon. The skeletons are not dismembered and show no convincing evidence that the amphibian was eaten, making a predator-and-meal interpretation unlikely.

Calling the pair “friends” is a modern story, not a fossil result. The evidence does not reveal whether both were alive at the same moment, whether the burrow owner tolerated the visitor or whether either animal was in a torpid state. The secure observation is an unusual joint burial in one burrow, which rules out some possible scenarios without recording social intent.

Diet and ecological role

The canines, multi-cusped postcanines and inferred jaw muscles fit the capture of small animal prey. Insects and small vertebrates are plausible foods, but no stomach contents or securely associated coprolites identify a precise menu. Tooth form indicates what the jaws could do more readily than it identifies every meal.

“Insect-eater” is therefore a possible summary, not a demonstrated dietary specialisation. Thrinaxodon may have taken several kinds of animal food that were available. Fossils do not fully exclude scavenging or occasional plant material. Its ecology should be reconstructed from the whole evidence rather than from one tooth feature.

Where it sits on the mammal line

Thrinaxodon was an epicynodont, more closely related to mammals than early synapsids such as Dimetrodon or gorgonopsians. The Permian synapsid guide explains how those branches relate without treating one fossil genus as a direct ancestor of another.

That position does not make Thrinaxodon a proven direct ancestor of mammals. It is a well-sampled member of an early side branch that preserves a useful mix of characters. Mammals arose later within more derived cynodonts. Like other early synapsids, it belongs in the site's ancient mammal catalogue as a close relative on the wider evolutionary story, not as a true mammal.

What a reconstruction can and cannot show

The skeleton constrains a large head, elongated trunk, long tail and short limbs. Limb placement and movement varied, so a single frozen pose simplifies its biomechanics. No skin, hair, whiskers or external ears have been recovered for Thrinaxodon.

Filamentous covering is possible by comparison with evolutionary relatives, but a dense coat is not directly known. Colour is entirely speculative. The burrow casts document use of underground shelter; they do not show hibernation, permanent pairs or care of young. A careful reconstruction separates these choices from the anatomy preserved in stone.