Domestic chick colour discrimination test. Photo credit: Peter Olsson.

Domestic chick colour discrimination test. Photo credit: Peter Olsson.

Few animals put on as good a show as the birds. Strutting about in their bright plumage, no one can fail to be impressed by the rich variety of colour. And, equipped with their large eyes and advanced retinas, birds are believed to have some of the best vision on the planet too. ‘We find the highest spatial acuity and the ability to detect very fine patterns [in birds]’, says Peter Olsson from Lund University, Sweden. Given the impressive visual abilities of birds, Olsson, Olle Lind and Almut Kelber were curious to find out just how good their colour vision is. Explaining that birds have four colour-sensitive photoreceptors (cones) – compared with our three – which are equipped with light-filtering oil droplets, the team wondered how well chickens are able to discriminate between colours and how their abilities are affected as the light fails (p. 184).

Fortunately, domestic chicks are relatively easy to train – they'll do anything for a peck of seed. Having first trained 48-hour-old birds to recognise a packet of seed that was covered in orange squares, Olsson then tested whether the birds could distinguish the pack that they had been trained to recognise from an empty one that was a different shade of orange or yellow. He also tested how well the birds could distinguish green from blue/green shades. ‘The main idea was to test very different types of colours, that way we can make sure that our conclusions are more general’ says Olsson. Then he tested how well the chicks were able to distinguish between the coloured packages when he dimmed the lights from levels simulating an overcast day (250 cd m−2) to moonlight (0.01 cd m−2).

After weeks of working patiently with the chicks, the team could see that in daylight, the birds were able to distinguish all but the most closely matched of colours. However, as the light failed, the birds began to struggle to distinguish the darkest and most similar colours. And, when the team compared the chicks' performance with that of human volunteers, they were surprised to find that even though the chicks are equipped with more types of photoreceptor than we are (in addition to the rods and colour-sensitive cones that we have, birds also have double cones), their colour discrimination is no better than ours: although Olsson points out that they probably see a broader palette of colours than we do.

Puzzled by the chick's unexpectedly poor performance, the team began to investigate the factors that might restrict their ability to distinguish similar colours in dim light. ‘We believe that colour vision is noise-limited,’ explains Olsson, who goes on to add that two types of noise are likely to dominate in low light conditions: photon-shot noise, which is due to the random arrival of photons at the retina; and dark noise, where photoreceptors sometimes fire off even though they have not picked up a photon. Adding the two types of noise to an already existing computer simulation of the bird's visual system, the team was able to successfully reproduce the birds' performance in dim conditions and show that both photon-shot and dark noise limit their ability to distinguish between colours when the light is low. ‘We needed to include dark noise to explain the intensity thresholds and as far as I know that has not been found before’, says Olsson.


Bird colour vision: behavioural thresholds reveal receptor noise
J. Exp. Biol.