Faculty Sponsor: Grace McKenzie-Smith
Hailey Hutchison
Hailey is a rising junior (‘27) originally from Guilford, Connecticut. At Wesleyan, she is majoring in Physics and working in the McKenzie-Smith lab to track the motion of ants in order to explore their social and nonsocial behavior patterns and how the differences in these phenotypes relate to their rate of evolution. In her spare time, she enjoys spending time outdoors and rowing with Wesleyan women’s crew.
Abstract: Ants are eusocial insects, meaning their colonies have a social structure characterized by multigenerational colonies, cooperative care of brood items, and division of labor. Scientists have also observed that ants exhibit emergent behaviors at the colony level, making groups of ants interesting systems to study at different social densities. We are interested in the patterns of locomotion observed in social (small groups of ants) and nonsocial (individual ants) contexts. Social selection and nonsocial selection, the mechanisms responsible for influencing an animal’s fitness, bring on evolutionary change through separate genetic pathways. This has led to the hypothesis that social behaviors evolve more quickly than nonsocial behaviors. Using species in the genera Leptothorax and Temnothorax, we video arenas that have either one ant or a small group of ants. These videos are then analyzed in the supervised, deep-learning software SLEAP (Social LEAP Estimates Animal Poses). We have used a collection of videos of various species to develop a skeleton that labels parts of the ant that we have deemed important for tracking data. I then use that skeleton to label ants on a set of frames before I train the SLEAP model on these labels and test the model’s ability to predict the position of the ants. This process is repeated until the user is satisfied with the accuracy of the model’s predictions and the user can now assign identities to each ant in the video in order to track and collect the postural data. Using these inferred postures, we can extract basic locomotive information like an individual’s speed and proximity to others. With preliminary tracking data from our first trials, we see evidence that in some species there may be differences in the behavior patterns, such as trajectories around the arena and the speed distribution of individuals, among ants placed in a solo context versus those placed in a group context. However, not much is known about how these locomotive phenotypes evolve and the rate of evolution when thinking about these individual and social behaviors. Our objective is thus to explore where differences in the evolutionary process, between individual contexts and social contexts, may lie.
Poster-Summer-25-1