Krishnamachari Prahalad¹ (), Martina Poletti¹; ¹University of Rochester

Microsaccades frequently occur during examination of complex foveal stimuli as they shift the preferred locus of fixation (PLF) and support fine spatial vision within the central 1°. Before a microsaccade, processing resources are allocated toward its target, improving discrimination at the upcoming landing site. Here we investigated whether this preparatory state alters object recognition at the start-site by testing whether crowding strength at the PLF changes immediately before microsaccade onset. Eye movements were recorded with a high-resolution Dual-Purkinje Image (dDPI) eye tracker, providing arcminute-level precision necessary to localize gaze and deliver retinally contingent stimuli. Ten observers reported the identity (4AFC) of a central Pelli-font digit presented either alone or with horizontal flankers while maintaining fixation or preparing a cue-directed microsaccade toward a 20’ off-center target. A simultaneous orientation-discrimination task at the microsaccade target location verified pre-microsaccadic allocation of processing resources. Crowding was robust in both conditions, but significantly stronger prior to microsaccade onset. Threshold elevation (flanked minus unflanked) increased from 0.06 ± 0.05 logMAR to 0.11 ± 0.03 logMAR in the pre-microsaccade interval (paired t-test: t(9) = –2.79, p = 0.021). Critical spacing also increased (center-to-center: t(9) = –2.59, p = 0.029), indicating an expansion of foveal crowding zones. These effects occurred despite no change in unflanked acuity (t(9) = 0.37, p = 0.72) and no difference in mislocalization rates (t(8) = 0.54, p = 0.60). These results suggest that microsaccade preparation, rather than simply increasing internal noise or positional uncertainty, selectively enlarged spatial integration zones at the PLF. This work reveals the cost of microsaccade preparation at the start-site; while discrimination at the microsaccade goal improves, signal pooling at the current fixation locus increases. This redistribution of perceptual resources highlights how foveal processing is reweighted before microsaccade onset not only at its landing site but also at the PLF.

Acknowledgements: This work was funded by NIH Grant EY029788 to MP and NIH grant EY001319 to the Center for Visual Science.