Using neuronal models to capture burst-and-glide motion and leadership in fish

Area of Science:

• Computational neuroscience
• Animal behavior modeling
• Collective motion

Background:

• Existing self-propelled particle models capture large fish schools but not smaller shoals or intermittent locomotion (burst-and-glide).
• Many fish species utilize burst-and-glide swimming patterns, which are not adequately represented in current models.

Purpose of the Study:

• To develop a novel model integrating neuronal dynamics (FitzHugh-Nagumo) with fish motion to simulate social burst-and-glide behavior.
• To investigate the emergent social dynamics and leadership behaviors in simulated fish interactions.

Main Methods:

• Combined the FitzHugh-Nagumo model of neuronal dynamics with a fish motion model.
• Simulated single fish movement in a channel.
• Developed a two-fish model where visual cues influence internal states, leading to social interactions.

Main Results:

• The model accurately replicates single fish motion.
• The two-fish model exhibits rich dynamics: leader-follower behavior, periodic and chaotic leadership changes, and tit-for-tat interactions.
• Demonstrated that leadership switching can arise from deterministic interactions, not requiring external randomness.

Conclusions:

• The proposed model offers a more comprehensive approach to simulating fish social behavior, particularly burst-and-glide locomotion.
• The findings suggest that complex social dynamics in fish can emerge from deterministic neuronal and behavioral interactions.
• Provides empirically testable predictions for fish interaction studies and links locomotion to brain activity.