"Sharks and rays do contain both rod and cone photoreceptors; however rays possess two cone opsin genes whereas sharks have only one cone. Sharks therefore were found to have lost the ability to see colors."
Surely if you have both rods and one type of cone, you do still have some degree of colour vision, from the difference between the rod and cone response curves? Either the rods and cones both respond (one end of a colour spectrum), or just the rods respond (other end of the spectrum), and colours in between can be distinguished to some degree by the relative differences as the cone response progressively drops off.
> Color vision requires opponent neural mechanisms that compare the output of different spectral photoreceptor types. Inhibitory interactions between photoreceptor channels can decrease the signal-to-noise ratio and, therefore, may degrade the ability to detect subtle differences in image contrast (Kelber and Roth 2006); in dim light, a monochromatic visual system may be capable of discriminating a greater number of distinct reflectance spectra than a di- or trichromatic one, purely based on intensity differences (Vorobyev 1997). Color vision also degrades spatial acuity, especially when the spectral peaks of the photoreceptor channels are widely separated, because the signals from different cones types will reflect both spectral and intensity differences in the image (Roorda et al. 2001). For this reason, many animals rely predominantly on a single spectral channel to convey high-resolution achromatic spatial information (Lind and Kelber 2011), but of course any other spectral cone types are then taking up valuable retinal “real estate” that could otherwise be used to improve resolution in bright light (i.e., more receptors per degree of visual angle) and/or reduce noise in dim light (i.e., through neural summation). Due to absorption, reflection, and scattering by the water itself and any dissolved or suspended substances, most aquatic habitats are characterized by low visual contrast. Moreover, many sharks are active both night and day and must therefore operate under a wide range of light intensities. It is likely therefore that sharks are often operating close to the threshold of their visual capabilities where the disadvantages of color vision circuitry might be detrimental to survival.
Surely if you have both rods and one type of cone, you do still have some degree of colour vision, from the difference between the rod and cone response curves? Either the rods and cones both respond (one end of a colour spectrum), or just the rods respond (other end of the spectrum), and colours in between can be distinguished to some degree by the relative differences as the cone response progressively drops off.