Irritator: a spinosaurid known chiefly from its skull

Irritator challengeri is a spinosaurid theropod from the Early Cretaceous of northeastern Brazil. Its strongest evidence is a largely complete skull from the Romualdo Formation. The fossil was damaged and artificially altered before scientists studied it, but later preparation and CT data revealed far more anatomy than is visible on the surface. The skull is unusually informative; the rest of the animal remains largely a reconstruction.

Quick facts

A skull altered before study

The holotype is SMNS 58022, now held by the State Museum of Natural History Stuttgart. It was collected from the Araripe Basin of Ceará, Brazil, and is usually placed in the late Aptian part of the Romualdo Formation. The exact circumstances of its excavation and early trade history are poorly documented. That gap is part of the specimen’s provenance problem, not a reason to invent a more precise field record.

The fossil reached researchers with its snout artificially lengthened. Added material and rearranged fragments changed the apparent outline of the upper jaw. Early CT scans exposed the alteration, although the technology at the time could not resolve every internal surface. Subsequent preparation and modern medical and micro-CT studies allowed scientists to separate bone, matrix and later additions and to build a digital reconstruction.

The genus name refers to the irritation caused by discovering how the skull had been modified. The species name honours Professor Challenger, a fictional character in Arthur Conan Doyle’s stories. Neither name describes a behavioural trait. The first account in 1996 assigned the skull to a maniraptoran; later examination of its anatomy established it as a spinosaurid.

What the holotype preserves

SMNS 58022 preserves most of the braincase and much of the skull roof, palate and rear portions of the jaws. The premaxillae are largely missing, as are the front ends of the maxillae and the very front of both lower jaws. Some bones are displaced or compressed, but the specimen retains internal surfaces that are hidden in most dinosaur skull fossils.

Researchers have used CT segmentation to identify the palate, braincase cavities, tooth positions and contacts between the bones. The resulting digital model is a reconstruction of the fossil’s anatomy, not a scan of an undamaged living skull. Missing pieces have to be mirrored from the opposite side or inferred from adjacent bones, and the reconstructed parts should remain distinguishable from what is directly preserved.

The holotype has been reported as the most complete known spinosaurid skull, but it is not a complete skeleton. No body vertebral column, pelvis, arms, legs or tail is securely connected to it. Other spinosaurid bones from the Romualdo deposits may belong to Irritator, Baryonyx-like relatives or a separate taxon. Without diagnostic overlap, assigning them to this genus is uncertain.

Long snout, teeth and jaw mechanics

The skull has a long, narrow rostrum and nostrils set back from the tip. Its conical teeth lack the deep serrations typical of many meat-slicing theropods. They are suited to piercing and gripping prey. The front end of the upper jaw is incomplete, so a full count and exact spacing of the teeth have to be estimated. The preserved positions show replacement teeth inside the jaws.

CT reconstruction revealed a substantial palate and a more complete braincase than could be seen externally. The 2023 osteological reappraisal also described the surangular shelf, the retroarticular region and how the lower jaws meet the skull. A proposed model suggests that the rami could spread and rotate slightly as the mouth opened. This is a mechanical inference from the joint surfaces and bone geometry, not an observed motion in a living animal.

The study’s biomechanical interpretation indicates a relatively quick closure but not the crushing force expected from a deep, massive skull. The teeth and slender snout are compatible with seizing small prey, including fish. The skull does not preserve a meal, however, and tooth shape alone cannot tell us the entire diet. A spinosaurid could take fish and other available vertebrates.

A narrow snout and forward-facing field of view may have helped the animal aim at moving prey. Such estimates depend on reconstructing the missing front of the skull and on comparing eye orientation with other theropods. They should not be expanded into claims about a precise hunting strategy or a fixed head angle.

Classification and the Angaturama question

Irritator belongs to Spinosauridae, a group of long-snouted theropods from the Cretaceous. It is usually placed among Spinosaurinae, though the exact branching order within the family varies between analyses. Those changes partly reflect incomplete fossils for other spinosaurs and different choices about which characters to score.

Angaturama limai, another named spinosaurid from the Romualdo deposits, is known from a front part of a snout that complements the missing region of the Irritator skull. Researchers have asked whether the two names might represent one genus or even different parts of one skull. Both preserve a tooth at a comparable position, and their preservation and size are not identical. The proposal that they are one individual is not accepted as established fact.

Without overlapping diagnostic bones, synonymy remains difficult to prove or reject. The conservative account keeps Angaturama separate while noting the hypothesis. If future study demonstrates that the specimens belong to the same species, nomenclatural priority would need to be assessed using the publication dates and formal rules. At present the skull of Irritator remains the secure basis for its own name.

Age and the Romualdo environment

The Romualdo deposits formed in a changing basin with lagoons, coastal waters and continental settings. Their fossil assemblage includes fishes, turtles, crocodile-line reptiles, pterosaurs and dinosaurs. The sequence is part of the Araripe Basin, where concretions helped preserve delicate fossils. Local stratigraphy and the age model have been revised, so the fossil should be described by its assigned late Aptian interval rather than an unsupported exact year.

Fish were abundant, and the skull of Irritator is compatible with catching slippery prey. That makes piscivory a reasonable part of its ecology, but “fish-eater” is not the same as “fish-only.” Nor does feeding on aquatic animals show that a predator regularly swam underwater. These are separate questions requiring different anatomical evidence.

The braincase and inner-ear region have been studied for head posture and balance. They provide anatomical constraints, not a record of a particular daily behaviour. A fossil brain endocast preserves the space inside the skull, not the complete soft brain or its activity. Conclusions about agility, habitual swimming and sensory performance therefore require careful qualification.

How much of the animal can be reconstructed?

Length estimates around six to eight metres are obtained by scaling the skull against more complete spinosaurids. No full backbone or limb series lets researchers measure Irritator directly. Mass estimates have an even wider margin because the depth of the body, length of the neck and tail, and proportions of the limbs are unknown.

A long snout and conical teeth are supported by the skull. A high sail, powerful tail, broad feet and long forelimbs are often borrowed from other spinosaurs. Those features may be plausible for a spinosaurid, but they are not preserved in SMNS 58022. A useful illustration can show a comparative body plan while identifying which parts remain hypothetical.

The skull offers no direct evidence for skin colour, feathers, display patterns, calls, social groups or parental care. Nor does the specimen reveal a dramatic fight or a preferred prey species. The evidence-led picture is narrower: a spinosaurid with a long, slender skull, piercing teeth and unusual jaw anatomy lived in the Araripe region during the Early Cretaceous.

In the dinosaur catalogue, Irritator is valuable because one skull preserves anatomy that is missing in many relatives. A fuller body skeleton or securely associated elements could change estimates of its size and locomotion. For now, the skull is the centre of the account, and the rest should stay visibly provisional.