Visually guided landing behaviour in the housefly, Musca Domestica

Abstract

Landing is a critical aspect of insect flight. The landing behaviour comprises of multiple components like deceleration, leg extension and orientation of the body towards the landing surface. We have used the housefly (Musca domestica) as a model system to study the initiation and control of the landing behaviour on plane surfaces at two different orientations: 1) vertical 2) inverted (ceiling). Nearly half of all flies performing inverted landings collided their head with the ceiling, despite extending their legs before touchdown. For the flies performing smooth inverted landings and vertical landings, the points of onset of deceleration, leg extension and body orientation were functions of both the flight velocity and distance from the landing surface. We observed greater variability in the order of initiation of the various components during vertical landings. Leg extension occurred before deceleration in a majority of the flies performing inverted landings. These flies also contacted the landing surface at greater velocities. Decelerating flies reduced their velocity as a logarithmic function of the distance from the landing surface while landing on both the surfaces. Our experiments suggest that flies utilize the rate of expansion of the landing object to initiate and control landings, and that the inverted landing behaviour can be more variable and error prone. Using a mathematical model we show that measuring the rate of expansions of either the entire landing target or a point on the landing surface, is not sufficient to initiate landings on all kinds of plane surfaces.