Spatiotemporal dynamics, neural circuitry, and environmental context of decision-making in Drosophila melanogaster

Abstract

My thesis examines the intricate dynamics of decision-making in Drosophila melanogaster across three different contexts. In Chapter 1, I investigate how locomotor decisions progress through time and space. I show that turning decisions in Y-mazes occur earlier than previously believed, that local geometry plays a critical role in shaping the course of a decision, and that future turns can be predicted by past locomotor behavior. Further, I show that this predictability is under control of both external and internal information processing, using sensory and information-processing mutants, and simulated model flies. Last, I present evidence that this novel approach of describing decision-making by its spatial and temporal dynamics is generalizable to human participants in a virtual Y-maze. In Chapter 2, I identify a key neuron pair involved in phototactic decision-making, and propose a neural circuit underlying modulable phototaxis. In Chapter 3, I explore decision-making of truly wild flies in nature using the ACORN robotic platform. l show that there exist behavioral idiosyncrasies in truly wild flies similar to lab- reared flies. Finally, I present evidence for day-to-day fluctuations and seasonal changes in higher order locomotor behavior features.