Morphological and physiological identification of medulla interneurons in the visual system of the tiger beetle larva.

The morphology of visual interneurons in the tiger beetle larva was identified after recording their responses. Stained neurons were designated as either medulla or protocerebral neurons according to the location of their cell bodies. Medulla neurons were further subdivided into three groups. Afferent medulla neurons extended processes distally in the medulla neuropil and a single axon to the brain through the optic nerve. They received their main input from stemmata on the ipsilateral side. Two distance-sensitive neurons, near-by sensitive and far-sensitive neurons, were also identified. Atypical medulla neurons extended their neurites distally in the medulla and proximally to the brain, as afferent medulla neurons, but their input patterns and the shapes of their spikes differed from afferent neurons. Protocerebral neurons sent a single axon to the medulla neuropil. They spread collateral branches in the posterior region of the protocerebrum on its way to the medulla neuropil. They received main input from stemmata on the contralateral side. Medulla intrinsic neurons did not extend an axon to the brain, and received either bilateral or contralateral stemmata input only. The input patterns and discharge patterns of medulla neurons are discussed with reference to their morphology.

Behavioural responses of the tiger beetle larva to moving objects: role of binocular and monocular vision.

The larva of the tiger beetle Cicindela chinensis is an ambushing hunter with a body length of 15-22 mm that lives in a tunnel in the ground. It ambushes prey, keeping its head horizontal at the opening of the tunnel. When prey approaches the tunnel, the larva jumps to snap at it. When an object moves beyond its jumping range (approximately 15 mm), however, the larva quickly withdraws deep into the tunnel. These responses are mediated by two of six pairs of stemmata. How does the larva judge the hunting range using such a simple visual system? A previous study suggested that both binocular and monocular vision are used for distance estimation. Range estimation by binocular vision was further confirmed in the present behavioural observations: larvae jumped towards objects beyond the normal hunting range when virtual images of such distant objects were formed close to the larva using prisms or a narrow window. A possible mechanism involved in range estimation by monocular vision was also examined in behavioural experiments. The depth of the image in the retina appears to play a role in distance estimation because a larva with one functional stemma, the other stemmata being occluded, changed its response to a very distant object from an escape to a predatory jump when a concave lens was placed above its head. Two alternative ideas, based on optical and morphological data, are proposed to explain this behavioural change by the one-stemma larvae. First, as for myopic people, the larva might see clearly only objects that are close. Second, an infinitely distant object might produce a focused image only on the central part of the retina, whereas an object within hunting range (<15 mm) might do so on surrounding regions of the retina. The latter idea implies that the region of the retina at which the larva perceives a clear image is concerned with which type of behaviour is released, a predatory jump or an escape. We conclude that visual information about hunting range in the tiger beetle larva is extracted both peripherally by the spatial pattern of image clarity and centrally by binocular vision.

Responses of medulla neurons to illumination and movement stimuli in the tiger beetle larvae.

Intracellular responses of medulla neurons (second-order visual interneurons) have been examined in the tiger beetle larva. The larva possesses six stemmata on either side of the head, two of which are much larger than the remaining four. Beneath the cuticle housing the stemmata an optic neuropil complex occurs consisting of lamina and medulla neuropils. Response patterns of medulla neurons to illumination and moving objects varied from neurons to neurons. For movement stimuli black discs and a black bar were moved in the rostro-caudal direction above the larva. Comparison of responses to the discs and the bar suggested a spatial summation of responses in some neurons, and tuning to small objects in some neurons. The majority of neurons responded to objects moving at heights of 10 mm and 50 mm with the same discharge pattern. A few neurons, however, showed distance sensitivities responding with an increase of spike discharges to moving objects only at either of the two heights. Such distance sensitivities still remained in one-stemma larvae, three of the four stemmata being occluded. These data are discussed in relation to distinct visual behavior of the larva and with special reference to perception of the hunting range.