Date of Graduation


Document Type


Degree Type



Eberly College of Arts and Sciences



Committee Chair

Gary Marsat

Committee Member

Clifton Bishop

Committee Member

Sadie Bergeron


Sensory systems are often uniquely tailored to encode behaviorally relevant signals and comparative studies across species can thus reveal how evolutionary changes shape sensory functions. The structure of communication signals varies widely between ghost knifefish species. Recent findings suggest that the nervous system co-adapted to the various signal structures observed across species to support different sensory behaviors. The aim for this thesis was to compare the sensory behavior of 3 species of ghost knifefish to contrast their behavioral performance with the known differences in neurophysiology. We hypothesize that for the different signal types and species, the ability to discriminate small variations in chirp properties will match the neural encoding method and accuracy employed in the sensory system. We used a habituation-dishabituation paradigm to reveal the ability to discriminate chirp variants. We found that in two species discrimination ability depended on the frequency of the background beat signal and thus on context. For low-frequency beat typical of same-sex interactions, A. albifrons discriminated chirp variants but not A. leptorhynchus but the contrary was observed for high frequency beats characteristic of male-female interactions. A third species, A. devenanzii accurately discriminated all chirp signal independently of the frequency of the background beat. These differences match the differences in sensory encoding previously characterized. We argue that these results thus establish a clear link between signal structure, neural coding strategy and perceptual tasks. These new findings provide an important step forward in our understanding of the co-evolution sender and receiver mechanisms.