Purchase this article with an account.
Paul Mitchell, Jay Edelman; Allocentric spatial information improves saccadic accuracy under task conditions that load spatial memory or limit saccade preparation time. Journal of Vision 2009;9(8):415. doi: https://doi.org/10.1167/9.8.415.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Many visual stimuli remain fixed spatially relative to the environment and/or to objects within it and the brain may exploit these regularities when localizing them. Recent work in our lab suggests the saccadic system can use distant stable landmarks and nearby shifting objects to guide saccades accurately to memorized locations (Mitchell & Edelman, SfN 2006). In this prior study saccades to environment-fixed targets were accurate regardless of landmark target distance. Saccades to targets fixed relative to a shifting object were inaccurate with only large objects. We hypothesized that using a long instructed delay (∼750 ms) and requiring only 1 target location to be memorized freed up sufficient resources for distant stable or nearby shifting references to aid targeting. Increasing task demands by decreasing motor preparation time or increasing memory load would then make performance more dependent on distance from and size of the reference. We varied task demands and measured saccade error when saccades were directed to a memorized target visible only after the saccade; target location was environment-fixed or object-fixed across a trial block. Reference objects were circles of 3 sizes (radius = 3.25°, 6.5°, 13°). Task demands were increased by 1) requiring 3 targets fixed to the environment to be memorized (3-targ) or 2) using a reactive saccade task to a target fixed relative to a shifting reference object. Saccade endpoint error was substantially higher in the 3-targ task than in the 1-targ task when no reference objects were present and moderately higher when the largest (13°) reference object was present. As reference object size decreased performance improved, approaching that of visually-guided saccades. In the reactive task we found a similar dependence of endpoint error on reference object size. These results demonstrate the saccadic system can use allocentric references to improve saccadic targeting under increased task demands.
This PDF is available to Subscribers Only