Apparent Motion Perception in the Praying Mantis: Psychophysics and Modelling

Apparent motion is the perception of a motion created by rapidly presenting still frames in which objects are displaced in space. Observers can reliably discriminate the direction of apparent motion when inter-frame object displacement is below a certain limit, Dmax. Earlier studies of motion perception in humans found that Dmax scales with spatial element size, interpreting the relationship between the two as linear, and that Dmax appears to be lower-bounded at around 15 arcmin. Here, we run corresponding experiments in the praying mantis Sphodromantis lineola to investigate how Dmax scales with element size. We used moving random chequerboard patterns of varying element and displacement step sizes to elicit the optomotor response, a postural stabilization mechanism that causes mantids to lean in the direction of large-field motion. Subsequently, we calculated Dmax as the displacement step size corresponding to a 50% probability of detecting an optomotor response in the same direction as the stimulus. Our main findings are that mantis Dmax appears to scale as a power-law of element size and that, in contrast to humans, it does not appear to be lower-bounded. We present two models to explain these observations: a simple high-level model based on motion energy in the Fourier domain and a more detailed one based on the Reichardt Detector. The models present complementary intuitive and physiologically-realistic accounts of how Dmax scales with element size in insects. Author Summary Computer monitors, smart phone screens and other forms of digital displays present a series of still images (frames) in which objects are displaced in small steps, tricking us into perceiving smooth motion. This illusion is referred to as “apparent motion”, and for it to work effectively the magnitude of each displacement step must be smaller than a certain limit, referred to as Dmax. Previous studies have investigated the relationship between this limit and object size in humans and found that larger objects can be displaced in larger steps without affecting motion perception. In this work, we investigated the same relationship in the praying mantis Sphodromantis lineola by presenting them with moving chequerboard patterns on a computer monitor. Even though motion perception in humans and insects are believed to be explained equally well by the same underlying model, we found that Dmax scales with object size differently in mantids. These results suggest that there may be qualitative differences in how mantids perceive apparent motion compared to humans.