| It is known that monkeys are able to spatially update saccade targets across an intervening saccade or smooth pursuit (SP) movement. Less is known about the dynamics of the underlying neural signals during this updating, and how they might depend on instruction. Recently we found that neural network models developed different types of spatial remapping strategies depending on the predictability of the intervening SP. Here, we reexamined the ability of primates to update saccade targets across SP movements, specifically testing the hypothesis that if animals are cued when and where SP will end, memory guided saccades will have lower latency. The animal learned all aspects of the spatial updating task, but made idiosyncratic direction-dependent errors in non-cued conditions. When the cue was incorporated, saccade latencies were significantly lower, (p<0.01) however, there was no change to accuracy or precision. This suggests that the mechanism for spatial updating requires that the SP be executed prior to the formation of the anticipatory updating signals for the saccade plan. |
