Optomotor response studies of insect vision

Abstract

Precise methods have been developed to study the optomotor responses of insects in correlation with microprobing and histology. These include stimulus patterns whose noise power is controlled to one part per million and a directly-connected digital computer system that permits the accurate reduction and analysis of responses. A range of flight and head torque reactions was studied for Musca domestica . Patterns producing gross temporal changes of sudden transients were found to elicit more complex reactions than hitherto reported. However, the steady flight torque responses to precision cylindrical patterns of various wavelengths, rotating at precisely constant velocity, provided detailed information on the optical properties of the ommatidia and the motion-detection capabilities of the eye. These studies indicate that the photoreceptors can be characterized by visual fields whose full width to 50% of effective light flux sensitivity is about 5° in the dark-adapted state. It is, however, about 40% less in the light-adapted case. Histological investigations show that this is caused by a shielding pigment migration. Correlated temporal changes in acuity can be detected in the optomotor responses with an adaptation period of about 10 min. Detection of constant velocity motion was found to involve the interaction of only a few adjacent ommatidia with total responses given by a simple saturable summation of these unit interactions. This process of parallel summation from many unit areas of the total field provides a very high signal to noise ratio for the nervous system, enabling the insect to detect patterns of quite low intensity and contrast when the wavelength is greater than 15 or 20°. Thresholds for such patterns are about three orders of magnitude below those of light response signals normally detectable by microprobing in the retinular cells. However, the relatively high ratio of ommatidia field to interommatidial spacing results in a rapid increase of threshold with reduced wavelength. The optomotor response threshold for a 3° pattern exceeds that of a 15° pattern by a factor of more than 1000. This is the minimum wavelength detectable with the precision patterns that were used. Many of the apparent reactions reported in other investigations can be attributed to inaccuracies in the stimulus patterns.