Individual leg tracking reveals the basic building blocks of behavior. James S. Kain¹, Chris Stokes¹, Quentin Gaudry², Xiangzhi Song^1,3, James Foley¹, Rachel Wilson², Benjamin de Bivort^1,4. 1) The Rowland Institute at Harvard, Cambridge, MA; 2) Department of Neurobiology, Harvard Medical School, Boston, MA; 3) College of Chemistry & Chemical Engineering, Central South University, Changsha, China; 4) Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA.

   All animals have evolved unique sets of behavior, but how did these behaviors evolve? To examine this question we engineered a high-resolution technique to track, for the first time, all six legs of a fruit fly behaving spontaneously while tethered to a floating ball. We developed custom software to automatically identify and categorize all the behaviors of an individual walking fly. Our sensitive methods easily identified walking, turning and postural adjustments, and can discriminate several distinct types of grooming. With this setup we are exploring several important aspects of basic locomotor behavior: 1) Recording the behavioral fingerprints of flies revealed that individuals possess unique behavioral personalities that persist between trials. 2) We can probe circuit-level properties of locomotor behavior since all of our leg tracking was designed to be compatible with electrophysiology, optophysiology, two-photon and fluorescent microscopy. Additionally, we can present visual stimuli to the fly to study more complex locomotor behavior (e.g. phototaxis or optomotor responses). Using moving bar displays, we found flies exhibited less variation in their behavior under closed-loop conditions relative to open-loop. 3) Finally, we developed methods to uncover the basic behavioral building blocks. We suspect these motifs or atoms of behavior are the fundamental units that the animal strings together to generate complex behaviors such as phototaxis or a head groom. We found that there were approximately 20 motifs underlying locomotor behavior. Equipped with these techniques, we can now ask whether differences in behavior across age, ecological conditions, sex, and species have corresponding changes in the types of motifs or in the way the motifs are arranged? This knowledge will be vital to understanding the evolution of behavior.