Get access

Superior colliculus inactivation alters the relationship between covert visual attention and microsaccades


  • Ziad M. Hafed,

    Corresponding author
    1. Animal Physiology, Institute of Neurobiology, University of Tuebingen, Tuebingen, Germany
    • Werner Reichardt Centre for Integrative Neuroscience, Tuebingen, Germany
    Search for more papers by this author
  • Lee P. Lovejoy,

    1. Department of Psychiatry, Columbia University and New York State Psychiatric Institute, New York, NY, USA
    Search for more papers by this author
  • Richard J. Krauzlis

    1. Systems Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA, USA
    2. Laboratory of Sensorimotor Research, National Eye Institute, Bethesda, MD, USA
    Search for more papers by this author

Correspondence: Ziad M. Hafed, as above.



Microsaccades are tiny saccades that occur during gaze fixation. Whereas these movements have traditionally been viewed as random, it was recently discovered that microsaccade directions can be significantly biased by covertly attended visual stimuli. The detailed mechanisms mediating such a bias are neither known nor immediately obvious, especially because the amplitudes of the movements influenced by attentional cueing could be up to two orders of magnitude smaller than the eccentricity of the attended location. Here, we tested whether activity in the peripheral superior colliculus (SC) is necessary for this correlation between attentional cueing and microsaccades. We reversibly and focally inactivated SC neurons representing peripheral regions of visual space while rhesus monkeys performed a demanding covert visual attention task. The normal bias of microsaccade directions observed in each monkey before SC inactivation was eliminated when a cue was placed in the visual region affected by the inactivation; microsaccades were, instead, biased away from the affected visual space. When the cue was placed at another location unaffected by SC inactivation, the baseline cue-induced bias of microsaccade directions remained mostly intact, because the cue was in unaffected visual space, and any remaining changes were again explained by a repulsion of microsaccades away from the inactivated region. Our results indicate that peripheral SC activity is required for the link between microsaccades and the cueing of covert visual attention, and that it could do so by altering the probability of triggering microsaccades without necessarily affecting the motor generation of these movements.