Project Title
High-frequency accelerometer monitoring of foraging and movement behavior in a secretive predator (Central Rat Snakes, Pantherophis allegheniensis)
Faculty Mentor(s) Name(s)
Dr. Dominic DeSantis
Abstract
Accelerometers (ACTs) are becoming increasingly common in studies of animal behavior in nature. ACTs are small (<1 g) piezo-electric (spring-like) sensors capable of measuring three-dimensional acceleration derived from subject-motion, and advanced machine learning techniques enable automated classification of distinct behavioral states from these data. The goal of this project is to expand upon a recently validated framework for ACT monitoring of behavior in snakes, which are traditionally challenging study subjects for ethological field studies given their highly cryptic and secretive habits. We are conducting captive validation trials to enable development and application of a machine learning model for accurate classification of key behaviors in Rat Snakes (Pantherophis alleghaniensis). Targeted behaviors for classification include full-body movement, immobility, predatory strikes, constriction, and ingestion. All classification model procedures will be conducted with open-source software, demonstrating the increasing accessibility of ACT studies. Following captive validation, this method will be applied to wild-ranging P. alleghaniensis to remotely quantify continuous activity patterns and predatory behaviors in nature. We envision validation of this technique carrying significant conservation and management implications. Real-time monitoring of foraging efficiency in the field facilitates improved interpretation of the causes of variation in individual behavior and performance, and its effects on population trajectories.
High-frequency accelerometer monitoring of foraging and movement behavior in a secretive predator (Central Rat Snakes, Pantherophis allegheniensis)
Accelerometers (ACTs) are becoming increasingly common in studies of animal behavior in nature. ACTs are small (<1 >g) piezo-electric (spring-like) sensors capable of measuring three-dimensional acceleration derived from subject-motion, and advanced machine learning techniques enable automated classification of distinct behavioral states from these data. The goal of this project is to expand upon a recently validated framework for ACT monitoring of behavior in snakes, which are traditionally challenging study subjects for ethological field studies given their highly cryptic and secretive habits. We are conducting captive validation trials to enable development and application of a machine learning model for accurate classification of key behaviors in Rat Snakes (Pantherophis alleghaniensis). Targeted behaviors for classification include full-body movement, immobility, predatory strikes, constriction, and ingestion. All classification model procedures will be conducted with open-source software, demonstrating the increasing accessibility of ACT studies. Following captive validation, this method will be applied to wild-ranging P. alleghaniensis to remotely quantify continuous activity patterns and predatory behaviors in nature. We envision validation of this technique carrying significant conservation and management implications. Real-time monitoring of foraging efficiency in the field facilitates improved interpretation of the causes of variation in individual behavior and performance, and its effects on population trajectories.