No abstract

Emotional disturbances constitute a major health issue affecting a considerable portion of the population in western countries. In this context, animal models offer a relevant tool to address the underlying biological determinants and to screen novel therapeutic strategies. While rodents have traditionally constituted the species of choice, zebrafish are now becoming a viable alternative. As zebrafish gain momentum in biomedical sciences, considerable efforts are being devoted to developing high-throughput behavioral tests. Here, we present a comparative study of zebrafish behavioral response to fear-evoking stimuli offered via three alternative methodologies. Specifically, in a binary-choice test, we exposed zebrafish to an allopatric predator Astronotus ocellatus, presented in the form of a live subject, a robotic replica, and a computer-animated image. The robot's design and operation were inspired by the morphology and tail-beat motion of its live counterpart, thereby offering a consistent three-dimensional stimulus to focal fish. The computer-animated image was also designed after the live subject to replicate its appearance. We observed that differently from computer-animated images, both the live predator and its robotic replica elicited robust avoidance response in zebrafish. In addition, in response to the robot, zebrafish exhibited increased thrashing behavior, which is considered a valid indicator of fear. Finally, inter-individual response to a robotic stimulus is more consistent than that shown in response to live stimuli and animated images, thereby increasing experimental statistical power. Our study supports the view that robotic stimuli can constitute a promising experimental tool to elicit targeted behavioral responses in zebrafish.

In this work, we explore the feasibility of regulating the collective behavior of zebrafish with a free-swimming robotic fish. The visual cues elicited by the robot are inspired by salient features of attraction in zebrafish and include enhanced coloration, aspect ratio of a fertile female, and carangiform/subcarangiform locomotion. The robot is autonomously controlled with an online multi-target tracking system and swims in circular trajectories in the presence of groups of zebrafish. We investigate the collective response of zebrafish to changes in robot speed, achieved by varying its tail-beat frequency. Our results show that the speed of the robot is a determinant of group cohesion, quantified through zebrafish nearest-neighbor distance, which increases with the speed of the robot until it reaches . We also find that the presence of the robot causes a significant decrease in the group speed, which is not accompanied by an increase in the freezing response of the subjects. Findings of this study are expected to inform the design of experimental protocols that leverage the use of robots to study the zebrafish animal model.