Behavioural Brain Research, 19 (1986) 133-146 133 Elsevier BBR 00546 THE EFFECTS OF HIPPOCAMPAL LESIONS UPON SPATIAL AND NON-SPATIAL TESTS OF WORKING MEMORY J.P. AGGLETON l, P.R. HUNT 1 and J.N.P. RAWLINS 2 lDepartment of Psychology, University of Durham, Durham DH1 3LE and 2Department of Experimental Psychology, University of Oxford, Oxford OX1 3UD (U.K.) (Received August 28th, 1985) (Revised version received December 9th, 1985) (Accepted December 9th, 1985) Key words: hippocampus - recognition memory - working memory - spatial memory - interference - rat A series of experiments examined the proposal that the primary effect of hippocampal damage in rats is to disrupt working memory. Although extensive hippocampal lesions produced a severe impairment in forced-choice alternation - a test of spatial working memory - the same lesions did not impair the acquisition of a non-spatial test of working memory - delayed non-matching-to-sample. This test of object recognition required the rats to select that arm in a Y-maze which contained unfamiliar stimuli. Rats with hippocampal lesions were able to learn and perform this task at normal rates, even with retention delays of as long as 60 s. Two additional experiments helped confirm that the animals had indeed learnt a non-spatial test of working memory. The final experiment examined whether hippocampal lesions resulted in an increased sensitivity to proactive interference. It was found that repetition of test stimuli within a session, which increased interference, did attenuate recognition performance but there was no evidence that the animals with hippocampal lesions were differentially affected. INTRODUCTION Considerable interest has been aroused by the proposal that hippocampal damage selectively disrupts working memory in rats 19. This proposal argues that, the hippocampus and its primary afferent and efferent connections are necessary for learning flexible stimulus-response asso- ciations such as information which is useful for only one trial of an experiment but is irrelevant, or even misleading, on subsequent trials (working memory). In contrast, the learning of constant response associations (reference memory) does not require the hippocampus. Thus a hippocam- pal lesion should not affect learning always to turn left in a T-maze (reference memory) but will dis- rupt discrete-trial alternation (working memory) which requires remembrance of the previous trial 19,26. The emphasis of this hypothesis on the mnemonic functions of the rat hippocampus has proved particularly attractive in the light of the human amnesic and monkey experimental litera- ture which has repeatedly shown that the hippo- campal formation has a key role in memory 5,29. Much of the evidence for the importance of the hippocampus in working memory has come from studies of spatial learning by rats 19. There is little doubt that hippocampal system lesions severely disrupt tasks such as discrete-trial spatial alter- nation, complex maze learning and the radial arm maze devised by Olton and Samuelson21. It has been strongly argued, however, that the observed deficits are a consequence of the spatial nature of the tasks rather than the requirement for working memory 16'~s. One direct way of distinguishing between these alternatives would be to examine Correspondence: J.P. Aggleton, Department of Psychology, University of Durham, South Road, Durham DH1 3LE, U.K. 0166-4328/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)