This study implements and evaluates passive interaction using an autostereoscopic display and ultrasound haptics technology, simulating Garra rufa ("doctor fish") nibbling. When the virtual doctor fish touches the user's hand, ultrasound tactile sensations are presented by spatio-temporal modulation (STM) as an ultrasound focal point orbits around the contact points. A user study evaluated parameters affecting realism, including STM frequency, the number of focal points, ultrasound amplitude, and hand moistening. Comparing several combinations of parameters revealed that representing contact with fewer representative points and setting the frequency of the STM to 10 Hz produced the most realistic experience.