Post-saccadic following in the marmoset monkey as a read-out of pre-saccadic attention

Abstract

Primates move their eyes 2-3 times per second to bring objects of interest to central, high-resolution vision. For moving objects, they use a combination of rapid saccadic eye movements along with smooth following movements to track targets continuously. Each saccadic eye movement produces perceptual enhancements for the target. And for saccades that are made to moving targets, or stationary apertures that contain motion, there is a smooth post-saccadic following response (PFR) of the target motion (Kwon et al, 2019). This PFR occurs involuntarily even when the motion is task irrelevant and could provide an automatic behavioral read-out of the targets motion. However, PFR movements are small, so it is unclear how reliable they would be as a trial-by-trial read-out. Here we examined PFR in marmoset monkeys performing a foraging task that requires almost no training and which has been shown to involve pre-saccadic neural enhancements of motion selective responses in visual area MT (Coop et al., 2024). We found that PFR in marmosets is highly consistent with humans and could be used to read-out the target motion. More so, we found that the motion in non-target apertures also influenced PFR but to a lesser extent than the target. The gain of PFR was distributed equally between target and non-target apertures before the saccade, and then was rapidly enhanced only for the target motion in driving the post-saccadic following. Thus, PFR provides a behavioral measure of target enhancement relative to distracters in addition to providing a read-out of the targets motion. Significance StatementUsing a saccade foraging paradigm in marmoset monkeys we measured visual motion integration and pre-saccadic enhancement based on the smooth following eye movements made after a saccade to a motion aperture. We find that marmosets exhibit post-saccadic following behaviors akin to humans, underscoring the evolutionary continuity in visual processing across primates. This following response provided an estimate of target motion that was half as accurate as explicitly trained reports but required minimal training to achieve. It also provided an estimate of the pre-saccadic enhancement for the target relative to distracters in the visual field.