EVENT-RELATED POTENTIALS IN THE RETINA AND OPTIC TECTUM OF FISH

Compound field potentials were recorded with up to 18 microelectrodes in comb, brush, or spear arrays on and in the optic tectum and with suction electrodes from the distal stump of the cut optic nerve and from the optic nerve head in the opened eye in elasmobranchs and teleosts. Diffuse light flashes of different durations and submaximal intensities were delivered in trains with regular or irregular interstimulus intervals (ISI). Event-related potentials (ERPs) are visible in single trials and begin at 50-200 ms after an 'oddball' flash, especially one that is slightly weaker, briefer, or delayed by as little as 6% of ISI, compared with the more frequent stimulus. ERPs to the opposite condition are not of the same form or size. One or more stimuli were omitted from a train or the train terminated after various conditioning times. Deflections occur beyond the expected visual-evoked potentials (VEPs) to the last flash and are called omitted-stimulus potentials (OSPs). They occur on schedule-~100 ms after the next flash would be due-almost independent of intensity, duration, or conditioning time. They are considered to be ERPs without any necessary implication or denial of a temporally specific expectation. Three components of OSP occur alone or in combination: an initial fast peak, a slow wave, and an oscillatory spindle up to 1s or more in duration. This resembles the OFF response to steady light. All these components are already present in the retina with optic nerve cut. The same mean ISI with a high proportion of jitter gives OSPs with only slightly longer latencies and smaller amplitudes; the OSP acts as though the retina makes an integrated prediction of ISI, intensity, and duration. During a conditioning train the equilibrium between excitation and inhibition after each flash changes according to frequency, intensity, duration, and conditioning time; the VEP reflects this in a shape unique to the ISI; inhibition increases rapidly after each flash and then decays slowly according to the recent mean ISI. This allows rebound disinhibition after missing, weak, or delayed flashes (OSP or ERP) or causes an altered VEP after a longer or stronger oddball. It seems unlikely that the OSP or oddball ERP in fish tectum is equivalent to mammalian ERPs under the same regime or signals higher cognitive events, because they are already present in the retina, require flash frequencies >1 Hz, and grow with frequency up to and beyond flicker fusion. The possibility, even in mammals, that some ERP components are rebounds or the like, at lower precognitive levels, must be considered. Low-level sensory ERPs can have long latency, complex forms, with evoked rhythmicities, and can integrate stimulus history over many seconds.