RT Journal Article SR Electronic T1 Drosophila learn efficient paths to a food source JF bioRxiv FD Cold Spring Harbor Laboratory SP 033969 DO 10.1101/033969 A1 Rapeechai Navawongse A1 Deepak Choudhury A1 Marlena Raczkowska A1 James Charles Stewart A1 Terrence Lim A1 Mashiur Rahman A1 Alicia Guek Geok Toh A1 Zhiping Wang A1 Adam Claridge-Chang YR 2015 UL http://biorxiv.org/content/early/2015/12/09/033969.abstract AB Elucidating the genetic, and neuronal bases for learned behavior is a central problem in neuroscience. A leading system for neurogenetic discovery is the vinegar fly Drosophila melanogaster; fly memory research has identified genes and circuits that mediate aversive and appetitive learning. However, methods to study adaptive food-seeking behavior in this animal have lagged decades behind rodent feeding analysis, largely due to the challenges presented by their small scale. There is currently no method to dynamically control flies’ access to food. In rodents, protocols that use dynamic food delivery are a central element of experimental paradigms that date back to the influential work of Skinner. This method is still commonly used in the analysis of learning, memory, addiction, feeding, and many other subjects in experimental psychology. The difficulty of microscale food delivery means this is not a technique used in fly behavior. In the present manuscript we describe a microfluidic chip integrated with machine vision and automation to dynamically control defined liquid food presentations and sensory stimuli. Strikingly, repeated presentations of food at a fixed location produced improvements in path efficiency during food approach. This indicates that improved path choice is a learned behavior; the other foraging metrics examined showed no evidence of learning. Active control of food availability using this microfluidic system is a valuable addition to the methods currently available for the analysis of learned feeding behavior in flies.