It is a great age for dinosaur studies, with more and more fossil evidence coming to light, including finds from China, South America, and other areas sparsely recorded until this century.
One of the most exciting developments has been the discovery of non-skeletal remains, which was considered unimaginable when I was a kid (i.e., when I first became interested in dinosaurs and fossils).
In recent years we have seen remarkably preserved feathers, not only proving that dinosaurs wore feathers, but giving us a detailed view of what the feathers look like. Even more, this gives an idea of the color scheme of the animal.
This month an international team led by Jakob Vinther reported a remarkably preserved Psittacosaurus with considerable amounts of skin closely associated with the skeleton. The residues contain melanin, and the distribution was used to infer the possible visible appearance of the animal. Overall, it appears to be a (not too surprising) “countershading” pattern, dark on top, and light underneath, as well as a black or dark face.
Moving rather farther from the fossil remains, they also conducted a computational study, projecting the inferred colors onto a 3D model of the reconstructed body. The resulting model was studied by applying different lighting, representing conditions in possible habitats (open desert, deep forest, etc.) They conclude that the pattern is an effective visual camouflage for a forest. The pattern might also suggest bipedal walking, though that is more speculative.
I note that Psittacosaurus has a rather mysterious bristles on the tail. These don’t resemble familiar avian species, and it is not clear what function they may have had. Dinosaurs are so weird, that’s part of what is so fun!
This is an outstanding find and it is so cool to get some real notion of what dinosaurs looked like.
It is also an interesting example of “computational” argument, sliding farther and farther from the fossil data, piling inferences upon inferences, and comparing models (theoretically based) with reconstructions from the data. The headline conclusions about color scheme and inferred habitat are based on a rather long chain of inference, as well as arguments based on computations.
I call this “computational” argument, though the computations were not fully digital as would be expected. Actually, the modelling in this case was surprisingly “analog”. For some reason they eschewed the obvious digital modelling in favor of a much less convenient and accurate plaster model. The lighting was done in situ in a botanical garden, rather than using simulations that could explore a broad and continuous range of environments.
The final conclusions are scarcely controversial (countershading is a common adaptation, the fossil was found in a fossil forest environment). Still, I think it will be important to replicate this study, revisiting the assumptions and inferences explicit and implicit in the modelling and computer analysis. I’m not sure whether the published materials are sufficient for such a review or not.
This replicability problem is a perennial challenge for digitally enabled science of all kinds.
- Jakob Vinther, Robert Nicholls, Stephan Lautenschlager, Michael Pittman, Thomas G Kaye, Emily Rayfield, Gerald Mayr, and Innes C Cuthill, 3D Camouflage in an Ornithischian Dinosaur. Current Biology, http://www.sciencedirect.com/science/article/pii/S0960982216307060