Compensatory head rolling during corrective flight steering in locusts

Abstract

1. Locusts (Locusta migratoria) flying under open-loop conditions respond to simulated course deviations (movements of an artificial horizon around the roll axis) with compensatory head movements and with steering reactions of wing muscles (Figs. 3, 4). Steering was quantified as shifts of the relative latency between spikes in the left and right M97 (first basalar muscle). For practical reasons these shifts are a more useful measure than corrective torque itself, to which they are linearly proportional over much of the range (Fig. 2). 2. Steering in M97 is elicited visually (horizon movement) and by proprioceptive input reporting head movements (neck reflexes). Compensatory head movements reduce the strength of steering because the reduction in visual information signalling deviations is only partially balanced by proprioceptive input from the neck (Fig. 4C). 3. Under closed-loop conditions, flying locusts stabilize the position of an artificial horizon against a constant bias (Figs. 5–7), the horizon oscillating slightly along the normal orientation. Head movements do not follow the horizon movements as closely as under open-loop conditions, but on average head movements are compensatory, i.e. the mean mismatch between head and horizon is less than the mean mismatch between body and horizon. 4. The horizon position is stabilized when the head is free to move, but also when the head is immobilized. In the latter case the oscillations along the straight flight path are more pronounced (Fig. 7), indicating that the reduction of steering by compensatory head movements (as seen under open-loop conditions, Fig. 4C) reduces overshoot. 5. The control and the significance of (compensatory) head movements for course control are discussed.