Two-Dimensional Perisaccadic Visual Mislocalization in Rhesus Macaque Monkeys

Perceptual localization of brief, high-contrast perisaccadic visual probes is grossly erroneous. While this phenomenon has been extensively studied in humans, more needs to be learned about its underlying neural mechanisms. This ideally requires running similar behavioral paradigms in animals. However, during neurophysiology, neurons encountered in the relevant sensory and sensory-motor brain areas for visual mislocalization can have arbitrary, noncardinal response field locations. This necessitates using mislocalization paradigms that can work with any saccade direction. Here, we first established such a paradigm in three male rhesus macaque monkeys. In every trial, the monkeys generated a saccade toward an eccentric target. Once a saccade onset was detected, we presented a brief flash at one of three possible locations ahead of the saccade target location. After an experimentally imposed delay, we removed the saccade target, and the monkeys were then required to generate a memory-guided saccade toward the remembered flash location. All three monkeys readily learned the task, and, like humans, they all showed strong backward mislocalization toward the saccade target, which recovered for later flashes from the saccade time. Importantly, we then replicated a known property of human perisaccadic mislocalization, as revealed by two-dimensional mislocalization paradigms: that mislocalization is the strongest for upward saccades. For horizontal saccades, we additionally found stronger mislocalization for upper visual field flashes, again consistent with humans. Our results establish a robust two-dimensional mislocalization paradigm in monkeys, and they pave the way for exploring the neural mechanisms underlying the dependence of perisaccadic mislocalization strength on saccade direction.