Lighting conditions and retinal development in goldfish: absolute visual sensitivity.

Goldfish (Carassius auratus) were reared from hatching in constant light (340 lux), cyclic light (12 hr 320 lux, 12 hr dark) or constant dark. Absolute visual threshold was determined psychophysically in animals that still responded to visual stimuli after 1-3 years of exposure, by means of a classically conditioned respiration suppression technique wherein animals were presented with different intensities of large diffuse flashes of monochromatic light. Fish reared in constant light and fish reared in cyclic light responded reliably to stimuli above threshold, but fish reared in constant light were on average 0.58 log unit less sensitive at 532 nm, near the peak of the rod action spectrum. Two of the four fish reared in darkness did not respond to the stimuli, and thus could not be conditioned, and another fish reared in darkness responded only occasionally; threshold could not be measured in these three fish. The one fish reared in darkness that responded consistently enough to be conditioned was more than 5 log units less sensitive than normally reared fish on the first day of testing, and became progressively less sensitive over the next 2 days. Rearing under constant dark or constant light had no obvious effect on spectral sensitivity at absolute threshold. The effect of rearing in constant light on absolute threshold correlates with morphological changes in rod density, but the effect of rearing in constant darkness does not.