Liz Weerdmeester: Electrocommunication in Gymnorhamphichtys rondoni in the wild

BCCN Berlin / TU Berlin

 

Abstract

In neuroethology, neuronal control of natural behaviors is studied in a wide range of animal models. Especially from an evolutionary point of view, quantifying behavior of an animal in its natural habitat is crucial to uncover adaptations of the nervous system to the animal's everyday life and challenges. For this master thesis, a weakly electric fish was studied which has been electrophysiologically well characterized, although little is known about the natural behavior of this animal in the wild. To bridge this knowledge gap, we went to the Amazon jungle in Leticia, Colombia, where we submerged a grid of electrodes in the Yahuarcaca terra firme stream, which is inhabited by a variety of weakly electric fish species. Data was recorded for five consecutive days, and machine learning techniques were exploited to extract electric fish activity from this complex spatiotemporal dataset. A novel tracking algorithm was developed to track the activity of one specific weakly electric fish: the sandfish Gymnorhamphichthys rondoni. Some newly obtained insights in the life of the G. rondoni include its natural resting and foraging behavior with corresponding electric organ discharges and movement patterns. Other insights include the response of the G. rondoni to encounters with the predatory electric eel, Electrophorus varii. Most importantly, although not perfected yet, the developed tracking algorithm provides a useful tool for future research into electrocommunication of various weakly electric fish species in the wild.

 

Additional Information

Master Thesis Defense

 

Organized by

Tilo Schwalger