Do animal eyespots really mimic eyes?

The diversity of anti-predator adaptations in the natural world has long been an active area of research in evolutionary and behavioural biology. A common visually-obvious feature found on prey are 'eyespots', being approximately circular markings often with concentric rings and conspicuous colours. These are found on a range of animals, especially adult and larval Lepidoptera and fish. One of the most widespread functions of eyespots seems to be to intimidate or startle predators: delaying, preventing or halting an attack. However, while the fact that they can influence predators in this way is uncontroversial, the mechanism(s) behind why they are effective is debated. Traditionally, they have been assumed to work by mimicking the eyes of the predator's own enemies, and much research in this field is conducted under the implicit or explicit assumption that this theory is correct. However, eyespots might work simply by being highly salient stimuli that promote sensory overload, biases, or neophobic reactions in predators. A range of recent studies has aimed to test these alternatives. Here, we critically evaluate this work and what it tells us about the mechanisms underlying eyespot function. We conclude that although eye mimicry is plausible, there remains a lack of evidence to support it and most observations are at least equally consistent with alternative mechanisms. Finally, we also discuss how the debate can be resolved.