ABS 2023
Search
Characterizing (un)natural vocal interactions between female zebra finches
Benjamin Hoffman1, Logan James2, Jen-Yu Liu1, Marius Miron1, Sarah Woolley2. 1Earth Species Project, Berkeley, CA, United States; 2McGill University, Montreal, QC, Canada

Interactive vocal communication is important for a range of social interactions. At a basic level, interactions require one individual to respond when another vocalizes. However, in some species (e.g. humans) there are real-time adjustments to the acoustic properties of vocal signals that are shaped by the interaction. These processes remain poorly understood in the vocal interactions of animals, particularly for vocalizations produced outside the most commonly studied contexts. Here, we characterize interactive, dyadic communication between pairs of female zebra finches, a colonial species that produces numerous contingent vocalizations. We parameterize these interactions (n>1e6 calls), and find acoustic features of a bird's call that could predict aspects of her partner’s response. We test the degree to which these changes are dependent on live social interactions through comparison with the vocal behavior of birds in four playback regimes. Together, these data lend insight into the plasticity of vocal signals and importance of social feedback during vocal interactions and may provide the foundation for using machine learning (ML) systems that emulate interactions between birds.