Lots of animals eat rocks. Or at least they swallow rocks; they don't eat them in the sense of digesting them. This is true of many reptiles and birds that are alive today, and a few mammals such as seals and whales. It was also true of some extinct animals, including herbivorous dinosaurs and marine reptiles. Rocks that have been in the digestive system of an animal are called gastroliths, which literally means "stomach stones." Any kind of rock can become a gastrolith — all it has to do is be swallowed by something.

Many kinds of extinct animals have been found with groups of small, rounded rocks inside their rib cages, and similar stones are often found by themselves in rock formations that also contain dinosaur fossils. The presence of gastroliths in the fossil record raises a couple of important questions: what are gastroliths for, and how do we know that a gastrolith is really a gastrolith? New discoveries in the past few years are forcing paleontologists to reconsider both of these questions.

What are gastroliths for?
Among living animals, gastroliths are most commonly found in birds, especially ground-living birds like chickens and ostriches and active swimmers like penguins, crocodiles, seals, and toothed whales. Extinct animals that have been found with definite gastroliths in their bodies include plant-eating dinosaurs like sauropods, primitive ceratopsians, and ostrich-mimics, marine reptiles like plesiosaurs and (rarely) ichthyosaurs, and crocodilians. These animals mostly fit into two categories: plant-eaters (herbivorous dinosaurs and many birds) and animals that swim (all the rest).

Many birds swallow sharp pebbles and grit and hold these rocks in a muscular part of their stomachs called the gizzard. The gizzard contracts and grinds the gastroliths against each other and against the food that the bird has swallowed (remember that birds have to swallow each bite whole). The rocks grind down the food — essentially, the bird is using the gastroliths to chew the food in its gizzard — and the rocks grind each other down, too. Eventually the sharp, jagged chunks of rock become smooth, rounded pebbles, and they are not much good for grinding anymore. So the bird will vomit them out and find new, sharp rocks to swallow.

Definite gastroliths have been found in the rib cages of sauropod dinosaurs, primitive ceratopsians such as Psittacosaurus, and in toothless theropods like ostrich dinosaurs and the feathered Caudipteryx. All of these animals seem to have been herbivores. Perhaps they all had gizzards, like birds, and used gastroliths to grind their food. To test this idea, paleontologist Oliver Wings fed different kinds of rocks to ostriches to study the function and characteristics of gastroliths. He found that in ostriches, gastroliths make up between one-fifth and one-half of the stomach contents and about 1% of body mass. Similar amounts of gastroliths have been found with Psittacosaurus, Caudipteryx, and the ostrich dinosaurs, so the hypothesis that these dinosaurs used gastroliths to aid their digestion is supported. However, in all cases of sauropod gastroliths, the rocks would have been less than 0.1% of body mass — too little to be an important part of the digestive process. Furthermore, the vast majority of sauropod skeletons are found without gastroliths. So the presence of gastroliths in some sauropods is mysterious.

The function of gastroliths in aquatic animals is also unclear. For a long time, scientists have assumed that swimming animals swallow rocks to help control their buoyancy. For example, many crocodilians like to float in the water with just their eyes and nostrils showing so they can ambush their prey. A stomach full of rocks might help them keep their bodies under the water and out of sight. Many seals and whales dive deep in the ocean to feed, and the gastroliths might help them dive deeper (of course, they also have to haul the gastroliths back up when they come back to the surface).

But some recent research casts doubt on this hypothesis. Using computer models of crocodiles, a paleontologist named Don Henderson found that gastroliths in living crocodiles and in fossil marine reptiles typically make up less than 2% of body mass. Gastroliths would have to be more than 6% of body mass to have much of an effect on buoyancy. Below that level, the filling and emptying of the lungs with air has a much greater effect of buoyancy. However, the small amounts of gastroliths found in aquatic animals might help stabilize their bodies in the water, so they have less of a tendency to roll from side to side. Also, many seals and whales find their food on the bottom of the ocean, and they might accidently swallow rocks in the process of feeding. These rocks are gastroliths, but they are not "for" anything, they are just a consequence of how the animals feed.

Other researchers have looked at gastroliths found with elasmosaurs (long-necked plesiosaurs) and found that they have signs of impacts from being rolled around and banged against each other. These gastroliths have been found with ground up fish bones in the stomachs of the elasmosaurs. Using the criteria of Henderson's crocodile work, these gastroliths are not numerous enough to have an effect on the animals' buoyancy. But by comparison with Wings's ostrich data, the gastroliths were present in large enough numbers to help grind up food. So elasmosaurs may have swallowed gastroliths to mash up the fish they swallowed and not to help them swim or dive.

Recognizing gastroliths
Sometimes a fossil skeleton is found with a pile of pebbles or small rocks inside the rib cage, or scattered around the body. These rocks are usually pretty easy to identify as gastroliths. The only other thing they could be is stream pebbles that washed up against the body of the animal as it was being buried under sand and silt. Paleontologists compare the rocks to others in the same formation to see if they could have washed in from someplace else, or if they are really gastroliths. This is usually only necessary if the gastroliths are scattered. If the rocks are found in a nice tight pile in the rib cage, they are almost certainly gastroliths.

In some rock formations that contain dinosaur fossils, it is very common to find small, rounded rocks and pebbles in much finer-grained sediment. How did these rocks get there? Think about going to the beach, or to a muddy stream bank. Some of the sand and mud is coarse and some of it is fine, but for the most part it is pretty uniform. You probably don't find a lot of rounded rocks in the same place as you find sand or mud. On the other hand, if you go to a cold stream in the mountains, you find lots of rounded cobbles, but not much sand or mud. Moving water tends to sort rocks and pieces of rocks and group pieces of the same size together.

The same principle applies when these sediments are turned into stone. Sand becomes sandstone, mud becomes mudstone, and groups of pebbles are "glued" together to form conglomerates. So if you find rounded rocks — say, marble-sized up to fist-sized — in a deposit of sand or mudstone, there must be an explanation for how they got there, because moving water would tend to separate the sand grains from the big rocks.

For the past few decades, many paleontologists have assumed that these rounded rocks are gastroliths. Birds that use gastroliths occasionally regurgitate the rocks that have become smooth and rounded, and then they swallow new, sharp rocks. They probably do this because the sharper rocks are better for grinding up the food in their gizzards. It seems reasonable to assume that dinosaurs did the same. Maybe all these out-of-place gastroliths are vomited up by dinosaurs. Furthermore, some of these rocks can be traced back to the rock formations they originally came from, and some of them are quite a long way from where they started. Some are over one hundred miles from their sources. If these rocks are gastroliths, they might have been swallowed by dinosaurs in one place and vomited out in another, and in this way they could tell us about the movements of dinosaurs.

There's just one problem. New research has called the identity of these rocks into question. Oliver Wings found that ostrich gastroliths are rarely polished (i.e., smooth and shiny). Many rounded rocks found in dinosaur-bearing formations are identified as gastroliths specifically because they are highly polished. But if the shiny, rounded rocks are not gastroliths, how did they get polished, and how did they get into the finer-grained sediments? One possibility is that they are remnants of old layers of conglomerate that have weathered away, leaving just a few stream-rounded pebbles sitting on the sandstones and mudstones below them. Most isolated "gastroliths" are found on the surface and not actually embedded in sandstone or mudstone. Maybe they are relics of old layers of coarser rocks.

Another possibility is that dinosaur gastroliths may have had different characteristics from bird gastroliths. A different team of paleontologists compared genuine dinosaur gastroliths (i.e., those found inside the rib cages of articulated skeletons) with some potential gastroliths found on finer sediments. They used an electron microscope to compare the wear on the surface of the stones. They concluded that many of the isolated gastroliths have the same wear pattern as genuine gastroliths.

So what are the isolated, rounded stones found in dinosaur-bearing rocks? No one knows for sure. They seem to be very different from the gastroliths of living birds, but very similar to confirmed gastroliths in dinosaurs. The gastrolith research described here is at the cutting edge of paleontology, and there are strong arguments on both sides. It will probably take more time and more work before a clear answer emerges.

For a long time, the idea that many dinosaurs and other extinct animals swallowed rocks, either to help grind up their food or to help control their buoyancy, has been widely accepted. Smooth rocks found in dinosaur-bearing formations have been assumed to be "used" gastroliths vomited up by dinosaurs. New research suggests that much of what we thought we knew about gastroliths may be wrong. Some plant-eating dinosaurs didn't have enough gastroliths to grind up their food, and many aquatic animals don't carry enough gastroliths to really affect their buoyancy. Many animals may have swallowed gastroliths infrequently (like sauropods) or accidentally (like whales). The identification and function of gastroliths is in many cases uncertain. However, many paleontologists are taking a fresh look at gastroliths, and they are developing new ways to test their hypotheses.

If this story has a moral, it is that we shouldn't take anything for granted in science. Even a seemingly simple phenomenon, like crocodiles swallowing rocks to help them sink, may turn out to be more complicated that it looks. And even "obvious" hypotheses should be tested.


Text by Matt Wedel, 5/2007. Photo credits to come.