Majungasaurus: the abelisaurid of Cretaceous Madagascar

Majungasaurus crenatissimus was an abelisaurid theropod of Late Cretaceous Madagascar. It lived during the Maastrichtian, approximately 70–66 million years ago, in what is now the island's north-west. Several skulls, partial skeletons and isolated bones make it far better known than many other abelisaurids. Its short, high skull carried a low rounded ornament, while its four-fingered arms were extraordinarily small. Tooth marks on sauropod and Majungasaurus bones show that it fed on large carcasses, including members of its own species. They do not reveal whether the animals eaten were hunted or already dead.

Comparing it with Abelisaurus helps separate family traits from what is actually preserved in one genus. The Madagascan fossils document the reduced hands, legs and much of the skull directly. Abelisaurus is known only from one incomplete head, so a similar body for it is inferred rather than found.

Quick facts

The names and the long history of discovery

Majungasaurus means lizard from Majunga, an older spelling associated with the city and province now called Mahajanga. The species epithet crenatissimus refers to the strong serrations along its tooth edges. French military personnel collected the first relevant bones in Madagascar in 1895. Charles Depéret described two teeth, a claw-bearing bone and vertebrae in 1896 under the name Megalosaurus crenatissimus. At that time, Megalosaurus was often used as a broad holding category for poorly understood large theropods. Depéret later moved the species to Dryptosaurus, another assignment that did not reflect its true relationships.

René Lavocat studied an incomplete right dentary from the same deposits in 1955. Catalogued MNHN.MAJ 1, it became the basis for the genus Majungasaurus. Because the original material was insufficiently diagnostic, this jaw was later selected as the neotype of M. crenatissimus, a replacement name-bearing specimen. The International Commission on Zoological Nomenclature formally conserved that solution in 2011.

Another naming error began in 1979. A dome-like piece of skull roof was named Majungatholus atopus and interpreted as a pachycephalosaur. The discovery of the much more complete skull FMNH PR 2100 in 1996 showed that the “dome” was part of the skull ornament of an abelisaurid. Majungatholus atopus is consequently treated as a junior synonym of Majungasaurus crenatissimus, not as a separate thick-headed dinosaur.

Classification and species

Majungasaurus is a theropod within Ceratosauria, Abelisauroidea and Abelisauridae. Abelisaurids generally had deep, shortened skulls with coarsely textured facial bones, robust necks, much-reduced arms and strong hind limbs. The Madagascan genus is one of the best sources of direct anatomical evidence for that combination. Its position within the family has varied in detailed analyses, and some place it in a branch called Majungasaurinae alongside Indian and other Gondwanan forms.

The membership and exact branching of that narrower group depend on the taxa and characters analysed. It is firmer to call the animal a derived abelisaurid than to treat one particular diagram as permanent. The classification guide explains why ranks and evolutionary branches are not identical. Only M. crenatissimus is considered a valid species of the genus; old combinations with Megalosaurus and Dryptosaurus and the name Majungatholus belong to its study history.

What the fossils preserve

The main finds come from the Maevarano Formation, particularly the Anembalemba Member near Berivotra. Material includes several nearly complete or partial skulls, jaws, teeth, vertebrae from several regions, ribs, shoulder and pelvic elements, forelimb bones, almost complete hind limbs and portions of the tail. At least three partial postcranial skeletons complement one another. No single individual preserves every bone, but the combined material covers much of the skeleton.

Skull FMNH PR 2100 is especially informative. Its bones are largely separated but not badly crushed, allowing detailed study of the snout, palate, braincase and inner ear. Other specimens include younger and mature animals. Across growth, the skull became relatively taller and more robust, sutures between bones interlocked more strongly, and the eye socket became smaller relative to overall head size. That pattern would be hard to identify from one specimen alone.

Museum mounts often combine casts from several individuals and reconstructed missing regions. They convey the well-supported body plan but are not replicas of a complete fossil found articulated in the ground. Some vertebrae and foot bones have healed injuries or abnormal growth. These demonstrate that individual animals survived damage. They do not identify the cause, and a healed lesion cannot automatically be turned into a story about combat between members of the species.

Size, skull and teeth

Better represented specimens are commonly restored at roughly 5.5–7 metres long and around 750–1,100 kilograms. Fragments from larger individuals allow larger estimates, but values above seven or eight metres are particularly sensitive to assumed proportions. No one complete vertebral column fixes total length. A range with an explanation is more useful than one exact number for the species.

Adult skulls could reach roughly 60–70 centimetres. They were short, high and wide, with a blunt snout. Many outer bone surfaces carried coarse sculpture. The nasal bones thickened and fused, while a low rounded median projection rose from the frontals. It may have supported keratinous tissue, but neither its precise soft-tissue outline nor its colour is known. The projection should not be depicted as a tall, proven horn or used as evidence for head-butting.

Teeth were fairly low, laterally compressed and strongly serrated. Reconstructions of the full tooth complement count four premaxillary, seventeen maxillary and seventeen dentary positions on each relevant side. Teeth were replaced during life, so all positions did not necessarily contain equally mature crowns at one instant. The robust neck, with substantial muscle attachments on vertebrae and ribs, could have helped stabilise the head against feeding forces. Long-term gripping with the jaws is a functional proposal, not behaviour directly observed in the fossils.

Tiny arms and powerful legs

The forelimbs were exceptionally short. The forearm was only about a quarter of upper-arm length, wrist bones were not ossified, and each hand retained four very short fingers with few phalanges. Finger mobility was probably limited. Muscle attachment marks show that the arms need not have been entirely inert, but they were not suited to grasping large prey. They should not be treated as miniature versions of the more flexible hands of many other theropods.

The hind limbs bore the animal's weight, and a long tail balanced the head and trunk. It walked on two legs. Robust limb proportions alone do not yield a reliable numerical running speed. Neither the image of an especially sluggish predator nor that of a fast sprinter can be established from these bones. The skeletal structure constrains movement, but precise performance requires models and assumptions about muscles and soft tissues.

Maevarano habitat and feeding traces

Near the end of the Cretaceous Period, Madagascar was already an island. Maevarano sediments accumulated on a low river plain with pronounced seasonality. Long dry intervals alternated with heavy rain, flooding and sediment-rich flows. Rapid burial in channels and floodplain deposits helped preserve many vertebrates. The wider fauna included the sauropods Rapetosaurus and Vahiny, the smaller theropod Masiakasaurus, birds, snakes, turtles and diverse crocodylomorphs.

Majungasaurus was the largest securely known terrestrial predator in that assemblage. Tooth marks matching its dentition occur on sauropod bones. They show that it processed sauropod carcasses, not whether it killed those animals. Comparable grooves and punctures also occur on bones of several Majungasaurus individuals. Their size, shape and placement are consistent with a large local theropod, and no other known terrestrial predator of matching size occurs in the deposits.

These marks provide strong evidence that it consumed members of its own species. “Cannibalism” here describes feeding, not necessarily active killing of another adult. Scavenging, taking an occasional opportunity or killing a competitor could all leave similar traces. The fossils cannot reveal frequency, circumstances or a preferred social strategy. Nor does a set of individuals in one formation prove pack hunting, permanent groups, territoriality or parental care.

What remains reconstruction

The old pachycephalosaur identification of Majungatholus was overturned by better skull material. Claims that the median projection served as a battering ram remain unsupported; display, recognition and other functions are possibilities rather than findings. Its sex differences, mating rituals, sounds and exact hunting style are unknown. No securely described skin impressions establish the precise covering, scales, colours or any filaments of this genus.

The fossils give a detailed skeletal picture and a rare direct record of feeding on its own species. They do not tell us how those individuals died. In the dinosaur catalogue, Majungasaurus illustrates how skulls, complementary skeletons and marked bones can support stronger conclusions than one isolated specimen, while still leaving behaviour and appearance partly open.