Feathers are built to be light, flexible and replaceable, which also makes them poor candidates for surviving deep time. Yet a fossil collected in northeastern Montana preserved one feather on its broken surface and several more inside. The container was not amber or fine lake sediment. It was the fossilized dung of a meat-eating dinosaur.
Jingmai O’Connor and 13 colleagues describe the specimen in a paper published September 10 in Current Biology. It comes from the uppermost Maastrichtian rocks of the Hell Creek Formation, close to the end of the Cretaceous about 66 million years ago. The authors call the exposed specimen arguably the best-preserved feather yet recovered from Mesozoic rock.
This is one study of one coprolite, not settled consensus about how all late Cretaceous birds were feathered or why some birds survived the mass extinction. Its unusually complete contents nevertheless connect feather anatomy, an extinct aquatic bird and a predator’s meal in a way ordinary bones rarely can.
A lucky break in Hell Creek
David DeMar Jr., a researcher and collections manager at the University of Washington’s Burke Museum, found the dark reddish-brown nodule in 2016 while collecting fish fossils from a rocky outcrop. It was about half the size of a golf ball. Looking through a hand lens, he saw a tiny feather exposed where the object had broken.
The setting made that observation particularly unusual. Paleontologists had searched the Hell Creek Formation for more than 150 years without reporting feathers there. Early Cretaceous sites in China and elsewhere sometimes preserve feathers beside articulated skeletons. Comparable soft tissue is almost absent from upper Cretaceous rock, leaving a thin record for the period immediately before the mass extinction.
The fossil’s discovery was therefore literal luck. An intact outer surface might have offered no hint of the material hidden within. The specimen is now displayed at the Burke Museum, according to a University of Washington account of the work.
How an unremarkable stone became fossilized dung
A feather on a reddish lump did not by itself prove that the object was a coprolite. The team studied its mineral composition and scanned it with micro-computed tomography. Thousands of X-ray slices were digitally stacked, allowing the researchers to inspect the interior without cutting the specimen apart.
The scans revealed additional feathers, tiny scales from a gar and fragmentary leg bones from a bird. The mix of digested animal remains within the phosphatic mass supported the conclusion that this was fossilized feces produced by a medium-sized theropod.
The exact predator is not known. A young Tyrannosaurus rex or the disputed animal called Nanotyrannus are possibilities discussed around the find, but the dropping cannot be assigned confidently to either. “Meat-eating dinosaur” is the narrower claim that the specimen can support.
A food chain compressed into one fossil
The bird bones resemble those of a hesperornithiform, a group of specialized aquatic birds that lived during the Cretaceous. Many members were flightless and used their feet to propel themselves underwater, occupying a role loosely comparable to modern loons.
The gar scales add another level. The paper interprets them as the bird’s last meal rather than direct prey of the theropod. A small fish would have been a difficult, low-return target for a much larger predator, while it fits the diet of a diving bird. The likely sequence is fish eaten by bird, followed by bird eaten by dinosaur.
That reconstruction remains an inference, but it is grounded in the association of scales, bird bones and feathers inside the same coprolite. ScienceBlog has previously covered how fossil dung can reconstruct ancient diets and ecological change. This specimen compresses a particularly vivid set of trophic relationships into an object smaller than a golf ball.
A modern shaft on an extinct kind of bird
The exposed example is a pennaceous feather, meaning its branching barbs form a firm vane rather than loose down. Its central shaft, or rachis, has a square cross-section and a spongy, medulla-filled interior. That design combines stiffness with low weight and is familiar from anatomically modern feathers.
The feather also has densely packed barbs and what the authors describe as a low structural index. In living aquatic birds, comparable construction helps reduce water penetration. Those features independently fit the hesperornithiform identification suggested by the nearby leg bones.
Its modern-looking construction does not make the bird a direct member of the group that survived the end-Cretaceous extinction. Hesperornithiforms sat outside Neornithes, the crown lineage from which every living bird descends, and they disappeared at the boundary. The discovery instead shows that an advanced, light and stiff feather design had evolved in a close relative before the asteroid impact.
Fossil feathers can retain information beyond outline alone. In other deposits, microscopic structures have helped researchers reconstruct aspects of color and iridescence, a subject explored in earlier ScienceBlog coverage of feather evolution. What makes the Hell Creek specimen different is its three-dimensional preservation in ordinary rock rather than a flattened film or amber inclusion.
The hidden feathers complicate the plumage
Micro-CT exposed more than copies of the visible feather. It revealed additional feather forms, including morphologies that do not occur in living birds. The researchers interpret the collection as evidence that the same aquatic bird carried modern-looking, water-resistant feathers alongside smaller, more primitive body-feather types.
That mosaic resists a simple story in which one feather design neatly replaced another throughout the body. Evolution can preserve old structures while modifying other parts of the same animal. It also means the fossil offers information about plumage diversity that the single exposed feather would never have supplied by itself.
An extinction clue, not an extinction answer
The authors suggest that plumage could have influenced which birds endured the cold interval after the asteroid impact. Body feathers provide thermal insulation. If extinct lineages retained feather forms that held heat less efficiently than the plumage of Neornithes, that difference might have mattered during an “impact winter.” Molting strategy could also affect survival if feather replacement temporarily reduced insulation or flight performance.
This is a hypothesis, not a demonstrated cause. One coprolite cannot reconstruct an entire coat, a molt cycle or the physiology of a lineage. It cannot separate plumage from habitat, diet, body size, reproduction and geography. Aquatic hesperornithiforms went extinct despite living near water, which weakens the idea that a waterside habitat alone explains survival, but it does not make feathers the sole alternative.
Genetic evidence adds another, very different view of the boundary. Previous ScienceBlog reporting on bird genomes described rapid evolutionary change around the impact. The feather fossil contributes anatomy from an extinct branch, but the broader question of survival will require several kinds of evidence to agree.
Coprolites become a new soft-tissue archive
Fossil dung has long been used to recover bone fragments, plant remains, pollen, parasites and chemical traces. The new specimen suggests that at least some carnivore coprolites may also preserve feathers, fur or other delicate structures in three dimensions.
This one was recognized only because it happened to split across a feather. Systematic CT scanning could expose comparable contents in specimens whose surfaces give nothing away. That prospect may be the most durable result of the paper: collections already sitting in museums could contain soft-tissue evidence that no one had reason to seek.
A coprolite is evidence of behavior, digestion and environment at once. In this case it also became a protective capsule, carrying the anatomy of an aquatic bird through roughly 66 million years. The find does not solve the extinction of ancient birds, but it changes where paleontologists can reasonably look for the feathers needed to investigate it.