SNELL-ROOD, Emilie C; PAPAJ, Daniel R; University of Arizona; University of Arizona: Detecting signals against heterogeneous noise: do butterflies learn locally constant noise?

Organisms must detect and interpret signals against background noise. While much research has focused on organisms learning particular signals or signal-background combinations, little research has considered the role of learning the background environment itself. Characteristics of background environments are often unpredictable, but can be locally or seasonally constant. Thus, there may be benefits to learning to perceive signals against particular backgrounds. In a series of experiments, we investigate the role of the background environment in learning: given a temporally or spatially constant background, do individuals learn to ignore that background, and how does such learning affect performance in novel tasks or against novel background environments? Females of Battus philenor (Papilionidae: Lepidoptera) were trained to different colored oviposition models, against different background colors. When individuals were tested on a green background following training on a brown background, they performed significantly worse than control individuals, which were trained on a green background. While this result implies female swallowtails learn to ignore green backgrounds, we can not exclude alternative hypotheses such as changes in crypticity. To test whether individuals are learning to ignore green backgrounds, females are trained to a particular task on a green or brown background and then trained to a novel task on a green background. If females learn to ignore a particular background independent of the signals themselves, learning should be faster on the second task for those individuals with previous experience on that background. Our results indicate that experience with certain noise characteristics may affect an organism’s future performance in the same or novel noisy environments.