Erlikosaurus: a beaked therizinosaur from Mongolia

Erlikosaurus andrewsi was a therizinosaur from Late Cretaceous Mongolia, probably living about 96–90 million years ago. Its nearly complete, three-dimensionally preserved skull is among the most informative therizinosaur heads known. The rest of the skeleton is limited to neck, shoulder and foot material, so its familiar body shape is partly reconstructed from relatives.

The skull preserves a toothless front to the jaws, evidence for a keratinous beak, and many small teeth behind it. CT scans revealed internal spaces, nerve canals and clues to the jaw muscles. Together these features support mainly plant feeding, while leaving the exact diet and many details of the soft tissues uncertain.

Quick facts

Name and discovery

Erlikosaurus joins the name Erlik, a ruler of the underworld in Turkic and Mongolian mythology, with the Greek word for “lizard”. The species name andrewsi honours Roy Chapman Andrews, the American explorer who led well-known Central Asian expeditions.

A Soviet–Mongolian expedition discovered the fossils at Bayish Tsav in 1972. The type specimen is commonly catalogued as IGM 100/111 or MPC-D 100/111, depending on the museum numbering system. It includes an almost complete skull with lower jaws. Associated material includes neck vertebrae, an almost complete humerus and an incomplete foot.

Altangerel Perle named Erlikosaurus andrewsi in 1980. A more detailed account appeared the following year using the spelling Erlicosaurus, but the original spelling Erlikosaurus has priority. These names do not refer to two animals; they reflect a nomenclatural correction in the early literature.

At the time, therizinosaurs were still often treated as a separate, puzzling group. A toothless beak, small cheek teeth, unusual jaws and a robust foot did not fit the standard picture of a predatory theropod. Erlikosaurus supplied one of the key skulls that helped show how these animals fit within theropod evolution.

Classification and disputed relationships

Erlikosaurus is a theropod maniraptoran within Therizinosauria. In most analyses it belongs to Therizinosauridae, although its exact position inside that family changes with the anatomical characters and taxa used. The overall classification of dinosaurs places it among specialised relatives of other maniraptorans, not outside Theropoda. The dinosaur catalogue lets readers compare its profile with those of related genera.

Some analyses place Erlikosaurus close to Segnosaurus, which lived in the same formation; others recover a different arrangement among derived therizinosaurs. A robust shared result is that Erlikosaurus was not an early transitional form at the base of the group. It was already a specialised member of its branch.

Only one species is generally recognised, Erlikosaurus andrewsi. A proposed synonymy with Enigmosaurus mongoliensis has not gained broad acceptance. Erlikosaurus is known mainly from a skull, whereas Enigmosaurus is known chiefly from pelvic bones. Because the two taxa preserve few directly comparable parts, their synonymy cannot be tested well. Keeping the names separate is the cautious choice.

The skull and its evidence

The type skull is about 25 centimetres long. It is damaged and deformed, yet it retained enough three-dimensional structure for computed tomography to separate bones virtually, correct some distortions and reconstruct its overall form. That makes it unusually valuable among therizinosaur skulls, which are often fragmentary.

The front ends of the upper and lower jaws lacked teeth. Numerous openings and canals in the bones indicate blood supply and innervation compatible with a keratin covering. A beak is therefore a well-supported inference, although its precise outline and the full extent of the soft tissues are not fossilised.

Behind the beak, the lower jaw carried 31 relatively small teeth. They were closely spaced and unlike the large slicing teeth of a typical active predator. The rear of the lower jaw was deep, and its joint and muscle attachments provide clues to feeding mechanics. A digital model estimated bite forces of about 43–65 newtons at the snout and 90–134 newtons at the rear teeth on each side. Those values are calculated from reconstructed muscles, not directly measured in a living animal.

The model nevertheless supports a relatively weak bite better suited to cropping or processing plant material than restraining large prey. It does not prove the whole diet. The skull lacks stomach contents and cannot tell us which plants were eaten or whether occasional animal food supplemented them.

Size and body reconstruction

The vertebral column, pelvis and femur are missing, so body length is estimated by comparison. A common range is roughly 3–4.5 metres. Estimates reaching six metres rely more heavily on the proportions of a chosen relative. Body mass is similarly uncertain, with published figures around 150–500 kilograms. The wide range reflects the incomplete skeleton, not measurement error in a complete animal.

The skull was low and elongated, with a large opening in front of the eye and extensive internal air spaces. CT work has examined the nasal passages, inner ear and brain cavity. The relative size of the olfactory bulbs suggests a developed sense of smell, while the elongated semicircular canals relate to balance and stabilising gaze. Estimates of hearing sensitivity resemble those of some other coelurosaurs.

These anatomical observations do not measure intelligence or reveal complex behaviour. Similarly, the almost complete humerus, neck elements and right foot confirm therizinosaur anatomy without supplying a complete body outline. Long hand claws, a broad pelvis, deep torso and feathers in artistic reconstructions are inferred mainly from relatives.

Feeding and habitat

The combination of a beak, numerous small teeth and a relatively weak reconstructed bite supports a mainly plant-based diet. Erlikosaurus may have clipped leaves and soft shoots with the front of the jaws, then held or partly cut food with the teeth. It did not chew with the side-to-side motion seen in mammals.

Some functional models suggest that neck movement could help pull vegetation while reducing the load on the jaw muscles. That is a mechanical hypothesis, not an observed feeding action. Invertebrates, eggs or other nutrient-rich foods cannot be ruled out, so “mainly herbivorous” is more defensible than “strictly herbivorous”.

The Bayn Shiree Formation belongs to the Cretaceous Period. Its deposits record river channels, floodplains, lakes and ground with seasonal changes in moisture. Other therizinosaurs shared the region with ornithomimosaurs, tyrannosauroids, dromaeosaurids, ankylosaurs, early hadrosauroids, turtles and crocodylomorphs.

Erlikosaurus and the larger Segnosaurus may have divided feeding resources. Their jaw proportions and tooth shapes differ, allowing the possibility that they selected plants of different toughness or fed in different ways. No stomach contents or securely associated coprolites prove that ecological interpretation, and the fossils do not show the two species interacting.

There is no direct evidence for herds, nests or parental care in Erlikosaurus. These behaviours cannot be borrowed from other maniraptorans. Feathers are plausible because of its position in the family and are directly known in the earlier Beipiaosaurus, but the density and distribution of an Erlikosaurus coat remain unknown. Colour is not preserved.

Reconstruction limits and scientific value

Illustrations often give Erlikosaurus the huge claws, broad pelvis and deep, feathered body familiar from larger relatives. This is a reasonable reconstruction of a therizinosaur body plan, but specific proportions of the hand, pelvis, tail and torso are not established for Erlikosaurus itself. The skeleton does not preserve all those parts.

The proposed synonymy with Enigmosaurus is another unresolved issue. Sharing a geological formation is not enough to merge two taxa when their known bones barely overlap. New material preserving comparable skull and pelvic features would be needed to test the idea.

A weak bite also does not mean the animal was defenceless or restricted to tender leaves. Bite mechanics describe one part of its anatomy. Body size, forelimbs and behaviour may have mattered, but the incomplete skeleton cannot establish how it protected itself. The strongest conclusion remains that the skull combines a beak, small teeth, complex sensory anatomy and a jaw better suited to plant processing than to subduing large prey.

Erlikosaurus is especially valuable because it records the head of a specialised theropod in unusual detail. The skull helps explain how a branch of maniraptorans evolved away from the typical predatory feeding apparatus while retaining the ancestry and broader anatomy of theropods. Its importance comes from the fossils actually preserved, not from a fully known body.