1903 | South Dakota, USA
Dinosaur gastroliths
What exactly is a tool? Seems a simple question, but I’ve talked before about how difficult definitions can be. In borderline cases we may even have to fall back on the old story about the judge and pornography: you know it when you see it. But what if you can’t see it? What if it’s, well, inside a dinosaur’s stomach?
Today’s post takes us back further than we’ve ever gone, into a vanished world of extinct monsters and skin-winged beasts. The secrets of that world are only available to us through a creative meld of scarce evidence and speculation. And yet, as awe-inspiring and incomprehensible as some of those creatures may be, they were in many ways just regular old animals. Each of them still had to eat, just like you and me (and like ostriches and sharks, which will be useful information later on). That’s where we’ll be looking today, at dino-digestion, and the stone markers that it sometimes left behind. And I’ll make the case for these actually being considered dinosaur tools.
So let’s start up our time machine and begin our journey to the Jurassic and beyond. But first, a quick first stop in South Dakota, at the turn of the twentieth century, where we find Barnum Brown swimming among the plesiosaurs.
The stomach for it
In the summer of 1903, Brown was in the Niobara shales of South Dakota, fossicking for fossils. He was in his early thirties, and a rising palaeontological star. Just the year before, he’d found the first example of a new type of giant carnivorous dinosaur out in Hell Creek, Montana. When he’d excavated the specimen and sent it back to his sponsors at the American Museum of Natural History, it was officially given a scientific name that emphasised its size and presumed predatory dominance: Tyrannosaurus rex.
This first reconstruction of the T. rex, using Brown’s information, was published by Henry Osborn in 1905. Only the shaded bones were known at the time, but already it towered over a comparative human:
Barnum Brown wasn’t hunting Tyrannosaurs in the Niobara shales, though. Those rocks were laid down as part of the Western Interior Seaway, a vast inland sea that split North America from top to bottom during the Cretaceous period. Rather than land-based predators, this was the realm of large marine reptiles such as the plesiosaurs, long-necked and flippered underwater hunters that mythology suggests gave rise to Scotland’s Loch Ness monster. Suitable territory, then, for a man named after the greatest showman, P.T. Barnum.
But I digress. We’re actually not interested in any ancient reptiles right now, but rather what Brown noticed scattered among the fossils in those shales. I’ll let him tell it:
In nearly every instance a large number of siliceous stones were found associated with the bones, often embedded in the matrix en masse. In one specimen…there was at least half a bushel of these stomach stones, ranging from the size of a walnut to four inches across.
The Greek for stomach stones is gastroliths; you may also hear them called gizzard stones. Some modern animals—including the dino-descendent birds—have a muscular stomach or gizzard that serves as a kind of mill to grind down food after it’s swallowed. To help with this grinding, a selection of these animals also deliberately swallow stones, which then crush against each other and the food as the stomach contracts. Along with birds, gizzards are today found in crocodiles and alligators, plus a few invertebrates.
Once an animal has died, gizzard stones are not always easy to recognise. They need to be distinguished from all the other bits of rock that surround a skeleton. Useful indicators include finding a group of stones that don’t geologically match any in the immediate area, those stones being rounded or polished from their time in the mill, and their direct association with what would have been the stomach area of the animal. Even with these criteria, though, it remains a matter of interpretation. For example, apparent gizzard stones associated with the extinct New Zealand giant moa have been found there since the nineteenth century, along with remains of plant matter, although the possibility remains that those stones were also or instead used as part of moa nest building or display.
Getting stoned
What’s certain is that at least some dinosaurs, and their extinct reptile relatives, did swallow stones to help grind their daily meal. Take specimen LDNHMF2008 in the Lande Museum of Natural History, China, as described in 2020 by a team led by Shundong Bi of Yunnan University.
This remarkable collection of bones, eggs and stones preserves the remains of a ‘medium-sized oviraptorid’—a two-legged dinosaur—still sitting on a nest of at least two dozen eggs. Some of the eggs, which are on average more than 20cm long, still contain the tiny bones of unborn dino-embryos inside. And right in the centre of it all is a small collection of pebbles, just where the stomach of the brooding adult would have been. As the baby oviraptors emerged from their eggshells, one of the first sounds they heard may well have been the grinding lithic gizzard of their protector.
In the first image below you can see the nest as reconstructed by Bi and colleagues, while the second image shows the fossilised remains from a top-down perspective, dino-tail heading off to the upper right and preserved forelimb at the bottom. I’ve circled the gastroliths for you:
In this case there are very few explanations for how the stones could get there, other than as gizzard stones. They didn’t wash in later. However, dinosaur gastrolith expert Oliver Wings has suggested that theropods like this nesting one were less likely to have gizzard stones than the plant-eating sauropods (think of those bus-sized land giants with the thick legs and long necks and tails). And even in sauropods, this method of food processing might be quite rare.
In his work, Wings calls out the need to distinguish true gastroliths from other ‘exoliths’, or stones that don’t belong in the surrounding sediment but may have arrived there some other way than via a gizzard. He’s backed up this call with experiments on modern animals, such as this slowly rounded series of granite cubes that spent up to 50 days inside ostrich (Struthio camelus) gizzards:
Surveying a series of North American sites with fossils of beasts such as Camarasaurus, Diplodocus, Apatosaurus, Barosaurus and more, Wings estimated that only 2-4% of those animals had gizzard stones. In exercising caution and emphasising other explanations, Wings is directly echoing Barnum Brown from a century earlier, who compared stones found in fossil beds with examples of naturally wind-worn, polished pebbles. We are on firmer ground/water, though, among the swimming dino-relatives that attracted Brown to South Dakota.
In that same 1903 expedition, Brown reported that:
The conclusion seems evident that invertebrate animals formed a large part of the food of plesiosaurs and that, in default of crushing teeth, the breaking up of the food was effected by the aid of these stomach stones, the presence of which further implies a thick-walled, gizzard-like arrangement in the alimentary canal.
In the intervening decades, this picture has been backed up. In the Pierre Shale of Kansas, various members of a type of plesiosaur known as an elasmosaur have been found with gastroliths clustered in their abdomen. This example is specimen NJSM 15435, as reported by David Cicimurri and Michael Everhart in 2001, along with two of the larger of the 95 gastroliths found with it (totalling 6.8kg of stones):
Ancient gastroliths have even made it into popular culture—where would a dinosaur post be without an obligatory mention of Jurassic Park? In the 1993 film, the first dinosaur that our heroes encounter is a sick Triceratops. Harding, the park vet, is stumped as to why the animal gets ill periodically, and Dr Ellie Sattler tries to help out by examining the animal’s eyes, tongue and droppings. The film never gets around to telling us what actually was wrong, other than plants being involved somehow, because at that point a storm shows up and chaos (theory) takes over.
In the original 1990 novel, however, with a Stegosaurus playing the role of the sick dinosaur, Ellie does solve the mystery:
It was an interesting puzzle, she thought. How to explain the periodicity of the poisoning? She pointed across the field. “You see those low, delicate-looking bushes?”
“West Indian lilac.” Harding nodded. “We know it's toxic. The animals don't eat it.”
“You're sure?”
“Yes. We monitor them on video, and I've checked droppings just to be certain. The stegos never eat the lilac bushes.”
…
“Finding anything?” Grant said, coming up to join her.
Ellie sighed. “Just rocks,” she said. “We must be near the beach, because all these rocks are smooth. And they're in funny little piles.”
“Funny little piles?” Grant said.
“All over. There's one pile right there.” She pointed.
As soon as she did, she realized what she was looking at. The rocks were worn, but it had nothing to do with the ocean. These rocks were heaped in small piles, almost as if they had been thrown down that way.
They were piles of gizzard stones.
The stegosaur was accidentally eating the toxic lilac berries each time it discarded and replaced its worn down gastroliths. Unfortunately, switching the sick animal for a Triceratops in the film makes less sense, because recent research has shown that those horny beasts actually have quite sophisticated chewing and slicing teeth. Gizzard stones are therefore less likely to have been necessary for their digestion than for the stegosaur (although who knows what might have gone wrong during the DNA extraction and repair process to recreate those animals in Costa Rica).
Not in Kansas anymore
We’ve seen how finding gastroliths is easier if they’re a type of stone not found in the surrounding sediment. But in some cases tracing the origin of the stones shows that they were very far from local. At the same time as informing us on dino-digestion, those foreign rocks can give us unusual insights into extinct reptile behaviour.
In 2013, Laura Codorniú and her team at the Universidad Nacional de San Luis in Argentina published the first report of gastroliths in two pterosaurs (Pterodaustro guinazui). Like the swimming plesiosaurs, the flying pterosaurs aren’t technically dinosaurs, although they were around at the same time. Taking care to rule out alternative explanations, Codorniú noted that the stones were concentrated in the animal’s abdominal cavity, and had likely been packaged inside a single organ, while the surrounding fine-grained claystone didn’t contain any similar pebbles.
I’ve circled the dark gastroliths again in this image (of specimen MIC-V263). The elongated fingers that supported the pterosaur’s wings point off to the left, while the curved jaw with its thin teeth is folded back across the centre, and a lone foot sticks downwards:
Those thin teeth, and many more now lost, formed a kind of basket or sieve that allowed the Pterodaustro to filter-feed. However, Codorniú and her colleagues suggest that the stronger teeth at the front of its jaw would have been suitable for picking up gastroliths when needed. They posit this flying reptile as something akin to modern flamingoes, which also filter feed and have gizzard stones to help break down their diet. While plesiosaur gastroliths have sometimes been considered as ballast—extra weight to help control buoyancy in the water—that clearly doesn’t apply to pterosaurs, leaving digestion as the most plausible explanation.
It’s not clear where exactly the pterosaurs collected their rocky snacks. But in other cases the geological source of the stones is nowhere near where a skeleton is found. For example, in early 2021 Joshua Malone and fellow researchers from the USA reported pink quartzite gastroliths from Wyoming that most likely came from the gut of a large sauropod such as Barosaurus or Diplodocus. When they crushed the stones to examine and get ages for the minerals inside, the closest geological match they could find was around 1000km to the east of the discovery site. Malone and his team concluded that these huge dinosaurs may have carried the stones from at least that far away, as part of a long-distance migration.
One of the world’s most famous dinosaurs, Dippy the Diplodocus, was from Wyoming originally. In the decades his cast presided over the entrance hall to London’s Natural History Museum (and before then, such as the 1905 photo below), who knew that his display was perhaps a few rocks short of a dinner?
Gastroliths and gizzard stones are a fundamental part of an animal’s digestion, for those that need them. It shouldn’t surprise us then that they are also part of the wider circle of this-eats-that life. This was nicely demonstrated by Kenshu Shimada of the University of Illinois in a 1997 study that examined the remains not of extinct dinosaurs, or even reptiles, but the Late Cretaceous shark Cretoxyrhina mantelli.
Among the fossilised remains of one such shark from Kansas (KUVP 68979), Shimada reports several polished black pebbles. Since sharks both modern and ancient have no history of ingesting stones for any reason, the most plausible conclusion was that the top predator had chewed on a tasty plesiosaur stomach. Echoing Dorothy from the Wizard of Oz, the report notes that:
If the pebbles are in fact plesiosaur gastroliths, their presence possibly indicates the wide geographic migratory habits of either the plesiosaur, the shark, or both, because the lithology of these pebbles suggests that they are not native to Kansas.
Playing inside the house
So here we are. Dinosaurs, especially some large sauropods, as well as plesiosaurs and pterosaurs all have evidence for stomach stones. The majority seem directly linked to digestion, although those found in the swimming reptiles, like gastroliths in modern crocodiles or penguins, may also or instead have played a role in helping the animal balance in the water as ballast.
These stones are environmental objects, used by extinct animals for a specific purpose that they cannot achieve by their biology alone. Does this mean that we have evidence for dinosaur tool use? As I noted at the start of this piece, definitions are tricky, and internal objects are explicitly excluded by all modern definitions of tools. The slippery-slope risk is that other things like ingested food, or even inhaled air, might get swept up into a definition so broad that tool use becomes a meaningless and intractable topic. But I think there’s a case to be made for gastroliths in particular.
For one, these aren’t food. They do mechanical work inside the body, akin to capuchin nut cracking or sea otter shellfish pounding, but using internal muscles rather than limbs to drive the process. They may end up getting worn down or polished, just as happens to accepted tools like the oyster-cracking stones of long-tailed macaques (in fact I’ve previously used the wear patterns on macaque tools as a diagnostic device to reveal just what the monkeys were processing). So far gastroliths pass the test.
The ‘internal use’ criterion remains a sticking point. But even if the distinction is a useful one, I’d argue that the gut is the most external of internal places. It’s essentially a sophisticated tube running through an animal that opens to the outside world at both ends—anything indigestible that goes in is only a temporary resident, analogous to an object held in a clenched fist. Or a closed mouth.
And while animals do ingest stones accidentally while they forage (just as most humans end up with at least some unintended grit in their diet), the evidence points towards these extinct reptiles deliberately ensuring that their gastric mill is kept in working order. Whether it’s a pterosaur selecting pebbles to break down its filtered meal, a sauropod wandering vast distances as its stomach churns away, or an oviraptor perched on its nest awaiting the next generation, these animals need to be able to find, recognise, choose and maintain a stony component to their behaviour that is critical to, but separate from, their actual diet.
A stronger objection may be that these dinosaurs and their relatives aren’t holding and positioning the stones as they’re used. But that seems like a fine and unnecessary distinction to me. This isn’t a case of a bear scratching against a tree, or even a wrasse striking a bivalve on a coral outcrop, where the external environment is fixed and used for its resistance. Gastroliths are separate pieces of the environment, regularly taken far from their original source, and only able to do the work of breaking down food because of the musculature of the animal holding them. Clumsy tool use is still tool use, and precision isn’t part of any definition.
Modern animals hold tools in their beaks, claws, rostrums, trunks, lips, paws, snouts, fingers and more. I think we should add stomachs to that list, and tool-using dinosaurs to our overview of nature’s bounty. Any objections?
Sources: Osborn, H. (1905) Tyrannosaurs and other Cretaceous carnivorous dinosaurs. Bulletin of the American Museum of Natural History 21:259-265. || Brown, B. (1904) Stomach stones and food of Plesiosaurs. Science 20:184-185. || Hamilton A 1891. Notes on moa gizzard-stones. Transactions and Proceedings of the New Zealand Institute 24:172–175. || Wings, O. 2007. A review of gastrolith function with implications for fossil vertebrates and a revised classification. Acta Palaeontologica Polonica 52: 1–16. || Bi, S. et al. (2020) An oviraptorid preserved atop an embryo-bearing egg clutch sheds light on the reproductive biology of non-avialan theropod dinosaurs, Science Bulletin doi: 10.1016/ j.scib.2020.12.018. || Wings, O. (2015) The rarity of gastroliths in sauropod dinosaurs – a case study in the Late Jurassic Morrison Formation, western USA. Fossil Record 18:1-16. || Wings, O. & P.M.. Sander (2007) No gastric mill in sauropod dinosaurs: new evidence from analysis of gastrolith mass and function in ostriches. Proceedings of the Royal Society B 274:635–640. || Mateus, O (1998) Lourinhanosaurus antunesi, a new Upper Jurassic allosauroid (Dinosauria: Theropoda) from Lourinhã, Portugal. Memórias da Academia de Ciências de Lisboa 37:111–124. || Brown, B. (1907) Gastroliths. Science 25:392. || Cicimurri, D. & M. Everhart (2001) An Elasmosaur with Stomach Contents and Gastroliths from the Pierre Shale (Late Cretaceous) of Kansas. Transactions of the Kansas Academy of Science 104:129-143. || Crichton, M. (1990) Jurassic Park. Alfred A. Knopf. || Erickson, G. et al. (2015) Wear biomechanics in the slicing dentition of the giant horned dinosaur Triceratops. Science Advances 2015;1:e1500055. || Codorniú, L. et al. (2013) First Occurrence of Stomach Stones in Pterosaurs. Journal of Vertebrate Paleontology 33:647-654. || Shimada, K. (1997) Paleoecological Relationships of the Late Cretaceous Lamniform Shark, Cretoxyrhina mantelli (Agassiz). Journal of Paleontology 71:926-933.
Main image credit: American Museum of Natural History, AMNH 6254 (Psittacosaurus mongoliensis) || Second image credit: Osborn (1905) || Third image credit: Oxford University Museum of Natural History || Fourth/fifth image credits: Bi et al. (2020) || Sixth image credit: Wings & Sander (2007) || Seventh/eighth image credits: Cicimurri & Everhart (2001); http://oceansofkansas.com/Gastro.html || Ninth image credit: Jurassic Park, Universal Pictures, 1993 || Tenth image credit: Codorniú et al. (2013) || Eleventh image credit: Natural History Museum, London.