Targetting and optomotor space in the leaf-hopper Empoasca vitis (Gothe) (Hemiptera: Cicadellidae)

Abstract

The accuracy with which the visuomotor system of Empoasca vitis can prescribe a jump trajectory towards a bright, uniformly illuminated target shape was investigated. The targets consisted of discs, rings or bars cut out of black card and mounted in front of a source of stroboscopic light diffused through translucent paper. The targets varied in size, subtending angles of 7-110 ° to the insect's eye. Target resolution was measured independently along the horizontal and vertical axes and was defined in terms of the standard deviation of the horizontal (s.d.x) or vertical (s.d.y) angular coordinates of the take-off paths measured with respect to the geometric centre of the target. In all cases involving single connected images, the point towards which jumping was directed coincided with the geometric centre of the image area (the light centroid). This meant, for example, that the insects specifically targetted the dark core of a ring of light. No formal features of the target shapes, such as light-dark boundaries, were targetted. Vertical axial resolution was independent of image size (mean s.d.y for all image sizes 6.9 °), but horizontal axial resolution only achieved this minimal value at intermediate image sizes (approximately 30-60 °). When two separate but identical target shapes (vertical bars) were presented at different horizontal separations, the dark area between them was accurately targetted up to target separations of approximately 30-50 °. At greater separations, there was an increasing tendency to target one or other of the two shapes. From these experiments it is concluded that, in the frontal visual field, spatial integration can be performed to an accuracy (as defined in these experiments) of 7 °x7 °. Performance deteriorates outside an optimal range of image sizes, and this is faithfully reflected in the decline in horizontal axial resolution when jumping. However, this decline in visual performance is not reflected in target resolution along the vertical axis because it is masked by a biomechanical constraint, i.e. the inability of the legs to set take-off angles outside a narrow range of 30-60 ° to the horizontal.