Nonlinear Processing of Sensory Interference Drives Social Behavior in Weakly Electric Fish

Weaky electric fish;electrosensory;interference;jamming;modulation;envelopes;feedback;behavioral;Mechanical Engineering

Sasidharan Madhav, ManuCowan, Noah J. ;

Johns Hopkins University

关键词: Weaky electric fish;    electrosensory;    interference;    jamming;    modulation;    envelopes;    feedback;    behavioral;    Mechanical Engineering;   

Others  :  https://jscholarship.library.jhu.edu/bitstream/handle/1774.2/37059/SASIDHARANMADHAV-DISSERTATION-2014.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: JOHNS HOPKINS DSpace Repository

【 摘 要 】

Sensory systems in animals and robots can receive information from the environment via modulations of sensory signals.Environmental features and conspecifics generate these modulations, and effective computational algorithms in sensory systems can demodulate signals to perceive the features. However, interaction of multiple modulations can themselves generate emergent modulations, which can encode relative information between environmental features. These emergent modulations create sensory interference which can be detrimental or beneficial to sensing. Organisms that probe their environment using autogenous sensory signals are prone to such interference from other nearby individuals in the same social group.Some animals have evolved behaviors that allow them to retain a high level of sensory performance in such a social context. For example, the weakly electric glass knifefish, Eigenmannia virescens, produces a pseudo-sinusoidal oscillating electric signal used for electrolocation and social communication.In my thesis, I explore how the electrosensory system of weakly electric fish use nonlinear processing to respond to interference from conspecifics.In Eigenmannia and other wave-type electric fishes, the interaction between two individuals produces first-order modulations termed beats, and interaction between three or more individuals produce ``beats of beats;;;;, termed envelopes. In this thesis, I provide an analytical basis for the emergence of beats and envelopes in sensory systems. I derive conditions where these emergent modulations arise, and show how biologically plausible nonlinear mechanisms can extract information embedded in them.Experimentally, I characterize the Jamming Avoidance Response (JAR) in Eigenmannia in terms of an input-output model.The JAR is a widely studied response to beats. The JAR is locally unstable; it is an escape response. An experimental closed-loop around the electrosensory system of the animal ;;stabilized;; the behavior, facilitating system identification analysis. The experimental input--output responses were parsed into a global nonlinear model of the JAR.I also describe our discovery of a new behavior in these fish, termed the Social Envelope Response (SER). The SER is a behavioral response to envelopes. We experimentally determined the properties of the SER for Eigenmannia.