The retina of the leopard frog projects topographically to the superficial neuropil of the entire contralateral tectum. In the rostromedial neuropil of the tectum, there is a map of the binocular region of the visual field seen from the ipsilateral eye that is in register with the map of the binocular region of the visual field seen from the contralateral eye. The ipsilateral eye projects indirectly to the tectum through nucleus isthmi (n. isthmi), a midbrain tegmental structure. N. isthmi receives input from the ipsilateral optic tectum and sends projections bilaterally that cover both tectal lobes. Previous workers have not been able to find visual activity from the ipsilateral eye in the caudolateral optic tectum, representing the monocular visual field of the contralateral eye. We show electrophysiologically that across the entire extent of n. isthmi there are two superimposed maps, one map representing the entire visual field of the contralateral eye, the other map representing the binocular visual field of the ipsilateral eye. We also studied the behavioral consequences of localized lesions to n. isthmi and compared them to the behavioral consequences of localized lesions to the optic tectum representing equivalent areas of the visual field. Lesions to the optic tectum produce scotomas in the corresponding portion of the visual field. Lesions to n. isthmi, even medial n. isthmi representing the superior visual field, lead to scotomas in the temporal-most portion of the contralateral ground level visual field. Thus, the representation of visual space in n. isthmi is not a simple copy of the tectal representation of visual space.

Administration of a single subcutaneous dose of 5,7-dihydroxytryptamine (5,7-DHT) to newborn hamsters results in a significant increase in the density of serotoninergic (5-HT) fibers in the superficial layers of the superior colliculus (SC) and marked abnormalities in the uncrossed retinotectal projection when these animals reach adulthood (Rhoades et al., 1993). The present study was undertaken to determine whether elevation of 5-HT in the developing SC altered the visual representation in SC. Multi-unit recordings from SC cells demonstrated that the overall organization of the visual map in the superficial SC laminae was normal and that the receptive-field sizes for unit clusters were unchanged in the 5,7-DHT-treated animals. However, when a combination of CNQX and MK-801 was directly applied to the SC to block postsynaptic activity, the receptive fields of unit clusters (presumably retinotectal axon terminals) in the 5,7-DHT treated animals were significantly larger than those in the normally reared hamsters. These results are consistent with the conclusions that elevation of 5-HT levels in the developing SC reduces the postnatal refinement of the crossed retinotectal axons, and that mechanisms operating within the SC may act to maintain normal sizes for the receptive fields of its constituent neurons.