The effects of scopolamine, diazepam, and lorazepam on working memory in pigeons: an analysis of reinforcement procedures and sample problem type.

Two groups of pigeons were trained on a delayed-matching-to-sample (DMTS) task with both identity and symbolic problems, that had either a) specific outcomes correlated (differential group) or b) outcomes uncorrelated (nondifferential group), for each correct sample-choice sequence. After reaching a criterion of 90% correct at the 0 s delay, subjects were tested under saline, methylscopolamine (0.03 mg/kg), scopolamine (0.007, 0.015, 0.03 mg/kg), diazepam (0.0, 1.0, 1.75, 2.5 mg/kg), and lorazepam (0.0, 0.5, 0.75, 1.0 mg/kg) at delays of 0 to 8 s. Scopolamine, diazepam, and lorazepam at all doses impaired performance in the nondifferential group; however, in the differential group, the medium and high doses of both scopolamine and lorazepam, and only the high dose of diazepam impaired performance. The differential outcomes procedure, relative to the nondifferential procedure, enhanced retention in the non-drug state and under these amnestic drugs. Impairments observed in the differential group were a result of decreased performance only on samples correlated with a secondary reinforcer (flashing hopper light); there was no decreased performance on samples correlated with a primary reinforcer (grain). Neither group showed any differences in performance as a function of identity versus symbolic problems in a nondrug or drug state.