Did Plant-Eating Dinosaurs Really Only Eat Plants?

An ankylosaur fossil with fish in its belly provides ancient evidence that herbivore diets are more flexible than they’re assumed to be.

Ankylosaurs commonly are depicted as prey to larger, carnivorous dinosaurs. But one species may have been the hunter, not the hunted. (Vuk Kostic / Shutterstock)

Ask a non-biologist about the differences between an herbivore and a carnivore, and they’ll tell you that it’s written like destiny in an animal’s bones. Carnivores have sharp fangs and claws; herbivores blunt teeth and hooves or pads. There are some animals that dabble on both sides of the line, of course, but the categories of “predator” and “prey” are usually seen as inviolate.

When it comes to dinosaurs, known only from bare bones, the line between the two modes seems especially concrete. Take ankylosaurs: few herbivorous dinosaurs seem as dedicated to their role as plant eaters as these armored, stubby-legged, broad-bellied creatures. Their teeth were tiny; their beaks blunt; they were unable to chew. With their spike-adorned backs, they have traditionally been depicted as hunkering down in the face of ravening predators. Like modern livestock, their place in Mesozoic ecosystems seems set: Always the hunted, never the hunter.

But according to an August announcement by the paleontologist Ji Qiang and his colleagues, at least one species of Chinese ankylosaur, Liaoningosaurus, might have bucked the trend. Discovered in 2001, Liaoningosaurus is already odd as far as ankylosaurs go. The largest known specimens are about a foot long, far smaller than the species’ more famous club-and-spike-tailed kin. The first Liaoningosaurus fossil found preserved a swatch of petrified tissue that Ji claims represents a plastron, a bony structure on the stomach common in aquatic reptiles like turtles. The new specimen described has no plastron. Instead, it has a belly full of fish fossils, in addition to what appears to be the tail of some unfortunate reptile. Ji’s team also noted other odd anatomical markers: a loosely fused pelvis, relatively sharp claws, odd teeth, and a head similar to that of a turtle.

Taken together, Ji and his colleagues argue that Liaoningosaurus was something truly unexpected: a small, swimming carnivore, from a lineage previously considered to be entirely herbivorous. In essence, an animal that didn’t just look superficially like a turtle, but perhaps acted like one as well.

The evidence for an aquatic, carnivorous lifestyle isn’t quite that cut and dried, though. The fish fossils in the Liaoningosaurus skeleton’s stomach, for example, present a bit of a conundrum. In a paper in the Journal of Geology, Ji and the others present three possibilities for how the fish got there. There’s the possibility that the little ankylosaur was indeed swimming after and devouring large numbers of little freshwater fish. But it’s also conceivable that the creature’s carcass settled atop fish that were already dead. Or perhaps the fish were sheltering or feeding inside the body of the dinosaur when it was buried, so that they were all preserved together. After all, the fish are scattered inside the body, not tightly concentrated in the gut as might be expected from a meal.

The authors stand by a predatory interpretation. “Fish specimens collected from the same rock of other places nearby show a very clear outline, in sharp contrast to those inside the ribcage,” says Xiao Chun Wu, a paleontologist with the Canadian Museum of Nature who worked with Ji on the paper. “All fish skeletons inside the ribcage are vague in body outline, with the tail part missing in most cases, which would have resulted from digestive acid in the stomach.”

Yet even the anatomical evidence for Liaoningosaurus’ aquatic lifestyle—the plastron and unfused bones—is a bit ambiguous. Victoria Arbour, an akylosaur specialist with the Royal Ontario Museum, examined the supposed plastron on the original specimen back in 2010 and came away unconvinced. “It has a very fine scaly texture on it that really looks a lot more like a skin impression,” she says. “And no other specimen I know of has a plastron, including this new specimen. It seems more like a weird chunk of skin that got preserved.”

The skeleton’s unfused bones, which Ji claims are similar to those of marine reptiles, could simply indicate that the animal is a juvenile, Arbour says; dinosaur bones tend to fuse as they get older. And given the very, very small size of the animal, she thinks that youth is the better explanation. Ji and his colleagues offer as evidence of adulthood that multiple other specimens of Liaoningosaurus are known in Chinese collections, all of them tiny, but these specimens have yet to be published in scientific journals, which makes the claims hard to verify at the moment.

Part of the reason behind this whole debate comes from the fact that the paper’s photos—the main method by which other paleontologists can easily examine a paper’s evidence—have poor resolution, which leaves a lot of room for doubt. “I think the author’s conclusion is sound,” says Jordan Mallon, a dinosaur paleoecologist and colleague of Wu, “but the case could have been put forward more strongly in the paper.” Nonetheless, he thinks the idea that Liaoningosaurus was a fish eater is a good explanation for the odder features of the skeleton, like the tooth shape and sharp claws.

So, are the Liaoningosaurus specimens adults or juveniles? If they’re juveniles, does that mean that they weren’t aquatic? Were they really eating fish? There are a couple of ways to answer these questions, Arbour says. One of them would be to grind up a segment of bone and look at the cell growth (the histology), which can help answer how old the animals were when they died. Another would be a more thorough scientific description of the other Liaoningosaurus specimens Ji and his colleagues referenced in their paper. Until then, the turtle-like habits of Liaoningosaurus aren’t a sure thing.

If Ji and his team’s hypothesis does prove true, Liaoningosaurus offers a chance to revise a common assumption. Many of those modern animals that are considered herbivores are less wedded to plants than people think. Deer, cattle, and horses have all been known to devour everything from baby birds to dead fish. While the remains of most ankylosaur species suggest that they primarily ate soft plants, it’s always possible that some of them did occasionally indulge in meatier fare. Fossils, even incomplete or mysterious ones, are often good reminders of how unpredictable nature can be. But modern animals can offer similar insights into the past as well. “Natural history is more complicated than what we have thought of today,” Wu says.