What the dog really saw
“Reason tells me, that if numerous gradations from a simple and imperfect eye to one complex and perfect can be shown to exist, each grade being useful to its possessor, as is certainly the case;…then the difficulty of believing that a perfect and complex eye could be formed by natural selection, though insuperable by our imagination, should not be considered as subversive of the theory.” from On The Origin of Species by Charles Darwin (1809-1882)
If ever you need to be reminded that we each have our own way of looking at the world, take a dog for a walk at night.
At night it is clear that dogs experience life in a way that is beyond the reach of our faculties. It’s not merely their astonishing sense of smell. Nor their remarkable hearing. It is their sight.
At night the human eye is challenged. We evolved as a diurnal species, active during daytime hours and largely inactive at night. As diurnal hunter-gatherers, our survival was contingent on seeing details in the light of day – both those near at hand – like the objects and foodstuffs we manipulated – and those at a distance – like the prey we pursued and the landscapes we navigated. Our eyes are well suited for these purposes in daylight. We see objects within a “useful” distance with great clarity – objects held at arms reach have extremely well-defined edges, and those at a running distance away can be discriminated from each other.
What’s more, we have the capacity to discern a variety of colours from each other. Humans, and many other primates, are unique amongst placental mammals in that we have trichromatic vision. Trichromatic vision is characterised by the ability to see colours based on three types of colour-detecting cells. These colour detecting cells, cones, are found in the retina at the back of the eye. The three types of cone cells are distinguished by the wavelengths of light they detect. Our retinas have a region that is enriched in these three types of cone cells to detect colour, the fovea. Human retinas also have another type of cells, known as rods, which function to perceive light. They enable us to determine shade, to make our objects our or periphery, and to make judgements about dimension and distance. The human eye has around 120 million rod cells, and only 6 to 7 million cone cells, but these cone cells are concentrated within the fovea so as to comprise almost 100% of the cells there – the centre for colour detection.
Working together collectively, the three cone cell types together with the rods, literally provide colour and shape to our world. In daylight, they allow us to see red fruit in a background of green leaves, and to determine how far they are away from us, the depth they occupy relative to each other. Many of our fellow mammals would not have as nuanced picture of the world as we do.
Our eyes are fantastic devices for our daytime existence.
At night, they don’t fare so well.
Our small pupils let in very little light, so that at night we loose our capacity to see not only colour, but also to work out the details of objects – both near and far. For us, night literally becomes the land of nightmares. At night, our vision doesn’t allow us to see what predator might be lurking, what uneven ground may topple us, or what objects are safe or dangerous. Night is a dangerous time as seen with human eyes.
Not so for dogs.
Dogs are thought to have their origins as crepuscular species. That is, they are at their most active at dusk and dawn – ideal for an opportunistic omnivore residing in higher latitudes – where dusk and dawn make up a significant proportion of the day. Of course, dusk and dawn are also times when both nocturnal and diurnal prey are vulnerable. It was against this backdrop that dog senses evolved.
An acute sense of smell is like a dog’s long-distance vision – it picks up cues from far away, providing an odour-shaped picture of the landscape. Navigation over nearer distances requires vision honed for dusk and dawn.
In order to navigate the low-light environment at dusk and dawn, the eye must capture as much light as possible. The canine eye is exquisitely suited for that purpose. The canine eye has a very large pupil – the region of the eye that functions as a window to let the light in. In fact, in some dogs, it looks like the entire eye is pupil – with the coloured iris surrounding it a mere ring at the outer edge of the eye.
Light is transmitted to the retina by the eyes’ lenses. The lens system has two components – the cornea and the crystalline lens. The cornea is that transparent part of the eye that bulges at the front. The crystalline lens sits behind the cornea, where it serves to focus the light on the retina. Dogs have large corneas, which allows much light to be gathered and transmitted through the crystalline lens of the pupil.
The light focused by the crystalline lens makes an image on the retina. As with humans, the canine retina functions to perceive light and colour. But, relative to human eyes, dog eyes have a higher proportion of rod cells, and only two of the three types of cones cells. As such, dogs are able to perceive light more readily than humans.
Crucially, light that enters the canine eye is not merely captured by the retina. We humans can see that some of the light must be reflected, in what we call “eyeshine” – the glow of a dog’s eyes when light is cast on them at night. What is happening here?
Eyeshine is due to a special layer of cells found in the canine eye. Behind the upper half of the retina sits a special layer of cells that work like a mirror. Collectively these cells form what is known as the tapetum lucidum, a name derived from Latin for “bright tapestry”.
Woven of specialised cells, the tapetum lucidum functions to reflect light back through the retina. What’s more, the tapetum lucidum is constructed in such a way that the light reflected back through the retina has had its wavelength shifted so that it is more readily detected by rod cells. Functioning in this manner, the tapetum lucidum works to amplify the light that is perceived by the retina. The dog sees more of the light, so to speak.
Beyond their light capturing capability, dog eyes are situated in their head so that they capture a large field of view. The human field of view is approximately 180o. The human field of view is effectively the half circle defined from one outstretched arm to the other outstretched arm. Because of the position of their eyes, offset from the front of their face, slightly to each side of their face, the dog field of view extends further than a humans – to almost 240o. It is the equivalent of you being able to see over your shoulder all the time.
So, on an evening walk, dogs are able to see a much more comprehensive world, full of light and motion. Our nightmare is their wonderland.
But the dog’s night-time vision comes at a cost – and also perhaps a strange benefit.
The preponderance of rod cells in the canine human means that there are fewer cones. The equivalent to the human fovea in canine eyes is depleted in cone cells – only 10% of the cells in this region are cone cells, in comparison to the near 100% concentration in humans. What’s more, the canine retina has only two of the three types of cones cells. Taken together, this means that dogs are impaired in their fine discrimination of some colours – notably shades of green from shades of red. Dogs see more light in the world, but it is not as richly coloured.
The capacity to capture more light also means that dogs have a lower ability to focus that light to create a clear image at multiple distances. Humans are able to focus on objects nearby and at a distance. We are able to accommodate distances in our focusing – consequently, this flexibility in focusing ability is known as accommodation. Dog eyes are unable to do this to as great an extent. Their large pupils make focusing on nearby objects difficult. Where we could focus on an object held 10 cm from our noses, dogs would need to have that same object held many tens of centimetres away to focus on it.
What’s more, the tapetum lucidum is relatively non-selective when it reflects light back into the eye. The light reflected by the dogs’ tapetum lucidum is not all captured by the retina, but rather is reflected in many different directions within the eye. As a consequence, with light for the image coming from multiple angles, the image that is formed is blurry.
So, the net consequence of being able to see objects under low light is that dogs are compromised in their ability to see the fine details in the way that we are. In some ways, dogs are able to see both more and less. Their night-time world is richer than ours, but it lacks the details we would be able to see in daylight. This said, it may be that dogs’ night-time vision provides them with the capacity to see things in the daytime that we can only imagine.
Humans perceive and respond to light in the so-called “visible” wavelengths. These are the wavelengths with which we are all familiar – comprising the colours of the rainbow – red, orange, yellow, green, blue, indigo and violet – and, of course, all the shades in between. Red light has the longest wavelength of visible light, while violet has the shortest. There are wavelengths of light that are shorter than violet, in the 280-400nm range, the ultraviolet (UV) wavelengths. We are unable to see these wavelengths. What’s more, these short wavelengths are damaging to human eyes. Prolonged exposure of a human retina to UV light will destroy the cells, bringing on blindness. Consequently, our eyes have evolved to block out as much UV as possible.
Humans, and our primate relatives, allow less than 5% of UV light that strikes the eye to be transmitted the retina. The rest is absorbed by other parts of our eye, notably our lenses, so as to protect the retina. We share this in common with many other diurnal species that have high visual acuity, including other primates, and also cattle, horses, and grey squirrels. These species share in common a high density of cone calls – and with them, the capacity to see objects in detail.
Dogs, on the other hand, are not as effective at filtering out this ultraviolet light. More than 60% of UV light can be transmitted through the canine eye to the retina. Dogs share this in common with many crepuscular and nocturnal creatures, like cats, mice, flying squirrels, and hedgehogs. All of these species have a high concentration of rod cells, and the capacity to see under dim light.
A simple hypothesis to explain the high UV light transmission in the eyes of crepuscular and nocturnal creatures is that they don’t need to worry about the high levels of UV from midday day light. Consequently, the retinas of these animals needn’t be protected from UV light. This is a reasonable hypothesis for nocturnal animals, but is less convincing for crepuscular animals. Crepuscular animals are frequently opportunistic, and will pursue prey and travel during daylight hours. Why risk the damaging effects of UV light? Why not evolve the capacity to protect the retina that diurnal species possess? Could there be a value to allowing the UV light to reach the retina?
There are many examples of animals that actually make use of cues derived from the reflection and absorption of UV light. Insects and some birds are well known to derive cues from UV light transmission. Flowers have evolved to make specific UV light absorbing compounds in specific patterns so as to attract pollinators. In fact, in some plant species, the pattern in which the UV-absorbing compounds is laid down in flower petals induces strong pollinator preferences, and behaviours that encourage cross-pollination.
Might it be that crepuscular animals allow UV light to pass through to the retina are similarly using UV light as useful cue? Clearly, dogs are not pollinators that derive a benefit from identifying UV-light related patterns in flower petals. However, there may be other cues that dogs could derive from UV-light mediated signals. For example, UV absorbing or fluorescing compounds in the urine of prey could be useful for tracking. Alternatively, for an omnivore unable to clearly discriminate between red and green, there may be advantages in identifying edible plant matter, like fruit or young shoots, from other foliage on the basis of UV signals. An animal that is not trichromatic might actually derive considerable benefit from seeing UV wavelengths.
Of course, as Ed Yong recently pointed out, the big problem with the UV vision hypothesis is that transmission of UV light to the retina doesn’t mean that the light is perceived to induce a behaviour. So, the hypothesis must remain a hypothesis, until it is tested. Testing this particular hypothesis mightn’t be difficult. Dogs have already demonstrated incredible ability for visual discrimination. It might be possible to use the same methods to determine if differential responses were invoked in a discriminatory problem involving a UV light absorption or fluorescence. Until that time, we are left merely hypothesising that the dog’s view of the world includes UV light.
What we do know is that the way in which dogs see the world is very different from ours. Despite our shared history together we have remarkably different ways of seeing the world. This applies not only to the differences in perception between dogs and humans, but more generally to our fellow creatures on this planet.
We have constructed a world, and a worldview, that is founded on human vision – the way we see and interpret the world. Frequently, we are baffled when those we share the planet with are confused or don’t respond “appropriately” to our vision. But we need to remember that our vision is our vision. It may not be the way others see the world. Despite our shared existence on this planet, and despite sometimes striking similarities between us, we each have a different way of envisioning the world. We have our own lenses, our own field of view, our own way of integrating that information that makes sense in our lives. The walk in the dark that is a nightmare world for some is a world of incredible discovery and reflection for others. It is the same world, just seen with different eyes.
Images: All photographs by Malcolm M. Campbell. Special thanks to Renabe, Zooey, Brindle & Kelso for modelling their eyes.
Douglas RH & Jeffery G (2014) The spectral transmission of ocular media suggests ultraviolet sensitivity is widespread among mammals. Proceedings of the Royal Society B: Biological Sciences 281: 20132995
Koski MH & Ashman TL (2014) Dissecting pollinator responses to a ubiquitous ultraviolet floral pattern in the wild. Functional Ecology (early online)
Lind O, Mitkus M, Olsson P, & Kelber A (2014) Ultraviolet vision in birds: the importance of transparent eye media. Proceedings of the Royal Society B: Biological Sciences 281: 20132209