Meeting Abstract

S1-2.2  Wednesday, Jan. 4  Computational analysis of hummingbird flight ALTSHULER, D.L.*; SEGRE, P.S.; STRAW, A.D.; University of British Columbia; University of British Columbia; Institute for Molecular Pathology doug@zoology.ubc.ca

A central challenge to the study of maneuverability is complexity because animal movement occurs with coordination of multiple rotational and translational degrees of freedom, each with complicated time histories. One solution is to break down extended periods of motion into a hierarchy of movement at different temporal scales. For example, the decomposition of movement into simpler building blocks has been used in robotic control applications and to describe kinematic and neural control features in both humans and animals. Recent work in flying insects suggests that specific combinations of rotation and translation form discrete units of movement that are then assembled to form more complex motion trajectories. We tested the generality of this model for flying animals by examining Anna’s hummingbirds (Calype anna) during periods of solitary and paired (competitive) flight in a large arena. The three dimensional positions and orientations were determined using a custom-designed automated tracking system capable of high temporal and spatial resolution. Unlike the previous description of insect flight modes, hummingbirds varied axial and torsional velocities continuously and no discrete units of movement were detected. However, the range of motions was constrained for both pure velocity components and for combinations of torsional and axial velocities. Comparing solitary to competitive flights across multiple days further revealed that the limits on movement vary by individual, behavioral manipulation and day of measurement. This suggests that a computational framework for describing free flight maneuvers can be used to study the limits to maneuvering performance.