First- and second-order conditioning in goldfish and their relation to the telencephalon

Previous work has shown that normal and telencephalon-ablated goldfish ( Carassius auratus) acquire trace classical conditioning at equal rates, suggesting that first-order conditioning is not affected by the ablation. However, since it is known that the length of delays and the spacing of trials are important factors with regard to deficits engendered by telencephalic ablation, these were investigated in the present work. The first two experiments investigated extended (15-sec) trace intervals using massed trials and intermediate (10-sec) trace intervals using spaced trials. No differences were observed in the rates of acquisition of control and telencephalon-ablated fish. In addition, extinction rates were found to be similar. In a third experiment, control fish, trained classically with light as CS and shock as UCS, were found to reverse their side preference in a T maze on presentation of the same light as “reinforcement” for a turn to the previously preferred side. Pseudoconditioned control fish and pseudoconditioned and conditioned telencephalon-ablated fish did not show such a reversal. It is considered probable that the telencephalon processes secondary or derived reinforcement but is not responsible for the processing of primary reinforcement.