One of the most enigmatic aspects of dinosaurs are their colour. Aside from what we've gathered from only a handfull of fragments of fossilized skin, anything about the colouration or patterning of dinosaurs and ancient birds is really unknown. Artists go to great lengths to illustrate "realistic" pictures of ancient creatures, but they simply are just an artist's rendition, something pulled from the imagination. However, it's possible that taking an "educated guess" about the plumage of ancient birds might become as extinct as the creatures themselves.
An article published in the most recent issue of Biology Letters by Vinther et. al details a new discovery regarding the colouration of fosilized feathers (Vinther et al. Biology Letters 10.1098/rsbl.2008.0302). The article discusses a known fossil of a feather from the Crato
Formation in Brazil (Figure 1a in the article). The reason the fossil is impressive is twofold: firstly, it is very well perserved and shows amazing detail, including the barbs and barbules. Secondly, and more importently, it shows a striking banding pattern. Previous studies on the fossil showed the black bands to be made mainly of tiny carbon structures (Davis and Briggs, 1995). This is hardly a surprise, because fossil feathers are usually composed mainly of carbon. Use of scanning electron microscopes (SEM) showed the structues to be oblong bodies about 2 microns across. These were thought to simply be fossilized feather-degrading bactera from our very distant past.
Vinther et. al. took another look at this fossil. They took a wing from a Red-winged Blackbird , ground it up, froze it and then compared SEM images of the feather to the fossilized bacteria. What they discovered was that the bacteria weren't bacteria at all! Instead, they showed an incredible similarity to eumelanosomes, the structures in feathers that give them dark pigmentation. The authors concluded that this was no mere similarity; the structures are ancient eumelanosomes! This would mean that the dark bands on the feathers were actually dark-pigmented bands on the living specimen. The lighter parts of the feathers showed none of these structures, and would then mean that they were lighter (if not white) bands on the original feather. This makes much more sense than the original bacteria hypothesis; after all, why would bacteria only perserve in the dark areas of the fossil and not the light ones? It is more likely that, since eumelanin (the pigment-prodicing chemical in melanosomes) is resistant to degredation to chemicals, the black feathers themselves are more resistant to bacterial decay than white feathers. All this means that, by examining fossilized feathers for eumelanosomes, we can determine what the patterning of ancient birds looked like!
But it gets even better. Melanosomes don't just give a dark, black pigmentation. Melanosomes of different shapes, arrangements and distributions confer different colours like red, black, or brown. Other types of melanosomes, like Phaeomelanosomes, can give other colours like yellow, though the authors caution that it is unknown if these melanosomes preserve the same as eumelanosomes. But, if they do, this would suggest that not only can we determine the patterning of ancient birds but also their colouration! This could potentially extend to determining the colouration of feathered dinosaurs as well. Also, it has implications in understaning the behaviour and ecology of these ancient creatures. As the authors state, birds have the ability to see a broad range of colours; this is one good reason why birds have evolved elabourate and colourful plumages. Thus, knowing the plumages of ancient birds and dinosaurs could give us some insight into the way they lived (assuming, of course, that colour vision didn't evolve after birds had sufficently diverged into their modern-day descendants). Perhaps the illustrations we see in books, magazines and on TV will soon be alot more accurate!