Everything You Thought You Knew About Dinosaur Colours is Wrong
The new science of paleocolour is shining a light into what dinosaurs looked like and how they lived.
Paleoartist Robert Nicholls is modelling the head of a Tyrannosaurus rex. A fossil skull dictated the shape of the head but the colour is, at present, informed guesswork. Should he evoke the sorts of earth tones associated with many modern reptiles and amphibians? Or look to the bright, flamboyant plumages of some modern birds – the only dinosaurs to survive into the twenty-first century? Up until now we have only been able to guess, but soon that may change.
Long thought an impossible dream, the emerging field of palaeocolour is revolutionising our view of the prehistoric world, turning it from black-and-white into glorious technicolour. So far only a handful of dinosaurs, insects and reptiles have been studied but, as Johan Lindgren, a scientist from the University of Lund, says, “We’re only just scratching on the surface.”
Finding evidence of colour in the fossil record will do much more than simply tell us what hue to paint a T. rex. Bones can fossilise. but behaviour does not. “When we look at the animals and plants we see in the world around us we see striking colours and colour patterns,” says Maria McNamara from the University of Cork. “Animals use colour for camouflage, for avoiding predators, for mating signals and also for signalling within their social group. So evidence of colour in animals has the potential to tell us about this very enigmatic aspect of the biology of ancient organisms.”
Despite this, it is only in the decade that palaeocolour has become an area of serious research. Scientists have known for decades that some fossil shells and insects appear to preserve colour but no one had been sure whether that colour was real or some by-product of the fossilisation process. In exceptional circumstances skin and feathers will fossilize – in much less than one per cent of all fossils discovered – but the colour of the fossil (typically black or brown) is not a guide to the colour of the living creature.
In 2006, Jakob Vinther – then a PhD student at Yale, now a lecturer at the University of Bristol – sat down to study the fossilized ink sac of a 200-million-year old squid relative. Since the nineteenth century fossil hunters have known about the preservation of such ink sacs – some even wrote letters using the ink - but for Vinther it proved a eureka moment. “I was looking at this ink and realizing that it’s identical to ink in living squid and so it must be composed of melanin. That’s the same pigment that we have and dinosaurs have.”
Pigments produce colour by the selective absorption of certain wavelengths of visible light. Typical pigments include melanins, carotenoids (bright reds and yellows) and porphyrins (greens, reds and blues). Other colours – known as structural colours -- are produced by light scattering nanostructures. (The brilliant iridescent colours on a peacock’s tale, for example.) Melanin controls, among other things, our hair and eye colour. It is produced and stored in tiny cellular bags called melanosomes, and these come in two forms. A sausage-shaped type produces black shades; a round-shaped variety creates rusty reddish hues.
Combinations of these melanins alongside absence of pigment create grey, brown and white colours. If pigments had been preserved in the ink sacs, Vinther reasoned, then melanin – or the melanin-bearing melanosomes - might also be found in fossilised skin and feathers.
He found his answer in the fossil skull of a small, 55-million-year-old bird from his home country of Denmark, preserved with a dark halo of feather impressions and two stains where the eyes use to be. “I was sat there [looking for evidence of melanosomes] zooming in with the microscope, and suddenly I was like, blimey, they’re there! We can put colours in fossil dinosaurs.”
His supervisor Derek Briggs was sceptical. The structures Vinther described were already well known and classified – as bacteria. “They are the same size and shape as bacteria and they’re found on these rotten carcases where you would expect to find decay bacteria,” explains MacNamara. “It all seemed very plausible.” Seeking further evidence, Vinther and Briggs looked at a fossil feather from the Cretaceous period with distinct black and white colour bands. Where the feather was black there were sausage-shaped melanosomes, where it was white there were no melanosomes (white indicates an absence of pigment). Had the melanosomes been bacteria they should have been seen on both the black and white parts of the feather.
Vinther published his initial findings in 2008, and then race to produce the first coloured dinosaur ensued, using the shape of the melanosomes to deduce hue and pattern. In 2010, two teams at the University of Bristol, one led by Vinther and one by Michael Benton, published within days of each other. They showed, respectively, that the birdlike Anchiornis huxleyi was crowned with a red crest, and the feathered dinosaur Sinosauropteryx prima had a reddish-brown striped tail.
Since then, further studies have built on Vinther’s original hypothesis, including work by Lindgren, who used a technique called time-of-flight secondary ion mass spectrometry (ToF-SIMS) to analyse the chemical composition of various fossils and found direct evidence of the chemical signature of melanin pigments. Elsewhere, McNamara has explored the preservation of non-melanin coloration methods, including the first systematic investigations of the fossil record of structural colour. In 2016 she published the first paper showing evidence of the preservation of carotenoid pigments, reconstructing a 10 million-year-old green-and-brown patterned fossil snake from northeastern Spain.
The are also concerns that the new field is developing too quickly, that claims are being made which, Lindgren suggests, state more than we actually know.
One issue, as he sees it, is that a fossil may originally have contained not just melanin but other pigments or structural colours that didn’t survive the fossilisation process. “Then we would just get a false image. So the colour reconstructions don’t necessarily show what the animal actually looked like.” Another issue is that extrapolating the colour of a bird or dinosaur from a single isolated feather – or from the minute samples that are used in the ToF-SIMS technique – could be misleading. Imagine trying to determine the colouration of a modern-day peacock from pigments taken from just a few spots, Mary Schweitzer of North Carolina State University cautioned in a recent PNAS article.
For McNamara, a key point is that melanin itself is not yet properly understood. “We need to learn more about melanin in modern animas before we go near fossils. We really did jump in with two feet.” Melanin provides for much more than just colour – UV protection, for instance, and mechanical strength (the reason some birds have dark wing tips). She is interested in the fact that melanin exists not just in hair and skin but also in internal organs. “We’re going to try and find out what is controlling melanin evolution. We’ve always thought oh its for colour, oh its for sexual selection and camouflage, but if it’s in all these internal organs then maybe it evolved for a completely different purpose.”
Lindgren agrees. “In general there are still so many unknown factors,” he says. “My feeling is that as palaeontologists we have a tendency to oversimplify, to think A must point to B. But if you were to ask a biologist, we know that in the modern world there is never ever a single factor that leads to one thing.”
But for Vinther, the continual arguments about palaeocolour go too far. Every time he publishes a new paper, he wonders what people are going to say, what he will have to respond to. “It’s tiring and stressful. I wish people would spend more time advancing the field rather than holding it back with objections. Let’s move forwards. Let’s try and figure out what the frontiers and the limitations are in a quantifiable way.”
This July scientists in Australia announced the discovery of the world’s oldest colour, which turns out to be a rather fetching shade of bright pink. These porphyrin pigments, derived from microscopic sea-dwelling organisms, date back an astonishing 1.1 billion years, long before the existence of the insects and animals Vinther, McNamara and others are studying.
As our knowledge of ancient colour increases, scientists may be able to chart its progress through deep time, answering questions such as, what is it that drives colour evolution? Is it natural selection – the desire to hide yourself – or sexual selection – the desire to advertise? “Or was there ever a time when competition pressures were less and sexual selection wasn’t happening?” McNamara wonders. “What would colour patterns look like if they weren’t controlled by those factors? Would you just get really crazy patterns? We can’t assume that the world we see today has always been like this.”
And already palaeocolour can bring new insights into the daily lives of long-dead creatures. For instance, it had long been presumed that the small, four-winged Microraptor was nocturnal, based on the large size of its eye sockets. Then Vinther, Quanguo Li from the Beijing Museum of Natural History, and colleagues, discovered that the dinosaur possessed iridescent plumage (an example of structural colour) – something that would make no sense if this dinosaur were active only at night.
It can also tell us about the environment an animal lived in. Typically scientists gather clues by looking at other fossil animals and plants found nearby -- but this technique falls down if the animal’s body has been transported – by, for example, a river – away from the place where it had lived.
Vinther studied the fossil of a small, plant-eating dinosaur called Psittacosaurus, a relative of Triceratops, and concluded that it had a dark back and pale belly – a colour arrangement known as counter-shading. Common among modern animals ranging from whales to deer, both predators and prey use it to blend in with their surroundings. (Parts normally in shadow are light; parts normally exposed to the sky are dark.) The amount and distribution of the light and dark areas typically corresponds to different habitats e.g. open plains or dark forest floor. The counter-shading on the Psittacosaurus suggests that it lived in a habitat with diffuse light, such as a canopy forest.
In his workshop Nicholls shows me the model of the Psittacosaurus that he developed with Vinther. About the size of a Labrador, this rather cute looking creature has a distinctive parrot-billed beak and a dark brown and orange mottled back that becomes progressively paler, down to a creamy underbelly. “What I really like about this colour reconstruction work is that you’re the one defining what an animal looks like, down to the colour pattern, for the first time,” Nicholls says. “Being able to show people something that no one has ever seen before. That is the best.”