Most bony fish have remarkably good colour vision. Not only can they see many of the colours that we take for granted, but their visual spectrum extends into the ultraviolet and they even see the effects of polarisation. But other species may not be so lucky. Sarah Van-Eyk from The University of Queensland, Australia, explains that little is known about the visual system of cartilaginous fish. Even though giant shovelnose rays were recently discovered by Nathan Hart to have three types of cones – the cells in the retina that detect colour – she adds, ‘You need to do a behavioural experiment to confirm that they use colour vision and the logistics of doing behavioural studies on rays is difficult.’ However, Ulrike Siebeck had recently successfully trained damselfish to distinguish between different colours, so maybe Van-Eyk could do the same with shovelnose rays to find out whether they also have colour vision (p. 4186).

Walking out onto the Moreton Bay sand banks with a seine net, Van-Eyk and her colleagues successfully caught two juvenile rays and transported them back to the lab. Then the hard work began. ‘We had to learn to train the rays. The training is important to be sure the animals understand the task before you start the experiment’, says Van-Eyk. She explains that there could be two reasons why a ray cannot distinguish between two colours: either it cannot see the difference and so does not have colour vision, or it does not understand the task. Van-Eyk had to be sure that the rays knew exactly what they were being asked to do before she could test their vision.

First, she trained the rays to eat from a feeder dispensing tasty fish-mash, and once the animals were comfortable with the feeder she positioned a bright blue laminated sheet of paper next to the feeder, so that the ray had to touch the blue object before the feeder would dispense a snack. Next, Van-Eyk took away the feeder and only gave the ray a food reward when it touched the blue object. Finally, Van-Eyk trained the ray to distinguish between the blue object and another that was grey, rewarding the ray when it correctly touched the blue object.

After 3 months of painstaking training, Van-Eyk was convinced that the rays understood the question and would tap the paper when they saw a colour that they had been trained to recognise. However, she had not proved that the ray had colour vision; they could have been using some other visual cue, such as brightness or contrast, to distinguish the blue object from the grey. To be sure that the rays see in colour, Van-Eyk had to set the animal a really hard puzzle: could the ray pick out the blue object from three grey objects, one of which was equally as bright as the blue object and would be indistinguishable from blue if the animal lacked colour vision?

Teaming up with Connor Champ and Justin Marshall, Van-Eyk designed grey colours that were as bright as, dimmer than or brighter than the blue object, and then offered the ray a choice between the four to see whether the ray could still pick out the blue one. Remarkably, it did: the ray successfully tapped the blue paper. So, shovelnose rays do use their three cones for colour vision and Van-Eyk is keen to find out if the ray's close cousins, bamboo sharks, also have colour vision.

S. M.
U. E.
C. M.
N. S.
Behavioural evidence for colour vision in an elasmobranch
J. Exp. Biol.