Ayumi Mizuno , Malgorzata Lagisz, Pietro Pollo, Yefeng Yang, Masayo Soma, Shinichi Nakagawa

Eyespot patterns have evolved in many prey species. These patterns were traditionally explained by the eye mimicry hypothesis, which proposes that eyespots resembling vertebrate eyes function as predator avoidance. However, it is possible that eyespots are not the mimicry of eyes: according to the conspicuousness hypothesis, eyespots are just one form of vivid aposematic signals where only conspicuousness matters. To test these hypotheses and explore factors influencing predators’ responses, we conducted a meta-analysis with 33 empirical papers focusing on bird responses to lepidopterans having conspicuous patterns (eyespots and non-eyespots). Supporting the latter hypothesis, the results showed no clear difference in predator avoidance efficacy between eyespots and non-eyespots. When comparing geometric pattern characteristics, bigger pattern sizes and smaller numbers of patterns were more effective in preventing avian predation. This finding indicates that paired concentric patterns have weaker deterring effects than single ones. Taken together, our study supports the conspicuousness hypothesis more than the eye mimicry hypothesis. Due to the number and species coverage of published studies so far, the generalisability of our conclusion may be limited. The findings highlight that pattern conspicuousness is key to eliciting avian avoidance responses, shedding a different light on this classic example of signal evolution.