Conservation lecture: Ultraviolet vision in mammals

How good is his camouflage?

I have not posted for a while on the monthly conservation lectures we have at Bristol, but this week’s one was too interesting not to share. It was given by Professor Ron Douglas, who is Professor of Visual Science at City University, London, and was on the often misunderstood nature of colour vision and ultraviolet perception in mammals.


First, a brief outline of how colour vision works. Vertebrates have two types of light sensitive cells in the retina of the eye. Rod cells respond to more or less all frequencies of light, and when a photoreceptor cell is struck by a photon a chemical reaction occurs which generates a nerve impulse to the visual centres of the brain. Rod cells are more sensitive, and are used for vision in low light levels. However, they cannot respond differently to different frequencies, and so enable colour vision. This is carried out by the other cells, called cone cells, which occur in several varieties each containing a receptor pigment that only responds to a specific range of frequencies, from short wavelengths (= blue or UV light) through to long wavelengths (= red or infrared light). They are not tuned to a single wavelength, but have a peak absorption in different parts of the spectrum. The brain interprets the different strength of response from different cone cells to distinguish colours.

Ask most people what a dog sees and they will generally say they are colour blind. Most people will understand this to mean that they see the world in shades of grey, and in fact this is not true. Technically, dogs (and most if not all mammals except primates) are dichromatic, which means they have two photoreceptor pigments in their eyes. By contrast, most if not all non-mammalian vertebrates are tetrachromatic, with four different pigments. Most primates have three pigments, and are termed trichromatic. What seems to have happened in the evolution of mammals is that the original vertebrate tetrachromatism was reduced to a dichromatic state by deletion of two of the visual pigments. Most early mammals seem to have been nocturnal, and in low light levels cone cells do not work so well, which explains how a serious visual defect managed to survive. In primates the sole remaining long wave receptor was duplicated, and subsequent mutations have shifted the absorption frequency of one of the copies towards the blue end of the spectrum, improving colour discrimination somewhat. Even so, the colour discrimination even in primates is not nearly as good as in non-mammalian vertebrates, whose colour receptors are spread more evenly across the visual spectrum.

There is however an additional wrinkle to colour vision. In order to respond to a frequency the cone cell must actually receive the appropriate photon, and in many mammals (including humans) the lens of the eye is actually opaque to UV frequencies. The human short-wavelength receptor is actually quite responsive to UV frequencies, but humans still cannot see in IV light because the photons are absorbed by the lens of the eye before reaching the retina. If the lens is destroyed, for example to remove a cataract, humans can see UV light (apparently it looks a whitish blue), but in modern cataract surgery a UV-opaque artificial lens is inserted as a replacement, so patients will not experience the same changes to colour perception these days.

Prairie Dog (UV-blind)

So, how does this affect colour vision in mammals? It turns out that how transparent the lens is to UV varies considerably across mammalian groups. Primates and sciurid rodents such as squirrels have UV-opaque lenses, whereas hedgehogs, murine rodents such as mice and some others have UV transparent lenses and see UV perfectly well. Other mammals, such as many deer, are somewhere in the middle and can see at least some UV light.

Wood Mouse - UV-sensitive

This has very interesting implications for behaviour and ecology of mammals. Reindeer for example can see UV fairly well, and snow reflects UV. The hair of Polar Bears however, although it looks white to human eyes, actually absorbs UV, so against a bright UV background a Polar Bear looks quite dark to an animal with the appropriate vision. How various Arctic mammals actually look to UV-sensitive eyes is a topic for further investigation – I suspect that at least some species have UV-reflective fur to aid concealment.

Reindeer - partial UV-vision

Why some mammals can see UV and others do not is not at all clear. UV light scatters easily, and makes it hard to focus, so an animal reliant on accurately judging distances may find it a distraction. A nocturnal animal may need eyes that let in as much light as possible, so being able to see UV may be a side effect of this. Even so, a lot of research still needs to be done on this area, and there are considerable implications for animal welfare, especially in an indoor environment where the majority of the lighting is artificial. It is possible that as UV can damage DNA, protecting the retina from UV may be advantageous, especially in long-lived mammals. However, many birds are also long-lived and they do not appear to suffer from any special visual problems as they age.

Artificial light sources have improved considerably in recent years, with new forms of lighting such as LED lights and new forms of fluorescent lights for captive reptiles being the best known. The issue of lighting for mammals has traditionally taken a back foot, as it was assumed that mammals saw things much the same way people do. This is plainly not the case however, and needs to be taken into account if a captive mammal is to be maintained in an environment that appears (to it) as natural as possible.

There is another lighting-related issue that also needs to be considered. Florescent lights actually flicker at a fairly high frequency of around 100 Hz. Humans can only distinguish flickering up to around 60 Hz – faster than that and the light appears continuous. Poultry, and probably most birds, can distinguish up to 120 Hz, and being kept under normal fluorescent light is rather like living permanently under strobe lighting. This cannot be good for them, and probably goes some way to explaining why birds often breed better in aviaries which are naturally lit. More modern, specially designed, lighting may flicker at frequencies high enough for the flicker to be invisible to birds. What the sensitivity of reptiles is is not clear – it probably varies across various groups in any event.

Also, another issue of importance to conservation breeding occurs to me, in the possibility of mutations affecting only UV patterning and therefore not normally visible to humans. A lot of breeding planning for conservation purposes involves selection of appropriate breeding stock, and animals with obvious visible mutations are not bred from. However, the key word there is ‘obvious’. Human visual acuity and colour discrimination is so poor compared to birds and reptiles that it would be surprising if animals that have significant variations from normal appearance (at least to each other) are not at least sometimes included in breeding stock. In the captive setting this has no real affect, but if animals like this are then returned to the wild there could be serious implications for their fitness to survive.

Comments would be much appreciated.
(images from wikipedia)