Edentosuchus

Edentosuchus tienshanensis was a small crocodyliform from Lower Cretaceous deposits of the Tugulu Group in the Wuerho area of Xinjiang, northwestern China. It is best known for a short, high skull and an unusually varied tooth row: pointed front teeth and broad, multicusped rear crowns occurred in the same jaws. The genus is included in the ancient crocodylomorph catalogue.

For decades the animal was reconstructed from two incomplete skulls and a few neck vertebrae. Two specimens reported in 2026, including one with a much more complete palate and skull base, add important cranial anatomy and CT-based comparisons. They improve knowledge of the head but do not turn the historical sample into a complete associated skeleton.

Quick facts

A name based on fragmentary fossils

Chinese palaeontologist Yang Zhongjian, also known as C. C. Young, established the genus and species in 1973. The holotype, IVPP V 3236, preserves part of the skull roof, the front of both lower jaws and the beginning of the neck. Its incomplete condition made the animal difficult to place. Early comparisons suggested a relationship with primitive crocodyliforms, but the available characters did not settle its position.

A joint expedition of the Natural History Museum of Los Angeles County and the National Geological Museum of China collected another skull in 2000. GMPKU-P 200101 comes from Lower Cretaceous beds of the Tugulu Group roughly 10–20 kilometres southeast of Wuerho. Its front and temporal parts of the skull are joined to the front halves of the lower jaws. The assignment to Edentosuchus is supported by overlapping anatomy: the tooth-bearing bones, long lower-jaw symphysis and tooth shapes can be compared directly with the type.

A separate specimen, GMPKU-P 200102, preserves articulated vertebrae, a pelvis and much of a hind limb. It is a small basal crocodyliform from the same general setting, but it lacks a skull and shares no diagnostic bones with the known Edentosuchus specimens. It cannot be confidently assigned to this genus. Its limbs therefore should not be used as if they were the proven legs of Edentosuchus.

New material changes what can be seen

In 2026, Yuxin Tan and colleagues described two new specimens from the Tugulu Group at Wuerho. The more complete one preserves the palate, the back of the skull and the part of the lower jaw behind the tooth row, regions that earlier fossils had left poorly known. The second adds further rostral material. Computed tomography allowed the team to reconstruct obscured internal surfaces and compare the skull digitally.

The new specimens show that variation among the known skulls is not automatically evidence for separate species. The authors interpret some differences as related to ontogenetic stage, meaning that the individuals were at different points in growth. Their updated phylogenetic analysis supports a position within Protosuchidae. This is a result of the character matrix and the fossils currently included; it remains a scientific hypothesis that later discoveries can test.

The 2026 paper improves cranial knowledge, especially of the palate and occipital region. It does not establish a complete skeleton or resolve every question about the postcranial body. The distinction matters because older reconstructions sometimes borrowed a body from the unrelated GMPKU-P 200102 specimen found in the same region.

A short skull with unusual proportions

The better-known skull, GMPKU-P 200101, has a relatively short, tall and moderately broad snout. Large eye sockets face upward and outward. Much of the outer bone surface carries close-set pits, while the sides of the jaws beside the tooth rows are smoother and bear deep openings for vessels and nerves. The separate antorbital opening found in many archosaurs is absent, and the lower temporal opening is reduced.

The lower-jaw symphysis, where the two dentaries meet, extends unusually far back and involves much of the tooth-bearing region. Behind the enlarged lower canine the jaw narrows sharply before widening again toward the eye region. These details help distinguish the skull from other early crocodyliforms. They should not be confused with a direct record of bite force or feeding behaviour, which would require additional functional evidence.

The 2026 specimens add the palate and back of the skull to this picture. CT-based reconstruction makes hidden surfaces available for comparison and provides more characters for phylogenetic analyses. It does not remove deformation or incompleteness from every fossil, so conclusions still depend on which parts are preserved and how they are scored.

Three kinds of upper tooth

Each maxilla carries five teeth in the described skull. The first two have three cusps: one large central cusp and two smaller accessory cusps. The third and fourth are nearly square in cross-section. Their chewing surfaces carry curved rows of small cusps around three larger central elevations. The fifth is the largest and roundest, with a prominent central cusp and two smaller additions. Rather than sitting in completely separate sockets, the upper teeth occupy a continuous groove.

Nine teeth are preserved in each lower jaw, although the front of the tooth row is incomplete and the original count may have been higher. The second lower tooth is enlarged into a canine-like form. When the jaws closed, it fitted into a notch between the premaxilla and maxilla. Seven smaller teeth follow it and increase in size towards the rear; some echo the complex crown pattern of the upper cheek teeth.

This condition is called heterodonty: teeth in different positions have different shapes. A front tooth could pierce or hold an object, whereas broad rear crowns could crush or process it. Those are mechanical possibilities, not a proven diet. No stomach contents, coprolites or diagnostic feeding traces have identified a meal for Edentosuchus.

What the teeth suggest about food

The multicusped rear crowns are poorly suited to simply slicing large pieces of flesh. They may have processed hard-bodied prey or a varied diet. Small vertebrates, arthropods and other animals available around floodplains are plausible foods; some plant material cannot be ruled out. The shape of a tooth indicates how it could interact with food, but does not identify every item the animal ate.

Similar multicusped teeth evolved in other crocodyliforms, including Simosuchus, but the two genera do not group together in phylogenetic analyses. Complex crowns therefore arose independently in more than one branch. This is a useful warning against treating one conspicuous feeding adaptation as a reliable family-tree marker.

Where it fits among crocodyliforms

A 2004 redescription placed Edentosuchus within Protosuchidae, beside an informally named form from the Jurassic Kayenta Formation. That arrangement would extend the known history of protosuchids from the Early Jurassic into the Early Cretaceous. The deep branches in the analysis had weak support, however, and some slightly less parsimonious trees moved the genus closer to Mesoeucrocodylia.

The 2026 analysis again recovers the species within Protosuchidae using a revised character set and the new specimens. This is stronger evidence than the older, sparse sample could provide, but placement still depends on the taxon sample and character scoring. “Protosuchid” also does not mean a direct ancestor of living crocodiles. It describes an early branch, not a demonstrated line of descent.

The Tugulu landscape and limits of reconstruction

Near Wuerho, mudstones, shales and fine-grained sandstones record rivers and lakes. The region had a hot climate with marked dry seasons, broad floodplains and temporary water bodies that expanded during wetter intervals. Pterosaurs, turtles, psittacosaurids, theropods, sauropods and stegosaurs are also known from the broader fossil assemblage. Their presence describes the community represented in the rocks; it does not prove that Edentosuchus encountered each animal.

The fossils support a small crocodyliform with a distinctive skull and specialised teeth. A precise body length, limb proportions, number of armour rows and tail shape have not been established from securely associated remains. Illustrations of a complete animal necessarily use comparisons with relatives for those missing regions. The most defensible reconstruction keeps the skull evidence separate from the unassigned postcranial skeleton found nearby.