Live fish learn to anticipate the movement of a fish-like robot ^*

Abstract

The ability of an individual to predict the outcome of actions of others and to change own behavior adaptively is called anticipation. There are many examples from mammalian species - including humans - that show anticipatory abilities in a social context, however, it is not clear to what extent fishes can anticipate the actions of their interaction partners and what are the underlying mechanisms for that anticipation. To answer these questions, we let live guppies (Poecilia reticulata) interact repeatedly with an open-loop (non-interactive) biomimetic robot that has been previously shown to be an accepted conspecific. The robot performed always the same zigzag trajectory in the experimental tank that ended in one of the corners, which gave the live fish the possibility to learn both the location of the final destination as well as the specific turning movementof the robot over three consecutive trials. The live fish’s reactions were categorized into a global anticipation which we defined as relative time to reach the robot’s final corner and a local anticipation which was the relative time and location of the live fish’s turns relative to robofish turns. As a proxy for global anticipation, we found that live fish in the last trial reached the robot’s destination corner significantly earlier than the robot. Overall, more than 50% of all fish arrived at the destination before the robot. This is more than a random walk model would predict and significantly more as compared to all other equidistant, yet unvisited corners. As a proxy for local anticipation, we found fish to change their turning behavior in response to the robot over the course of the trials. Initially the fish would turn after the robot, which was reversed in the end, as they began to turn slightly before the robot in the final trial. Our results indicate that live fish are able to anticipate predictably behaving social partners both in regard to final movement locations as well as movement dynamics. Given that fish have been found to exhibit consistent behavioral differences, anticipation in fish could have evolved as a mechanism to adapt to different social interaction partners.

Experienced social partners hinder learning performance in naive clonal fish

“…Clonal Amazon mollies can learn in an operant conditioning paradigm within a few days and a low number of repeated training sessions to associate food with a location in their laboratory environments. This is in line with current research on fish cognition, which shows that fish are avid learners and have sophisticated cognitive abilities [Brown et al, 2008, Kohda et al, 2019, Bshary and Triki, 2022, Bierbach et al, 2022]. Further, Fuss and Witte [2019] and Fuss et al [2021] found similar learning capabilities in both parental species of the Amazon molly, P. latipinna and P. mexicana, and also in the closely related guppy ( P. reticulata ).…”

Fish-inspired robotic algorithm: mimicking behaviour and communication of schooling fish

“…Aside from using particle models and observation studies, robotic fish embedded within a school of fish are being used to observe and influence the behaviors of fish [44,45]. Researchers have demonstrated that the appearance and locomotion of a robot play crucial roles in its successful assimilation within a school of biological fish [46,47].…”

Proactivity of fish and leadership of self-propelled robotic fish during interaction

“…At present, most studies choose collective parameters to analyze the interactive experimental results according to their experimental purpose. For example, the preference for robotic fish [11,20], time spent in robotic zone [14,17], probability of a collective avoidance and following [23], collective decision difference [2], the average angular distance used to characterize the different degree of individual distribution on the ring tank [9], and reaction time delay between robot and fish [22]. In this study, the collective parameters are put forward from two aspects: fish interaction proactivity and relative position relationship after interaction occurs.…”

“…Considering that fish have a certain potential for learning [22], in order to limit the effect of fish learning on robotic fish behavior, at least 30 min apart between each set of trials, and no more than ten experiments were conducted each day. At the start of each experiment, we turned off the pool lights to minimize the presence of humans in the fish's sight, and gently placed the robotic fish in the pool to avoid stimulating fish.…”

“…Chemtob et al interacted a robot with a zebrafish school and modulated the behavior of the schooling fish based on observed behaviors [21]. Bierbach et al conducted an open-loop interaction experiment between a biomimetic robotic model and guppies, showing that live fish could predict the social partner's final motion position and motion predictable behavior in terms of dynamics [22]. Papaspyros et al compared reactive and imposing decision interaction models for robotic fish, proving that the reactive model had higher leadership characteristics than open-loop models [9].…”