Behavior is often an elusive component of dinosaur biology. Scientists cannot observe dinosaur behavior directly, so instead they must rely on information from bones and other types of fossils left behind by living dinosaurs.

Feeding

Some types of behavior are quite general, and present less difficulty for scientists. General feeding behavior can often be deduced from the shapes of teeth and jaws. Herbivores tend to have flat or leaf-shaped teeth, many of which may be worn down by constant grinding of food. Herbivore jaws have low joints, which provide better leverage for precise chewing. In contrast, carnivore teeth tend to be curved, pointed, and serrated. The teeth usually do not meet each other precisely, and there is often little difference between teeth in different parts of the jaw.

More specific inferences about feeding behavior are problematic. It is often difficult to identify exactly which plants or animals a particular dinosaur fed on, or how this feeding took place. Occasionally dinosaurs are preserved with stomach contents (that is, their last meal) intact, but such examples are rare. Some specimens preserve tooth marks that can be matched to the teeth of particular predators. Dinosaur feces can be preserved as coprolites, but it can be difficult to match up a coprolite with the dinosaur that produced it.

Dinosaurian herbivores show varying degrees of chewing ability. Sauropods performed perhaps the least oral processing of food, using their teeth to crop or strip plants from their branches. Indeed, the small heads of sauropods may be likened to the front part of the mammalian skull, which houses the incisors and canines only. Food grinding would have been accomplished in the gut, which housed a collection of swallowed stones called gastroliths for this purpose. Most ornithischians appear to have possessed cheeks and beaks along with their usual complement of teeth, implying a more sophisticated degree of oral processing. Cheeks would have permitted food to remain in the mouth for extensive chewing, as indicated by the wear surfaces on the teeth.

In some dinosaur herbivores—ceratopsians and hadrosaurs in particular—these teeth are specialized into a tightly packed battery with a single, large surface for cutting and grinding plants. This surface probably functioned like the molars of mammals, allowing these dinosaurs to extensively chew their food.

Dinosaur predation has been the subject of great speculation in recent years. Some theropods are thought to have been pack-hunters, using the benefits of group behavior to bring down prey that were larger than themselves. This suggestion is controversial because there is no direct evidence of this type of behavior. Some associations of fossils suggest that certain theropods may have lived or hunted together, but the evidence remains equivocal.

Paleontologists also debate whether predatory dinosaurs actively hunted their prey, or merely scavenged. Evidence of healed injuries on herbivorous dinosaurs indicates that at least some theropods attacked living prey items. The information from theropod anatomy itself is often more difficult to interpret. For example, with structural adaptations such as large olfactory bulbs in the brain (for processing the sense of smell), robust teeth (for puncturing bones), and reduced forelimbs, tyrannosaurs may have been well-adapted for a scavenging lifestyle. On the other hand, tyrannosaurs also possess unusually long legs with interlocked foot bones (similar to those of modern ungulates) and some degree of binocular vision, implying that these giants may have been fully capable of hunting live prey. It seems more likely that dinosaurian predators were similar to modern large carnivores, feeding on dead or live prey opportunistically. Smaller theropods show a variety of body designs that suggest a wide range of predatory habits. Slender, long-limbed troodontids, for example, may have been lithe predators, but the bulkier dromaeosaurs were much more powerfully built.