I have always had a fascination with paleontology – the reconstruction of extinct species and landscapes. I still enjoy documentaries that do a good job of transporting me to another world in the Earth’s past.

One disappointing aspect of such reconstructions, however, is that we have very little idea what many extinct creatures really looked like. You can only tell so much from fossilized bones. Occasionally we get skin impressions as well. But artists have largely guessed when it comes to color. We can make educated guesses, from looking at extant species and basic principles such as camouflage, but this has significant limitations. Could we ever infer the coloration of a tiger from it’s fossils? Probably not.

In addition, there are many specialized soft-tissue features, sometimes even defining features of animals, that do not fossilize – a camel’s hump, for example, or an elephant’s trunk. What dramatic features of dinosaurs or ancient mammals have we simply no clue about from their fossils?

That is why I love breakthroughs that offer us a new window into the past – ways of inferring things about extinct species that may have seemed impossible before. There have already been a number of studies published looking at melanosomes, fossilized pigment protiens that allow us to reconstruct at least part of the color of extinct birds.

Another group has now published a different technique that appears to give us more information about feather color, whether on bird ancestors or non-avian theropod dinosaurs.

This also raises another excellent example of the limitations of reconstructions based entirely on bones. Our picture of theropod dinosaurs, like velociraptor, were likely completely off because we pictured them with lizard-like skin – typical dinosaurs. But with the discovery of new specimens with feather impressions, it now seems probable that many theropod species sported dramatic feathers, for display or other purposes.

The new technique uses synchrotron X-rays to look for and map the chemical residues of melanin pigments. While there are other pigments in feather color, melanin pigments seem to be dominant. Specifically they are looking for trace metal (such as copper) organic compounds that are related to specific types of melanin.

This chemical approach is complementary to the prior techniques, which used melanosome morphology to infer color – were they rod-like or more rounded.

The authors conclude from the current results:

This suggests that Confuciusornis sanctus most probably had darkly shaded regions, with the most intense eumelanin pigmentation in the downy body feathers and in the lengthy retrices.

The primary limitation is that there may be non-melanin pigments, like carotenoids, contributing to color. They also used their technique on feathers from living birds to show that the results are valid. So while there are limitations to the information, it does appear to be telling us something real about the color of these animals.

These techniques are really cool – who would have thought previously that we would be able to tell anything about the color of extinct animals. Hopefully this is just the beginning also. This and similar techniques of finding trace remnants in fossils that can be used to infer much more about the specimens than just bone structure are beginning to take off. Already we can infer what animals likely ate, how much oxygen was around, and similar things.

I hope researchers discover similar techniques by which to infer more about the soft tissues of fossil species. Slowly, a more and more clear picture of extinct animals is emerging.