Shifting balances in the weighting of sensory modalities are predicted by divergence in brain morphology in incipient species of Heliconius butterflies

Integrating and weighting sensory perception across modalities is crucial to how animals adapt to their environment. Divergence in brain structure is often in sensory processing regions, suggesting that investment reflects ecological needs. Here, we use two parapatric closely related species, Heliconius erato cyrbia and Heliconius himera, to test the hypothesis that divergence in sensory brain regions affects foraging decisions. These butterflies are isolated across an ecological gradient, which is linked to differences in brain morphology, with H. e. cyrbia investing more in visual centres and H. himera investing in olfactory centres. Here, we demonstrate that these two species vary in how they associate visual and olfactory cues with positive food rewards. We found that when individuals were trained on paired olfactory and visual stimuli, then presented with these stimuli in conflict, they showed distinct behavioural responses. Heliconius himera was more ..., This data was collected by scoring observations of butterfly behaviour in common garden assays. The raw number and proportions of choices are provided in the full data set file. All statistical analysis are in the R scripts., R Software required to open and run the R scripts., # Data from: Shifting balances in the weighting of sensory modalities are predicted by divergence in brain morphology in incipient species of Heliconius butterflies

[https://doi.org/10.5061/dryad.dbrv15f0h](https://doi.org/10.5061/dryad.dbrv15f0h)

## Description of the data and file structure

All experiments were recorded using an action camera (GoPro©) during the first 30 min in which flowers were presented. The colour assay was recorded every day, while the odour and conflict assays were recorded on the first, fourth and fifth day due to the reduction of work during the early stages of the Covid-19 pandemic. Using the video footage, the total number of feeding attempts (defined by proboscis extensions) were counted. The data are a cumulative result of all individuals divided in groups (individuals which were tested together). To calculate differences between species, we used linear mixed-effect models (‘glmer’) implemented with the ‘lme4’ package (Bates et al., 2015) using a Poisso...