The interaction between working and reference spatial memories in rats on a radial maze.

The interaction of reference and working memory was studied in rats on an eight-arm radial maze. Each trial involved a two-phase procedure in which a rat was forced to enter four arms on the maze in a study phase and then was allowed to choose among all eight arms in a test phase given 5-s later, with choice of only the previously unvisited arms rewarded. For each rat, two arms on the maze were designated as reference memory arms because they were never entered in the study phase and were always rewarded in the test phase. The other two arms never entered in the study phase and rewarded in the test phase were working memory arms and varied randomly from trial to trial. In Experiment 1, rats showed acquisition of equivalent preference for entering the reference and working memory arms in their first four choices of the test phase. Subsequent tests carried out in Experiment 2 compared performance at 5-s, 1-h, and 24-h retention intervals when reference memory and working memory were congruent and incongruent. Higher accuracy for choice of reference memory arms than working memory arms appeared at the 1-h and 24-h retention intervals on congruent tests but not on incongruent tests. A process dissociation procedure analysis indicated that working memory but not reference memory declined over the 24-h retention interval. The interaction of working and reference memory was shown by superior choice of reference memory arms on congruent tests than on incongruent tests at 1-h and 24-h retention intervals but not at the 5-s retention interval. These findings suggest that working and reference memory are independent systems that can facilitate and compete with one another. This article is part of a Special Issue entitled: Tribute to Tom Zentall.