What did troodontid dinosaurs actually eat? When the scattered evidence is pieced together, many of them may have been omnivorous

Troodontid dinosaurs were bird-like in appearance, small-bodied, and agile. Their enlarged pedal claws and serrated teeth make them seem, at first glance, like typical small predators. In reality, however, their diet may have differed somewhat from what we have long imagined. Theropod dinosaurs have traditionally been pictured as primarily flesh-eating predators, such as tyrannosaurs, but to understand troodontids we first need to set aside that stereotype. These animals lived from the Late Jurassic to the Cretaceous and were distributed mainly across Laurasia, that is, what is now Asia and North America, and possibly Europe as well. Phylogenetically, they belonged to Paraves, so their ecology, especially their diet, may help us understand how the ancestors of birds gradually moved toward a high degree of ecological diversity.

Phylogenetic tree of Troodontidae and their tooth types（Image source：Fan YC et al. (2026)， CC BY 4.0 ）

Geographic distribution of troodontid fossil localities（Image source：Fan YC et al. (2026)， CC BY 4.0 ）

If we want the most direct evidence for diet, we need to look at what was preserved inside the body cavity. Unfortunately, fossils of this kind are extremely rare. There is, however, one well-known case from China: Jinfengopteryx elegans, which preserves several reddish-yellow oval structures in the abdominal region. The true identity of these structures remains controversial. They have been interpreted as seeds, nuts, eggs, or even ovarian follicles. Naturally, the last meal of a single individual cannot directly represent the habitual diet of an entire species, and there is considerable room for interpretive bias, so this evidence can only be discussed as an isolated case.

Reconstruction of Jinfengopteryx elegans（Image source：Matt Martyniuk， CC BY 2.5 ）

Another indirect line of evidence comes from gastric pellets. In living birds, indigestible material such as bones, hair, and feathers is often regurgitated as a compact mass. At Egg Mountain in North America, researchers reported two masses that may have been gastric pellets produced by Troodon formosus, and these contained mammalian bones. Even so, these masses were not found directly associated with troodontid skeletal remains, so the inference that they were produced by troodontids is based mainly on elimination of other likely candidates.

Reconstruction of Troodon formosus（Image source：Atlantis536，CC0 1.0 ）

Studies of dental microwear provide yet another clue. After comparing microwear on the teeth of several theropod dinosaurs, researchers found that the teeth of Stenonychosaurus inequalis show a wear pattern consistent with a "puncture-and-pull" feeding motion, one that is fairly similar to the way some living birds of prey process food. At the same time, those teeth rarely show abundant pits produced by repeated impacts against hard materials. This suggests that they probably did not regularly bite through bone, nor did they process very hard food items in the way some specialised bone-crushing carnivores do. The resulting impression is that they did eat meat, but were probably more focused on soft tissues, small prey, or swallowing prey whole, rather than processing carcasses the way a large apex predator would.

Reconstruction of Stenonychosaurus inequalis（Image source：Tomopteryx， CC BY-SA 4.0 ）

In some North American fossil localities, researchers have also found abundant shed teeth referred to troodontids together with the remains of newly hatched or very young hadrosaur individuals. This association immediately invites a predatory interpretation, namely that troodontids entered dinosaur nesting grounds to feed on eggs or hatchlings. That is a reasonable explanation, but there is another, more ecological interpretation as well. Nesting areas may themselves have been resource-rich environments. Not only would young hadrosaurs have been concentrated there, but troodontids may also have been drawn to the same places because they offered abundant soft vegetation, easily captured small animals, and even food resources that could be excavated from underground. Their co-occurrence with juvenile hadrosaurs, then, does not necessarily indicate only a predator-prey relationship. It may instead reflect attraction to the same high-resource area, or perhaps both situations occurred.

So were troodontids anatomically more like hunters, or more like omnivores? One of their most striking features is the enlarged second pedal digit. This claw is commonly interpreted as a tool for grasping and restraining prey, broadly comparable in function to the talons of modern raptorial birds. However, some researchers have suggested that it may also have served other purposes, such as digging for small animals in underground burrows or excavating subterranean plant structures. In evolutionary terms, a single structure often does not serve only one function. As long as it was useful, the same claw could probably have been used both for predation and for digging, depending on what opportunities the environment offered.

The teeth themselves tell a similarly complicated story. Many troodontids retained serrations, and such structures are indeed well suited to slicing flesh. Yet several Asian forms, including Byronosaurus, Mei, Jinfengopteryx, Xixiasaurus, and Urbacodon, have little to no serration. The absence of serrations does not automatically mean herbivory, but it does suggest that their feeding behaviour and food-processing strategies differed from those of more typically serrated forms. In other words, Troodontidae was not a clade with a single dietary pattern. In different regions and at different times, different troodontid lineages may have followed distinct ecological trajectories. If we simply take the diet inferred for a particular Late Cretaceous North American troodontid and apply it wholesale to all Early Cretaceous Asian troodontids, we are likely to be very wrong.

Serrated teeth of Troodon formosus（Image source：Joseph Leidy，CC0 1.0 ）

Some studies have used the chemical composition of fossil tooth enamel to estimate the trophic position of troodontids within food webs. The results place troodontid values between those of herbivorous dinosaurs and carnivorous theropods, suggesting that they occupied a trophic level intermediate between primary and secondary consumers. These chemical signals support the interpretation that troodontids were mixed feeders, perhaps even plant-dominated omnivores. At that point, the otherwise scattered evidence begins to fit within a single logical framework: the possibility of plant consumption, the ability to take small animals, the multifunctional role of the pedal claw, and the mixed signal coming from the teeth. Troodontids may therefore have been following an ecological pathway that differed from that of more typical carnivorous theropods from the outset.

Taken a step further, this perspective suggests that some ancestors of Deinonychosauria in a broader sense may also have been omnivorous. Modern birds show extraordinary dietary diversity, ranging from raptorial and fish-eating birds to fruit-eating, nectar-feeding, and omnivorous crows. That ecological flexibility may not have appeared suddenly only in later bird evolution. It may have had deeper roots, already established at the stage of troodontids and their close relatives. Omnivory does not tie an animal to a single resource base, and when environments change, it allows an easier shift to another way of life.

Of course, our current knowledge remains limited. Troodontid fossils are not especially abundant, and specimens that can be used directly to reconstruct diet are rarer still. In addition, Troodon formosus itself remains taxonomically controversial, and some researchers have argued that it may be a historical name that lumps together several similar species, to the point that part of the literature has regarded Troodon formosus as a nomen dubium. If that is true, then fossils currently referred to Troodon formosus may actually represent several animals with somewhat different diets, making dietary reconstruction even more complicated. For that reason, a clearer understanding of troodontid dietary evolution will require much more than additional isolated teeth. It will depend on further evidence from the skull, body proportions, biogeochemistry, and broader comparisons among specimens from different parts of Asia and North America.

Author: Shui-Ye You

Reference:

Fan YC et al. (2026). Diet of bird-like troodontid dinosaurs: synthesis of a contentious clade. Biological Reviews.