Mimicry in coral reef fish: how accurate is this deception in terms of color and luminance?

Batesian and aggressive mimics are considered to be under selective pressure to resemble their models, whereas signal receivers are under selection to discriminate between mimics and models. However, the perceptual ability of signal receivers to discriminate between mimics and models is rarely studied. Here we examined 15 model–mimic coral reef fish pairs using nonsubjective methods to judge the accuracy of mimics in terms of color and luminance. We then investigated the potential ability of fish with various visual systems to discriminate between model and mimic colors using theoretical vision models. We found the majority of mimics closely resembled models in terms of color and luminance from a nonsubjective perspective. However, fish that have potentially trichromatic (3 distinct cone photoreceptors) visual systems with ultraviolet sensitivity had a much better capacity to discriminate between models and mimics compared with fish with midrange sensitivity or dichromatic (2 cone photoreceptors) fish. The spectral reflectance of color patches reflected by models and mimics became more similar with an increase in depth, indicating that signal receivers may be more likely to distinguish mimics from models in habitats located closer to the surface. There was no such change in luminance contrast with depth. The selection pressure on mimics to accurately resemble their model is therefore predicted to vary depending on the visual system of the signal receiver and the light environment.