A case for mutualistic deceptive mimicry

It has long been understood that species that are profitable for predators to attack can gain protection if they resemble unprofitable species (Batesian mimicry), and that unprofitable species may face selection to evolve a common warning signal (Müllerian mimicry). Here we suggest that there may be widespread selection for another form of protective mimicry, so far unrecognized, that can arise even among profitable prey. Specifically, when predators adopt species-specific attack strategies, then co-occurring prey species that are caught in different ways may be selected to resemble one another. This is because the mimicry may increase the chance that the predator deploys an inappropriate attack strategy, thereby increasing the probability the prey will escape. We refer to this phenomenon as “mutualistic deceptive mimicry”, since the mimicry misleads the predator yet potentially benefits all co-mimics. We show that this hypothesis is quantitatively plausible. We then provide an empirical ‘proof of concept’ demonstrating that predators can learn to attack distinct prey types in specific ways and that this behaviour readily generates selection for mimicry. Finally, we discuss how this unrecognized form of mimicry fits into an earlier classification of protective mimicry and suggest a number of potential examples in the natural world.