Is temporal synchrony necessary for effective Batesian mimicry?

Batesian mimicry occurs when palatable mimics gain protection from predators by evolving a phenotypic resemblance to an aposematic model species. While common in nature, the mechanisms maintaining mimicry are not fully understood. Patterns of temporal synchrony (i.e. temporal co-occurrence) and model first occurrence have been observed in several mimicry systems, but the hypothesis that predator foraging decisions can drive the evolution of prey phenology has not been experimentally tested. Here, using phenotypically accurate butterfly replicas, we measured predation rates on the chemically defended model species and its imperfect Batesian mimic under four different phenological conditions to understand the importance of temporal synchrony and model first occurrence in mimicry complexes. We predicted that protection for mimics increases when predators learn to avoid the models' aposematic signal right before encountering the mimic, and that learned avoidance breaks down over time in the model's absence. Surprisingly, we found that asynchronous model first occurrence, even on short time scales, did not provide increased protection for mimics. Mimics were only protected under conditions of temporal synchrony, suggesting that predators rely on current information, not previously learned information, when making foraging decisions.