Paleontologists are looking beyond bones to reveal the hues of prehistoric animals that vanished millions of years ago. But the young field has its share of disagreements.

Michael Benton used to tell his paleontology students that they would never know the true color of a dinosaur. After all, even fossils that sport light or dark patches may not indicate the creature’s original hue. But in recent years, the vertebrate paleontologist at the University of Bristol in the United Kingdom has had to revise those lectures.

Researchers reconstructed the plumage color of a Jurassic birdlike dinosaur called A. huxleyi using color-imparting melanosomes, the morphology of which had been preserved in fossil feathers. Image courtesy of Michael DiGiorgio (artist).

In 2010, Benton and colleagues found evidence that the feathered dinosaur Sinosauropteryx prima had reddish-brown stripes on its tail (1). The same year, another group claimed that birdlike Anchiornis huxleyi bore a red crest on its head (2). Since then, others have discovered that a marine ichthyosaur was dark-colored, and that the early bird Confuciusornis sanctus possessed dark feathers with light wing tips (3, 4) (Table 1). Suddenly, the color of prehistoric animals has become an active topic for research rather than speculation, turning Benton into an optimist: “If you’re a scientist, never say anything is impossible,” he now tells students.

Most of these colorful revelations have emerged from fossils that contain evidence of melanin pigments—responsible for earth tones, such as red, black, brown, and buff—or the tiny cellular bags, called melanosomes, which produce and store melanins. But some scientists are already identifying the brighter hues of ancient snakes and insects. “Probably all of the colors can eventually be identified,” predicts Benton.

The work is not only helping to repaint the colorful pictures of dinosaurs that charm schoolchildren and museum-goers. Color can provide clues about how animals lived, mated, or died. Bright feathers suggest that animals used those hues to select a desirable mate, for example, whereas green coloration would offer camouflage amid foliage, indicating there must have been predators about. “Color has roles in lots of different parts of animal behavior and social interactions,” says Nick Edwards, a paleontologist at the University of Manchester in the United Kingdom.

But the fledgling field of paleo-color has also generated controversy. Some scientists dispute whether melanosomes could be preserved in fossils, and argue that researchers are being fooled by what are in fact fossilized bacteria. Others question the use of certain metals as proxies for pigments. “I think what’s happening is what often happens when people come up with new ideas,” says Luis Chiappe, Vice President for Research and Collections at the Natural History Museum of Los Angeles. “We got super excited because we had figured out ways of potentially devising the colors of these extinct animals… now we’re starting to realize it’s not as simple.”

The Race to Color a Dinosaur In retrospect, it’s not so surprising that pigment traces can survive for so long. Melanins, derived from the amino acid tyrosine, are remarkably tough polymers, and they tightly cross-link with proteins, such as the keratin in feathers. This strengthens tissues, and could even be responsible for helping to preserve fossil feathers, says Benton. Nineteenth century fossil hunters certainly knew that melanins could stand the test of time. Some even wrote letters with the melanin pigment ink that they found preserved in the ink sacs of fossilized squids. And it was during his own studies of squid fossils in 2006 that Jakob Vinther, then a graduate student at Yale University in New Haven, Connecticut, began to wonder whether melanins could persist in other fossils. A handful of scientists, including his supervisor Derek Briggs, had previously seen minuscule, sausage-shaped formations in a variety of fossils, and labeled them fossil bacteria (5, 6). Vinther suggested these formations could be melanosomes instead. Melanosome shape, he reasoned, could even give clues to color: sausage-shaped melanosomes contain the black pigment eumelanin, whereas round melanosomes hold reddish-brown pheomelanin. Briggs, a paleontologist, was skeptical, but willing to test the idea. He suggested that Vinther take a closer look at a fossil feather that had dark and light stripes. If the blobs were melanosomes, they should be found only in the dark sections. If they were bacteria, they ought to be all over the feather. When Vinther examined the light sections with a scanning electron microscope, he saw nothing more than the imprint of the feather itself. But in the dark sections he found the oblong bodies, along with their impressions in the rock (7). His hypothesis appeared to be correct. Once Vinther’s idea went public in 2008, the race was on to color in a dinosaur. It ended in a dead heat: Benton’s and Vinther’s teams unveiled the colors of stripe-tailed S. prima and red-headed A. huxleyi within days of each other in early 2010 (1, 2). Other discoveries followed fast, and Benton says that the majority of paleontologists have come to accept that the preserved blobs are melanosomes. Even Michael Wuttke, the paleontologist recently retired from the General Department of Cultural Heritage Rhineland-Palatinate in Germany, who originally thought they were bacteria (5), says he is now “absolutely convinced” that melanosomes are preserved in hairs and feathers. However, evolutionary biologist Mary Higby Schweitzer, of North Carolina State University in Raleigh, calls the conclusions “overstated,” a position she admits has made her unpopular among most paleo-color aficionados. “I’m not against melanosomes preserving,” says Schweitzer. But, she adds, it’s “more parsimonious to assume a microbial origin until disproven with data other than pictures.” Illustrations courtesy of Lucy Reading (graphic artist).

Rainbow of Possibilities Researchers have started to collect such data. Johan Lindgren, a paleontologist at Lund University in Sweden, uses time-of-flight secondary ion mass spectrometry (ToF-SIMS) and other methods to analyze the chemical composition of fossils. By exciting the fossil surface with an ion beam, Lindgren can measure the masses of the atoms and molecules that fly off, and he has discovered black eumelanin pigment itself in fossils, such as a fish eye and an A. huxleyi specimen (8, 9). Lindgren concludes that at least some of the microbodies in fossils are indeed melanosomes. Vinther, now a paleontologist at the University of Bristol, also uses ToF-SIMS; he and other scientists say this direct evidence of the chemical signature of melanin pigments clinches their argument. “As far as I’m concerned, case closed,” says Briggs. However, some critics point out that these analysis methods may offer a misleading picture of pigmentation. Essentially, they can only look at a limited number of spots from a given specimen. Scientists have to chip off a small bit of the fossil to subject it to mass spectrometry; to minimize damage they typically only look at a handful of chips. And Vinther and colleagues studied only one isolated feather to decide that Archaeopteryx lithographica was probably black (10). This kind of extrapolation could be wrong, say Schweitzer and others. Imagine trying to determine the coloration of a modern-day peacock from pigments at just a few dozen spots, she cautions: you might not come up with the right pattern. To avoid this problem, Edwards and colleagues use X-ray analysis (which causes no damage) to probe pigments across an entire fossil. Along with collaborators at the Stanford Synchrotron Radiation Lightsource (SSRL) in Menlo Park, California, these researchers use a technique called synchrotron rapid-scanning X-ray fluorescence that causes different elements to emit a characteristic burst of light. In fossil organisms such as C. sanctus, the scientists see copper, which is one of several metal ions that can bind to melanins, and is also found in tyrosinase, an enzyme that controls the production of melanins (4). Another technique, X-ray absorption spectroscopy, revealed that the copper was bound to organic molecules, which were “most probably derived from precursor pigment molecules,” says Roy Wogelius, a University of Manchester geologist on the team. Based on these techniques, the researchers concluded that the tips and one side of A. lithographica feathers were darkly pigmented, whereas the other side was lighter (11). Elucidating Structural Color Pigments are not the only source of animal colors. The bright hues of a butterfly’s wing and the metallic sheen of a beetle are structural colors, produced when light bounces off multilayer, nanoscale structures that scatter the rays. Maria McNamara, of the University College Cork in Ireland, has shown that these nanostructures can occasionally survive the fossilization process (15), allowing researchers to decipher the color of insects that lived long ago. McNamara and her team examined the greenish-blue fossils of 47-million-year-old forester moths by using electron microscopy and spectrophotometry to analyze how they scatter light (16). The researchers concluded that the moths’ wings contained multilayer structures that would have produced color in life, but the wings probably weren’t originally greenish-blue. That’s because the fossilization process can change the architecture of the nanostructures, and therefore the color they produce. To simulate this process, McNamara exposed modern-day beetles to high temperatures and pressure in an autoclave. She found that the highest temperatures turned their wings black, whereas lower temperatures shifted the hue toward the blue end of the spectrum (17). Based on their measurements, McNamara and colleagues decided that the forester moth would have actually been a matte yellowish-green, likely matching leaves in its environment. This implies that these prehistoric moths had already evolved camouflage mechanisms to deter predators. Meanwhile, other researchers have identified structural color in a 50-million-year-old beetle, and McNamara and others have found it in fossil feathers (18⇓⇓–21). “It is an important area,” says Nick Edwards, a paleontologist at the University of Manchester in the United Kingdom. “The structural aspects to coloration … can have a profound effect on how coloration can be perceived by another organism.” Vinther and Ryan Carney, a paleontologist at the University of South Florida in Tampa who worked with him on the A. lithographica feather, are not convinced. The researchers point out that copper binds to all kinds of organic compounds, including the humic acids in decomposing biomaterials. Briggs and Schweitzer add that copper compounds could have moved around between fossilizing organisms and the surrounding rock. Nevertheless, Wogelius and Edwards argue that because they map the entire fossil, they can correlate organic copper complexes with the places they expect to see pigmentation, such as in feathers and fossil melanosomes (4, 11). “That’s why the whole image is so important,” says Edwards. “It’s another way to check whether we’re being fooled by some sort of geochemical process.”

From Color to Behavior Armed with these discoveries, paleontologists are now considering how color fits in with their suspicions about prehistoric animal behavior. For example, Vinther suggests that the findings of colorful fossil feathers could help scientists understand why feathers evolved. The very first proto-feather filaments were probably no good for flight, and paleontologists think they helped keep dinosaurs warm. But it’s a big leap from single filaments to the branched, flat-feathers needed to soar into the prehistoric skies. Pigments seem to appear around the same time as flat feathers, and Vinther suggests they served as a “billboard” for displays—to attract a mate or indicate one’s species—before they ever took creatures airborne. Scientists already knew that birds and dinosaurs had ornamental feathers or even bony headgear, notes Chiappe, but the color findings certainly support the idea that those structures were for show. Pigmentation isn’t always about flash, though. In one study, Lindgren found dark eumelanin pigment in three fossil reptiles: a leatherback turtle, a lizard-like aquatic mosasaur, and an ichthyosaur (3). Lindgren suspects the ichthyosaur was uniformly dark, like modern deep-diving sperm whales, which are camouflaged by their coloration in the ocean’s depths. This theory is supported by other evidence that the icthyosaur might have been a deep diver, such as its large eyes (12). Back at the surface, the dark pigment might also have helped the reptiles absorb heat while basking in the sun, as it does for modern leatherbacks. Recently, Vinther and colleagues used melanosomes to discern the identity of the mysterious Tullimonstrum gregarium fossils, first found in Illinois in the 1950s by amateur paleontologist Francis Tully. The creature’s squishy finned body, elephantine trunk ending in eight sharp teeth, and bizarre, barbell-like midbody organ earned it the nickname “Tully monster,” and paleontologists have long debated whether it might have been a mollusk, a worm, or a vertebrate. Using electron microscopy and ToF-SIMS, the team identified melanosomes in the creature’s eyes, arranged in layers typical of vertebrates. The team concluded that the Tully monster was a vertebrate, albeit an unusual one (13). These kinds of studies demonstrate the impact that paleo-color is having on the wider field, says Carney: “It allows us to gain more information than just what they looked like; we can infer things about their function and behavior.” Based on samples taken from a leatherback turtle fossil (A) (scale bar, 10 cm), this ion image (B) (scale bar, 3 μm) shows the spatial distribution of peaks characteristic of eumelanin (green), silicon oxide (blue), and sulfate (red) superimposed onto a scanning electron microscopy image of the “skin.” An enlargement (C) (scale bar, 300 nm) of the demarcated area in B (white box) shows a melanosome-like microbody. Reproduced from ref. 3, with permission from Macmillan Publishers Ltd.: Nature, copyright 2014.

Beyond Melanins Melanins are only part of the natural color palette. Other pigments, such as the plant-derived orange carotenoids, are commonly found in modern species, from flamingoes to goldfish. (And pigments aren’t always the color source; see Elucidating Structural Color.) But these molecules are much less stable than melanin pigments, and lack the heavy elements needed for X-ray analysis at SSRL, making it challenging to find them in fossils. Still, it might be possible to deduce their presence. Maria McNamara, a paleobiologist at University College Cork in Ireland, has found a way to infer these other colors in a 10-million-year-old, exquisitely-preserved snake skin from Spain (14). The snakeskin is mineralized, its biological structures replaced by calcium phosphate. Nonetheless, the snakeskin preserves the characteristic microscopic structures of individual cells. In reptiles, three distinct cell types are largely responsible for producing color: melanophores make black; xanthophores hold red, yellow, and orange carotenoids; and iridophores contain crystals that scatter light. A combination of melanophores with iridophores or xanthophores produces green coloration. Each pigment cell type has a different shape, size, and location, so McNamara and her team used electron microscopy to identify them in the fossil snakeskin, and concluded that the creature once sported blotches of green, black, and yellow-green, with a pale underbelly. “These types of blotches are really good at breaking up the body outline for camouflage,” McNamara explains. She believes that the snake probably evaded predators during the day by hiding among leaves, as many modern snakes do. “It’s an amazing specimen, and it really opens the door to expanding our palette of color reconstruction,” says Carney. McNamara suspects other fossils might yield to her technique; fish and amphibians also have diverse types of pigment cells. But Matthew Shawkey, an evolutionary biologist at the University of Ghent in Belgium, is skeptical of the findings because, he says, the fossil cells that McNamara identified do not look much like the modern versions, meaning they may not be pigment cells at all. Briggs, who advised McNamara during her postdoctorate studies, suspects that the cells changed shape as the skin mineralized (as, for example, muscle cells have been shown to do in some fossils), and so he believes the results are valid. Briggs doubts that the method will find wide use, though, because it requires fossils with finely preserved details. Although some scientists remain to be convinced about paleo-color studies, the public has embraced the discoveries, a side benefit of the work that’s not lost on the researchers. Both Edwards’ and McNamara’s team have taken part in the United Kingdom’s Royal Society Summer Science Exhibition in London, a prestigious outreach event, and McNamara also showcased her research at the World Science and Technology Fair in Bangkok in 2012. “Science can often be intimidating to students and kids,” says Carney. “Dinosaurs, especially colored flying dinosaurs, are a great way to get them excited.” The discoveries are certainly enticing, agrees Philip Donoghue, a paleobiologist at the University of Bristol. But then, it’s not too surprising to find that ichthyosaurs were dark or that early birds flashed ornamental feathers. He wants more. To achieve truly novel insights, scientists may need to identify colors in many more fossils. That could allow them to investigate the geographical distribution of different color patterns, how they vary within a species, or how colors evolved over time. Following the initial thrill of the first paleo-color studies, Donoghue says that researchers now need to pose some key questions: “What work can we put that data to?” he asks. “What hypotheses can we now test?” With any luck, as the paleo-color field matures, such inquiries will offer new insights into the lives of prehistoric animals, even as it brightens museum exhibits and children’s books with true-to-life colors from the distant past.