, , , , , , , , , ,

Following on from my previous post about how we colour species for which we do not (yet) have available colour data, I thought I’d talk a bit more about ‘playing it safe with colour’ – a topic we touched on briefly before. This deals with the likelihood of various colours in nature and how that likelihood changes in differing circumstances. First, a probability primer, which you are more than welcome to skip it if you are confident about the subject.


You’re probably all aware that there is a 50% chance of getting either heads or tails when a coin is flipped in the air. If the first flip resulted in heads, the second flip is still equally likely to be heads or tails. I mention this to dispell the widely-held belief that the 1 in 2 probability means the next flip is more likely to be tails due to some invisible process keeping track of how often you’ve been lucky/unlucky so far. If you were going to predict that both coins would result in heads, the probability of you being right is 1 in 4 (odds times number of times flipped). “Wait, isn’t that a contradiction?” not at all: it is less likely for 2 coins to land heads than for one coin to land heads but each coin flip is as likely as any other. It is equally likely to successfully predict first heads then tails as it is to predict both heads or both tails.

“Why do I need to know this? I just want to paint my herd of ceratopsians drowning in a flash flood!” well, herd diagrams are where probability gets really relevant: we’re looking at the cause of colouration and the cause will affect the probability of each colour occurring in your group. Suppose your male ceratopsian has a brightly coloured head crest. You need to identify how many of your herd are likely to be male. Start with a simplistic 50% chance and then start adding factors: if your colour display develops with sexual maturity, you need the percentage of adults in your male group, which can be calculated using size of egg clutch per season, age range of individuals in the herd, likelihood of any one individual surviving to maturity, and (where applicable) likelihood of adornment increasing the risk of predation – as is the case for peacocks where predators will pull a male out of the tree by its tail. If your colourful crest develops as a result of being an alpha male, such as the plated cheeks of Orangutans, there may only be 1 or 2 in your group with the markings. If it’s winter, and your gender-specific marking only develops around mating season, 0% of your herd will have it. There are often no definitive answers to these questions in the fossil record but analogies with living animals help and they are still important to consider – the ‘something is not quite right’ brain response can be a cruel taskmaster.

Most Frequent Colours

As mentioned in the earlier post, melanin is the most common form of pigmentation. Eumelanin is more common than Phaeomelanin so, to play it safe with probable colours, you should stick to the white to black, blonde to dark chocolate brown range with an occasional flourish of ginger here and there. Do bear in mind that pure white has disadvantages in practice (see camouflage section below) and is therefore much less common. That said, you are quite safe illustrating an occasional albino as the condition has been observed in numerous species spanning mammalia, reptilia, and aves. If you want to be extra safe, draw an albino new-born as they don’t often survive to adulthood.


Being able to hide from predators or ambush prey is a distinct evolutionary advantage over competitors that can’t. For this reason, camouflage markings are numerous in the animal kingdom. What works as camouflage in one place will stand out like a sore thumb in another – imagine placing a bengal tiger in the arctic, where his stripes no longer blend in with the patchy patterns of the jungle. This is why the habitat changes the probability of encountering specific colours: If I draw a green Hadrosaur I’d expect it to do well in forested, lush green areas; If I draw a mustard coloured Theropod, I am giving it an advantage in the desert but a disadvantage elsewhere; grey creatures would do well in rocky, mountainous regions and so on.

Don’t just think about whether your subject can hide from you – you want to think about whether it can hide from its predators or prey. Suppose you’re drawing a large herbivorous mammal, such as Macrauchenia – a South American hoofed beast – your subject must hide from some terrifying, ground-running birds of prey called Andalgalornis. Most birds can see colour (which includes seeing much further into the ultraviolets than we can). Much later in its evolutionary history, the nightmarish birds go away and Macrauchenia is left to face mammalian predators. Mammals tend to have worse colour vision than birds. This means that our herbivore could look a lot less like his background and still hide from his hunters.

No modern mammal is able to produce natural green pigment, despite most natural hiding places being at least partly green. Hairs are too simple for structural green colouration and mammals don’t seem to be able to process carotenoids into dietary pigmentation, so green doesn’t even feature as an option – even if it presents a significant increased likelihood of survival. Was that always the case, even when large mammals were regularly preyed upon by non mammals with remarkable vision? Chances are in favour of you getting the colour right by basing it on an extant mammal than by going out on a limb and making a green Macrauchenia.

Many species have camouflaged markings as infants that they lose later on when they’re more experienced. Find me one example where the infant is brightly coloured while the parent is camouflaged. The only way I see it working is if the infant hides under the parent but it still seems like an unnecessary waste of energy. Quality plumage comes with an energy cost. Many birds will grow poor quality, dingy feathers so they can grow and get airborne as fast as possible, then worry about the decent coat at the next molt. This is why fledgling crows look like brown, patchy adults. What does a baby animal need pristine bold colours for anyway? Sexual selection is not a factor yet and they have no business intimidating challengers from other herds. The only reason I can think of is if the colours deter predators by making them think the little ones are venomous or toxic (see the following section on danger colours).

Lastly on the subject of camouflage colours you should note that most species who wish to blend in will have a darker back than their belly. This is true regardless of colour, species, family, or class. Here’s a challenge for you: find a species whose belly is darker. If you found Hallucigenia there’s a very good reason for that: he’s upside-down. Most light sources in nature come from above. The ones that don’t (forest fires, for example) aren’t worth camouflaging against. If the light is above you and your back is darker, the shadow cast by the light will cancel out against your markings. If your back is lighter, you will stand out in contrast all the more (cue the dramatic Hadrosaur death painting with stark white belly flashing). As you can see, we don’t just have colour to think about – we need to play the odds with value too.

Danger Colours

When an animal has weaponry – venom or a nasty sting – it can afford to forego camouflage. As a result, animals soon learn that brightly coloured species mean danger. In some cases, the animal is bluffing and is merely coloured as though it were dangerous (hover flies pretending to be wasps, for example). Probability based on contemporary animals suggests that insects, arachnids, amphibians and reptiles are more likely to employ danger markings and mean it. Are bright yellow birds employing a form of threat bluff? It would need to be established whether their predators are aware of bright yellow being a danger colour and whether they are more or less likely to actively prey upon a duller coloured morph of the same species. I’m not sure whether such a study has been done. I do know that, in cases where birds are toxic to the touch or when consumed (such as the Hooded Pitohui), those birds are not significantly more colourful than non toxic birds from the same region. This suggests that the predators are less likely to learn that brightly coloured = danger. Toxic mammals do not employ bright colours, for reasons already stated.

Besides bright colours, another danger tactic is to show eyes on other body parts to resemble their predator’s predator. Butterflies, moths, and birds employ this tactic today. I have seen numerous illustrations of ceratopsian frills with big eye patterns. I find this quite hard to believe – mainly because anything that hunts large ceratopsians is already the apex predator and fears very little. What tyrannosaur is going to look down at a Torosaurus and fear a creature with eyes about 16 times as large as his own coming up at him? He hasn’t learnt to be afraid of that predator because no such predator exists. The ceratopsian is a dangerous animal in its own right and can make the tyrannosaur wary of whatever markings it has. It tends to be the case that animals which resort to intimidation markings like eye impressions are usually bluffing. Protoceratops: now there’s a candidate for eye decorations on head crests. It’s small and its natural predator (Velociraptor) had something bigger to fear (Tarbosaurus).

Sexual Selection

With birds, if the male has different plumage to the female, the male is the most brightly coloured. The male is also usually a little smaller. It’s an important conscious decision to make if your subject is going to be doing the hunting for a mixed gender social group, or nest sitting. Some males do tend the eggs (Penguins, for example) but odds are your nest sitter is the female.

One other thing to be wary of: nocturnal birds such as owls tend to have worse colour vision than diurnal ones. Try not to give your night birds sexual display colours that their partners would not be able to see.


We’ve already touched upon albinism, the condition which has the dramatic effect of leaving your art subject starkly white with pink eyes. This is not the only malady that leaves its mark on colouration: dietary pigment is only as good as your food supply, so a malnourished bird is a drab creature indeed. If your Baryonyx is revelling in an abundance of fish, you can draw his yellows vibrant and healthy. If he’s prowling the delta unsuccessfully, perhaps a more dreary yellow is in order. Equally, your Brachylophosaurus’ bright red nasal markings could get quite muted if you’re painting him in winter.

Structural colouration can be damaged, so you can demonstrate that your blue-tinged Troodon is a seasoned fighter with copious bronze/brown nicks in his plumage.

Mammals are equally colour afflicted with poor health: the glossy black of a cat’s coat can go quite brown with old age and/or kidney problems; the melanic lustre of your hair will fade to grey with age. Balding too can indicate stress in your bird or mammal, which will of course affect their colour (and overall shape).

Some birds, whose colours are supplemented by the oil in their preening gland, will look more faded if they are not washing.

Reptiles that regularly shed their skin will look a lot brighter and shinier after shedding than they did before-hand. This is not strictly an ailment so much as a seasonal bodily function but I’ve included it for completeness.

The Water Effect

If you were to ask me which prehistoric animal has been accidentally given the correct colours most often, I’d say the Plesiosaur. To my knowledge, there is no evidence for colour in Plesiosaurs yet (we live in hope). Despite this, when an artist paints a blue plesiosaur submerged in the middle distance, they are more-or-less guaranteed to be correct. The reason for this is a simple one: water drains colour from light the deeper you go. At 50 metres down, all the reds are gone, most of the oranges and a good chunk of the violets. By 100 metres, the yellows and violets can no longer be seen. Keep going to 150 metres and half your green spectrum is gone, along with the darker blues. Regardless of its colour at the surface, most things look a shade of cyan. How green or blue that cyan looks will give you a rough idea of the original tone. Apparently this topic is important to anglers as well as artists. Here’s a site with a nice spectrum illustration of this effect. Air does this too, by the way: this is why trees look bluer on the horizon. We don’t have any land-based giants at the moment, so we don’t get to appreciate the effect in relation to animal subjects but, in times gone by, the casual observer would have noticed pale blue brachiosaurs grazing in the distance.

Borrowed Colour

Some animals are not the colour they started off with. Some species of sloth become coated in green algae – appearing for all intents and purposes like a green mammal. Some fishing birds (the Marabou Stork among them) defecate on their own legs, giving them a white coat instead of the black they started off with. There are a few other examples but they are definitely in the minority when viewing ecosystems as a whole… Unless you count dietary pigment, in which case borrowed colour is positively abundant – at least in Archosaurs.

That’s all for now on the topic of extinct colour prediction. Tune in next week for part 3 including a fun game.