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 largefield 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 highlevel 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.

Apparent Motion Perception in the Praying Mantis: Psychophysics and Modelling