Skip to main content

Research projects

Table of Contents

In my research I apply advanced data analysis techniques to unstructured data from passive electric recorders in the field to learn about the social behavior of electric fish.

Communication and social behavior of electric fish in the wild
#

Why are some species social while others are not? How do animals communicate with each other, and what information do they convey? These questions drive my research, which lies at the intersection of ecology, behavior, and neurobiology. I am broadly interested in how these fields can be integrated to understand the development and function of social behaviors and communication in animals.

Electric fish offer a unique opportunity to study these phenomena as they offer a wide variety of social lives that can be explored through electric signals. We design and deploy electrode arrays and required analysis software to monitor electric fish in their natural habitats. This non-invasive technology captures electrolocation- and communication signals and tracks movement, providing detailed, quantitative observation of behavior in the wild—an approach that is unique to electric fish.

Using data analyses and machine learning, we process the large-scale datasets generated from field recordings, which allows us to detect subtle patterns in communication and movement. I aim to observe, compare, and understand social interactions in electric eels and other species of electric fish in their natural environments.

By combining the study of communication signals and movement patterns, I seek to uncover how and why animals interact during courtship, competition, and foraging. This way, we might start to understand how communication mediates complex social behaviors in species where little is currently known, let alone observable in the wild.

Electric fish have already been a focus during my master’s studies, which resulted in the following two projects:

Chirp detector
Electric fish produce fast frequency sweeps to communicate, called chirps. Here, I detected these chirps with a CNN-classifier on spectrograms that was trained on data I simulated.

Electric duet
Some electric fish synchronously modulate their frequencies in duets. I detected these events using a custom covariance-based event detector on frequency estimates from spectrograms.