Noninvasive Realistic Stimulation/Recording of Freely Swimming Weakly Electric Fish: Movement Detection and Discharge Entropy to Infer Fish Behavior

TL;DR

This study presents a noninvasive system for long-term stimulation and recording of electric signals in freely swimming weakly electric fish, revealing insights into their behavior, movement, and stimulus discrimination capabilities.

Contribution

Developed a novel apparatus enabling real-time, noninvasive stimulation and recording of electric pulses in freely moving fish, allowing detailed behavioral analysis over several days.

Findings

Fish exhibit long exploratory behavior in new environments.

Fish can discriminate real from random stimuli based on pulse timing.

Behavioral sleep correlates with stimulus randomness.

Abstract

Weakly electric fish are unique models in Neuroscience allowing experimentalists to access, with non invasive techniques,a central nervous system generated spatio-temporal electric pattern of pulses with roles in at least two complex and not yet completely understood abilities:electrocommunication and electrolocation. We developed an apparatus to allow realistic stimulation and simultaneous recording of electric pulses in freely moving Gymnotus carapo for very long periods-several days. Voltage time series from a 3dimensional array of sensitive dipoles that detects electric field in several positions underwater were digitized and home made real-time software allowed reliable recording of pulse timestamps,independently of the fish's position,and also to infer fish movement. A stimulus fish was mimicked by a dipole electrode that reproduced the voltage time series of real conspecific pulses,but according to timestamp sequences previously recorded that could be chosen by the experimenter. Two independent variables were used to analyze fish behavior:the entropy of the recorded timestamp sequences and the movement of the fish inferred from pulse amplitude variability at each detection dipole. All fish presented very long transient exploratory behavior (about 8hours) when exposed to a new environment in the absence of stimuli. After the transient there were several intervals(5min-2hours),in which entropy vanished and no movement was observed, that could be associated with behavioral sleeping. Our experiments also revealed that fish are able to discriminate between real and random stimuli distributions by changing the timing probability of the next discharge. Moreover,most fish presented behavioral sleep periods when the artificial fish timestamp sequence was random,but no fish showed any behavioral sleep period when the artificial fish fired according to a real fish timestamp series.