Spatial learning data in a foraging task for Heliconius erato butterflies

This dataset contains data from experiments described in the manuscript: Moura, P. A., Cardoso, M. Z., and Montgomery, S. H. (submitted). Heliconius butterflies use landscape features, but not local landmarks, during spatial learning.In this study, we use a four-choice spatial learning experiment (one rewarded and three unrewarded feeders) and adjacent visual cues. We performed three sets of experiments: 1) Experiment 1 tests whether Heliconius can learn the position of a rewarding feeder associated with a landmark, both of which were fixed in place within a transparent-walled cage; 2) Experiment 2 tests whether Heliconius can learn an association between a rewarding feeder and a mobile landmark when both were randomly positioned together, such that within a transparent-walled cage the only reliable cue was the position of the landmark; 3) Experiment 3 tests whether Heliconius can learn the position of a rewarding feeder in an environment without surrounding landscape features. Main results of the experiments are: (1) Heliconius butterflies do not need a local landmark to learn the position of a reliable food source; (2) Heliconius butterflies did not associate a food source to an adjacent visual cue; (3) Heliconius butterflies learn positional information but do not use local landmarks even in the absence of external visual cues. Our research suggests that Heliconius butterflies do not rely on local visual landmarks to pinpoint the location of a food source but most likely rely on the coupling of general visual landmarks, or landscape cues, and odometric and compass information. We hypothesize that in the visually complex environments in which Heliconius occur, general landscape cues coupled with egocentric cues may provide a more accurate and temporally stable strategy for identifying patchily distributed pollen sources where large-scale environmental features are more stable than local vegetation.