Abstract
Generally speaking, saccadic eye movements are both very quick, covering up to 500°/s, and very accurate, usually within an error of 1% of target eccentricity. Despite persistent pressure by growth, fatigue, and aging, saccades are believed to remain accurate throughout life by a form of motor learning termed saccadic adaptation. Robust saccadic adaptation to induced post-saccadic visual error has been demonstrated by trans-sacadically displacing targets. Somewhat recently, Wallman et al. (VSS, J. of Vis., 2001) demonstrated that when both a large and small sectioned circle were presented simultaneously and participants attended the smaller, saccadic adaptation was four times greater than when participants attended the larger. However, the observed effect in this study could possibly be accounted for by the concentration of post-saccadic visual information available to the participant while attending the smaller target. In the study, the size of the rings, 8.5° and 0.8° respectively, could allow more information for a visual error computation to be available earlier in the occipitotemporal pathway, as the small target can be contained by neurons as early as V1 (RF around 1.5°) as compared to post-V4 (RF around 4°) with the larger target. We tested this hypothesis by having participants search a real-world image for a pre-displayed cutout using a recently presented gaze-contingent paradigm to study saccadic adaptation (Garaas et al., J. Neusci. Meth., in press). In four separate conditions, the cutouts were either large (4° × 4°) or small (1° x 1°) and were chosen from either the first or fourth quartile of possible cutouts ranked by contrast and spatial frequency. Results from the task provide evidence that saccadic adaptation is primarily influenced by the concentration of visual information available at the attended location, as participants show a nearly two fold increase in the amount of adaptation while searching for the high-information targets.