Developmental and neurophysiological determinants of ocular motor behavior

The results presented in this dissertation comprise two studies examining abnormalities associated with early disruption of binocular vision on oculomotor behavior and neural systems involved in producing smooth pursuit in normal primates. For study one, we reared monkeys under daily alternating monocular occlusion (AMO) conditions designed to eliminate binocular sensory fusion and induce strabismus (ocular misalignment). We found that strabismic monkeys made disconjugate saccades (i.e., unequal saccades in each eye). The amount of disconjugacy varied linearly with the orbital position of the fixating eye. If the saccadic disconjugacy was large, there was an immediate postsaccadic drift and steady-state ocular misalignment was usually different from misalignment immediately after the saccades. Our results show that AMO rearing produces oculomotor deficits similar to those observed in some strabismic humans. The source of the saccadic disconjugacy could be due to miscalibration of motor areas in the brain or due to nonlinearities in the oculomotor plant.; In study two, the role of middle superior temporal (MST) neurons in generation of smooth-pursuit was probed. We report the results from 44 smooth-pursuit cells in MST cortex of two normal monkeys during the following paradigms: (1) Double-target task—While the animal was fixating a laser spot, a second spot was oscillated (2 4 Hertz ± 1°) in the preferred and non-preferred directions. Eighty-one percent of the units were not modulated indicating that most cells did not carry a retinal slip signal. (2) Sinusoidal tracking with blink—The sinusoidally moving target spot (0.25Hz, ± 10°) was blinked off for a period of 250ms to 400ms during smooth-pursuit. Ninety percent of the units continued firing without decrement during the blink suggesting eye movement related encoding. (3) Step-ramp with blink: the target spot was extinguished for a period of 100ms during step-ramp tracking (10°/s). Neuronal response could be better fit with a model based on eye parameters (position, velocity, acceleration) rather than a model with retinal error parameters. Based on the results, we suggest that most MST cells encode a smooth-pursuit command or an efference copy of the smooth-pursuit eye movement rather than a pure retinal error signal.