'Seen in part, feared in full': conditional visibility enhances eyespot defences without experience-dependent decay

Predation is a significant ecological pressure that shapes species evolution and community dynamics. Eyespots, as common anti-predator signals, are widely observed in various animals, particularly among lepidopteran species. These markings function by either mimicking the eyes of natural predators or acting as conspicuous warning signals to deter potential threats. However, eyespot effectiveness depends on environmental context (e.g. microhabitat structure) and predator experience. Here, we examined how partial eyespot concealment and short-term predator experience influence early-stage predation behaviour using artificial prey and domestic chicks. We used artificial prey and manipulated eyespot visibility across microhabitat conditions. Our results show that partially occluded eyespots consistently delayed predation, with chicks taking longer to approach and exhibiting prolonged hesitation compared with prey with fully visible or absent eyespots. Short-term experience did not alter these behavioural responses, indicating that immediate learning plays a limited role in modulating reactions to salient deceptive signals. Together, these findings demonstrate that the concealment of the eyespots themselves, rather than habitat complexity alone, which produces a measurable deterrent effect, likely through a combination of neophobia towards conspicuous stimuli and eye-mimicry processes. Our study highlights how microhabitat-mediated visibility shapes the functioning of visual defences and offers new insight into the ecological and perceptual factors influencing the evolution of eyespots in avian predator-prey interactions.