Ekrixinatosaurus: an abelisaurid from the Candeleros Formation

Ekrixinatosaurus novasi was a large abelisaurid theropod known from one partial, scattered skeleton in Argentina's Candeleros Formation. Its name refers to the modern blasting work that led workers to notice the bones, not to a volcanic eruption or the animal's death. The holotype, MUCPv-294, preserves parts of the skull, backbone, pelvis and hind limbs, but no forelimbs and no complete skull.

The genus was once assigned to the Early Cenomanian, around 100–97 million years ago. U–Pb dates published in 2026 place ash beds in the Candeleros and overlying Huincul formations in the Albian, earlier in the Early Cretaceous. The holotype's exact position relative to those dated beds has not been published, so its age remains approximate. The profile is part of the dinosaur catalogue, where comparisons do not turn formation-level dates into a precise birthday for one fossil.

Quick facts

Name and discovery

Ekrixinatosaurus is usually translated as “explosion-born lizard”. The name commemorates present-day blasting during pipeline construction after which workers noticed fossil material. It does not describe an ancient explosion, volcanic event or cause of death. The species epithet novasi honours palaeontologist Fernando Novas.

The holotype MUCPv-294 was recovered from red beds of the Candeleros Formation in Neuquén Province. The bones were scattered across an area of roughly 15 square metres rather than lying as a connected skeleton. The original description named a new abelisaurid in 2004. Since then, comparisons with more complete relatives have helped refine its size and phylogenetic position, while several large parts remain missing.

Classification and relatives

Ekrixinatosaurus belongs to Abelisauridae, a family of ceratosaurian theropods widely represented across Gondwana during the Cretaceous. Its family placement is well supported by the skull and postcranial features. Most analyses place it within Brachyrostra, a radiation of predominantly South American abelisaurids with relatively short, deep skulls.

Early analyses grouped it near taxa then called Majungatholus, now Majungasaurus, and carnotaurins. Later work commonly placed it near Skorpiovenator and Ilokelesia, outside the more derived Furileusauria that includes Carnotaurus and several related forms. The exact branching order changes between phylogenetic datasets. “Early or basal brachyrostran” is more cautious than calling it a direct ancestor of Carnotaurus.

Only one reliable species is recognised, Ekrixinatosaurus novasi. The genus is not based on a single tooth. The holotype has multiple diagnostic parts, but no second specimen is currently available to test normal variation, age-related change or differences between individuals.

What the fossil includes

The skull material includes the left maxilla, part of the right maxilla, the dentary bones of the lower jaw, elements from the skull base and isolated teeth. Much of the skull roof, the premaxillae and the braincase are missing. A complete head in an illustration must therefore be assembled by combining preserved bones with comparisons to other abelisaurids.

Vertebral material includes cervical, one dorsal, sacral and caudal vertebrae, along with ribs and haemal arches. The pelvis preserves iliac and pubic elements and proximal parts of the ischia. The hind limb is better represented: the holotype includes a complete left femur, part of the right femur, a left tibia, parts of the right tibia and left fibula, ankle bones, metatarsals, toe bones and a claw-bearing phalanx.

No forelimbs are included in the type material. Very short arms in reconstructions are inferred from better-preserved relatives, not measured in MUCPv-294. Calling the specimen “well preserved” does not make it complete. It is disarticulated and lacks extensive parts of the skull, trunk, tail and limbs.

Size and anatomy

The original authors estimated a total length of about 7–8 metres. A later estimate based on allometric comparison is close to 7.4 metres. The often repeated figure of 10–11 metres comes from scaling a large reconstructed head and is controversial. The femur and tibia provide direct measurements, but the full head, trunk and tail are not preserved, so a complete length must be modelled.

The left maxilla is about 42 centimetres long and contains 16 tooth sockets. The skull has been estimated at around 83 centimetres by comparison with Carnotaurus and Majungasaurus. That number is not a measurement of an intact head. Its jaws held laterally compressed, serrated teeth, and the preserved upper jaw suggests the deep snout typical of abelisaurids.

The left femur is about 77 centimetres long and the tibia about 69 centimetres. The shorter lower leg and robust bones suggest a strongly built animal rather than a long-legged cursorial specialist. The fourth trochanter of the femur provided attachment for major tail muscles. These features inform how the leg could extend and the body balance, but they do not yield a measured maximum speed.

The cervical vertebrae were short from front to back and carried strong muscle-attachment surfaces. They supported a powerful neck and a relatively deep head. Caudal vertebrae had large transverse processes associated with the tail musculature. Body mass is much less secure than the lengths of individual preserved bones. A broad estimate around 1–2 tonnes is preferable to a false exact figure.

Revised age and the Candeleros landscape

Red beds of the Candeleros Formation were traditionally assigned to the Early Cenomanian, roughly 100–97 million years ago. In 2026, U–Pb dating of zircon crystals from an ash bed in the Candeleros yielded 111.4 ± 1.1 million years. An ash bed in the overlying Huincul Formation yielded 106.8 ± 2.2 million years. These dates place both units in the Albian, earlier than the traditional assignment.

The ash beds date volcanic material at particular stratigraphic levels. The exact position of MUCPv-294 relative to the dated Candeleros bed has not been published. The formation-level result therefore revises the geological framework without giving this animal an exact age of 111 million years. The fossil remains Albian in the updated framework, but its position within that interval is unresolved.

Candeleros sediments record rivers, channels, floodplains and areas shaped by wind in a seasonal, relatively dry continental setting. Watercourses supported diverse vertebrates. The formation has yielded sauropods, theropods, crocodylomorphs, turtles, amphibians and trackways. Some are known from separate locations and horizons. Their presence in the same formation does not prove that every named animal shared one local habitat at the same time.

Diet and behaviour

Serrated teeth, a theropod skeleton and the animal's placement among abelisaurids support a carnivorous diet. Its likely prey included vertebrates available in its environment, and it may also have scavenged. No stomach contents, coprolites or diagnostic bite marks connect a particular meal to Ekrixinatosaurus.

A short, deep skull and strong neck could have helped the animal resist forces while feeding. However, no full biomechanical bite study establishes the force or one preferred technique for this genus. Claims that it crushed bone, used its head as a battering ram or regularly attacked adult sauropods go beyond the known material.

Large abelisaurids and carcharodontosaurids both lived in the wider region. Their presence allows hypotheses about niche differences, perhaps in prey size, capture method, habitat or timing. It does not show how those differences worked. Group hunting, speed, colour, vocalisations, reproduction and parental care are not directly known.

Popular reconstructions and limits

The main size dispute centres on the 10–11 metre reconstruction. It uses a large estimated head and scales the rest of the animal from it. The measured hind-limb bones fit more comfortably with the earlier 7–8 metre range and the later estimate near 7.4 metres. Ekrixinatosaurus was a large predator, but the fossil does not establish it as the largest abelisaurid.

Paired horns like those of Carnotaurus are not present in the holotype. Much of the skull roof is missing, so the exact upper profile and any soft-tissue ornament remain uncertain. The forelimbs, tail outline, skin, colour and muscle shape also require comparison or artistic choice.

The species did not die in an explosion. Modern blasting led workers to the fossils. Nor is its skeleton complete: it preserves a useful set of bones, but many parts of the head and body are absent. A careful reconstruction should show what relatives suggest while keeping those missing features distinct from direct evidence.

What can be reconstructed

The fossil directly documents an abelisaurid jaw and teeth, parts of its backbone and pelvis, and robust hind limbs. Comparative anatomy supports a deep skull, strong neck and reduced forelimbs, but some parts of that outline are inferred. The best-supported total length is around 7–8 metres, not a precise record-breaking value.

New dating places the Candeleros Formation in the Albian, although the type fossil's exact level remains unresolved. Ekrixinatosaurus is important because it shows that large abelisaurids shared Patagonia with other theropod groups earlier than previously thought. The evidence supports a substantial predator with a partial skeleton; it does not supply a complete scene of its life.