Spatial memory and learning in Heliconius butterflies
Heliconius; Cognition; Site fidelity; Spatial learning; Landmark learning; Social learning
Animals that encounter a certain degree of variation throughout their lives should benefit from the ability to learn and memorize where and when to find food. Although it is clear that learning and memory are important for navigation, habitat exploration, and memorization of landmarks in many species of insects, these behaviours and their neural basis are largely well studied in social insects. Neotropical butterflies of the genus Heliconius (Heliconiinae, Nymphalidae) are emerging system for the study of learning, memory, and adaptive brain elaboration. Many of their unique ecological and behavioural characteristics (pollen feeding, limited home range, trap-lining behaviour, nocturnal gregarious roosting sites) suggest that spatial memory in particular may be essential for their survival and reproduction, particularly linked to the recognition of long-distance visual stimuli, as landmark-learning is thought to play a central role in navigation during foraging. In this context, this doctoral research aims to answer the central question “Is there evidence spatial memory in Heliconius butterflies?” Therefore, in chapter one, we use a mark-release-recapture experiment to test whether individuals of two Heliconius species exhibit true site fidelity. We further test this fidelity by measuring flight orientation during a translocation experiment, and by recapturing translocated butterflies to identify whether individuals return to their site of origin. We found that 214 non-translocated butterflies display extreme stability in site choice, and 144 translocated butterflies consistently return to their site of origin, rapidly orientating towards their home site upon release. This suggests site fidelity in Heliconius is not solely explained by low dispersal, but is a response to the distribution and stability in ecological resources. In chapter two, we found experimental evidence for spatial learning in a foraging context at different spatial scales (9m 2 e 100m 2 ). In chapter three, we used a set of three experiments to test for evidence of spatial learning associated with the presence of landmarks. We tested whether they learned to find food in an experiment where (1) the location of both the reward and the landmark is fixed; (2) the location of both changes throughout the experiment; and (3) the location of both is fixed, but in an experimental cage with covered walls to avoid interference from external visual cues. We found that butterflies learn the location of the reward when it is predictable (fixed), regardless of the presence of landmarks, and that learning appears to be aided by the use of external visual cues. Finally, in chapter four, we provided new evidence contrary to the long-standing hypotheses that Heliconius butterflies may use social information to learn the location of new resources. We found that experimental butterflies from two groups with demonstrator butterflies with different colour preferences presented a similar learning rate, demonstrating that learning of a foraging task is not facilitated in a social context. This supports the contention that foraging decisions in Heliconius butterflies are influenced by innate biases and individual experience, rather than social information from conspecifics.