What the Dinosaurs Left Us

Their fossilized poop shows what they ate.

By Karen Wright|Saturday, June 01, 1996
The sun has just set over the tranquil Santa Barbara campus of the University of California, and the crisp evening air is redolent of warm sand and eucalyptus. Scores of students are jogging or cycling under the rosy gold autumn sky; a few stroll back from the beach with surfboards under their arms.

But in a low white building on the east side of campus, in a cavelike room that smells of wet stone, Karen Chin is hard at work. Chin is hunched over a cluttered bench, her dark hair fanning halfway down her lab coat, her slender fingers holding a small gray rock against the motionless blade of a circular saw. She has repositioned the rock several times, in search of the right cut, when her concentration is shattered by a colleague entering the lab.

Hey, Karen, calls the colleague in greeting. You still messing around with poop?

The short answer is yes. Karen Chin was, is, and probably always will be messing around with poop--petrified, prehistoric poop, the poop of ages past. She’s a pioneer in a specialty so peculiar it’s not taught in any university. It doesn’t even have a formal name, though one does suggest itself: paleoscatology. It is safe to say that Chin is the world’s leading paleoscatologist. Also the world’s only paleoscatologist.

For the past six years this doctoral student has been analyzing and categorizing hundreds of the fossilized leavings that go by the polite name of coprolites. The specimens come from around the world and across the epochs. They include 300-million-year-old fish feces; dinosaur dung from the Triassic, Jurassic, and Cretaceous; and a sloth stool issued during the last ice age. Some of the fossils have been ravaged by time and are nearly unrecognizable. But others have survived more or less intact, their humble morphologies uncannily familiar in spite of their antiquity.

It is Chin’s dream to bring order to this exocolonic chaos. In coprolites, she hopes to find evidence of feeding habits and behavior available from no other fossil source. Most important, she expects to discover the diets of ancient creatures so that paleontologists may one day reconstruct ecological webs from the very bowels of prehistory.

So far, however, Chin’s results aren’t much more impressive than her subject matter. If she is to tease out the secrets of coprolites, Chin must first devise a way of grappling with the daunting variety and anonymity of her specimens. On this particular evening, she has set out to section a fragment of putative T. rex turd from the Royal Saskatchewan Museum. The whole specimen was 15 inches long and this big around, she says, putting the tips of her thumbs and forefingers together in a disconcertingly large O. The fragment, which was cleaved from its fecal parent with a pair of wire cutters, resembles a chunk of light-colored concrete with darker, elongate inclusions that Chin recognizes as bone. Pieces of bone are common in carnivore coprolites, she says; she’s cutting open the fossil to see what else she can find out.

Karen’s the first person in the history of coprolites who’s had the technology and the will to treat them to such a detailed analysis, says paleobotanist Bruce Tiffney, Chin’s doctoral adviser. And she’s only at the very beginning of that. It’ll take her lifetime’s worth of work, plus some other people’s, to make a reasonable picture out of this.

But no one else has volunteered.

I don’t like to be associated with the nasty aspects of defecation, Chin declares some time later from the carpeted floor of her office, where an assortment of aged scat surrounds her like a most unappetizing picnic lunch. I’m interested in coprolites from a biological standpoint. I’m trying to develop an overview of what they can tell us about the past. And that means that I need to look at all different types of coprolites and all different types of preservation.

Hence the protean display on the office floor. There are fractured loaves of dark gray rocks, chalky palm-size crescents, thumb- shaped orange nuggets, irregular pebbles in stunning aquamarine, and numerous variations on the common brown sausage. Time and geology have bestowed upon these specimens a flamboyance they undoubtedly lacked when they first saw daylight. Yet many still bear a signature of their provenance in the form of faint longitudinal striations. Those are probably marks from the anal sphincter, Chin explains.

Coprolites form much the way bone fossils do, when minerals invade the microscopic interstices in organic matter and grow into crystals there. Sometimes mineralization helps preserve the living material itself; other times the crystals replace the organic template. In either case, the more readily a substance decomposes, the less likely it is to remain intact long enough to become fossilized. Dung is at a definite disadvantage there, and because of that disadvantage coprolites are rarer than bone fossils. But they are still plentiful: hundreds have been collected from the field, and millions more languish in fossil beds the world over.

Until Chin came along, nobody much cared. Coprolites haven’t gotten much respect, says Chin. A lot has been published on them, but mostly the authors just described the appearance of coprolites from a given locality, then put them on a shelf in a museum. Paleontologists considered the information coprolites might provide far too dubious to warrant the time it would take to decipher it. And there was a certain stigma attached to the enterprise.

Trained as a botanist and ecologist, Chin doesn’t share her colleagues’ prejudice. Where others see ambiguity and guffaws, she sees opportunity. Chin thinks the main problem is that there is no context in which to evaluate coprolites. So she has decided to provide that context, by devising schemes for identifying and classifying the phenomena of interest.

In the past, most coprolites have been identified on the basis of shape, she says. Here’s one from Nebraska. Chin offers a pale, slightly bowed cylinder with rounded ends. This is probably mammal, about 31 million years old. It looks like feces, right? I mean, you look at this and go, yeah, this is fecal material. Right?


This is dinosaur, from Montana, says Chin, handing over an innocuous gray-brown individual with blocky edges. Darker speckles in the rock give it a heathered texture, but mostly it’s just evocative of other rocks. Now, this lacks an identifiable shape, Chin notes. It’s because I’ve been collecting these for so long that I recognized it easily.

These two samples are very different. And that is a problem when you’re trying to interpret paleobiological information from coprolites. How can we recognize coprolites with atypical shapes? How can we compare coprolites from different ages, from different depositional environments, from different animals? Those are the kinds of questions I’m trying to address in my research.

In an ideal world, of course, coprolites would be classified and compared according to their organism of origin, just as fossil bones are. Unfortunately, because of what Chin calls the detached nature of feces, it’s almost impossible to match droppings with droppers.

You can, however, narrow the range of possible culprits. The criterion of shape does sometimes say a little bit about who did what. Spiral coprolites, for example, are thought to be the exclusive province of primitive fish, which include sharks and lungfish as well as many extinct taxa. Because their intestinal valves are spiral shaped, these fish produce (and produced) distinctive oblong coils, many of which have turned up in sediments from the Paleozoic (570 million to 245 million years ago) and Mesozoic (245 million to 65 million years ago).

But shape isn’t usually a reliable indicator of source. Anyone who’s emptied a litter box or flushed a toilet knows that the issue of a single individual can change shape dramatically over time. Conversely, the droppings of different species can look very similar. Logs, pellets, piles, pinched ends--these morphologies are generously distributed by and among all manner of vertebrate life-forms.

The size of a fecal deposit may also suggest something about its maker: many large Mesozoic coprolites are attributed to dinosaurs because paleontologists assume that nothing else alive at the time could have manufactured mounds of such breadth or slugs of such girth. But size, too, can confound. A 1,000-pound moose leaves morsels no more than an inch long. Rodents that share a community latrine can generate heaps of fused waste several feet high.

Even pristine dung in modern ecosystems can be ambiguous, says Chin. And then it gets rained on, stepped on, decomposed--eaten, even. Imagine, then, what confusion can be wrought by a few million years of geologic pressures. In Yorkshire, England, one paleontologist found Jurassic droppings that were nearly two-dimensional. We’re talking soft material, Chin says, and soft material is subject to deformation even under the best of circumstances.

So, with a characteristic disregard for appearances, Chin is exploring some of the less obvious features of her specimens. She’s cut them open, sliced them up, and pulverized them. She’s examined their insides with electron microscopes and made exhaustive inventories of their contents--animal, vegetable, and mineral. She’s run geochemical analyses to characterize organic matter in the fossils, and elemental and mineral profiles to examine the processes by which they became fossilized. Her subject matter may lack sophistication, but her methods have it in spades.

Chin’s strange scatological journey began in 1989 in Bozeman, Montana, where she had a job making thin sections of fossil bones for paleontologist Jack Horner at the Museum of the Rockies. Chin had recently turned to dinosaurs after more than a decade studying extant ecosystems. For 15 summers she’d worked as a Parks Service ranger and naturalist. On the job, in nearby Glacier National Park, she’d come to appreciate the informational value of feces. Though she might see elk, mountain lions, and grizzlies only rarely, their stools were comparatively easy to find and much more approachable. From such samples, researchers could deduce the animals’ numbers, territory, and diet. Chin bought field guides to Rocky Mountain scat and began assembling her own personal photo collection. When Horner told her that he’d found some suspected dinosaur coprolites, it seemed fated that she and they should meet.

The coprolites came from a site in northwest Montana called the Two Medicine Formation, where Horner was unearthing the bones and nesting grounds of Maiasaura, a duck-billed dinosaur. The first thing Chin noticed about these fossils was that they were not the discrete orbs, pellets, or cylinders commonly described in the literature. These were more indiscrete: vast and seemingly formless, like massive cow pies turned to stone and then broken by a nasty fall. And the broken pieces were big: some of the blocks measured more than a foot on a side.

The Two Medicine specimens were full of fibrous bodies large enough to be seen with the naked eye. Chin sectioned them, just as she had Horner’s bone fossils. Under the microscope the dark fibers revealed themselves to be the mineralized remnants of shredded wood. The fossils, Chin decided, were the by-product of a large vegetarian--most likely the herbivorous Maiasaura. They didn’t look like the coprolites she’d seen in the literature, because the majority of those were the more firm and robust doings of carnivores.

Chin visited the Two Medicine site again and again, gathering samples with various compositions from different locations. She wanted to find a way to characterize the members of her 76-million-year-old suite, to tease them apart, expose their contents, and force them to yield information. Her preoccupation soon took on the proportions of a doctoral dissertation. She found a sponsor in Tiffney, whose theories about the role of herbivory in the evolution of plants were in need of empirical substantiation.

When she first came to me with the idea of this project, Tiffney remembers, I said, ‘Coprolites?’ My basic response was ‘Prove it.’

It seemed like a reasonable request. But as it happened, very few dinosaur coprolites had ever been demonstrated unequivocally to be in fact what they appeared to be. Chin carefully amassed the evidence for the scatological nature of her Two Medicine specimens. First, she argued, the presumed coprolites occurred as scattered aggregations rather than as one continuous layer, a distribution inconsistent with geologic deposition but consistent with the cow-pie hypothesis. Second, they occurred in the same sediments in which the duckbills had been preserved--so a connection between the two was chronologically kosher. Third, angular breaks in the woody fibers suggested that the plant material had been chewed up rather than stepped on or weathered by water.

And then there were the dung beetle burrows. Chin had noticed that several of her fossils were riddled with smooth channels, some measuring more than an inch in diameter. On a hunch, she’d shown these specimens to a noted dung beetle specialist in Ontario.

I had a look at these things and it was like, ‘Wow. I don’t know anything about paleontology or fossil rocks, but these look like the perfect soil trace of a modern-day dung beetle,’ says Bruce Gill, an entomologist at Agriculture Canada who became her collaborator. What clinched the ID, says Gill, was the presence of backfilled burrows: tunnels in the shadowy black rock that had been plugged with sand-colored sediments. Any number of invertebrates can dig through a dung pile, but only dung beetles fill their tunnels back up, using the soil displaced from brood cells hollowed out in the ground beneath the flop. And when you find a dung beetle backfilling, you know you’ve found dung.

They wouldn’t go to just any old rotting plants, says Gill. They’d go to the rotting plants that had passed through the gut. Very rich. Very enticing.

The presence of dung beetle burrows also revealed something of the perpetrator’s toilet habits: the animal had simply left its waste where it landed, rather than burying it as a cat would.

It just kept getting better and better--I mean, if you’re into this kind of thing, says Chin. We were able to say that what we were looking at in the Two Medicine Formation was the largest verifiable dinosaur coprolite. And now we had evidence for an ecological relationship between dinosaurs and some of the insects that lived at the time. We had the beginnings of a Mesozoic food web.

Chin had proved that her coprolites could live up to their name. She had discovered the oldest evidence of dung beetle activity. She had found the only evidence of dinosaur-insect interactions. In her Two Medicine fossils, Chin could even look forward to finding some dung beetle dung.

After the beetle breakthrough, Chin broadened her mission. She scouted museums for coprolites from other animals and other eras. She talked up her vision of a grand classification scheme at diggers’ meetings. As word got around, other researchers started sending their suspects to Chin. As always, she would do the dirty work.

Truth be told, the work isn’t all that dirty. The thin-section lab is a bit dusty, but that’s because it’s full of rocks getting cut up and ground down. Chin uses saws with diamond-studded blades and grinding wheels with abrasive slurries to reduce her samples to slices thousandths of an inch thick. Mounted on square glass slides, the sections project the semiopaque jumble of a kaleidoscope image.

What I’m trying to do now is classify the fossils in terms of fabrics, says Chin. I’ve looked at so many of these slides, I’m beginning to recognize patterns. You can have fabrics full of bioclasts--fragments of organic origin, such as bits of plants or clamshells--and you can classify the kinds of mineral grains, their size and distribution. Chin picks up a slide in comely tones of gray, brown, and off-white. The victim was one of her Two Medicine fossils. I could say, for example, that this type of fabric has larger bits of woody material in a fine-ground mass. Now, that fine-ground mass, under magnification, is a series of disaggregated tracheids--water-conducting plant cells.

Chin has been able to group the coprolites from 14 Two Medicine sites into four distinct categories based on the proportion of woody fibers and the type and quantity of plant cells they contain. Those categories, she believes, represent diets with different fiber contents and probably different nutritional values as well. A high-fiber diet, full of stems and bark, might keep an animal regular, but the animal would have to eat a lot to meet nutritional needs. The low-fiber samples may correspond to nutritionally rich diets of ferns, young leaves, and flowering plants, which would require more selective foraging. At one time Chin thought that her low-fiber coprolites might be the work of juvenile duckbills, because they seemed to be concentrated in a Maiasaura nesting area. Later excavations turned up a specimen too big to have come from a juvenile, and Chin went back to the drawing board.

While the Two Medicine investigation is still in progress, Chin has been examining coprofabrics from other locations. She’s found that coprolites offer a cornucopia of intestinal itinerants in addition to wood and shellfish. Sections of specimens have revealed whole and fragmented teeth, bones, seeds, leaves, stems, spores, fish scales, snail shells, and shards of volcanic ash. Some of the inclusions represent the animals’ intended diet; others were probably passengers therein.

Some people say, well, you wouldn’t get all that coming through the digestive tract, says Chin. But we all know that some things go through us, and you can see them if you look: tomato seeds, corn, lettuce. In bear scat, whole berries will come out.

Still, Chin concedes, some of the items she finds in her specimens may have arrived ex post facto. She’s begun examining some of the bioclasts in her samples for signs of gastric etching, evidence that they had passed through the gut.

Other techniques have helped Chin reconstruct the original composition. With organic geochemist Simon Brassell at Indiana University, Chin has used biogeochemical analysis to detect the carbon skeletons of organic compounds that can persist in coprolites even when gross structures such as plant cells have degraded. Some of the compounds isolated in this way implicate a particular source. Oleanane, for example, indicates the presence of angiosperms, or flowering plants; certain diterpanes are peculiar to plants called gymnosperms (which include conifers, such as pine trees). Biogeochemical analysis has provided the only evidence that Maiasaura ate flowering plants, since all the plant material in the Two Medicine coprolites that could be identified under the microscope was coniferous.

Chin also relies on X-ray diffraction to determine the mineral content of her fossils. Mineral analysis can reveal the preservational environment of a coprolite as well as what manner of manure was preserved. Carnivore leavings, for example, tend to be high in calcium phosphate, or apatite, a principal constituent of bone. Meat eaters’ dung fossilizes more readily than plant feeders’ because it is richer in minerals from the get- go.

I still can’t say for sure which animal did what, says Chin. But these techniques release you from the confines of size and morphology as criteria for classification. You’re not dependent on having an intact specimen anymore. It’s a place to start.

In time, Chin says, she may try to develop bile-acid markers that could help her match species with feces, and imaging techniques that would disclose a coprolite’s contents without its having to be sliced up first. She might even discover unknown species among the bioclastic mélange of her specimens. But in spite of the progress she’s made and her ambitions for the future, Chin hasn’t yet inspired a larger movement. And there’s only so much one woman can do.

There’s no renaissance in paleoscatology at the moment, says Adrian Hunt, director of the Mesalands Dinosaur Museum in Tucumcari, New Mexico, and a former collaborator of Chin’s. But maybe there will be. The laughter’s subsided.
Comment on this article

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar

Log in to your account

Email address:
Remember me
Forgot your password?
No problem. Click here to have it emailed to you.

Not registered yet?

Register now for FREE. It takes only a few seconds to complete. Register now »