Dinosaurs shared Mesozoic habitats with mammals, crocodile-line reptiles, turtles, amphibians, insects and many other animals. Their coexistence is clear from fossils in the same formations. Direct evidence that two animals fed on, attacked or physically encountered one another is much rarer. A shared landscape is not automatically a preserved interaction.
A mammal with dinosaur remains
One Early Cretaceous fossil from Liaoning, China, preserves a skeleton of the mammal Repenomamus robustus with remains of a juvenile Psittacosaurus in its abdomen. The 2005 study described this as direct evidence that a Mesozoic mammal had eaten a dinosaur. The specimen establishes that at least some mammals were predators of small vertebrates, including young dinosaurs.
A second, exceptionally preserved pair from the Yixian Formation brings a different kind of evidence. A 2023 study described an entangled Repenomamus and Psittacosaurus. The mammal's position, grasping limbs and jaws led the authors to interpret it as the aggressor in an attack. Both animals appear to have been buried suddenly, perhaps by a volcanic debris flow. The preserved association is compelling, but details of the final seconds and the reason both were buried remain reconstructed.
Two dinosaurs locked together
The famous “Fighting Dinosaurs” fossil preserves a Velociraptor and Protoceratops in close contact in the Late Cretaceous Djadokhta Formation of Mongolia. The theropod's claw is lodged near the herbivore's neck, while the herbivore's beak is near the predator's arm. The specimen documents a dangerous encounter. Its exact sequence, whether it began as predation or defence, and how the pair came to be buried are not directly recorded.
What other traces can tell us
Tooth marks on bones, healed wounds and contents of fossil digestive remains reveal other interactions. A tooth embedded in a bone or a wound that healed before death can strengthen an interpretation of attack. A coprolite containing bone or plant tissue records ingestion. Even then, a feeding trace may not distinguish hunting from scavenging, and the species that produced a mark may not be identifiable.
Context changes how strongly a trace can be read. A puncture beside a matching tooth, repeated marks in anatomically plausible locations, or damage with signs of healing can narrow the possibilities. A mark on a detached bone has less context, and similar damage can arise during trampling, transport or preparation. Researchers therefore describe the observed feature first, then explain which behaviours are consistent with it. This keeps a plausible ecological story from being mistaken for something the fossil directly records.
Trackways can show that animals crossed the same surface or moved in similar directions. They rarely prove a chase by themselves because the tracks may have formed at different times. Researchers compare track shape, sediment, bone marks and body fossils before proposing an encounter.
Rare events, not daily behaviour
Exceptional fossils attract attention precisely because preservation of an interaction is uncommon. They demonstrate possibilities: some mammals ate dinosaurs, some predators attacked prey, and some encounters ended in injury or death. They do not measure how often these events happened or establish a universal rule for an entire group.
The more cautious question is not whether dinosaurs interacted with other animals, but which interactions are preserved well enough to identify. For more on the anatomical evidence and uncertainty behind hunting, see our guide to dinosaur combat fossils and our account of injuries preserved in bone.
Frequently asked questions
Did Mesozoic mammals eat dinosaurs?
Yes. Repenomamus fossils preserve dinosaur remains in the mammal's abdominal region, direct evidence that at least some mammals ate young dinosaurs.
Does the Repenomamus-Psittacosaurus fossil prove exactly how the attack happened?
It preserves an entangled association interpreted as predation, with the mammal positioned as the aggressor. The precise sequence and burial event cannot be observed directly.
Were Velociraptor and Protoceratops fighting when they died?
Their fossil preserves close physical contact and likely combat, but it does not record every movement or prove the encounter began as a hunt rather than defence.
Can bite marks distinguish hunting from scavenging?
Sometimes other context helps, but many marks can be made on a carcass either before or after death. A bite mark alone rarely tells the full story.