Investigation of Multisensory Behavioral Control in Zebrafish During Target Tracking

Abstract

Animals capture several signals of different speeds and propagation patterns from their environment, such as light, sound, and pressure, through various sensory organs and interpret these signals in their central nervous systems (CNS) to represent a sensory expression of their environments. In addition, the same source can stimulate different sensory organs simultaneously. In such cases, multisensory integration is employed to accelerate information processing, increase the accuracy of perception, and reduce noise. Consequently, animals perform their behavior in the presence of feedback signals by generating the necessary motor signals. The main objective of this thesis is to design and build a novel experimental setup and stimuli system that allows the investigation of multisensory integration in animals during their instinctive behavior. In this study, we used free-swimming (unconstrained) Danio rerio (zebrafish), which exhibits swimming behavior against the water flow called positive rheotaxis and is referred to as a model organism in the literature. Rheotaxis is an instinctive behavior to station keeping by spending minimal energy to avoid drifting in the water flow, and zebrafish use both visual and mechanosensory cue signals to perform this behavior. Our new experimental setup based on the purpose of rheotaxis is a kind of virtual reality environment in the form of a speed-controlled laminar flow tunnel for the fish. First of all, to induce instinctive rheotaxis behavior in the zebrafish, we place it in our experimental setup where we obtain laminar flow. A low-gradient regime occurs behind it when we place an obstacle in the water. The fish can perceive this regime, and it tends to stay in this regime in accordance with the low energy cost purpose of rheotaxis. After that, when we move the obstacle perpendicular to the water flow and horizontally according to the fish's perspective, the low-gradient regime also moves, and, more importantly, the zebrafish follows this regime. Zebrafish detect the low-gradient regime behind the obstacle by combining mechanosensory and visual information simultaneously for target tracking behavior. The new experimental setup and the two-actuator stimulation system we built aim to decouple the multisensory integration that the zebrafish performs for target tracking during rheotaxis and allow us to understand how the zebrafish integrate the signals. In this context, we placed a D-shaped semi-cylindrical transparent plexiglass tube in the water flow and placed a blue neon strip LED inside this tube, which can move completely independently of the transparent tube to stimulate the zebrafish. The transparent D-shaped plexiglass tube is invisible in the water due to light refraction, stimulating only the zebrafish's lateral line organ. The neon strip LED inside the tube does not induce vibration in the water as it is surrounded by the tube and only provides visual cues to the zebrafish. We can provide mechanosensory and visual stimuli to the fish both synchronously and asynchronously, thanks to our unique high-resolution two-actuator stimulation system. We evaluate the closed-loop sensory-motor control processes of zebrafish during target tracking behavior with our determined sensor conflict scenarios. In our study, we repeated target tracking experiments for N=5 zebrafish and estimated the frequency responses of multisensory integration dynamics. Using our unique experimental setup, the two-actuator stimulation system, and the protocol, we found that mechanosensory stimuli dominate over visual stimuli, and visual cues alone are not sufficient to trigger the fish's tracking behavior when contrasted with mechanosensory cues, but zebrafish can integrate mechanosensory signals with visual signals to improve target tracking performance. Our results suggest that the dynamics of sensory integration cannot be explained by simple superposition methods.