Development of an automated system for long-term observation of navigational behavior in D. melanogaster larvae.

Navigational behavior in an animal can be modified due to learning, and can exhibit variations between individuals to create behavioral phenotypes. To achieve high precision and reliability for either of these phenomena, observations of individual animals must be over long timescales (10-100% of a lifetime). This can be difficult to carry out manually, so we implement and demonstrate an automated approach to achieve new orders of magnitude in precision and experimental duration with both free and biased roaming Drosophila melangaster larvae. We modify a 3D printer system to automatically maintain exploratory behavior while confining the animals within the experiment area via a “pick-and-place” mechanism. Here, we demonstrate the ability of this system to quantify and observe significant changes in larval run speed, turn rate, and turn handedness over experimental durations of several hours, indicative of some previously unexamined behavioral adaptations and mechanisms in learning and memory. We will continue to further exploit its high customizability to explore other creative applications for the robotic system, such as adapting the system for experiments with C. elegans and other model organisms.