Perisaccadic perceptual mislocalization strength depends on the visual appearance of saccade targets

We normally perceive a stable visual environment despite repetitive eye movements. To achieve such stability, visual processing integrates information across saccades, and laboratory hallmarks of such integration are robustly observed by presenting brief perimovement visual probes. In one classic phenomenon, perceived probe locations are grossly erroneous. This phenomenon is believed to depend, at least in part, on corollary discharge associated with saccade-related neuronal movement commands. However, we recently found that superior colliculus motor bursts, a known source of corollary discharge, can be different for different image appearances of the saccade target. Therefore, here we investigated whether perisaccadic perceptual mislocalization also depends on saccade-target appearance. We asked human participants to generate saccades to either low (0.5 cycles/deg) or high (5 cycles/deg) spatial frequency gratings. We always placed a high contrast target spot at grating center, to ensure matched saccades across image types. We presented brief perisaccadic probes, which were high in contrast to avoid saccadic suppression, and the subjects pointed (via mouse cursor) at their perceived locations. We observed stronger perisaccadic mislocalization for low spatial frequency saccade targets, and for upper visual field probe locations. This was despite matched saccade metrics and kinematics across conditions, and it was also despite matched probe visibility for the different saccade target images (low versus high spatial frequency gratings). To the extent that perisaccadic perceptual mislocalization depends on corollary discharge, our results suggest that such discharge might relay more than just spatial saccade vectors to the visual system; saccade-target visual features can also be transmitted.