Working memory tasks in five-choice operant chambers: use of relative and absolute spatial memories.

Rats were trained to nose poke into illuminated holes to perform 1 of 2 different spatial working memory tasks (relative recency or reward history) in a 5-choice operant chamber. A series of experiments indicated that choice accuracy on both tasks depended on (a) the holes' spatial separation, and (b) their relative rather than absolute positions. The results suggest that accurate choice depended on using a motor mediation strategy to turn, so as to encounter the target (correct) hole before encountering the alternative (wrong) hole. The drugs administered to the rats, d-amphetamine, scopolamine, and CGP-37849 impaired choice accuracy on these tasks, even though task performance had not appeared to depend on explicit memory for the sample responses. This suggests that parallel drug effects obtained on other operant matching- or nonmatching-to-position tasks may not have reflected truly amnesic effects of the drug treatments.