Free-viewing of static natural images and movies in marmoset monkeys.

Previous work (Mitchell, Reynolds, & Miller, 2014) with marmoset monkeys viewing natural images showed that, compared to larger primates, they have a restricted oculomotor range with fixations remaining largely within +/- 10 degrees. However, that previous study used images drawn from internet sources that may have been biased towards centrally placed objects of interest. It thus remains an open question whether or not the restricted range is due to motor-based constraints, i.e. inability/physical discomfort with moving eyes beyond this range, or due to sensory factors, i.e. lack of relevant/salient stimuli in the periphery driving eye movements. To test if eye movements might expand their range for salient peripheral stimuli, we allowed two head-fixed marmosets to free-view either static natural images or videos of their own marmoset colony, containing highly salient scenes of neighbors interacting. Further, they were presented at two different viewing distances (covering either +/-20 degrees or +/-30 degrees on the horizontal). The normalized density of fixations did not differ from previous work that used static natural images drawn from the internet. Further, there was also no difference between the distributions when videos and images were shown at the closer viewing distance where salient stimuli appeared more peripherally than at the farther viewing distance. This supports the conclusion that the oculomotor range is bounded by motor constraints and is robust to manipulations of stimulus salience or eccentricity. We therefore suggest that estimation of head-gaze in marmosets, even in dynamic contexts with salient social stimuli, could provide a reasonable estimate (+/- 10 degrees) of their gaze direction. This paradigm also allows a variety of extensions to study behaviors like social monitoring and gaze-following with head-fixed free viewing primates.