{"Affiliation":[{"label":"Affiliation","value":"Arts, Faculty of","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","classmap":"vivo:EducationalProcess","property":"vivo:departmentOrSchool"},"iri":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","explain":"VIVO-ISF Ontology V1.6 Property; The department or school name within institution; Not intended to be an institution name."},{"label":"Affiliation","value":"Psychology, Department of","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","classmap":"vivo:EducationalProcess","property":"vivo:departmentOrSchool"},"iri":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","explain":"VIVO-ISF Ontology V1.6 Property; The department or school name within institution; Not intended to be an institution name."}],"AggregatedSourceRepository":[{"label":"Aggregated Source Repository","value":"DSpace","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","classmap":"ore:Aggregation","property":"edm:dataProvider"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","explain":"A Europeana Data Model Property; The name or identifier of the organization who contributes data indirectly to an aggregation service (e.g. Europeana)"}],"Campus":[{"label":"Campus","value":"UBCV","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","classmap":"oc:ThesisDescription","property":"oc:degreeCampus"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","explain":"UBC Open Collections Metadata Components; Local Field; Identifies the name of the campus from which the graduate completed their degree."}],"Creator":[{"label":"Creator","value":"Mumby, David Gerald","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/creator","classmap":"dpla:SourceResource","property":"dcterms:creator"},"iri":"http:\/\/purl.org\/dc\/terms\/creator","explain":"A Dublin Core Terms Property; An entity primarily responsible for making the resource.; Examples of a Contributor include a person, an organization, or a service."}],"DateAvailable":[{"label":"Date Available","value":"2008-12-23T22:31:02Z","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"edm:WebResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"DateIssued":[{"label":"Date Issued","value":"1992","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"oc:SourceResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"Degree":[{"label":"Degree (Theses)","value":"Doctor of Philosophy - PhD","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","classmap":"vivo:ThesisDegree","property":"vivo:relatedDegree"},"iri":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","explain":"VIVO-ISF Ontology V1.6 Property; The thesis degree; Extended Property specified by UBC, as per https:\/\/wiki.duraspace.org\/display\/VIVO\/Ontology+Editor%27s+Guide"}],"DegreeGrantor":[{"label":"Degree Grantor","value":"University of British Columbia","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","classmap":"oc:ThesisDescription","property":"oc:degreeGrantor"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the institution where thesis was granted."}],"Description":[{"label":"Description","value":"The nonrecurring-items delayed nonmatching-to-sample (DNMS) task is an integral part of\r\ncontemporary monkey models of brain-damage-produced amnesia. This thesis began the development\r\nof a comparable rat model of brain-damage-produced amnesia. First, a DNMS task for rats was\r\ndesigned by adapting key features of the monkey task. Then, the rat DNMS task was studied in three\r\nexperiments; each assessed the comparability of the rat DNMS task to the monkey DNMS task.\r\nExperiment 1 determined the rate at which the rat DNMS task is learned and the asymptotic level at\r\nwhich it is performed, Experiment 2 assessed the memory abilities that it taps, and Experiment 3\r\ninvestigated the brain structures that are involved i n its performance.\r\nIn Experiment 1, rats were trained on the DNMS task and their performance was assessed at\r\nretention delays of 4, 15, 60, 120, and 600 s. All of the rats learned the DNMS task, and their\r\nperformance was comparable to that commonly reported for monkeys in terms of both the rate at\r\nwhich they acquired the nonmatching rule at a brief retention delay and their asymptotic accuracy at\r\ndelays of up to 120 s. These results establish that rats can perform a DNMS task that closely resembles\r\nthe monkey DNMS task and that they can approximate the level of performance that is achieved by\r\nmonkeys.\r\nExperiment 2 examined the effects of distraction during the retention delay on the DNMS performance of rats. Rats were tested at retention delays of 60 s. On half of the trials, the rats\r\nperformed a distraction task during the retention delay; on the other half, they did not. Consistent with\r\nfindings from monkeys and humans, distraction during the retention delay disrupted the DNMS\r\nperformance of rats. This suggests that similar memory abilities are involved in the DNMS\r\nperformance of rats, monkeys, and humans. Experiment 3 investigated the effects of separate and combined bilateral lesions of the\r\nhippocampus and the amygdala on DNMS performance in pretrained rats. Rats were tested both\r\nbefore and after surgery at retention delays of 4, 15, 60, 120, and 600 s. Each experimental rat received\r\nbilateral lesions of the hippocampus, amygdala, or both. There were no significant differences among\r\nthe three experimental groups, and the rats in each of the three experimental groups were significantly\r\nimpaired, in comparison to no-surgery control rats, only at the 600-s delay. In contrast, rats that had\r\nsustained inadvertent entorhinal and perirhinal cortex damage during surgery displayed profound\r\nD N M S deficits. These results parallel the results of recent studies of the neural basis of DNMS in\r\nmonkeys. They suggest that, in contrast to one previously popular view, neither the hippocampus nor\r\nthe amygdala play a critical role in the DNMS of pretrained animals and that the entorhinal and\r\nperirhinal cortex are critically involved.\r\nOn the basis of these findings, it appears that the rat DNMS task may prove to be a useful\r\ncomponent of rat models of brain-damage-produced amnesia. This conclusion is supported by the\r\npreliminary results of several experiments that are currently employing the task.","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/description","classmap":"dpla:SourceResource","property":"dcterms:description"},"iri":"http:\/\/purl.org\/dc\/terms\/description","explain":"A Dublin Core Terms Property; An account of the resource.; Description may include but is not limited to: an abstract, a table of contents, a graphical representation, or a free-text account of the resource."}],"DigitalResourceOriginalRecord":[{"label":"Digital Resource Original Record","value":"https:\/\/circle.library.ubc.ca\/rest\/handle\/2429\/3293?expand=metadata","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","classmap":"ore:Aggregation","property":"edm:aggregatedCHO"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","explain":"A Europeana Data Model Property; The identifier of the source object, e.g. the Mona Lisa itself. This could be a full linked open date URI or an internal identifier"}],"Extent":[{"label":"Extent","value":"4495793 bytes","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/extent","classmap":"dpla:SourceResource","property":"dcterms:extent"},"iri":"http:\/\/purl.org\/dc\/terms\/extent","explain":"A Dublin Core Terms Property; The size or duration of the resource."}],"FileFormat":[{"label":"File Format","value":"application\/pdf","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/elements\/1.1\/format","classmap":"edm:WebResource","property":"dc:format"},"iri":"http:\/\/purl.org\/dc\/elements\/1.1\/format","explain":"A Dublin Core Elements Property; The file format, physical medium, or dimensions of the resource.; Examples of dimensions include size and duration. Recommended best practice is to use a controlled vocabulary such as the list of Internet Media Types [MIME]."}],"FullText":[{"label":"Full Text","value":"T H E D E V E L O P M E N T O F A R A T M O D E L O F B R A I N - D A M A G E - P R O D U C E D A M N E S I A by D A V I D G E R A L D M U M B Y B . S c , Universi ty o f A lbe r t a , 1986 M.Scc, Universi ty of A lbe r t a , 1988 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y i n T H E F A C U L T Y O F G R A D U A T E S T U D I E S D E P A R M E N T O F P S Y C H O L O G Y W e accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A A p r i l , 1992 \u00a9 D a v i d G e r a l d M u m b y , 1992 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. (Signature) Department of Psychology The University of British Columbia Vancouver, Canada Date DE-6 (2\/88) A B S T R A C T T h e nonrecurring-items delayed nonmatching-to-sample ( D N M S ) task is an integral part o f contemporary monkey models of brain-damage-produced amnesia. This thesis began the development of a comparable rat mode l of brain-damage-produced amnesia. Firs t , a D N M S task for rats was designed by adapting key features of the monkey task. Then , the rat D N M S task was studied i n three experiments; each assessed the comparabili ty of the rat D N M S task to the monkey D N M S task. Exper iment 1 determined the rate at which the rat D N M S task is learned and the asymptotic level at which it is performed, Experiment 2 assessed the memory abilities that it taps, and Exper iment 3 investigated the brain structures that are involved i n its performance. In Exper iment 1, rats were trained on the D N M S task and their performance was assessed at retention delays of 4, 15, 60, 120, and 600 s. A l l o f the rats learned the D N M S task, and their performance was comparable to that commonly reported for monkeys i n terms of both the rate at which they acquired the nonmatching rule at a br ief retention delay and their asymptotic accuracy at delays of up to 120 s. These results establish that rats can perform a D N M S task that closely resembles the monkey D N M S task and that they can approximate the level of performance that is achieved by monkeys. Exper imen t 2 examined the effects of distraction during the retention delay on the D N M S performance of rats. Rats were tested at retention delays of 60 s. O n half o f the trials, the rats performed a distraction task during the retention delay; on the other half, they d id not. Consistent wi th findings f rom monkeys and humans, distraction during the retention delay disrupted the D N M S performance of rats. This suggests that similar memory abilities are involved i n the D N M S performance of rats, monkeys, and humans. Exper iment 3 investigated the effects of separate and combined bi lateral lesions of the hippocampus and the amygdala on D N M S performance i n pretrained rats. Rats were tested both before and after surgery at retention delays of 4, 15, 60, 120, and 600 s. E a c h experimental rat received bilateral lesions of the hippocampus, amygdala, or both. There were no significant differences among the three experimental groups, and the rats i n each of the three experimental groups were significantly impaired, i n comparison to no-surgery control rats, only at the 600-s delay. In contrast, rats that had sustained inadvertent entorhinal and perirhinal cortex damage during surgery displayed profound D N M S deficits. These results parallel the results of recent studies of the neural basis o f D N M S i n monkeys. They suggest that, i n contrast to one previously popular view, neither the hippocampus nor the amygdala play a cri t ical role i n the D N M S of pretrained animals and that the entorhinal and per i rhinal cortex are critically involved. O n the basis of these findings, it appears that the rat D N M S task may prove to be a useful component o f rat models o f brain-damage-produced amnesia. This conclusion is supported by the prel iminary results of several experiments that are currently employing the task. T A B L E O F C O N T E N T S Abstract i i Table of Contents iv L is t o f Figures v i Acknowledgements v i i G E N E R A L I N T R O D U C T I O N 1 1. Characteristics o f the Amnes i c Syndrome 2 H . M . ' s amnesia 2 A r e there mult iple dissociable memory systems? 6 2. E a r l y At tempts to M o d e l Bra in-Damage-Produced A m n e s i a i n Labora tory A n i m a l s 7 E a r l y studies of the effects of hippocampal lesions i n laboratory animals 8 W h y early attempts to model brain-damage-produced amnesia failed 9 3. T h e Preeminent M o n k e y Mode l s of Bra in-Damage-Produced A m n e s i a 11 The nonrecurring-items delayed nonmatching-to-sample ( D N M S ) task 12 T h e origins of the preeminent monkey models of brain-damage-produced amnesia 14 T h e effects o f bi lateral medial-temporal-lobe lesions on D N M S performance i n monkeys 15 The effects o f medial-diencephaUc lesions on D N M S performance i n monkeys 20 G E N E R A L R A T I O N A L E 21 A N E W D N M S T A S K F O R R A T S 23 1. Aggle ton 's Y - m a z e D N M S Task 23 2. Rothbla t and Hayes 's D N M S Task 25 3. Requirements o f an Effective Rat D N M S Task 27 4. T h e N e w R a t D N M S Task 27 T h e apparatus 28 G e n e r a l t raining and testing procediu-es 28 E X P E R I M E N T 1: D N M S P E R F O R M A N C E I N R A T S 35 L M e t h o d s 35 Subjects 35 D N M S training 35 M e a s u r i n g the retention function 36 2. Resul ts 36 Habi tua t ion 36 Object discriminat ion 36 A c q u i s i t i o n of D N M S 37 Re ten t ion functions 37 3. Discuss ion 38 E X P E R I M E N T 2: T H E E F F E C T S O F D I S T R A C T I O N O N D N M S I N R A T S 45 1. M e t h o d s 46 Subjects 46 Procedure 47 2. Resul ts 48 3. Discuss ion 48 E X P E R I M E N T 3: T H E E F F E C T S O F S E P A R A T E A N D C O M B I N E D L E S I O N S O F T H E H I P P O C A M P U S A N D A M Y G D A L A O N D N M S I N R A T S 52 1. M e t h o d s 54 Subjects 54 Habi tua t ion , object-discrimination, D N M S training, and measxiring the presurgery retention function 54 Surgery 55 Reacquis i t ion of D N M S 56 Postsurgery retention functions 57 Stage-two lesions and testing 57 2. Resul ts 57 Habi tua t ion 58 Object-discrimination training 58 D N M S training 58 Presurgery retention functions 59 His tology 59 Reacquis i t ion of D N M S 65 Postsurgery retention functions 65 3. Discuss ion 71 G E N E R A L D I S C U S S I O N 75 1. The Correspondence Between the Ra t D N M S Task and the M o n k e y D N M S Task 75 The general protocol o f the rat D N M S task resembles that o f the monkey D N M S task 77 The D N M S performance of rats is comparable to that of monkeys 77 The D N M S performance of rats, humans, and monkeys appears to involve s imilar memory abilities 78 2. Evidence F o r Correspondence Between the Neu ra l Systems That U n d e r l y Object Recogni t ion i n Rats, Monkeys , and Humans 78 3. Other D N M S Experiments i n Rats 80 D o the rat and monkey D N M S tasks involve similar memory abiUties? 81 D o similar neural systems underUe D N M S performance i n rats and monkeys? 83 L e s i o n experiments 84 Experiments designed to assess the etiological validity o f the rat D N M S task 86 T h e development of a test battery for use i n rat models of brain-damage-produced amnesia 89 3. Summary and Conclusions 91 R E F E R E N C E S 93 L I S T O F F I G U R E S Figure 1. The D N M S apparatus 29 Figure 2. M e a n percent correct on the first and last session at each delay dur ing acquisit ion training on D N M S 39 Figure 3. M e a n percent correct on mixed-delay D N M S sessions 41 Figure 4. M e a n scores on no-distraction and distraction trials 49 F igure 5. The presurgery (left) and postsurgery (right) retention functions that were determined on mixed-delay sessions 60 Figure 6. Reconstructions of hippocampal, amygdalar, and amygdalo-hippocampal lesions 63 Figure 7. Presurgery and postsurgery retention functions for 2 rats that sustained inadvertent damage to entorhinal, per irhinal , and temporal association cortex 67 Figure 8. Lesions i n 4 rats with hippocampal lesions that sustained inadvertent damage to entorhinal, per i rhinal , and temporal association cortex 69 A C K N O W L E D G E M E N T S I wish to thank John Pine l , for his advice, encouragement, support, and friendship. John's ideas provided the original impetus for the work described i n this dissertation. E m m a W o o d has my undying gratitude, not only for helping with many of the experiments described i n this dissertation, but also for being a truly loyal friend. I also wish to thank the other members of my supervisory committee, D o n W i l k i e and Peter Graf, whose excellent advice enabled me to consoUdate my thoughts on this thesis. I am also very grateful to Laur ie Carswel l and M i c h a e l Smeele, both of w h o m helped collect some of the data reported i n this dissertation, and to the following people who helped collect data for related experiments: Danie l le Anzarut , Ka te Banks, M a r i o n Buday, T i m Bussey, Noushine Dadgar , F a r h a d Dastur , Ey t an Dav id , Shayesta Dha la , T a m a r a Gai t , M i c h e l l e H i l l , D a n a Hronek , K a r i Journeaux, L i s a Kalynchuk, T o d K i p p e n , T o m Kornecook, A r d i s Krueger , Jeimifer K w o n g , S a l l y - A i m L e e , M i k e M a n a , G r e g Payne, V a n Re id la , M a r t i n Shen, K a r i n Steichle, K a r e n Tang, B i h n T ran , Louiset te Tremblay, and War r en Wart t ig . A b o v e a l l , I thank L i s a Pr i tchard for her enduring love and support; this dissertation is dedicated to her. G E N E R A L I N T R O D U C T I O N Bi la te ra l damage to medial-temporal-lobe or medial-diencephalic structures causes an amnesic syndrome i n humans that is characterized by an impairment o f the ability to form new memories . Over the past decade, the study of monkey models of brain-damage-produced amnesia has begun to shed light on the nature of human brain-damage-produced amnesia and its neural bases. T h e development of comparable rodent models would benefit the study of brain-damage-produced amnesia i n two general ways: (1) it would facilitate the conduct o f large-scale parametric experiments - the cost of large-scale monkey research is prohibitive for most researchers - and (2) it w o u l d provide a broader comparative basis for drawing inferences about the anatomical bases of brain-damage-produced amnesia. Th i s thesis constitutes the first stages i n the development of a rat m o d e l o f brain-damage-produced amnesia. A n integral feature of the monkey models o f human brain-damage-produced amnesia is the nomecurr ing-i tems delayed nonmatching-to-sample ( D N M S ) memory task. H u m a n amnesics have difficulty performing the D N M S task (Squire, Z o l a - M o r g a n , & Chen , 1988), and so do monkeys wi th bi lateral medial- temporal- lobe (Mishk in , 1978; M u r r a y & M i s h k i n , 1984; Z o l a - M o r g a n & Squire, 1985a; Z o l a - M o r g a n & Squire, 1986; Z o l a - M o r g a n , Squire, A m a r a l , & Suzuki , 1989) or media l -diencephalic (Aggle ton & M i s h k i n , 1989a, 1989b; Z o l a - M o r g a n & Squire, 1985b) damage. T h e present experiments were conducted (1) to develop a D N M S task for rats that resembles the D N M S task for monkeys, and (2) to determine whether it parallels the monkey D N M S task in terms of the memory abilities and the brain structures that it engages. Accordingly , the G e n e r a l Introduction deals wi th the fol lowing three topics: (1) a description of human brain-damage-produced amnesia, (2) an historical account of attempts to mode l human brain-damage-produced amnesia i n laboratory animals, and (3) a description of the preeminent monkey models o f human brain-damage-produced amnesia. 1. CHARACTERISTICS OF THE AMNESIC SYNDROME In some brain-damaged individuals, impai red memory occurs i n a relatively pure form, that is, i n the absence of other pr imary symptoms. The major characteristic o f this classic \"amnesic syndrome\" (Baddeley, 1990) is an inability to form new long-term memories . It occurs fol lowing bi lateral damage to either one of two b ra in areas - the media l diencephalon or the media l temporal lobe. Accord ing ly , depending on the hypothesized location of their brain damage, amnesic patients are classified as either medial-diencephaUc or medial-temporal-lobe amnesics. H . M . ' s amnesia Remarkably , much of what is known about brain-damage-produced amnesia has come from the study of a single patient. This patient is H . M . , who has been amnesic since 1953, when he received bi lateral medial- temporal-lobe resections for the treatment of intractable epilepsy (Scoville, 1954). H . M . ' s surgery removed the anterior two-thirds of the hippocampus, the parahippocampal gyrus, the uncus, and the amygdala of both hemispheres (Mi lne r , 1959). Despite his amnesia, H . M . is normal i n many respects. H e displays normal intelligence and intact perceptual and attentional abilities, and he suffers no apparent emotional or personaUty disorders (Scoville & M i h i e r , 1957). H . M . ' s deficits are, for the most part, Umited to his memory functions, and this is why he has been a particularly useful case for the study of medial-temporal-lobe amnesia (Mi lne r , 1968). The core symptoms of H . M . ' s medial-temporal-lobe amnesia are typical o f the amnesia of other patients wi th bi lateral medial-temporal-lobe damage (e.g., M i l n e r , 1959; V i c t o r , Angevine , M a n c a l l , & Fischer, 1961) and of the amnesia of patients with bilateral medial-diencephaUc damage (see V i c t o r & Yakovlev , 1955; V ic to r , A d a m s , & Col l ins , 1971). Accordingly , the fol lowing description of H . M . ' s amnesia serves as a general introduction to the predominant symptoms of human brain-damage-produced amnesia. Impaired memory abilities. Fo l lowing his bilateral temporal-lobe resection i n 1953, H . M . exhibited a severe anterograde amnesia, which has not diminished to this day; he has extreme difficulty forming memories o f events that he has experienced since his bra in surgery. T h e devastating impact that H . M . ' s anterograde amnesia has had on his life is apparent from the anecdotal accounts of Scoville and M i l n e r (1957). F o r example, H . M . is unable to recognize doctors and nurses w h o m he has seen many times, and he often reads the same magazines over and over again without real izing it. H . M . can remember smal l amounts of information - such as numbers or word associations ~ for several minutes, as long as he is al lowed to maintain his attention on them. A s soon as he is distracted, the information is lost. H . M . also displays m i l d retrograde amnesia; he has difficulty remember ing events that occurred during the year pr ior to his surgery, but his memory for events that occurred earlier remains largely intact ( C o r k i n , 1968). H . M . ' s memory impairments include both verbal and nonverbal information, and information i n al l sensory modalit ies (Scoville & M i l n e r , 1957). The term \"global amnesia\" is often used to refer to amnesic syndromes, such as H . M . ' s , in which memory for information i n a l l sensory modaUties is affected. There have been many experimental demonstrations of H . M . ' s difficulty i n forming long-term memories . T h e fol lowing are three of them: 1. H . M . was severely impaired on a digit-span + 1 test. In this version of the digit-span test, one new digit is added to the previous sequence each time that the subject gets a sequence correct. F o r example, if a subject were able to correctly repeat the sequence \" 4, 6, 3, 8,\" on the next tr ial the sequence might be \" 4, 6, 3, 8, 5 \" - and this sequence wou ld be repeated on each tr ial unt i l the subject recalled it correctly, at which point another digit would be added. N o r m a l subjects recalled sequences of about 15 digits after only 25 digit-span + 1 trials. In contrast, H . M . was unable to progress beyond his in i t ia l 6-digit memory span (Drachman & A r b i t , 1966). 2. H . M . was also severely impaired on a block-tapping memory-span + 1 test ~ a nonverbal version of the digit-span + 1 test (Mi lne r , 1971). Several blocks were spread out i n front o f H . M . , who watched as the experimenter touched several of them i n sequence. H . M . was then asked to repeat the same sequence of touches. H . M . was unable to learn a sequence that was one block greater than his ini t ial block-tapping span. 3. H . M was impai red on a nonverbal matching-to-sample task. H . M . was presented with a sample i tem (i.e., one of several different ellipses), and then, following a retention delay, the same i tem and several other similar ones were simultaneously presented. H . M . ' s task was to select the i tem that he had seen before. H e was unable to per form this task, even when the retention delay was only 5 s (Sidman, Stoddard, & M o h r , 1968). However , i n the same study, H . M . performed normally on the matching-to-sample task at delays of up to 40 s when verbal st imuli were used (i.e., sequences o f three consonants). The experimenters concluded that H . M . can perform the matching-to-sample task only when it is possible for h im to rehearse the st imuli during the retention delay. Spared memory abilities. Patients with medial-temporal-lobe and medial-diencephalic amnesia can fo rm some kinds o f long-term memories. Schneider (1912) conducted the first systematic studies o f preserved memory abilities in amnesic patients (see Park in , 1982). Schneider showed his amnesic subjects a picture of an object and later tested their retention by presenting a fragment of the or iginal picture. The subjects displayed an enhanced ability to identify the object f rom the picture fragment even though they could not recal l having seen the picture before. Clear ly , the experience of seeing the picture had been retained in some sense although the subjects were not consciously aware of it. Despi te Schneider's early demonstration, the extent of spared memory abihties i n brain-damage-produced amnesia d id not start to be appreciated unti l the 1960s. The first widely cited studies of spared memory abihties were studies of H . M . , but it was subsequently shown that the abilities that were spared i n H . M . also tended to be spared i n other patients with medial- temporal- lobe amnesia and i n patients wi th medial-diencephalic amnesia (Brooks & Baddeley, 1976, C o h e n & Squire, 1980; Teuber , M i l n e r , & Vaughan, 1968). It was apparent at the outset that H . M . ' s short-term memory abiUties had been unaffected by the surgery ~ his postsurgery digit span was 6 (Drachman & A r b i t , 1966), wel l wi th in the no rma l range. In 1962, M i l n e r (cited in Mur ray , 1990) reported that some aspects of H . M . ' s long-term memory had also been spared. She reported that H . M . could learn a mir ror drawing task and retain it from session to session. Since then, this finding has been extended to several other perceptual and motor skills. F o r example, H . M . ' s performance on the rotary-pursuit task and on a b imanual t racking task improved at a n o r m a l rate wi th practice, although his absolute level o f performance on these tasks was inferior to that of normal subjects (Cork in , 1968). H . M . also learned and retained the pattern-analyzing skills that are involved i n reading mirror-deflected text (Cohen & Squire, 1980), and he performed almost as we l l as n o r m a l subjects on the incomplete-picture task (Mi lne r , Co rk in , & Teuber , 1968). H . M . ' s good retention over successive sessions on these tasks contrasted with his persistent inabiUty to recal l previous sessions. Are there multiple dissociable memory systems? T h e fact that H . M . and other patients wi th brain-damage-produced amnesia can display long-term memory impairments i n some situations but not i n others suggests that there are at least two kinds of long-term memory, one which is impaired i n brain-damage-produced amnesia and one which is not. Several descriptive schemes have been proposed to distinguish the long-term memory abilities that are impa i red in brain-damage-produced amnesia from those that are spared. O n e scheme distinguishes between explicit and implici t memory ( G r a f & Schacter, 1985). Explicit memory refers to the conscious recollect ion of previous experiences at the time of retrieval, whereas implicit memory refers to the retrieval and expression of information stored from previous experiences i n the absence of conscious recollect ion. The explicit-implicit distinction is not based on assumptions about the quality of the stored information (Schacter, 1987a, 1987b). Patients wi th brain-damage-produced amnesia are impa i red on expUcit memory tasks, but they exhibit normal or near normal performance on implici t memory tasks (e.g., G r a f & Schacter, 1985). A n o t h e r influential scheme for distinguishing between the long-term memory abilities that are impai red in brain-damage-produced amnesia and those that are spared posits separate systems for declarative memory, which is assumed to be impaired, and procedural memory, which is assumed to be spared. U n l i k e the explicit-impUcit distinction, the declarative-procedural dist inction is based on assumptions about the nature of stored information. Declarative memories aie said to be neural representations of previously experienced perceptions, thoughts, or facts, that can be described verbally by the individual who possesses them ( S q u \u00f9 e , 1986)^. Procedural memories are said to be inherent i n the performance of ski l led actions and revealed by changes i n the quaUty of those actions. Other dichotomies that have been appUed to the dissociation between lost and spared memory abilities i n amnesia include \"memory\" versus \"habit\" ( M i s h k i n et al. , 1984), \"episodic memory\" versus \"semantic memory\" (Schacter & Tulving, 1982), and \"working memory\" versus \"reference memory\" (Hon ig , 1978). A l t h o u g h different i n detail, these hypothetical dichotomies are conceptually s imi lar to the declarative-procedural distinction. T h e not ion that there are at least two distinguishable types of long-term memory, one of which is impa i red i n brain-damage-produced amnesia and the other of which is not, suggests that there are at least two anatomically distinct long-term memory systems. In the following section, I provide an historical account of attempts to uncover these systems through the conduct o f experiments on animal models. 2. EARLY ATTEMPTS TO MODEL BRAIN-DAMAGE-PRODUCED AMNESL4 IN LABORATORY ANIMALS T h e discovery of H . M . ' s amnesia i n the 1950s came at a t ime when most theorists accepted K a r l Lashley's proposi t ion that memory traces are widely distributed throughout the cortex (i.e., the concept of equipotentiality), rather than localized within particular structures. Lashley's experiments wi th 1 B y this definition, nonverbal animals cannot have declarative memory. Instead, the term \"representational memory\" is often used to refer to memory functions i n animals that correspond conceptually to a subtype of declarative memory. Representational memories are neural representations of the attributes of a stimulus (Murray, 1990). animals had suggested that the degree of impairment on complex learning tasks is propor t ional to the amount o f cort ical damage (i.e., the concept of mass action), but is umela ted to the particular area of cort ical damage. H . M . ' s devastating and selective memory impairment fol lowing the removal of his 2 med ia l temporal lobes challenged Lashley's equipotentiality and mass action views . H . M . ' s case resulted i n renewed support for localizationist views of memory function, and many experimenters began to search for the neural substrate of long-term memory. They began by focusing on the structures of the media l temporal lobes. Early studies of the effects of hippocampal lesions in laboratory animals H . M . ' s amnesia was originally attributed to the removal of his hippocampus for three reasons: (1) There were previous reports of patients who had suffered from amnesia fol lowing bi la teral h ippocampal damage (Bechterev, 1900, cited i n V i c t o r et al. , 1961; Glees & Gri f f i th , 1952, cited i n V i c t o r et al. , 1961), (2) there appeared to be a correlation between the extent o f h ippocampal damage and the severity of amnesia i n a group of eight amnesic patients wi th bi lateral medial- temporal- lobe resections (Scoville & M i l n e r , 1957), and (3) patients with bilateral damage that was largely l imi ted to the amygdala d id not have amnesia (Scoville & M i l n e r , 1957). However , i n the 1950s and 60s, h ippocampal lesions were found to have inconsistent effects on the performance of learning and memory tasks by rats and monkeys. 2 A l t h o u g h there was already evidence from postmortem examination of Korsakof f amnesics that damage i n relatively small areas of the brain, namely the mammil la ry bodies and the walls o f the third ventricle, could cause impaired memory, both the suddermess and the severity o f H . M . ' s amnesia made his case more compell ing. Experiments with monkeys. M i s h k i n (1954) observed normal performance of preoperatively learned visual discriminations i n a monkey with bilateral hippocampal removals. Orbach , M i l n e r , and Rasmussen (1960) tr ied to dupUcate H . M . ' s medial-temporal-lobe lesions i n monkeys by removing the hippocampus and the amygdala from both hemispheres; however, the lesions produced no visual-discr iminat ion deficits, even when the trials were widely separated i n time and the monkeys performed irrelevant discriminations during the intertriai intervals. M i s h k i n (1954), M i s h k i n and P r i b r a m (1954), and Orbach et a l . (1960) found that monkeys wi th bilateral h ippocampal lesions were not impai red on a delayed-response task, and C o r r e l l and Scoville (1965) and D r a c h m a n and O m m a y a (1%4) found that monkeys wi th bilateral hippocampal lesions were not impaired on a delayed matching-to-sample task. B y the end of the 1960s, the only memory task on which monkeys wi th bi la teral h ippocampal lesions had been found to be impaired was the delayed-alternation task (Pr ibram, W i l s o n , & Conners , 1962; O r b a c h et al. , 1960). Experiments with rats. In the 1960s, rats with bilateral hippocampal lesions were shown to have difficulty performing successive-brightness-discrimination (Kimble , 1963), maze-learning (Kaada, Rasmussen, & K v e i m , 1961; K i m b l e , 1%3), and passive avoidance ( K i m b l e , 1963) tasks; however, they had no difficulty performing a simultaneous-brightness-discrimination task ( K i m b l e , 1963). T h e performance of rats on an active avoidance task was improved by bi lateral h ippocampal lesions (Isaacson, Douglas , & M o o r e , 1961). Why early attempts to model brain-damage-produced amnesia failed The lack of consistent learning and memory impairments i n laboratory animals wi th bilateral h ippocampal lesions led many investigators to conclude that different neural systems subserve memory i n humans than i n nonhumans. However , others questioned the adequacy of the tasks that had been used to assess the effects of hippocampal lesions on memory i n laboratory animals (e.g., D r a c h m a n & O m m a y a , 1964; O r b a c h et al. , 1960). Unfortunately, during the 1950s and 1960s human brain-damage-produced amnesia itself was not wel l understood, and therefore it was not clear which type of memory tasks would be suitable for model l ing it. B y the late 1960s, the study of H . M . and other amnesic patients was beginning to shed Ught on the nature of the spared memory abilities i n human brain-damage-produced amnesia. Some of these insights suggested possible explanations for why lesions that impair memory i n humans might not impai r the performance of certain memory tasks by laboratory animals. In particular, theories were proposed to explain why bilateral hippocampal lesions d id not disrupt the performance of the two kinds of tasks that had been used to assess their amnesic effects: (1) visual-discrimination tasks and (2) delay tasks. Visual-discrimination tasks. Evidence that procedural-learning abiUties are spared i n humans with brain-damage-produced amnesia provided an explanation of why bi lateral h ippocampal damage did not impai r visual discrimination learning in laboratory animals. O n visual-discrimination tasks, it is not necessary for the subject to remember what happened on individual trials because the information that is relevant to successful performance is presented repeatedly over many trials. Tha t is, visual discr iminat ion tasks are not tests of expUcit memory (or declarative memory) ; rather, they are similar to the tests of implici t memory (or procedural memory) that amnesic patients are capable of learning, and therefore, they are unlikely to be sensitive to brain-damage-produced amnesia. \"In everyday human learning there are no strict counterparts o f discr iminat ion tasks i n which the same piece of information is presented ad nauseam. In humans, motor learning perhaps comes closest to this...\" (Iversen, 1976; p . l 6 ) Thus , i n M i s h k i n ' s (1954) and Orbach et al.'s (1960) early monkey experiments, bi lateral h ippocampal lesions may have produced amnesia that went undetected by the visual-discrimination tasks that they used to assess it (see Z o l a - M o r g a n et al., 1982). Delay tasks. Evidence that short-term memory abiUties are largely spared i n patients wi th brain-damage-produced amnesia provided an explanation of why bilateral h ippocampal damage d id not impai r the performance of laboratory animals i n early experiments that employed various delay tasks. F o r example, the longest retention delays that C o r r e l l and Scoville (1965) and D r a c h m a n and O m m a y a (1964) used to test delayed matching-to-sample performance i n monkeys wi th h ippocampal damage were only 5 and 12 s, respectively, wel l within the range of short-term memory. Similar ly , M i s h k i n (1954), M i s h k i n and P r ib r am (1954), and Orbach et a l . (I960), who d id not observe delayed-response deficits i n monkeys wi th bilateral hippocampal lesions, a l l used delays of 10 s or less. B i la te ra l h ippocampal damage may have produced amnesia i n a l l of these delayed matching-to-sample and delayed-response experiments that went undetected because of the retention delays were too brief. In support of this interpretation, Z o l a - M o r g a n and Squire (1985a) recently found that monkeys wi th bi la teral lesions of the hippocampus and amygdala displayed a marked impairment on a delayed-response task at delays of 15 and 30 s, but performed normally when the delay was only 8 s. 3. THE PREEMINENT MONKEY MODELS OF BRAIN-DAMAGE-PRODUCED AMNESM B y the end of the 1960s, it was becoming apparent that the development of an imal models of brain-damage-produced amnesia would first require the development of memory tests for laboratory animals that amnesic patients would be expected to fail . In the m i d 1970s, such a task was developed -the nonrecurring-items delayed nonmatching-to-sample ( D N M S ) task by M i s h k i n and Delacour (1975). The D N M S task would later become a key component of monkey models o f brain-damage-produced amnesia. This section describes the monkey D N M S task and the effects o f medial- temporal-lobe and medial-diencephaUc lesions on D N M S performance i n monkeys. The nonrecurring-items delayed nonmatching-to-sample (DNMS) task O n each trial of the D N M S task, a sample object is presented to the subject. Then , fol lowing a delay, dur ing which the sample object is hidden from view, it is presented again, a long wi th an unfamiliar object. The subject is rewarded for selecting the unfamiUar object f rom this pair. M o n k e y s quickly learn to perform this task wi th few errors at retention delays of only a few seconds, and once they have done so, their performance is almost as good as that of humans at delays of up to several minutes. Media l - temporal - lobe and medial-diencephahc lesions disrupt D N M S performance i n both human and nonhuman primates (Squire, 1987). W h e n it was developed, the D N M S task was unique among memory tests for laboratory animals because it resembled human recognition-memory tests. In recognit ion-memory tests, subjects must decide which test items have been previously encountered. In typical human recogni t ion-memory tests, a Ust of items is presented to the subject (e.g., a Hst of pictures, words, or nonsense syllables). Later , a test list, which includes items from the first list and some new items, is presented, and the subject must identify the items that appeared in the first list (e.g., Postman, 1950; Strong, 1912). T h e D N M S task is identical i n principle to such tests of human-recognition memory; on each tr ial , the subject must distinguish between an object that was presented earher and one that was not. In most tests o f animal memory, the subjects learn stimulus-reward or response-reward associations, and the same stimuU recur over many trials. These tasks fall into two categories: (1) reference-memory tasks and (2) working-memory tasks. Reference-memory tasks are those i n which the relations among stimuli , responses, and reward remain constant over trials; working-memory tasks are those i n which the relations among stimuU, responses, and reward change from tr ial to t r ia l (O l ton et a l , 1979). T h e D N M S task is a working-memory task, but it differs from most other working-memory tasks for laboratory animals i n one important respect: It involves nonspatial stimuU. In this respect, it resembles most human memory tasks. P r io r to the development of the monkey D N M S task, laboratory animals had frequently been shown to have difficulty performing nonspatial working-memory tasks, such as delayed matching-to-sample, at delays of more than a few seconds. Th i s led to the view that nonhuman animals are poor at remembering nonspatial information (see Iverson, 1976; Nissen, R iesen , & NowUs, 1938). However , it is now clear that the poor performance of laboratory animals i n early studies of nonspatial working memory was a methodological artifact. In conventional nonspatial work ing-memory tasks, a small set of test stimuU are presented tr ial after t r ia l (e.g., A lex insky & Chapouthier , 1978). Accordingly , after a few trials, aU the test items are famiUar, and the recognit ion task i n effect becomes a recency-memory task ( M i s h k i n & Delacour , 1975) - o n each tr ia l , two famiUar objects are presented, and the subject must remember which of the two has been encoxmtered more recently. Labora tory animals appear to have difficulty making such recency discriminations after retention delays of more than a few seconds. The D N M S task is a test o f recognit ion memory because it makes use of nonrecurr ing items^; wi th nom-ecurring-items, the subject can solve the task by distinguishing between an object that it has seen before and one that it has not. Because the D N M S task is a test o f recognition memory that can be readily performed by norma l monkeys, it has provided 3 The term \"trial unique\" is typically used in place of \"nonrecurring-items\" i n reference to the monkey D N M S task. The advantage of the term \"nonrecurring-items\" is that a s imple antonym exists (i.e., \"recurring-items\"), which can be used to refer to versions of the task i n which the same stimuU are presented repeatedly over several trials. Throughout this dissertation, \" D N M S \" refers to the nonrecurring-items version of the delayed nonmatchaig-to-sample task. researchers wi th an appropriate test for determining whether medial- temporal- lobe and medial -diencephalic lesions cause amnesia in laboratory animals. The origins of the preeminent monkey models of brain-damage-produced amnesia The development of the monkey models of brain-damage-produced amnesia began with a serendipitous finding. M i s h k i n and Spiegler (cited i n M i s h k i n & Appenze l l e r , 1987) found that monkeys wi th bi lateral lesions of the amygdala had difficulty performing one-trial visual-discriminations, and they tr ied to accentuate the impairment by making bi lateral lesions that included both the hippocampus and the amygdala. Monkeys wi th amygdalo-hippocampal lesions were so severely impaired on the one-trial visual-discrimination task that the experimenters wondered whether they could remember the st imuli from one trial to the next. M i s h k i n tr ied to answer this question by assessing the effects of amygdalo-hippocampal lesions on D N M S ( M i s h k i n , 1978). H e found that monkeys wi th bilateral lesions to both the hippocampus and the amygdala were profoundly impai red on the D N M S task, whereas monkeys with bilateral lesions to the either the hippocampus or the amygdala alone were only mildly impaired. The severe impairment o f D N M S in monkeys wi th amygdalo-hippocampal lesions appeared to be a good animal mode l o f medial- temporal- lobe amnesia for two reasons: (1) because the brain damage i n monkeys wi th amygdalo-hippocampal lesions was similar to the bra in damage i n patients with medial-temporal-lobe amnesia, and (2) because accurate D N M S performance requires the kinds of memory functions that are impai red i n patients wi th medial-temporal-lobe amnesia. The effects of bilateral medial-temporal-lobe lesions on D N M S performance in monkeys Since M i s h k i n ' s demonstration that bilateral amygdalo-hippocampal lesions disrupt D N M S i n monkeys, several studies have been conducted to determine which medial- temporal-lobe structures must be damaged i n order to produce such a disruption. Fou r different hypotheses have received support: (1) the hippocampus, (2) the hippocampus and the amygdala, (3) the temporal stem, and (4) the rh ina l cortex. Hippocampus. There is controversy over whether lesions l imi ted to the hippocampus produce a recognit ion deficit i n monkeys that is comparable to the profound recognition deficit that is displayed by patients with medial-temporal-lobe amnesia. Mahut , Z o l a - M o r g a n , and M o s s (1982) and Z o l a -M o r g a n and Squire (1986) found severe D N M S deficits in monkeys wi th bi lateral h ippocampal lesions, whereas M i s h k i n (1978) and M u r r a y and M i s h k i n (1984,1986) found only m i l d D N M S deficits. Whe the r or not monkeys wi th lesions Umited to the hippocampus display severe D N M S deficits appears to depend upon whether or not they are trained on the D N M S task pr ior to surgery. I n experiments without presurgery training, bUateral hippocampal lesions have produced severe D N M S deficits (e.g. M a h u t , M o s s , & Z o l a - M o r g a n , 1981; Mahu t et al . , 1982; Z o l a - M o r g a n & Squire, 1986), whereas in experiments with presurgery training, bilateral h ippocampal lesions have produced only m i l d D N M S deficits (e.g., M i s h k i n , 1978; M u r r a y & M i s h k i n , 1984). Hippocampus-Amygdala. Bi la teral lesions of the hippocampus (Mahu t et al . , 1982; M i s h k i n , 1978; Z o l a - M o r g a n & Squire, 1985a; BachevaUer & M i s h k i n , 1989) or the amygdala ( M i s h k i n , 1978; M u r r a y & M i s h k i n , 1984) have been shown to produce m i l d D N M S deficits i n monkeys. C o m b i n e d bilateral lesions of both the hippocampus and the amygdala have been shown to produce a more severe impairment ~ one that is greater than would be expected from a simple summat ion of the effects of bi la teral lesions to either structure alone ( M i s h k i n , 1978; M u r r a y & M i s h k i n , 1984; Z o l a - M o r g a n & Squire, 1985a). The synergistic effect of bilateral lesions of the hippocampus and amygdala led to the proposal that these structures are critical links i n parallel neural circuits that are involved i n D N M S performance and that each cncuit can partially compensate for the loss of the other ( M i s h k i n , 1982; M u r r a y , 1990). These circuits are presumed to involve medial-diencephalic structmes. C o m b i n e d bi lateral damage to the pathways through which the hippocampus and amygdala communicate with medial-diencephalic structures \u2014 the fornix and the amygdalofugal pathway, respectively \u2014 caused a severe D N M S deficit i n monkeys, whereas bilateral damage to only one of these two pathways caused only a m i l d deficit (Bachevalier, Parkinson, & M i s h k i n , 1985). L i k e the memory impairments of patients wi th brain-damage-produced amnesia, the memory impairments of monkeys wi th medial-temporal-lobe lesions are not specific to a single sensory modahty. B i la te ra l hippocampal (Mahut et al., 1981) and amygdalo-hippocampal ( M u r r a y & M i s h k i n , 1984) lesions produced D N M S deficits in monkeys when either visual or tactual st imuU were used. Temporal stem. H o r e l (1978) pointed out that the surgical technique that was used for removing medial- temporal- lobe structures i n H . M . and other patients must also have damaged the temporal stem - a fiber pathway that links the temporal cortex wi th the amygdala and the orbi ta l frontal cortex ( C i r i l l o , H o r e l , & George, 1989). H e hypothesized that bilateral temporal-stem damage, not bilateral h ippocampal damage, was the cause of medial-temporal-lobe amnesia. H o r e l based his hypothesis on the fol lowing three Unes of evidence: (1) Bi la tera l temporal-stem lesions i n monkeys disrupt the performance of some memory tasks that are not disrupted by bUateral h ippocampal lesions (e.g., visual-discrimination tasks; H o r e l & Misantone, 1976). (2) The positive correlat ion between the extent of h ippocampal damage and the severity of memory deficits, which was reported by M d n e r (1974), can be accounted for by the fact that more extensive hippocampal resection is l ike ly to damage more of the temporal stem. (3) Tempora l cortex lesions i n monkeys produce symptoms similar to those produced i n humans by medial-temporal-lobe lesions. In fact, G o l and Faib ish (1967) reported that memory deficits in a group of amnesic patients were more highly correlated wi th the extent of temporal neocort ical damage than with the extent of hippocampal damage. C o u l d temporal-stem damage account for the severe D N M S impairment that M i s h k i n (1978) and M u r r a y and M i s h k i n (1984,1986) observed i n monkeys wi th bilateral amygdalo-hippocampal lesions? In a test of the temporal-stem and hippocampus-amygdala hypotheses of medial- temporal- lobe amnesia, Z o l a - M o r g a n , Squire, and M i s h k i n (1982) found that monkeys wi th bi la teral amygdalo-hippocampal lesions displayed a severe D N M S impairment, whereas monkeys wi th bi lateral temporal-stem lesions were unimpaired. However , C i r i l l o et al . (1989) recently found that temporal-stem lesions placed anterior to those made by Z o l a - M o r g a n et al . (1982) produced a severe impairment on a delayed matching-to-sample task i n monkeys. The anterior temporal-stem lesions made by C i r i l l o et a l . (1989) damaged the portions of the temporal stem that would be expected to be damaged i n monkeys and humans with amygdalo-hippocampal lesions; the more posterior temporal-stem lesions made by Z o l a - M o r g a n et a l . (1982) d id not. Rhinal cortex. Recent evidence suggests that the D N M S deficits that are produced i n monkeys by lesions of the hippocampus and amygdala may result from incidental damage to the rh ina l cortex (i.e., the per irhinal and entorhinal cortices). In monkeys, hippocampal and amygdalar lesions are usually made by aspiration, and thus, portions of the overlying cortex must first be removed to gain access to the hippocampus and amygdala. H ippocampa l lesions typically include the parahippocampal gyrus and the posterior half of the entorhinal cortex; amygdalar lesions typically include the p i r i fo rm and periamygdaloid cortex, the anterior half of the entorhinal cortex, and, i n some cases, the per irhinal cortex (Murray , i n press). In Mishk in ' s (1978) and M u r r a y and M i s h k i n ' s (1984) studies, the D N M S deficits may have been more severe i n monkeys with amygdalo-hippocampal lesions than i n monkeys wi th separate h ippocampal or amygdalar lesions because only the amygdalo-hippocampal lesions resulted i n the removal of the entire entorhinal cortex. M o n k e y s wi th bi lateral lesions of the entorhinal and perirhinal cortex (Murray , Bachevalier, & M i s h k i n , 1989) or of the per irhinal cortex (Meunier , Mur ray , BachevaUer, & M i s h k i n , 1990; Z o l a -M o r g a n et al. , 1989c) have been shown to be severely impaired on the D N M S task. Consistent wi th these findings, H o r e l , Pytko-Joiner, Voytko , and Salsbury (1987) observed a severe impairment i n the delayed matching-to-sample performance of monkeys following either ablat ion or reversible cooling lesions of the inferior- temporal gyrus, which includes much of the per i rhinal cortex. Z o l a - M o r g a n et a l . (1989a) have argued that the recent evidence that impHcates the rh ina l cortex i n the performance of D N M S suggests that the amygdala does not contribute to recognit ion memory. The i r conclusion is based on the following findings: (1) Radiofrequency lesions of the amygdala that spare the surrounding cortex do not produce an impairment i n D N M S ( Z o l a - M o r g a n et al. , 1989a), (2) aspiration lesions of the amygdala that do include the surrounding cortex produce an impairment i n D N M S ( M i s h k i n , 1978; M u r r a y & M i s h k i n , 1984), and (3) radiofrequency lesions of the amygdala do not exacerbate the D N M S deficits that have been produced by hippocampal aspiration ( Z o l a - M o r g a n et al. , 1989a). However , i n view of the report that bilateral ablation of the amygdala plus rhinal cortex produces a more severe D N M S impairment than does bilateral ablation of the rh ina l cortex alone ( M u r r a y & M i s h k i n , 1986), it is premature to conclude that the amygdala plays no role whatsoever i n recognit ion memory. Recently, M u r r a y ( in press) has suggested that the additional D N M S deficit that is produced when amygdalar lesions are added to rhinal lesions may result from damage to per i rh inal efferent fibers that course just lateral to the amygdala. This idea is consistent wi th C i r i l l o et al.'s (1989) fmding that anterior temporal-stem lesions cause a greater memory impairment than do posterior temporal-stem lesions. Poster ior lesions disconnect only the posterior half of the entorhinal cortex; anterior lesions disconnect most of the entorhinal cortex. Mur ray ' s hypothesis is consistent with Hore l ' s temporal-stem hypothesis o f medial- temporal- lobe amnesia - bo th predict that D N M S performance wi l l be impaired fol lowing lesions of the temporal-lobe white matter. However , Murray ' s hypothesis differs from Hore l ' s i n attributing the D N M S deficit fol lowing temporal-stem lesions to disruption of the projections from per i rhinal cortex, whereas H o r e l ' s hypothesis attributes the deficit to disruption of the projections from the anterior inferotemporal cortex. Synopsis. N o n e of the four interpretations of the effects of medial- temporal- lobe lesions on D N M S i n monkeys - hippocampus, amygdala-hippocampus, temporal stem, or rh ina l cortex ~ can be discounted on the basis o f existing evidence. The difficulty i n deciding among them is that they are not mutual ly exclusive: Damage to any one of the four areas may be sufficient to produce deficits imder the appropriate conditions, whereas maximal deficits may be produced through combined damage to some subset o f the four. H ippocampa l lesions appear to cause deficits that are more severe i n unpretrained monkeys than i n pretrained monkeys. This finding suggests that the hippocampus plays a role i n learning how to perform wel l on the D N M S task, but that it is less important once high levels o f D N M S performance have been achieved. Nevertheless, a model of the severe memory deficits suffered by medial- temporal- lobe amnesics, such as H . M . , should involve pretrained monkeys because H . M . is deficient i n the performance of everyday memory functions that he had naturally acquired and overlearned in the years pr ior to his surgery. Current evidence suggests those medial-temporal-lobe lesions that are most l ike ly to cause severe D N M S deficits i n pretrained monkeys involve the rhinal cortex. Severe D N M S deficits have been found i n pretrained monkeys with amygdalo-hippocampal lesions, but i n a l l cases, the amygdalo-hippocampal lesions have also involved rhinal cortex damage. Conversely, damage to the rhinal cortex that spares the hippocampus and amygdala can produce D N M S deficits i n both pretrained and unpretrained monkeys. A l t h o u g h addit ional damage to the hippocampus ( Z o l a - M o r g a n et al . , 1989c) or amygdala ( M u r r a y & M i s h k i n , 1986) can exacerbate the deficits produced by rh ina l cortex damage, this may occur because the amygdalar and hippocampal lesions disconnect remaining rh ina l cortex f rom other areas (Murray , in press). The anterior temporal stem lesions that have been shown to produce delayed matching-to-sample deficits in monkeys (C i r i l l o et al., 1989) disrupt many of the efferent connections of the rhinal cortex (Murray , in press). The effects of medial-diencephalic lesions on D N M S performance in monkeys There have been only a few studies of the effects of medial-diencephalic lesions on D N M S performance i n monkeys. The mediodorsal thalamic nuclei and the mammil la ry bodies are the two most consistently and extensively damaged brain areas i n Korsakoff amnesics (Vic to r et al . , 1971). A g g l e t o n and his colleagues found that bilateral lesions of the mammil la ry bodies d id not produce D N M S deficits i n monkeys (Aggleton & M i s h k i n , 1985), but that bi lateral lesions of the mediodorsal thalamic nuclei and the adjacent anterior nuclear complex (Aggleton & M i s h k i n , 1983a) or o f the mediodorsal nuclei alone (Aggleton & M i s h k i n , 1983b; Z o l a - M o r g a n & Squire, 1985b) did . Korsakoff amnesics have been shown to display similar bnpairments on a D N M S task (Squire et al. , 1988) and on a nonrecurring-items delayed matching-to-sample task (Aggleton, N i c o l , Hus ton , & Fai rba i rn , 1988). G E N E R A L R A T I O N A L E This thesis constitutes the ini t ial stages of an attempt to develop rat models of brain-damage-produced amnesia that are directly comparable to the monkey models. Such rat models could contribute to the study of brain-damage-produced amnesia i n two important ways: (1) They could facilitate the conduct o f large-scale parametric experiments - the cost o f large-scale monkey research is prohibit ive for most researchers. (2) They could provide a broader comparative basis for drawing inferences about the anatomical bases of brain-damage-produced amnesia. The first purpose of this thesis was to design and develop a rat version of the monkey D N M S task. T h e second was to assess the comparabili ty of the rat D N M S task to the monkey D N M S task i n terms of (1) the rate at which it is learned, (2) the asymptotic level at which it is performed, (3) the memory abilities that it taps, and (4) the brain structures that it engages. First , I designed a D N M S task for rats that resembles the monkey D N M S task. Then , I used it i n three experiments. Exper iment 1 assessed the abiUty of intact rats to learn and perform the D N M S task. Exper iment 2 examined the effects of distraction during the retention delay on the D N M S of intact rats - distraction interferes with D N M S performance i n monkeys ( Z o l a - M o r g a n & Squire, 1985a; Z o l a - M o r g a n , Squire, & A m a r a l , 1989a, 1989b) and humans (Squire et al. , 1988). Exper iment 3 examined the effects of separate and combined bilateral lesions of the hippocampus and the amygdala on the D N M S performance of rats. A l t h o u g h the pr imary purpose of this thesis was to assess the comparabil i ty of the rat and monkey D N M S tasks, the present experiments accomplished more. Because, together, they suggested that the rat D N M S task is a val id test of object recognition, each experiment also provided informat ion about the mnemonic abilities of rats or the neural bases of their abiUty to recognize objects. Exper iment 1 provided comparative data on the object recognition of rats; Exper iment 2 provided data on the effects of distraction on object recognition i n rats; Exper iment 3 provided evidence concerning the role of the hippocampus and amygdala i n object recognition i n rats. A N E W D N M S T A S K F O R RATS In preparat ion for designing a rat version of the monkey D N M S task, I analyzed two D N M S tasks that had already been developed for rats (Aggleton, 1985: Rothblat & Hayes, 1987). B o t h tasks bear some resemblance to the monkey D N M S task, but both differ from it i n major respects. M y general strategy was to incorporate features of the existing rat D N M S tasks that are part o f the monkey D N M S task and to eliminate features from them that either introduce a cognitive demand that is not present i n the monkey D N M S task or make the task particularly difficult for rats. Accord ing ly , the first two sections in this chapter describe the two previous rat D N M S tasks and discuss their strengths and weaknesses. T h e third section outlines the specific considerations that guided the design of my rat D N M S task. The fourth and final section of this chapter describes the task. 1. AGGLETON'S Y-MAZE DNMS TASK Aggle ton (1985) developed a Y - m a z e D N M S task for rats. In the Y - m a z e D N M S task, 40 distinctive goal boxes serve as the test st imuli . O n each trial, the rat is first enclosed for 20 s i n a sample goal box that is attached to one of the arms of a Y maze. Then , the rat is removed f rom the sample box, the sample box is removed from the Y maze and replaced by a featureless goal box, and the rat is placed i n the featureless goal box. Fol lowing a delay, the door to the featureless goal box is opened to provide the rat with access to the other two arms of the Y maze. A t the end of both of these arms are distinctive goal boxes; one of them matches the sample goal box, and the other one, an imfamil iar goal box, does not. T h e rat is rewarded i f it enters the unfamiliar goal box. That goal box then serves as the sample for the next tr ial . Aggleton's (1985) study was the first to demonstrate that rats can perform wel l on a D N M S task ~ once they had mastered the task at short delays, Aggle ton 's rats averaged approximately 8 0 % correct at delays of 120 s. A l t h o u g h the rat Y - m a z e D N M S task resembles the monkey D N M S task, it differs from it i n key respects. The fol lowing are five of them: 1. The rat Y - m a z e D N M S task uses a much smaller set of test s t imuli than the monkey D N M S task ~ 40 distinctive goal boxes i n the Y - m a z e D N M S task versus several hundred different objects i n the monkey D N M S task. Thus, repeated exposure to indiv idual StimuU is more frequent i n the rat Y - m a z e D N M S task than i n the monkey D N M S task. 2. In the rat Y - m a z e D N M S task, the 20-s duration of exposure to the sample is controUed by the experimenter. In the monkey D N M S task, the duration of exposure to the sample is control led by the subjects ~ the monkeys can respond to the sample wi th whatever latency they choose, which is typicaUy within 2 or 3 s. 3. In the rat Y - m a z e D N M S task, the experimenter handles the subjects dur ing trials; i n the monkey D N M S task, the subjects are not handled. 4. In the rat Y - m a z e D N M S task, the unfamiUar goal box on one tr ial serves as the sample goal box on the next tr ial ; in the monkey D N M S task, two new objects serve as the sample and unfamiUar items on each tr ial . 5. In the monkey D N M S task, the subjects must physically manipulate the test stimuU; in the rat Y - m a z e D N M S , the subjects enter the test stimuU. These differences between the rat Y - m a z e D N M S task and the monkey D N M S task might make the cognitive demands of the two tasks substantiaUy different, and thus, they make it difficult to generalize between them. 2. ROTHBLAT AND HAYES'S DNMS TASK The rat D N M S task that was developed by Rothblat and Hayes (1987) involves a straight runway wi th a start area at one end and a goal area at the other. The goal area contains three recessed food wells. The start area is separated from the runway by a door. O n each trial , the experimenter baits the central food we l l and positions a sample object over it. Then , the door is opened, and the rat runs down the runway to the goal area, where it displaces the sample object from the food we l l and retrieves the food. Then , the experimenter closes the door and returns the rat to the start area for the retention delay. D u r i n g the delay, the experimenter places the sample object and an unfamiliar object over the lateral food wells. A t the end of the delay, the door is opened, and the rat runs to the goal area. If it displaces the unfamiliar object, it is rewarded; i f it displaces the sample object, it is not rewarded. Different sample and unfamiUar objects are used on each trial wi thin a session. E a c h rat i n Rothbla t and Hayes's (1987) study received 12 trials per day at delays of 10 s. Over the first 10 sessions, their scores increased at a statistically significant, but unimpressive, rate; first-session scores averaged 68%, and tenth-session scores averaged only 75%. Af te r they reached the cri terion of at least 7 5 % correct over three consecutive sessions, each rat received addit ional sessions at delays of 120 s. The i r mean score at delays of 120 s was 63%. Monkeys typicaUy score between 8 5 % and 9 5 % at delays of 120 s (e.g., BachevaUer et al. , 1985; M i s h k i n , 1978; M u r r a y & M i s h k i n , 1984). Several features of Rothblat and Hayes's rat D N M S task make it more s imilar to the monkey D N M S task than is Aggleton 's Y - m a z e D N M S task. The following are three of them: 1. Rothbla t and Hayes 's task uses objects as test stimuU. 2. In Rothbla t and Hayes 's task, the subjects physicaUy manipulate the test objects; they displace the unfamUiar object from a food weU to obtain food. 3. Rothbla t and Hayes 's task employs a large pool o f test s t imuli (approximately 250 objects). A l t h o u g h Rothbla t and Hayes's task bears a closer resemblance to the monkey D N M S task than does Aggleton 's Y - m a z e task, it differs from the monkey D N M S task i n key respects. T h e following are three of them: 1. T h e subjects i n Rothblat and Hayes's task are handled during the retention delay. 2. Rats do not easily learn Rothblat and Hayes's D N M S task. 3. The asymptotic D N M S performance of rats on Rothblat and Hayes 's task does not compare favorably to that typically reported for monkeys. T h e slow learning i n Rothblat and Hayes's rats might have been partly due to the fact that in i t ia l D N M S sessions were conducted at retention delays of 10 s; shorter delays may have l ed to quicker learning. H a n d l i n g the rats during the retention delay might also have contributed to their poor performance. B e that as it may, the poor performance of Rothbla t and Hayes 's rats is a major shortcoming for two reasons. First, i f a rat D N M S task is to be considered to be comparable to the monkey D N M S task, then it is important that normal rats display similarly high baseUne levels of performance. Second, low baseline levels of performance make it difficult to demonstrate statistically significant deficits. 3. REQUIREMENTS OF AN EFFECTIVE RAT DNMS TASK M y first step i n designing a rat D N M S task that closely resembles the monkey D N M S task was to compi le the fol lowing list of desirable features, which was based on my analysis of Aggleton 's (1985) and Rothbla t and Hayes 's (1987) rat D N M S tasks and the monkey D N M S task: 1. The test s t imuli should be objects. 2. A large poo l of test objects should be used and two new objects should serve as the sample and novel objects on each trial within a session. 3. The rats should not be handled during sessions. 4. The operant response should be the displacement of an object from over a food we l l . 5. The durat ion of exposure to the sample object should be brief, and this durat ion should be control led by the subject, not by the experimenter. 6. It should be possible to train subjects at retention delays o f only a few seconds. Then , I designed a rat D N M S task that satisfied each of these requirements. 4. THE NEW RAT DNMS TASK This section describes the rat D N M S paradigm that I designed. The first subsection describes the apparatus, and the second outlines the general training procedure that was used i n each of the three experiments i n this thesis. The apparatus T h e apparatus, which was constructed of sheet a luminum (thickness = 0.127 cm) , was a straight runway that was mounted 70 cm above the floor (see Figure 1). The apparatus was 60 c m long, 20 c m wide, and 40 c m high. There were two identical goal areas, one at each end of the runway. E a c h of the goal areas was separated from the central starting area by an opaque guillotine door; bo th doors were located 30 c m from the nearest end wal l . B o t h goal areas had two recessed food wells, 3.5 c m i n diameter and 2.0 c m deep. The food wells were separated by a short divider wa l l (9 c m x 9 cm), which protruded at a 90-degree angle from the center of the end wal l . The food wells were centered 5 c m f rom the divider wal l , and 3 cm from the end wall . The sides of the goal areas were open to al low the experimenter to easily posit ion stimulus objects over the food wells and to quickly remove them. F o o d pellets (45 mg; Bio-Serv Inc., Frenchtown, N J ) were delivered to the food wells v ia funnels that were mounted on the outside of the apparatus and connected to the food wells wi th v inyl tubing. T h e test stimuU were 350 test objects of various shapes, textures, and colors, comparable to the \"junk\" objects that have been used i n the monkey D N M S paradigm (see M i s h k i n & Appenze l l e r , 1987). E a c h object was large enough to cover a food weU but small enough and Ught enough to be easily displaced by a rat. General training and testing procedures T h e general training procedure for the rat D N M S task comprised three phases: (1) habituation to the apparatus, (2) object-discrimination training, and (3) acquisition of D N M S . E a c h rat received no more than one session per day and no fewer than four sessions per week. The rats' body weights were reduced to 8 5 % of ad-lib values by l imit ing their daily ration of laboratory chow, and their weights were Figure 1. Tlie D N M S apparatus. (Pliotograpli by Jack Wong.) maintained throughout the experiment at a level that was 8 5 % of the typical weight of rats of the same age, sex, and strain that are maintained with continuous access to laboratory chow. Tra in ing commenced after the rats had been on the restricted feeding regimen for 14 days. They were housed individually wi th continuous access to water, and they were maintained on a 12:12 hr light-dark cycle, wi th light offset at 11:00 p.m. A l l testing occurred during the Ught phase of the light-dark cycle. The rats were not handled during a session once they had been placed i n the apparatus unless they urinated or defecated, m which case they were removed briefly so that the floor could be wiped clean. Habituation. T h e habituation phase consisted of six 20-min sessions. D u r i n g the first two sessions, the guil lotine doors remained open, and each of the four food weUs was baited wi th three or four food pellets. The wells were rebaited once the peUets i n a l l four wells had been consumed. O n the th i rd and fourth habituation days, the rats were shaped to run back and forth between the goal areas by alternately baiting a single weU i n each one. F o o d appeared equaUy often i n aU four weUs. T h e guiUotine doors remained open during these two sessions. The operat ion of the guiUotine doors was introduced during the fifth and sixth habituation sessions. The food wells were unbaited at the start of both of these sessions. W h e n the rat entered one of the goal areas, the experimenter lowered the door at the other end and baited one of the food weUs at that end with a single peUet. W h e n the rat approached to within about 3 c m of the lowered door, it was raised to provide the rat with access to the baited food weU. A s soon as the rat found the food, the far door was lowered, and one of the weUs behind it was baited. This door was opened as soon as the rat approached it. This cycle was repeated for the entire 20 m i n of the fifth and sixth habituation sessions. Object discrimination. Fo l lowing the six habituation sessions, each rat received four two-choice object-discrimination sessions; each session comprised 25 trials. These object-discrimination sessions were designed to accompUsh two goals: (1) to teach the rats to displace objects f rom over the food wells, and (2) to eliminate any side preferences, that is, preferences for either the right or left food wells (cf. Rothbla t & Hayes, 1987). F o r a particular subject, the same two objects served as the s t imuli for a l l 100 of its object-discr iminat ion trials. O n e of the objects was randomly designated the S+ (reward); the other object was designated the S- (no reward). T o begin each session, the rat was placed i n the center o f the apparatus. O n e door was open, and one was closed. The S + and S- were each posi t ioned over one of the food wells beh ind the closed door \u2014 the position of S + (left or right) varied from tr ial to t r ia l according to an irregular, but balanced, pattern. Then , the door was raised to expose the two objects. W h e n the rat approached and displaced an object, the far door was lowered behind it. If the rat displaced the S +, a food pellet was delivered to that food wel l ; i f it displaced the S-, no reward was delivered. Cor rec t ion was al lowed during the first object-discrimination session; i f the rat first chose S-, it was permit ted to then displace S + to obtain a reward before the experimenter removed the objects. D u r i n g the remaining object-discrimination sessions, correction was not a l lowed. A s soon as an object had been displaced, the experimenter removed the S + and S- and placed them over the wells at the other end of the apparatus i n preparation for the next tr ial . The duration of the intertr iai interval (i.e., the interval between the displacement of an object on one tr ial and the opening of the door to provide the rat wi th access to the objects on the next trial) varied, but it was typically 15 to 20 s. Acquisition of DNMS. F o r D N M S training, the pool of objects was divided into seven sets of 50 objects each. F o r each rat, a different set of objects was used on each consecutive session (i.e., a particular set was used, on average, once every seven sessions). Different pairs o f objects were used for each of the trials wi th in a session; one was randomly designated the sample, and the other the novel object. T o begin a D N M S session, the rat was placed i n the apparatus wi th the doors raised, and it was a l lowed to explore the apparatus for approximately 1 min . Then, the doors were lowered to enclose the rat i n the central starting area. Before each trial , a single food pellet was placed i n one of the four food wells, and the sample object was placed over it. The location of the pellet and the sample object varied according to an irregular, but balanced, schedule; they appeared at each of the four food wells wi th equal probabihty. Once the sample object was i n place, the novel object was placed over one of the food wells at the other end of the apparatus; its position, left or right, varied according to an irregular, but balanced, schedule. T o begin a trial , the experimenter raised the door to provide the rat wi th access to the sample object, which the rat approached and displaced from the food wel l . W h i l e the rat ate the food pellet, the experimenter removed the sample object and positioned it over the vacant food we l l at the other end of the apparatus. The other door was then raised, and the rat approached and displaced either the sample object or the novel object. A food pellet was delivered to the exposed food we l l i f the novel object was displaced; no pellet was delivered i f the sample object was displaced ~ rats were considered to have displaced an object only i f they moved it enough to expose the food wel l . A s the rat ate the pellet, the experimenter removed the objects and lowered the door farthest f rom the rat. W h e n the rat finished eating, it entered the central starting area and the experimenter lowered the other door to enclose it there. The next tr ial began as soon as the rat was enclosed i n the starting area, the new sample and novel objects were positioned, and the door was raised to provide the rat wi th access to the new sample object. M o s t intertriai intervals (i.e., the interval between the displacement of an object on one tr ial and the opening of the door to provide the rat with access to the sample on the next trial) were 30 to 40 s i n duration. If a particular rat was consistently slow to return to the starting area, it was occasionally rewarded wi th a food pellet as it entered the starting area. The rats were permit ted to make corrections dur ing the first two D N M S sessions, but not thereafter; i f the rat first chose the sample object, it was permit ted to displace the novel object to obtain a reward before the experimenter removed the objects. The retention interval, or delay, was the time between the removal o f the sample object and the raising of the second door to provide access to the sample and novel objects. The shortest delay that could be easily employed i n this D N M S task was approximately 4 s; this is the delay that was employed on each tr ial of the D N M S training phase. E X P E R I M E N T 1: D N M S P E R F O R M A N C E IN RATS The two ma in objectives of Experiment 1 were (1) to demonstrate that intact rats can learn my D N M S task and (2) to show that they can perform it over a wide range of retention delays. 1. METHODS Subjects The subjects were 14 experimentally naive, male Long-Evans rats (Charles River , Quebec) , 8 weeks o ld at the beginning of the experiment. D N M S training T h e rats were habituated and trained to perform the D N M S task as described i n the preceding section. F o r each subject, training continued at the 4-s delay unt i l it reached the cr i ter ion of at least 21 out o f 25 correct choices on two consecutive sessions, whereupon the delay was increased to 15 s. The delay was subsequently increased to 30 s, to 60 s, and to 120 s whenever a rat reattained the cri terion (two consecutive sessions of at least 21 correct trials) or completed eight sessions at a particular delay without reattaining the criterion. E a c h rat received four sessions at a 600-s delay after training at the 120-s delay was completed. 4 This experiment has been published (Mumby , P ine l , & W o o d , 1990). Measuring tlie retention function This phase of testing was designed to define each rat's retention function. It consisted of five mixed-delay D N M S sessions, each of which consisted of 25 trials. In each of these sessions, five trials were conducted at each of the following delays: 4 s, 15 s, 60 s, 120 s, and 600 s. These delays appeared i n the fol lowing order i n each session : 4 s, 15 s, 60 s, 120 s, 600 s, 600 s, 120 s, 60 s, 15 s, 4 s, 4 s, 15 s, and so on. 2. RESULTS A l l 14 rats progressed successfully through the habituation and object-discrimination phases of the training, and al l 14 learned the D N M S task. Once they had learned the task, they performed significantly better than chance at al l delays. Habituation B y the end of the final habituation session, a l l rats readily approached closed doors to gain access to food pellets on the other side. Object discrimination A l t h o u g h there was some init ial hesitation, a l l of the rats quickly learned to displace objects from food wells. O n the second object-discrimination session (i.e., the first session on which they were not permit ted to make corrections), the mean number of correct trials was 6 5 % , (ranging from 44% to 88%, SE = 3.30%), and on the fourth, and final, session the mean was 9 1 % (ranging from 7 2 % to 100%, SE = 2.08%). Acquisition of D N M S Figure 2 illustrates the performance of the D N M S task at each delay during the acquisit ion phase. Illustrated are the mean levels of performance on the first and last sessions at each delay. Performance dur ing the first training session at the 4-s delay was significantly above chance (M = 59%, f(13) = 4.14, p < .005, two-tailed; a two-tailed test was used because there is evidence that species differ i n their int ial propensities for selecting either the matching or the nonmatching stimulus. A t the 4-s delay, 13 of the 14 rats achieved cri ter ion within 16 sessions, not including the final two cr i ter ion sessions {M = 9.4 sessions or 235 trials). The 1 rat that d id not achieve cri ter ion at the 4-s delay scored as high as 8 8 % on some sessions, but was inconsistent; this rat was switched to the 15-s delay after 20 sessions. A s illustrated i n Figure 2, when rats were switched to delays longer than 15 s, their performance ini t ial ly declined and then improved over sessions at the new delay. The number of rats reattaining cr i ter ion within the max imum of eight sessions at the 15-, 30-, 60-, and 120-s delays was 8,12, 9, and 2, respectively. N o rats achieved the criterion within the four sessions that were administered at the 600-s delay. Retention functions Figure 3 illustrates the mean retention functions of the 14 rats; these were calculated f rom the rats' performance on the five mixed-delay D N M S sessions. It should be noted that the use of an appropriate log scale for the abscissa in Figure 3 would change the shape of the retention function. m a k i n g it more linear. However , I chose to use the present scale i n order to facilitate direct comparison wi th typical illustrations of monkeys' D N M S retention functions, many of which use the same delays and the same scale. The results o f a repeated measures analysis o f variance indicated that their performance declined significantly with increases i n the retention delay; repeated measures F(4, 52) = 75.72, p < .001. However , performance at the 600-s delay was st i l l significantly better than chance; f(13) = 2.77, p < .05, one-tailed. Performance at a l l delays was stable over the five sessions; that is, the mean scores on the first mixed-delay session were not significantly different f rom those on the fifth mbced-delay session. 3. DISCUSSION A l l of the rats i n this experiment learned the D N M S task. The rats' performance of the D N M S task was comparable to that commonly reported for monkeys in terms of both the rate at which they acquired the nonmatching rule at a br ief retention delay and their accuracy at longer delays. T h e rats required a mean of 235 trials to achieve the ini t ia l cri terion of 84% on two consecutive sessions, whereas, rhesus monkeys ( M i s h k i n & Delacour , 1975), cynomolgus monkeys (Aggle ton & M i s h k i n , 1983), and squirrel monkeys (Overman, M c L a i n , Ormsby, & Brooks , 1983) required means of 90,150, and 785 trials, respectively, to achieve a slightly more stringent cr i ter ion (e.g., at least 9 0 % correct on two consecutive sessions or at least 90 correct on 100 consecutive trials). D u r i n g the final mbced-delay test sessions, the rats i n the present experiment averaged 90%, 9 1 % , 8 1 % , and 7 7 % at delays of 4 ,15 , 60, and 120 s, respectively (see Figure 3). These levels of asymptotic performance compare favorably wi th the asymptotic levels observed in monkeys at comparable retention delays. T h e asymptotic scores of monkeys typically range between 90% and 100% at delays of about 10 s and between 8 5 % and 9 5 % at delays of 120 s (e.g., Aggle ton & M i s h k i n , 1983a, 1983b; M u r r a y & M i s h k i n , 1986). A t the 600-s Figure 2. Mean percent correct on the first and last session at each delay during acquisition training on D N M S . Performance improved between the first session and the last session at most delays (* p < ,05, ** p < .01, ** p < .001). Error bars show S E M s . Figure 3. Mean percent correct on mixed-delay D N M S sessions. Retention was statistically signiHcant at each delay (all ps < .01). Error bars show S E M s . delay, the rats scored 5 7 % correct; although this level of performance is significantly above chance, it is considerably lower than the 8 0 % that has been reported for monkeys ( Z o l a - M o r g a n , Squire, & A m a r a l , 1986). The better-than-chance first-session D N M S performance (i.e., 59%) conf i rmed previous observations i n both rats (Aggleton, 1985; Rothblat & Hayes, 1987) and monkeys ( M i s h k i n & Delacour , 1975). It presumably reflects the tendency of rats and monkeys to approach unfamiliar s t imul i (cf. Ennaceur & Delacour , 1988). The acquisition of the nonmatching rule was reflected i n the significant improvement m the rats' performance as the training sessions progressed. The improvement i n performance fol lowing the disruptive effect of lengthening the retention delay suggests that rats' D N M S abiUties continued to improve with additional training after they had reached the performance cr i ter ion at the 4-s delay. There are several cognitive or perceptual abiUties that may have continued to knprove; for example, it is possible that the rats gradually learned to avoid distraction for increasing periods o f t ime or that they became more efficient at encoding the physical attributes of the sample objects, or both. T o respond correctly, the rats had to learn the nonmatching principle and recognize the sample objects. They may have recognized the visual, tactual, or olfactory properties o f the sample objects, or they may have circumvented the mnemonic demands of the task by odor-marking the sample objects dur ing the sample phase of each tr ial . However , the foUowing observations suggest that they based their choices pr imari ly on their memory of the visual properties of the sample objects: (1) Rats rarely contacted objects without displacing them, which suggests that they were not responding on the basis of tactual differences between the test objects. (2) The rats frequently veered towards the correct object before reaching the goal area, which suggests that they were using visual cues. (3) D u r i n g a separate series of control tests, two identical objects were used as the sample on each tr ial , one dur ing the sample phase of the t r ial and the other dm-ing the choice phase of the same tr ial ; the rats performed as we l l on these trials as they d id on conventional trials i n which the same object served as the sample on both phases of the tr ial . This suggests that the rats were not performing the task by mark ing the sample objects. The apparatus has two notable features that may have accounted for the rats' excellent performance. First , the two separate goal areas permitted the experimenter to posi t ion the sample and novel objects before the start of a trial , so that ini t ia l training could be conducted at br ief retention delays (i.e., 4 s; cf. Rothbla t & Hayes, 1987). Second, the presence of a central starting area el iminated the need to handle the rats during the retention intervals (cf. Aggle ton , 1985; Rothbla t & Hayes, 1987); distraction during retention intervals has been shown to disrupt D N M S i n monkeys ( Z o l a - M o r g a n & Squire, 1985a; Z o l a - M o r g a n , Squire, & A m a r a l , 1989a, 1989b) and humans (Squire et al. , 1988). This experiment was the second to demonstrate high levels o f nonspatial work ing memory i n rats ~ the rats i n Aggleton 's (1985) Y - m a z e experiment performed almost as wel l . However , this experiment was the first to demonstrate impressive levels of object recognit ion i n rats. T h e potential utility o f this paradigm stems from the fact that it was expressly designed to m i m i c the widely studied monkey object-recognition D N M S task. E X P E R I M E N T 2: T H E E F F E C T S O F DISTRACTION O N D N M S IN RATS In order for the D N M S task to serve i n rat models of brain-damage-produced amnesia that are comparable to the monkey models of brain-damage-produced amnesia, it must be shown that rats solve the D N M S task using memory abiUties that are similar to those used by monkeys and humans. A comparative task analysis is a nonempirical way of determining that l ike l ihood. Such an analysis suggests that both the rat and monkey D N M S tasks could be solved using either of two different strategies - subjects could make their choices on the basis of (1) expUcit memory for the ini t ia l presentation of the sample object, or (2) the relative famiUarity of the two test objects. The former strategy would involve expUcit memory, whereas the latter strategy wou ld involve impUcit memory. The results of Exper iment 1 provide some empir ical evidence that rats, humans, and monkeys employ similar memory abiUties when performing the D N M S task; l ike monkeys and humans, rats perform the D N M S task better at br ief delays than at long delays. The purpose of Exper iment 2 was to further examine whether rats, monkeys, and humans employ similar memory abiUties when performing the D N M S task by assessing the effects on the D N M S of rats of another task manipula t ion (i.e., a task manipulat ion other than delay) that has been shown to influence the D N M S of humans and monkeys. Exper iment 2 assessed the effects of distraction during the retention delay on the D N M S performance of rats. In the late 1950s, B r o w n (1958) and Peterson and Peterson (1959) observed that humans display rap id forgetting of smal l amounts of verbal material i f they engage i n another cognitive task (e.g., counting backwards by threes) during the retention interval. M i l n e r (1972) and S idman et al . (1968) subsequently observed that introducing distraction during the retention delay produced a severe memory deficit i n patients wi th brain-damage-produced amnesia, even when no deficits were apparent i n the absence of distraction. There is controversy over whether distraction dur ing the retention delay has greater effects on the performance of amnesic patients than on the performance of normal subjects. Cermak , Butters, and More ines (1974) found that Korsakoff patients performed significantly worse than control subjects on the Brown-Peterson task, whereas Baddeley and War r ing ton (1970) found that they d id not. E t k i n (1972) observed that the performance of monkeys on a delayed matching-to-sample task was significantly better when the testing r o o m Ughts were extinguished during the retention delay than when they were not. H e concluded that these results reflected decreased retroactive interference from visual input dur ing the darkened retention delays. U s i n g an approach that was more s imilar to the Brown-Peterson paradigm, Z o l a - M o r g a n and Squire (1985) and Z o l a - M o r g a n et a l . (1989a, 1989b) observed that the D N M S of monkeys was poorer when distractor objects, which the monkeys could displace f rom food wells i n order to receive a food reward, were presented during the retention delay. Th i s distraction procedure was adapted for use with rats i n the present experiment. 1. METHODS The methods were identical to those of Experiment 1, except where otherwise noted. Subjects T h e subjects were 8 male Long-Evans rats, a l l of which had previously served as control rats i n other D N M S experiments - each had previously received between 1000 and 1700 D N M S trials. T w o of them had received bilateral sham lesions of the mediodorsal thalamus; electrodes had been lowered into the mediodorsal nuclei and withdrawn, without the passage of current. T w o of them had received a sham bi lateral hippocampectomy; a port ion of posterior parietal cortex and corpus cal losum overlying the dorsal hippocampus had been aspirated bilaterally, without damaging the hippocampus. O n e of them had been a control subject i n an ischemia experiment; hgatures had been placed around both carotid arteries, but they had not been constricted, and one of the femoral arteries had been caimulated but no b lood had been withdrawn. Three of them were intact control rats f rom Exper iment 3. N o n e of the 8 subjects had displayed any D N M S deficits pr ior to the commencement of Exper iment 2. Procedure E a c h rat received 10 D N M S sessions of 20 trials each. T h e delay was 60 s on each tr ial . A l l of the trials during odd-numbered sessions were ordinary D N M S trials (no-distraction trials), whereas a l l o f the trials during even-numbered sessions included a distraction task during the retention delay (distraction trials). O n each distraction trial , 20 s after the rat had displaced the sample object, the far door was raised to reveal a single distractor object over a baited wel l at the other end of the apparatus. A s the rat displaced the distractor object, the far door was lowered. Then , 20 s after the first distractor object had been revealed, the far door was raised to reveal another distractor object over a bai ted food wel l . A s the rat displaced the second distractor object, the far door was lowered beh ind it. Then , 20 s after the second distractor object had been revealed (i.e., 60 s after the rat had displaced the sample object), the door was raised to reveal the sample object and a novel object. A food pellet was delivered to the exposed food wel l i f the novel object was displaced. T h e two different distractor objects that were used on each tr ial were different than the distractor objects that were used on other trials within a session. Accordingly, 80 different objects - 20 samples, 20 novel objects, and 40 distractor objects - were used during each session. They were selected from the poo l of 350 objects. 2. RESULTS Figure 4 illustrates the mean scores for each of the 8 rats on the 100 distraction trials and on the 100 no-distraction trials. E a c h of the 8 rats obtained a lower average score on the distraction trials than on the no-distraction trials (p < .005). The i r mean score on distraction trials was significantly lower than on no-distraction trials; f(7) = 4.16, p < .01. The results of two repeated measures analyses of variance indicated that the scores on neither the distraction trials (F[4, 28] = 1.53,;? = .219) nor the no-distraction trials (F[4, 28] = 1.41, p = .255) changed significantly over the five sessions. The performance of the rats on no-distraction trials (M = 79%) was similar to that of the rats i n Exper iment 1, when they were tested at the same (i.e., 60-s) delay (M = 81%) . 3. DISCUSSION In this experiment, the D N M S performance of rats was disrupted by the inclusion of a distraction task during the retention delay. The present experiment was the first to assess the effects of an interpolated activity dur ing the delay on the D N M S performance of rats. M y observation of disruptive effects of distraction during the delay on the D N M S performance of rats is consistent wi th reports of s imilar findings in monkeys ( Z o l a - M o r g a n & Squire, 1985a; Z o l a - M o r g a n et al. , 1989a, 1989b) and humans (Squire et al. , 1988). Figure 4. Mean scores on no-distraction and distraction trials. Error bars show S E M s . Rat Rat Rat Rat Rat Rat Rat Rat 1 2 3 4 5 6 7 8 The rats' performance of the distraction task may have interfered with the mnemonic processing of the sample objects, the novel objects, or both. It is possible that the rats' D N M S performance was worse on distraction trials than on no-distraction trials because the novel procedure that was used on distraction trials confused them. T w o observations suggest that this was not the case: Firs t , the rats' D N M S performance on distraction trials d id not show any signs of improvement over the five sessions; i f the poor performance on distraction trials was caused by the novelty of the procedure used on those trials, then this effect should have lessened as the rats gained experience wi th it. Second, none of the rats showed any indication that they were confused on any of the distraction trials - a l l of them readily approached and displaced objects whenever they were revealed. It is l ikely that the distraction task that was used i n the present experiment and the distraction task that is used i n monkey D N M S experiments disrupted performance through more than distraction per se. Because the distractor st imuli were similar to the test s t imuli (i.e., they were objects of similar size), the performance of the distraction task may have disrupted mnemonic processing through proactive or retroactive interference. B e that as it may, the fact that the present findings paral le l those of monkey experiments that employed the same distraction task suggests that rats, monkeys, and hiunans employ s imilar memory abilities when performing the D N M S task. In humans, distraction during the retention delay disrupts performance of expUcit-memory tasks but has little or no effect on the performance of impUcit memory tasks ( G r a f & Schacter, 1987; S loman, Hayman , Ohta , L a w , & Tulving, 1988). Thus, the present fmding suggests that the memory abilities that rats use when performing the D N M S task resemble those that humans use when performing explici t-memory tasks, but not those that humans use when performing impUcit-memory tasks. E X P E R I M E N T 3: T H E E F F E C T S O F S E P A R A T E A N D C O M B I N E D L E S I O N S O F T H E H I P P O C A M P U S AND A M Y G D A L A O N D N M S IN RATS^ Bila te ra l damage to the media l temporal lobes has been shown to produce D N M S deficits i n both monkeys and humans, but there is controversy over which medial- temporal-lobe structures must be damaged to produce these deficits. Impaired D N M S has been found i n monkeys wi th h ippocampal ( M a h u t et al . , 1982; Z o l a - M o r g a n & Squire, 1985a; BachevaUer & M i s h k i n , 1989) or amygdalar ( M u r r a y & M i s h k i n , 1984) damage; however, in other studies, impairments i n the D N M S performance of monkeys have been observed only i f both the hippocampus and amygdala (e.g., M i s h k i n , 1978) or their pr imary efferents (BachevaUer et al., 1985) have been damaged. Recent evidence suggests that the impairments of D N M S following lesions of the hippocampus and amygdala i n monkeys may result f rom incidental cortical damage (Mur ray et a l , 1989; M u r r a y & M i s h k i n , 1986; Z o l a - M o r g a n et al. , 1989a, 1989c). Recently, the controversy over the role of the hippocampus and the amygdala i n recognit ion memory has extended to research on rats. O l ton and Feustle (1981) found impai red recurring-items D N M S i n rats with hippocampal lesions, whereas Aggle ton, Hunt , and Rawl ins (1986), Kesner (1991), and Rothbla t and K r o m e r (1991) found no D N M S impairment i n rats wi th h ippocampal lesions. Agg le ton , Bl indt , and Rawl ins (1989) found no impairment of D N M S fol lowing amygdala lesions, but they observed a substantial impairment foUowing amygdalo-hippocampal lesions; however, they emphasized that collateral pyriform-cortex damage may have contributed to the impairment displayed by the rats wi th amygdalo-hippocampal lesions. The amount of presurgery training on the D N M S task seems to influence the magnitude of postsurgery deficits. F o r example, hippocampal lesions in monkeys appear to have a less disruptive 5 This experiment is currently in press (Mumby, W o o d , & Pine l , in press) effect on their D N M S performance when they receive presurgery training (e.g., M i s h k i n , 1978; M u r r a y & M i s h k i n , 1984) than when they do not (e.g., Z o l a - M o r g a n & Squire, 1986). Presurgery training has advantages and disadvantages. O n one hand, postsurgery testing of subjects that have not had presurgery training is more l ikely to reveal an effect of the lesion. O n the other, presurgery training allows one to estabhsh stable baselines of performance i n individual subjects to which their postsurgery performance can be compared. M o r e importantly, presurgery training reduces postsurgery deficits that are due to impai red acquisition of the skills that are required for successful performance, rather than to impai red retention of the test objects, thus making postsurgery performance deficits more easy to interpret. The purpose of Exper iment 3 was to assess the effects of hippocampal , amygdalar, and combined amygdalo-hippocampal lesions on D N M S in rats. Rats were tested at retention delays of 4 ,15 , 60,120, and 600 s both before and after receiving either bilateral hippocampal , amygdalar, or amygdalo-hippocampal lesions. Af te r postsurgery testing, some of the rats wi th h ippocampal lesions and some of the rats with amygdalar lesions received an additional bi lateral lesion in order to give them combined amygdalo-hippocampal lesions; then, they were retested. Thus, some of the rats wi th amygdalo-hippocampal lesions received one-stage lesions, and others received two-stage lesions. I chose to make aspiration lesions of the hippocampus rather than neurotoxic lesions because the former method, while not as selective as the latter, enables more complete lesions of the hippocampal formation ~ i n my view, it was best to begin studying the effects o f h ippocampal lesions on the performance of the new D N M S task by making complete lesions. Moreove r , one of the major aims i n this study was to compare the effects of similar lesions i n rats and monkeys - h ippocampal lesions have been made by aspiration i n most of the comparable monkey studies. 1. METHODS T h e methods were identical to those of Experhnent 1, except where otherwise noted. Subjects The subjects were 22 experimentally naive, male Long-Evans rats that were between 8 and 12 weeks o ld at the beginning of training, and 7 of the rats that had served i n Exper iment 1. Habituation, object-discrimination, DNMS training, and measuring the presurgery retention function Habi tua t ion , object-discrimination training, D N M S training, and measurement o f the presurgery retention functions were conducted in the same way as i n Exper iment 1 ~ wi th two exceptions: (1) F o r D N M S training, the number of trials per session was 25 for the 7 subjects that were used i n Exper iment 1, but it was reduced to 20 for the other 22 subjects so that more rats could be run each day. The performance cri terion was 84% of the trials correct on two consecutive sessions (21 out o f 25 or 17 out of 20). (2) Some of the rats' received 5 mixed-delay sessions, whereas the others received 10 m k e d -delay sessions; although performance was stable over the 5 mked-delay sessions i n Exper iment 1,1 thought that it might improve i f the rats received more testing. The data from Exper iment 1 for the 7 rats that had served i n that experiment comprised their presurgery data for the present experiment. Surgery Fo l lowing presurgery testing, each rat received either bilateral aspiration lesions of the hippocampus (n = 11), electrolytic lesions of the amygdala (\u00ab =7), or combined amygdalo-hippocampal lesions (n =4), or it was assigned to a no-surgery control group (n=6). H i p p o c a m p a l aspiration required the removal of posterior parietal neocortex; a control group of rats wi th posterior parietal neocortex damage was not included because a pilot experiment had indicated that such lesions do not affect D N M S performance in pretrained rats. The 2 rats i n this pilot experiment were trained and tested i n the same way as those of the present experiment, except that the longest delay was 300 s instead of 600 s. The i r mean presurgery scores were 95%, 85%, 75%, 78%, and 78%, at delays of 4 ,15 , 60,120, and 300 s, respectively; their mean postsurgery scores were 92%, 8 5 % , 8 0 % , 82%, and 75%. A l l surgery was performed under pentobarbitol anesthesia (60 m g \/ k g ) . In preparat ion for h ippocampal aspiration, the scalp was incised and holes were dr i l led on each side of the skul l . E a c h hole extended from approximately 2 m m posterior to the coronal suture to 2 m m anterior to the l ambo id suture, and from 2 m m lateral to the sagittal suture to within 1 m m of the tempora l ridge. Then , the exposed dura mater was cut, and the underlying neocortex and white matter were aspirated with a glass Pasteur pipette to expose the dorsal hippocampus. Next, the dorsal hippocampus and part of the lateral hippocampus were aspirated and the cavity was filled with G e l f o a m ( U p j o h n C o . , D o n M i l l s , Ontar io) . The bilateral amygdalar lesions were made with a bipolar stainless steel wire electrode, which was insulated with Tef lon except for approximately 1 m m at its tip. The lesions of each amygdala were made at three sites; the following were the coordinates relative to bregma of the three sites: (1) A P -2.3, M L -4.5, D V -9.8; (2) A P -3.3, M L -4.5, D V -10.0; (3) A P -4.3, M L -4.5, D V -10.0. A t each site, 2 m A of current was passed for 20 s. T h e above procedures for hippocampal and amygdala surgery were combined for the rats that received one-stage amygdalo-hippocampal lesions. The amygdalar lesions were made first, then the h ippocampal lesions. Af te r surgery, the rats that received hippocampal or amygdalo-hippocampal lesions were placed under a heat lamp i n a recovery room for 1 day. Diazepam (10-15 m g \/ k g , IP) was administered as soon as they began to regain consciousness, and for the next 24 hr smaller doses were periodical ly administered to control convulsions that we have sometimes observed i n rats fol lowing hippocampal lesions. F e w convulsions were observed. The rats that received only amygdalar lesions were returned to their home cages immediately after surgery. A l l experimental rats were al lowed at least 14 days to recover f rom surgery before the commencement of postsurgery testing. Similarly, control rats were not tested for at least 14 days fol lowing their final presurgery mixed-delay session. D u r i n g the first 10 days each experimental and control rat had continuous access to food, after which they were returned to a restricted feeding regimen for at least 4 days before testing recommenced. Reacquisition of D N M S Fol lowing recovery, rats were tested on the D N M S task at 4-s delays unt i l they either reattained the original performance cri ter ion (i.e., at least 84% of the trials correct on two consecutive sessions) or completed 20 sessions without reattaining it. Postsurgery retention functions Next, each rat's postsurgery retention function was determined i n the same way that its presurgery retention function had been determmed - wi th either 5 or 10 mixed-delay sessions. Stage-two lesions and testing Some of the rats then received a second operation; 5 rats wi th amygdalar lesions received hippocampal lesions and 7 rats wi th hippocampal lesions received amygdala lesions. Fo l lowing recovery f rom their second lesion, these rats were treated and tested as they had been fol lowing their first lesion. 2. RESULTS The main result was that reacquisition of the D N M S task and postsurgery performance at delays of up to 120 s were unimpaired following either separate or combined bi lateral lesions of the hippocampus and amygdala. M i n o r , but statistically significant, impairments were observed i n a l l three experimental groups at delays of 600 s. Fou r rats wi th hippocampal lesions sustained inadvertent damage to entorhinal, peruhinal , or temporal association cortex; they were profoundly impai red at a l l delays. The results of the various phases of the experiment are described i n the fol lowing subsections. Habituation B y the end of the final habituation session, a l l rats readily approached closed doors to gain access to food pellets on the other side. Object-discrimination training A l l rats quickly learned the object discrimination task. O n Session 2, the first session on which they were not al lowed to correct their errors, scores averaged 6 8 % (ranging from 44 to 88%) , and on the final session, scores averaged 9 2 % (ranging from 76 to 100%). D N M S training O n the D N M S task, the rats required a mean of 280 trials (SE = 22.7) to reach the performance cr i ter ion at the 4-s delay. There were no significant differences i n the number of trials to cr i ter ion or i n the presurgery percent-correct scores between rats that received 20 trials per session and those that received 25 trials per session, so the data from all of the rats were pooled for the purpose of statistical analysis. M o s t rats' scores dropped each time that the delay was lengthened, and then recovered over subsequent sessions. Exc lud ing the first two sessions at each delay, the mean-percent-correct scores at each delay during acquisition training were almost identical to those obtained during the subsequent mked-de lay sessions. Presui^ery retention functions A s shown i n Figure 5, scores on the presurgery mixed-delay sessions decl ined significantly as the delay was lengthened (F[4,115] = 52.6,p < .001). There were no significant differences among the groups' presurgery scores at any delay. One-sample t tests revealed that the scores i n each group were significantly better than chance at a l l delays (allp& < .05; one-tailed). Performance at a l l delays was stable over the mixed-delay sessions regardless of whether rats received 5 or 10 of these sessions; that is, the mean scores from the fu-st mked-delay session were not significantly different f rom those from the final mixed-delay session. Histology Figure 6 shows the location and the extent of the largest and smallest amygdalo-hippocampal lesions. In several of the rats with two-stage amygdalo-hippocampal lesions, the amygdalar and hippocampal lesions overlapped to form a continuous lesion, thus making it impossible to determine the boundaries of the damage sustained during each surgery. Accordingly , F igure 6 also shows the damage sustained by one of the rats with hippocampal lesions that d id not subsequently receive amygdalar lesions and by one of the rats with amygdalar lesions that d id not subsequently receive h ippocampal lesions; the damage in these two rats was representative of that sustained by the other rats wi th h ippocampal and amygdalar lesions. M o s t of the hippocampal lesions included the entire dorsal hippocampus, most o f the lateral hippocampus, and a por t ion of the neocortex and corpus callosum that overiies the dorsal hippocampus. The ventral extent of most of the hippocampal lesions spared smal l portions of the Figure 5. The presurgery (left) and postsurgery (right) retention functions that were determined on mixed-delay sessions. Error bars show S E M s for presurgery scores. The S E M s for postsurgery scores were similar and ranged from 0.97 to 3.70. Delay in Seconds Delay In Seconds dentate gyrus and subiculmn. The caudal extent of these lesions varied i n the amount of presubiculum and parasubiculum that was removed. E a c h lesion extended rostrally to include the fimbria fornix, and two of them included slight unilateral damage to dorsal portions of the lateral septal nucleus. Some rats also sustained smal l unilateral lesions i n the caudate nucleus. S m a l l infarcts were also present i n the dorsal thalamus of some of the brains wi th hippocampal lesions. M o s t of these were unilateral and involved the habenular nuclei, lateral dorsal nucleus, lateral pulvinar nucleus, lateral geniculate, and media l geniculate nucleus; damage to these thalamic nuclei appeared to be unrelated to the behavioral results. Some of the brains wi th h ippocampal lesions also received part ial unilateral damage to the tectum. F o u r of the rats wi th hippocampal lesions sustained cort ical damage that was m u c h more extensive than i n the other rats m that group (see Figure 8). This damage was unintended and it included portions of entorhinal, perirhmal, and temporal association cortex (area Te2 ; Z i l l e s , 1985). Accord ing ly , the behavioral data from these 4 rats were excluded f rom the overall statistical analyses -they are dealt wi th separately at the end of this section. The extent o f damage to the thalamus and tectum i n these 4 rats was not unlike that found i n the other rats wi th h ippocampal lesions; however, one of them sustained large bilateral lesions of the lateral septal nucleus. T h e amygdala lesions varied i n the extent o f damage to specific nuclei , but the most consistent damage was to the media l two-thirds of the amygdaloid complex. N o specific amygdaloid nuclei were consistently spared. The caudal extent of some of the amygdala lesions included smal l portions of media l entorhinal cortex (see the largest amygdalo-hippocampal lesion i n Figure 6), but the presence of such damage d id not appear to be related to the behavioral results. Figure 6. Reconstructions of (a) tlie largest (grey) and smallest (black) amygdalo-hippocampal lesions, (b) the lesions from one of the rats with hippocampal lesions that did not subsequently receive amygdalar lesions, and (c) the lesions from one of the rats with amygdalar lesions that did not subsequently receive hippocampal lesions. The drawings were adapted from the atlas of Paxinos and Watson (1982). Reacquisition of D N M S O n the first few trials of the first postsurgery testing session, many of the lesioned rats were slow to approach the goal areas when the doors were raised, but by the end of that session, most o f them approached the goal areas without hesitation. E a c h rat readily displaced test objects from the food wells during the first postsurgery session and continued to do so thereafter. T h e mean number of postsurgery trials at the 4-s delay that was required to reattain the cr i ter ion by the no-surgery controls and the rats wi th hippocampal , amygdalcir, and amygdalo-hippocampal lesions was 66.7 (SE = 66.7), 53.6 (SE = 10.1), 40.7 (SE = 19.8), and 66.2 (SE = 22.4), respectively. N o n e of the differences among these means was statistically significant (\/'13,32] < 1). Postsurgery retention functions T h e mean postsurgery scores for each group on the mixed-delay sessions are illustrated i n Figure 5. Analys i s o f variance indicated that the postsurgery performance of rats wi th one-stage amygdalo-h ippocampal lesions was similar to that of rats with two-stage amygdalo-hippocampal lesions on mixed-delay sessions (F[l ,14] < 1). Therefore, the data from al l 16 rats wi th amygdalo-hippocampal lesions ~ 4 wi th one-stage lesions and 12 with two-stage lesions - were pooled for the purposes of statisrical analysis and the presentation of results in Figure 5. There were no statistically significant differences between the presurgery and postsurgery scores of any of the groups at any delay. F o r multiple plaimed comparisons between the groups' postsurgery scores at each delay, the crit ical confidence level was .01. Relative to the performance of the no-surgery control rats, postsurgery scores were significantly lower i n all three experimental groups at the 600-s delay; h ippocampal lesions t(ll) = 3.16, amygdalar lesions t(ll) = 3.34, and amygdalo-hippocampal lesions t(20) = 3.20, allps < .01). These differences occurred because the scores i n each experimental group at the 600-s delay were slightly lower after surgery than before, while the scores m the no-surgery cont ro l group were slightly higher after surgery than before. One-sample t tests revealed that the postsurgery scores at the 600-s delay in each of the three experimental groups were st i l l significantly better than chance (hippocampal, f(7) = 3.41; amygdalar, t(7) = 2.92; amygdalo-hippocampal , t{16) = 2.49; allps > .05; one-tailed). There were no statistically significant differences among the four groups at any of the delays less than 600 s. The 4 rats wi th hippocampal lesions that also sustained unintended damage to tempora l association, per irhinal , and lateral entorhinal cortex displayed profound postsurgery impairments . O n e of them (E3) averaged 80% correct over the last 5 postsurgery sessions at the 4-s delay (i.e.. Session 15 to 20) but could not achieve the criterion, and another one (E8) required 120 postsurgery sessions to reach the cri terion, which was more than was required by any of the hippocampal- lesioned rats without damage to these areas of cortex. The presurgery and postsurgery retention functions for these 2 rats are illustrated i n Figure 7, and reconstructions of their lesions are shown i n Figiu-e 8. B o t h of them displayed postsurgery deficits at a l l delays. The other 2 rats wi th unintended damage to entorhinal, per i rhinal , and temporal association cortex averaged less than 6 5 % correct over the last 5 postsurgery sessions at the 4-s delay, and thus they were not subsequently tested on mixed-delay sessions; their lesions are illustrated i n Figure 8. Figure 7. Presurgery (filled circles) and postsurgery (open circles) retention functions for 2 rats (E3 & E8) that sustained inadvertent damage to entorhinal, perirhinal, and temporal association cortex. The data are from mixed-delay sessions. Figure 8. Lesions in 4 rats with hippocampal lesions that sustained inadvertent damage to entorhinal, perirhinal, and temporal association cortex. Rat E8 Rat E3 3. DISCUSSION A l l rats i n this e x p e r \u00f9 n e n t received extensive presurgery training on the D N M S task. Then, the experimental rats received lesions of either the hippocampus, the amygdala, or both. A l l three groups displayed only m i l d postsurgery D N M S deficits; they were significantly impai red at delays of 600 s, but not at delays of 120 s or less. T h e observation of only minor D N M S deficits i n rats fol lowing bi lateral lesions of the hippocampus alone or the amygdala alone are consistent with previous reports i n bo th monkeys and rats. M i s h k i n (1978) and M u r r a y and M i s h k i n (1984,1986) observed only m i l d D N M S deficits i n monkeys wi th hippocampal or amygdalar lesions; Rothblat & K r o m e r (1991) observed no D N M S deficits i n rats wi th hippocampal lesions at delays of 30 s; Aggle ton et a l . (1986) observed no D N M S deficits i n rats wi th hippocampal lesions or amygdalar lesions at delays of 60 s; Kesner (1991) observed no D N M S deficits i n rats wi th hippocampal lesions on a D N M S task that was mode l l ed after the one used i n the present experiment. In contrast, the present finding of only m i l d D N M S deficits following combined amygdalo-hippocampal lesions is inconsistent wi th many previous reports. P rofound D N M S impairments have been reported following amygdalo-hippocampal lesions both i n monkeys (e.g., M i s h k i n , 1978; Z o l a - M o r g a n & Squire, 1985a) and i n rats (Aggleton et a l , 1989). The results of recent studies with monkeys suggest a way of reconcil ing the present results wi th the numerous reports of major impairments o f D N M S following combined lesions of the hippocampus and amygdala. They suggest that amygdalo-hippocampal lesions do not produce profound impairments on D N M S unless there is collateral damage to adjacent cortex. In primates, amygdalo-hippocampal lesions are usually created by aspiration, and thus they include damage to portions of the parahippocampal gyrus and the entorhinal cortex, and, i n some cases, the per i rh inal cortex ( M u r r a y & M i s h k i n , 1986). Impaired D N M S has been found in monkeys fol lowing bilateral lesions restricted to these cortical areas (Meunie r et al., 1990; M u r r a y et a l , 1989; Z o l a - M o r g a n et al. , 1989c). Fur thermore , monkeys wi th combined amygdalo-hippocampal lesions that spare the cortex surrounding the amygdala (i.e., entorhinal, perirhinal, and periamygdaloid cortices) perform no worse on a D N M S task than monkeys with lesions of the hippocampus and parahippocampal cortex ( Z o l a -M o r g a n et al. , 1989a). Thus, it is possible that the present rats wi th amygdalo-hippocampal lesions were only mi ld ly impai red on D N M S because associated cortical areas were spared \u2014 the dorsal approach that is used to aspirate the hippocampus in rats damages the posterior parietal association cortex ( K o l b , 1990), and it spares the entorhinal cortex and the perirhinal cortex. In the present experiment, 4 of the rats with hippocampal lesions sustained unintended damage to various amounts of entorhinal and perirhinal cortex. They were the only subjects that displayed severe postsurgery D N M S deficits. It is possible that the damage to entorhinal and per i rh ina l cortex was responsible for these deficits, but each of them also sustained damage to other structures that were not damaged in most of the rats with hippocampal lesions. M o s t notably, aU 4 of these rats received damage to the lateral geniculate nucleus. A l though this damage was not complete, it raises the possibility that visual impairment may have been responsible for their deficits. A n o t h e r possibility is that the D N M S deficits of these 4 rats resulted f rom damage to the temporal association cortex, which they a l l sustained. K o l b , Burhman , and M c D o n a l d (1989) found that rats wi th lesions of the temporal association cortex were unable to learn a Y - m a z e matching-to-sample task, whereas rats with lesions of the posterior parietal cortex displayed norma l learning. In monkeys, lesions in the homologous brain region \u2014 the inferotemporal cortex \u2014 produce impairments on visual discriminat ion tasks and impairments across all delays of a delayed matching-to-sample task ( H o r e l et al . , 1987). In the present experiment, the 2 rats with perirhinal, lateral entorhinal, and temporal association cortex damage that were tested on mixed-delay sessions both displayed postsurgery deficits across a l l delays. Accordingly , the possibility cannot be ruled out that their postsurgery D N M S impairments were due to perceptual rather than memory deficits, or to damage outside of the per i rhinal and entorhinal cortex. F o r example, these rats a l l sustained extensive bi la teral damage to the cingulum. St i l l , it seems unlikely that their hippocampal or amygdalar damage was a cri t ical factor i n thek impairments because their hippocampal and amygdalar damage was no more extensive than that sustained by the other rats, which displayed only m i l d impairments. There is a second f ac to r - i n addition to the lack of collateral damage to cr i t ical areas of c o r t e x -that could have contributed to the good D N M S performance of the present subjects wi th amygdalo-hippocampal lesions. This second factor is the extensive presurgery training that they received. In monkeys, impairments on D N M S tasks following hippocampal lesions tend to be greater in subjects that receive no presurgery training (e.g., BachevaHer & M i s h k i n , 1989; M a h u t et al . , 1982; Z o l a -M o r g a n & Squire, 1986) than i n subjects that do (e.g., Bachevalier et al . , 1985; M i s h k i n , 1978; M u r r a y & M i s h k i n , 1986). In the recent study of Rothblat and K r o m e r (1991), h ippocampal lesions d id not impai r D N M S performance i n rats that had received presurgery training. It is possible that the extensive presurgery training m the present experiment ( M = 1211 presurgery trials; range = 825 to 1500) made the rats' D N M S performance relatively insensitive to large hippocampal lesions. But , such an explanation cannot explain why Aggle ton et al . (1986) found no impairment on D N M S i n rats wi th h ippocampal lesions that d id not receive presurgery training. Thus, it is possible that amygdalar, h ippocampal , or amygdalo-hippocampal lesions would cause impairments i n rats wi th no presurgery training i n the D N M S paradigm that we used i n this study, but there is no direct evidence to support this view. A l t h o u g h the present findings suggest that lesions that are Umited to the hippocampus and amygdala do not cause severe impairments of D N M S , the statistically significant impairments at the 600-s delay suggest that D N M S performance is sensitive to such lesions i f the task is made sufficiently difficult. In support of this notion, other studies have found that m i l d D N M S impairments i n monkeys with hippocampal or amygdala lesions can be accentuated by requir ing them to remember lists o f several sample items at one time (e.g., Mahu t et al., 1982; M i s h k i n , 1978; M u r r a y & M i s h k i n , 1986). Similar ly , Raffaele and O l t o n (1988) found impaired D N M S in rats wi th h ippocampal lesions when only two StimuU were used throughout testing, thus increasing the degree of proactive interference over that which is present when different st imuli are used on each tr ial . Jagielo, Nonneman , Isaac, and Jackson-Smith (1990) reported a similar result using a two-stimuU matching-to-sample procedure. Overa l l , the findings from this experiment strongly suggest that object recognit ion i n rats does not require an intact hippocampus or amygdala, at least not when it is assessed by D N M S performance i n pretrained rats. There were indications that D N M S performance is severely impai red i n rats by damage to adjacent cortex, but this possibility requires systematic verification. G E N E R A L DISCUSSION This thesis constitutes the first stage i n the development of a rat mode l o f brain-damage-produced amnesia. Its first objective was to develop a D N M S task for rats: one that resembles the classic monkey D N M S task, one that can be readily learned by rats at short delays, and one that can be performed by rats, once learned, wi th a high degree of accuracy. This objective was clearly met. Firs t , I designed a task that closely resembles the monkey D N M S task. Then, i n Exper iment 1, rats learned this D N M S task at a rate not substantially slower than the rate at which D N M S is learned by monkeys, and once they learned the task, they performed almost as wel l as monkeys at delays of up to 2 m i n . This was the fu-st demonstration that rats can perform a nonrecurring-items object-recognition task at delays of more than a few seconds. T h e second objective of this thesis was to provide a prel iminary assessment of the potential of my rat D N M S task to serve as a component of rat models of human brain-damage-produced amnesia. The first two sections of this Genera l Discussion focus on this issue. The third section describes ongoing research that is ut i l iz ing my rat D N M S task. The final section summarizes the m a i n findings and conclusions of this thesis. 1. THE CORRESPONDENCE BETWEEN THE RAT DNMS TASK AND THE MONKEY DNMS TASK The development of monkey models of brain-damage-produced amnesia has revealed much about the anatomical bases of human memory and amnesia - see reviews by Squu-e (1987) and M u r r a y (1990). O n e reason for their success is that they appear to be isomorphic models; that is, they m i m i c both the underlying cause of human brain-damage-produced amnesia and its symptoms . F o r example, the D N M S deficits that are displayed by monkeys with bilateral amygdalo-hippocampal lesions are isomorphic wi th hmnan medial-temporal-lobe amnesia in the sense that (1) the b ra in damage i n monkeys wi th amygdalo-hippocampal lesions is similar to the bra in damage i n patients wi th medial-temporal- lobe amnesia, and (2) the accurate D N M S performance requires the kinds of memory functions that are impai red i n patients with medial-temporal-lobe amnesia. In fact, humans wi th medial- temporal- lobe amnesia display deficits on a D N M S task that is virtually identical to the classic monkey D N M S task (Squire et al. , 1988). In view of the success that monkey models o f brain-damage-produced amnesia have had, my strategy for developing a rat model of human brain-damage-produced amnesia was to dupUcate key features of the monkey models. Because the D N M S task is one of these key features, I chose to begin my attempt to develop a rat model of brain-damage-produced amnesia by developing a rat D N M S task. I began by assuming that the most useful rat model of brain-damage-produced amnesia would be one that includes a D N M S task that corresponds to the monkey D N M S task i n three general respects. (1) The test stimuU, the response requirements, and other key aspects of the pro tocol should be similar . Such similari t ies reduce the number of possible interpretations that could be made for differences i n the effects of brain damage on the D N M S of rats, monkeys, and humans. (2) The asymptotic D N M S performance of rats should be comparable to that of monkeys and humans over a wide range of delay durations. In order to interpret differences m D N M S deficits among species, it is important that their baseline levels of performance be similar. In addition, low baseUne levels of performance make it difficult to demonstrate statistically significant deficits. (3) The D N M S of rats must involve memory abiUties that are similar to those involved in the D N M S of monkeys and humans, 6 A n isomorphic model is one that simulates both the symptoms of the disorder and their underlying cause (cf. Kornetsky, 1977). that is, those involved in object recognition. A correspondence between the rat and monkey D N M S tasks i n terms of these three criteria would al low a broader comparative basis for studying the effects of bra in damage on memory ~ one that includes rats as wel l as monkeys and humans. The fol lowing three subsections summarize the evidence that my rat D N M S task is similar to the monkey D N M S task i n terms of the aforementioned three criteria. The general protocol of the rat D N M S task resembles that of the monkey D N M S task M y rat D N M S task mimics the monkey D N M S task in terms of the test s t imuli , the response requirements, and other key aspects of the protocol. A a large poo l of objects serve as the test s t imuli and two different objects are used on each trial within a session; the operant response is the displacement of the correct object to gain access to a food wel l ; the duration of exposure to the sample object is br ief and is controlled by the subject; it is possible to train rats at delays of only a few seconds and then to test them at a wide range of delays; and the rats are not handled dur ing sessions. The D N M S performance of rats is comparable to that of monkeys In Exper iment 1, the D N M S performance of rats compared favorably to that commonly reported for monkeys i n terms of both the rate at which they learned the task and their asymptotic performance levels at delays of up to 2 min . The rats required a mean of 235 trials to achieve the ini t ia l cr i ter ion of 84% on two consecutive sessions, whereas rhesus monkeys ( M i s h k i n & Delacour , 1975), cynomolgus monkeys (Aggleton & M i s h k i n , 1983), and squirrel monkeys (Overman et al. , 1983) required a mean of 90,150, and 785 trials, respectively, to achieve a slightly more stringent cr i ter ion (e.g., at least 9 0 % correct on two consecutive sessions or at least 90 correct on 100 consecutive trials). D u r i n g the final mixed-delay test sessions, the rats i n Exper iment 1 averaged 90%, 9 1 % , 8 1 % , and 7 7 % at delays of 4, 15, 60, and 120 s, respectively. The asymptotic scores of monkeys typically range between 9 0 % and 100% at delays of about 10 s and between 8 5 % and 9 5 % at delays of 120 s (e.g., Agg le ton & M i s h k i n , 1983a, 1983b; M u r r a y & M i s h k m , 1986). The D N M S performance of rats, humans, and monkeys appears to involve similar memory abilities The results o f Experiments 1 and 2 support the view that rats, monkeys, and humans employ s imilar memory abilities when performing the D N M S task. T w o variables - the durat ion of the delay (Exper iment 1) and presence of distraction during the delay (Experiment 2) ~ affected the D N M S of rats i n the same way that they affect the D N M S of monkeys and humans. The D N M S performance of rats decl ined when either the duration of the delay was increased or when a distraction task occurred during the delay (cf. Squire, 1987; Squire et al., 1988). These findings suggest, but do not prove, that similar forgetting processes occur in rats and primates during the delay. The observation of a disruptive effect o f distraction during the delay suggests that the D N M S task involves expUcit memory i n rats, as it does i n humans and monkeys. Distract ion during the retention delay has been shown to disrupt the performance of humans on expUcit-memory tests but not impUcit-memory tests ( G r a f & Schacter, 1987; S loman et al., 1988). 2. EVIDENCE FOR CORRESPONDENCE BETWEEN THE NEURAL SYSTEMS THAT UNDERLIE OBIECTRECOGNITION IN RATS, MONKEYS, AND HUMANS In order for rat models of brain-damage-produced amnesia to be isomorphic wi th human brain-damage-produced amnesia and with monkey models of brain-damage-produced amnesia, it is essential that s imilar neural systems mediate similar mnemonic functions i n rats, monkeys, and humans. F o r example, object recognit ion should be mediated by the same structures i n a l l three species. Accord ing ly , i f the D N M S task is a valid test of object recognition i n rats, as it appears to be in monkeys and humans, the D N M S performance of a l l three species should be sensitive to damage i n the same bra in areas. The results of Exper iment 3 suggest that the neural systems that are involved i n the D N M S of rats are s imilar to those involved in the D N M S of monkeys i n at least two ways: (1) Nei ther the hippocampus nor the amygdala appear to play a critical role i n the D N M S of pretrained rats or monkeys. The observation i n Experiment 3 of only mi ld D N M S deficits i n pretrained rats fol lowing bi la teral h ippocampal or amygdalar damage is consistent with reports of only m i l d D N M S deficits i n pretrained monkeys ( M i s h k i n , 1978; M u r r a y & M i s h k i n , 1984,1986) and rats (Aggle ton et al. , 1986; Kesner , 1991; Rothblat & K r o m e r , 1991) with bilateral damage to these structures. There have been reports of more severe D N M S deficits fol lowing hippocampal damage m monkeys (e.g., M a h u t et al. , 1982; Squire & Z o l a - M o r g a n , 1985a; Z o l a - M o r g a n & Squu-e, 1986); however, i n every one of those studies, the monkeys had not received presurgery training. It has not yet been determined whether h ippocampal lesions produce severe D N M S deficits i n rats without presurgery training. (2) The entorhinal cortex and perirhinal cortex appear to be critically involved i n the D N M S of pretrained rats and monkeys. M y observation i n Experiment 3 of severe D N M S deficits i n 4 rats wi th bi lateral entorhinal or per i rhinal damage is consistent with reports of severe D N M S deficits i n monkeys wi th bi lateral damage to these cortical areas (Meunie r et al., 1990; M u r r a y et al. , 1989; Z o l a - M o r g a n et al. , 1989c). It has not yet been determined whether lesions l imited to the entorhinal and per i rhinal cortex produce severe D N M S deficits in rats. The 4 rats in Exper iment 3 that displayed severe D N M S deficits fol lowing entorhinal and per i rh inal cortex damage had also received damage to portions of temporal association cortex, and therefore, it is possible that this damage contributed to their D N M S deficits. This , too, would be consistent wi th reports that bilateral lesions of the homologous cortical region i n monkeys ~ the inferotemporal cortex ~ produce matching-to-sample deficits ( H o r e l et al . , 1987). A t first glance, my observation i n Experiment 3 of only m i l d D N M S impairments i n rats wi th bi la teral amygdalo-hippocampal lesions appccirs to be inconsistent wi th reports of severe D N M S deficits i n monkeys wi th bilateral amygdalo-hippocampal lesions (e.g., M i s h k i n , 1978; Z o l a - M o r g a n & Squire, 1985a). However , this inconsistency may reflect differences i n the topography of telencephaUc structures i n rats and monkeys, rather than differences in their functions. In monkeys, amygdalo-hippocampal surgery usually involves the removal of most of the entorhinal cortex and parahippocampal cortex, and in some cases, parts of the per i rhinal cortex. It has been proposed that the severe D N M S deficits following bilateral amygdalo-hippocampal lesions i n monkeys are produced by this cort ical damage (Murray , in press). In the rats that received amygdalo-hippocampal lesions i n Exper iment 3, these cort ical areas were largely spared, and therefore, this could explain why their D N M S deficits were no worse following amygdalo-hippocampal lesions than fol lowing separate lesions of either structure. Th i s illustrates one advantage of being able to address questions about the neural bases of brain-damage-produced amnesia in both rats and monkeys - because many of the topographical relations among brain structures are different m rats and monkeys, the damage to structures other than the target structure that typically occurs during surgery may also be different m the two species. 3. OTHER DNMS EXPERIMENTS IN RATS A l t h o u g h the present findings suggest that my rat D N M S task has potential to serve in isomorphic rat models of human brain-damage-produced amnesia, they do not, by themselves, provide sufficient validation of such models. Accordingly , I have initiated several experiments to test the validity of the rat D N M S task as a component of isomorphic rat models of human brain-damage-produced amnesia and to use the models to answer questions about brain-damage-produced amnesia. These experiments are currently i n progress; however, several of them are complete enough to warrant br ie f discussion here. T h e first two subsections describe these experiments. The first subsection describes results that suggest that the rat D N M S task and monkey D N M S task employ similar memory abilities. T h e second subsection describes results that suggest that there is continuity among rats, monkeys, and humans i n terms of the neuroanatomy of their memory systems. The third and final subsection describes my current efforts to develop a battery of tests for use i n rat models of brain-damage-produced amnesia. Do the rat and monkey D N M S tasks involve similar memory abilities? T h e findings f rom Experiments 1 and 2 suggest that rats perform the D N M S task using memory abilities that are similar to those used by monkeys and humans. Task analyses also suggest that similar memory abilities are involved i n the performance of the rat and monkey D N M S tasks. Converging Unes of evidence f rom the foUowing experiments support this conclusion. DNMS with lists. In humans, the abiUty to later recognize items from a studied Ust o f items decreases as the number of items in the Ust increases (e.g.. Strong, 1912). A similar effect o f list length on D N M S and matching-to-sample performance has been reported i n monkeys (e.g., BachevaUer et al. , 1985; Gaffan, 1974; M a h u t et al., 1982; M i s h k i n , 1978; M u r r a y & M i s h k i n , 1984,1986). In this version of the D N M S task, a sequence of different sample objects is presented to the monkey. Af t e r a delay. each sample object f rom the list is paired wi th a different novel object. T h e D N M S of monkeys is negatively correlated wi th the number of sample objects i n the list. I adapted the D N M S - w i t h - l i s t s procedure for use wi th rats. O n each tr ial , a sequence of sample objects is presented to the rat at 20-s intervals. D u r i n g the test phase, which begins 20 s after the presentation of the last object i n the sequence, the sample objects are presented again i n the same sequence, and each one is paired with a difference novel object; the rat is rewarded for selecting the novel object of each pair. I have tested 5 rats on the D N M S task wi th lists of 3, 5, and 7 objects. Consistent wi th fmdings i n monkeys, their mean scores decUned substantially as the length of the list increased. T h e means were 72%, 6 5 % , and 62%, respectively. DNMS in anosmic rats. Monkeys appear to perform the D N M S task using visual cues. However , it is not as clear what sensory modality or modahties rats rely on when performing the D N M S task. In the Discuss ion of Exper iment 1,1 described some incidental observations that suggested that rats were also using visual cues when performing the D N M S task. O n e way to test whether rats use a particular sensory modality, or subset o f modali t ies , to solve the D N M S task is to make it impossible for them to use alternative modalit ies. In an ongoing experiment that is ut i l iz ing this approach, I have tested the D N M S performance of 4 anosmic rats. A l l o f the rats had received extensive D N M S training and testing pr ior to the induct ion o f anosmia, which was accomplished by flushing their nares wi th zinc sulphate; 2 of them were intact rats cind had served as controls i n other D N M S experiments, and 2 of them had bilateral amygdalo-hippocampal lesions and had served i n Exper iment 3. The posttreatment retention functions for these four rats were simUar to their pretreatment retention functions. Thus, olfaction does not appear to play a cr i t ica l role i n the D N M S performance of rats. Choice-response latency. The main dependent measure i n most studies of memory in laboratory animals is the accuracy of responding, and studies of D N M S are no exception. However , i n many contemporary studies of human memory, response latency is measured as we l l as accuracy, thus providing addit ional insights into the nature of the information-processing operations that underUe performance. In studies of recognition memory in humans, manipulations that increase the difficulty of the task tend to increase the subject's response latencies (e.g., Sternberg, 1966). Response latencies have not been measured i n studies of D N M S in monkeys, but it is believed that the D N M S task is a test of recognit ion i n monkeys. If the rat D N M S task involves cognitive processes that are s imilar to those involved i n human recognition tests, then similar relations among task manipulations, response accuracy, and response latency should occur i n both species. The prel iminary findings of an ongoing experiment suggest that response latency may be a useful dependent measure in studies of D N M S in rats. I videotaped several hundred of the D N M S trials of 3 rats at delays ranging from 4 s to 300 s. F o r each trial, I measured both response latency and response accuracy. Response latency - the amount of t ime between when the rat's snout first entered the goal area and when the selected test object began to move - was measured to the nearest .05 s. F o r each of the rats, mean response latency increased, and mean percent correct decreased, as the duration of the delay increased. These prel iminary findings are consistent wi th the hypothesis that the D N M S task is a test o f recognit ion i n rats, as it is i n humans and monkeys. Do similar neural systems underlie DNMS performance in rats and monkeys? A l t h o u g h the results of Experiment 3 suggest that there are some similarit ies in the neural systems that underUe the D N M S performance of rats, monkeys, and humans, many questions about the extent of this similarity remain. I have initiated experiments that are a imed at answering some of those questions. These experiments take two general experimental approaches: (1) L e s i o n experiments ask whether surgical lesions affect the D N M S of rats in the same way that simlar lesions affect the D N M S of monkeys and humans. (2) Experiments designed to assess the etiological vaUdity of the D N M S task ask whether nonsurgical treatments that produce amnesia in humans affect the D N M S of rats. Et io logica l ly val id models are often used to delineate the pathology and to elucidate the pathogenesis of a disorder (Rid ley & Baker , 1991). Lesion experiments The fol lowing experiments are being conducted to determine whether rats w i l l display D N M S deficits fol lowing surgical lesions that cause D N M S deficits i n monkeys and humans. Rhinal cortex lesions. The results of Experiment 3 suggested that damage to lateral entorhinal and per i rh inal cortex may cause severe D N M S deficits. However , i n that experiment, the 4 rats with damage i n these cort ical areas also had extensive bilateral hippocampal damage. Thus , any single contr ibution that the cort ical damage might have made to their deficits was unclear. I have begun an experiment to determine the effects of bilateral entorhinal and per i rh inal cortex damage on D N M S i n rats. Rats are receiving bilateral entorhinal and per i rhinal cortex lesions either alone or i n combinat ion with bilateral amygdalectomy or bilateral hippocampectomy. T h e training and testing procedures are identical to those of Experiment 3, so a direct comparison of the results with those of Exper iment 3 w i l l be possible. So far, I have tested 3 rats with bi lateral entorhinal and per i rh ina l cortex lesions, 2 rats with similar cortical lesions combined with bi la teral amygdalectomy, and 2 rats wi th similar cortical lesions combined with bilateral hippocampectomy. A n o t h e r 6 rats are currently undergoing presurgery D N M S training. So far, the two main results have been that (1) the rats i n a l l three experimental groups required considerably more postsurgery trials to reattain the cr i ter ion at 4-s delays than d id the rats i n any of the groups i n Exper iment 3, and (2) their postsurgery scores were substantially lower than those of the control rats in Exper iment 3 at delays of 120 s and 600 s, but not at shorter delays. These findings suggest that bi lateral lesions of the entorhinal and perirhinal cortex produce D N M S deficits i n rats, as they do in monkeys. It is not yet clear whether or not the effects of bi lateral entorhinal and perirhinal cortex damage are accentuated by the presence of addit ional damage to the hippocampus or amygdala. Hippocampal lesions in rats without presurgery training. A s mentioned i n the Discuss ion of Exper iment 3, the D N M S deficits that are displayed by monkeys wi th bilateral h ippocampal lesions are less severe i f they have had presurgery training than i f they have not (Murray , 1990). Consistent with these findings i n monkeys (e.g., M i s h k i n , 1978; M u r r a y & M i s h k i n , 1984,1986), the results of Exper iment 3 demonstrated that bilateral hippocampal lesions produce only m i l d D N M S deficits in rats that receive presurgery training. I have conducted a pilot experiment to determine whether the effects o f bilateral hippocampal lesions are greater in rats that have not had presurgery training. Rats wi th bilateral h ippocampal lesions ( \u00ab = 4 ) and rats with control lesions of posterior parietal cortex (n = 5) were trained on the D N M S task at 4-s delays unti l they reached the cr i ter ion of at least 17 correct trials out of 20 on two consecutive sessions. Af ter reaching the cri terion, each rat received 6 sessions at delays of 15 s and 6 sessions at delays of 60 s. The rats wi th h ippocampal lesions acquired the task at a normal rate, and they subsequently performed normally at delays of 15 s and 60 s. These findings suggest that bilateral hippocampal lesions do not produce severe D N M S deficits i n rats that have not had presurgery training, and thus, they appear to be inconsistent wi th the fmdings from monkey experiments (e.g., M a h u t et al., 1982; Squire & Z o l a - M o r g a n , 1985a; Z o l a - M o r g a n & Squire, 1986). However , they are consistent with Aggle ton et al.'s (1986) observation i n rats of normal acquistion of the Y - m a z e D N M S task and subsequent normal performance at delays of up to 60 s i n rats wi th bi lateral h ippocampal lesions. This experunent must be repUcated wi th more rats and wi th delays of longer than 60 s before it can be concluded that bilateral hippocampal lesions have similar effects on the D N M S of rats both wi th and without presurgery training. Medial-diencephalic lesions. The mediodorsal thalamic nucleus and the mammi l l a ry bodies are the two most consistently and extensively damaged brain areas i n Korsakof f amnesics (Vic to r et al . , 1971). B i la te ra l lesions of the mammil lary bodies have failed to produce D N M S deficits i n monkeys (Aggle ton & M i s h k i n , 1985) and rats (Aggleton, Hunt , & Shaw, 1990), but bi lateral lesions of the mediodorsal thalamic nucleus have produced D N M S deficits i n monkeys (Aggleton & M i s h k i n , 1983a, 1983b; Z o l a - M o r g a n & Squire, 1985b). Korsakoff amnesics display similar impairments on a D N M S task (Squire et al . , 1988) and on a nomecurring-items delayed matching-to-sample task (Aggleton et a l , 1988). I have init iated a series of experiments to determine the effects of bilateral mediodorsa l nucleus and mammi l la ry body lesions on the D N M S of rats. So far, I have tested only rats wi th bilateral mediodorsa l nucleus lesions. Relat ive to rats with sham lesions, rats wi th electrolytic bi lateral mediodorsal nucleus lesions displayed D N M S deficits whether they had received presurgery training or not. These results suggest that mediodorsal nucleus damage produces severe D N M S deficits i n rats, as it does i n humans and monkeys. The next stage in this study wi l l be to examine the effects of mammil la ry-body lesions and combined lesions of the mammil la ry bodies and mediodorsa l nucleus. Experiments designed to assess the etiological validity of the rat DNMS task T h e fol lowing experiments were designed to determine whether rats w i l l display D N M S deficits fo l lowmg nonsurgical treatments that simulate known etiological factors in human brain-damage-produced amnesia. Thiamine deficiency. A considerable amount of evidence Unks Korsakof f amnesia i n humans to chronic thiamine deficiency (Butterworth, 1989). In laboratory animals, a per iod of thiamine deficiency can produce subsequent impairments on memory tasks, including impaired acquisi t ion o f D N M S in monkeys (Wit t & Go ldman-Rak i c , 1983) and impaired spatial and nonspatial recurring-items delayed nonmatching-to-sample i n rats (Kno th & M a i r , 1991; M a i r , Ande r son , Langlais , & M c E n t e e , 1988). In an ongoing experiment that is ut i l izing my rat D N M S task , rats have been exposed to a per iod of thiamine deprivation lasting between 12 and 14 days, during which they were maintained on a thiamine-free diet and given daily injections of the antithiamine agent pyri thiamine. F o l l o w i n g recovery from this thiamine deprivation, the experimental rats that had received pretreatment training required more trials to reattain the criterion at delays o f 4 s than d id untreated control rats, and once they d id so, they displayed deficits at delays of 15 s, 30 s, 60 s, and 120 s. Exper imenta l rats that had not received pretreatment training displayed acquisition deficits as wel l as subsequent retention deficits. Pre lmmary histological findings indicated that al l of the experimental rats had damage to mid l ine thalamic regions, but no apparent damage to either the mammil lary bodies or the hippocampus. Chronic alcohol consumption. A l c o h o l consumption reduces the absorption of thiamine from the gastrointestinal tract (Butterworth, 1989) and it interferes with thiamine metaboUsm i n the bra in 7 M i k e M a n a , L i s a Kalynchuck, and John Pine l have been major collaborators i n this study. ( R i n d i , 1989). There is also evidence that chronic alcohol exposure can cause memory impairments i n humans (e.g., Osca r -Berman & Z o l a - M o r g a n , 1980a, 1980b) and laboratory animals (e.g., Beracochea & Jaffard, 1985), even i n the presence of a diet replete with thiamine. D o rats that chronical ly consume substantial amounts of alcohol alcohol exhibit neuropathology and memory deficits i f they are maintained on a thiamine-replete diet? I used a schedule-induced-polydipsia paradigm to induce experimental rats to dr ink an average of approximately 2.5 g of ethanol per day over a 156-day period; control rats were rendered polydipsic for water, but they received no ethanol. Throughout the experiment, a l l of the rats were fed nutritionally-balanced laboratory rat chow. W h e n later trained on the D N M S task, there were no clear differences between experimental and control rats in terms of either the rate at which they learned the task or their subsequent performance at delays of 15 s and 60 s. The brains of the rats that served i n this pilot experiment have not yet been studied. Transient forebrain ischemia. Cerebrovascular accidents often lead to impai red memory in humans (e.g.. Glees & Gri f f i th , 1952; Z o l a - M o r g a n , Squue, & A m a r a l , 1986). Z o l a - M o r g a n , Squire, and A m a r a l (1986) studied one patient who had suffered severe anterograde amnesia fol lowing a br ief pe r iod of cerebral ischemia. The only brain djunage that was observed m this patient and that could be reasonably l inked to his memory impau-ment was complete bilateral infarction of the C A l field of the hippocampus. In monkeys, BachevaHer and M i s h k i n (1989) found that a pe r iod of forebrain ischemia that produced damage to the C A l and C A 2 fields of the hippocampus also produced lasting D N M S deficits. In an ongoing study in rats, a two-vessel-occlusion procedure for inducing transient cerebral ischemia has been found to produce bilateral C A l lesions as wel l as severe D N M S deficits^. 8 E m m a W o o d has been the principal investigator in this study. Exper imenta l rats that had received pretreatment training required more trials to reattain the cr i ter ion at delays of 4 s than d id control (sham ischemia) rats, and once they d id so, they displayed deficits at delays of 15 s, 30 s, 60 s, 120 s, and 300 s; experimental rats that had not received pretreatment training displayed acquisition deficits as wel l as subsequent performance deficits at the same delays. The development of a test battery for use in rat models of brain-damage-produced amnesia A n y memory test can be failed for a variety of reasons, including some that have nothing to do with memory. Therefore, caution must be used when interpreting the findings f rom brain-damaged subjects on a single memory task. M o r e information is available about the severity and the range of brain-damage-produced memory deficits when subjects are tested on more than one memory test. Accord ing ly , recent studies of brain-damage-produced amnesia i n monkeys (Mahu t et al. , 1982; M u r r a y & M i s h k i n , 1986; Z o l a - M o r g a n & Squire, 1985a; Z o l a - M o r g a n et a l , 1989a, 1989b, 1989c) and humans (Aggleton et al. , 1988; Squire et al. , 1988) have employed a battery of tests, some of which are sensitive to brain-damage-produced amnesia i n humans and some of which are not. B y compar ing patterns of spared and impaired performance across several memory tasks i n subjects wi th discrete lesions to various parts of the brain, it may be possible to dissociate the mnemonic effects o f damage to different areas. I have recently developed a battery of memory tests for rats, a l l but one of which (i.e., a temporal-order recognition task) was designed to mimic a specific memory test that has been used i n recent studies of brain-damage-produced amnesia in monkeys. E a c h of these tasks is s imilar to its monkey counterpart i n terms of the test stimuli, the response requirements, and other key aspects of the protocol . E a c h of them uses the same apparatus and test stimuH as my rat D N M S task. Intact rats as wel l as rats wi th bilateral lesions of either the hippocampus, the amygdala, or the posterior parietal cortex are currently being tested on this test battery. The battery includes the D N M S task and five other tasks; the tasks aie administered in the following order: Task 1: Object discrimination. This task assesses the ability of rats to make object-reward associations. The methods are similar to those that were used i n the pretraining phase for the rat D N M S task - the only exception is that in this task, each rat is trained with a single pair of objects unt i l it reaches the cr i ter ion of at least 22 correct trials out of 25 on two consecutive sessions. Task 2: Reversal of object discrimination. In this task, the contingency that was operative in the the object discr iminat ion task, is changed; that is, the previous S- becomes S + and vice versa. The rats are trained to the cri terion of at least 22 correct trials out of 25 on two consecutive sessions. Task 3: Concurrent object discrimination. This task assesses the ability of rats to learn several object-reward associations concurrently. Eight pairs of objects are used. In each pair, one object is designated S + and the other S-. E a c h pair is presented five times per session i n an intermbced order. T h e rats are trained unt i l they reach a criterion of at least 36 correct trials out of 40 on two consecutive sessions. In an earUer pilot experiment, rats with bilateral amygdalo-hippocampal lesions required significantly more trials to reach this cri terion than d id rats wi th control lesions of the posterior parietal cortex. Task 4: DNMS. This task assesses the ability o f rats to learn the nonmatching principle, and to recognize, over a wide range of retention delays, objects that they have experienced i n a single recent episode. The general procedures are identical to those that have already been described for this task. Af te r reaching the cri terion at delays of 4 s, the rats receive six sessions at delays o f 15 s, 60 s, and finally 120 s. Task 5: DNMS with lists. This task was described on page 81. It assesses the abiUty o f rats to ho ld several items i n memory over a delay. The rats are tested on Usts of 3, 5, and 7 objects. Task 6: Temporal-order recognition. This task assesses the abiUty of rats to judge which of two previously presented objects was presented before the other. It is an adaptation of D N M S with a 5-object Ust. A list of 5 objects is presented to the rat, and then, 20 s after the presentation of the final object i n the Ust, 2 of the 5 objects are presented together; the rat is rewarded for choosing the object that appeared earUer. 3. SUMMARY AND CONCLUSIONS T h e study of monkey models of brain-damage-produced amnesia has begun to elucidate the neural bases of memory and amnesia. The development o f comparable rodent models wou ld benefit the study of brain-damage-produced amnesia in two general ways: (1) it would faciUtate the conduct of large-scale parametric experiments, and (2) it would provide a broader comparative basis for drawing inferences about the anatomical bases o f brain-damage-produced amnesia - one that includes rats as weU as humans and monkeys. Th i s thesis took the first steps i n the development of rat models of brain-damage-produced amnesia that are comparable to the monkey models. First , the monkey D N M S task, which plays a central role i n monkey models of brain-damge-produced amnesia, was adapted for use wi th rats; the rat D N M S task that was designed is similar to the monkey D N M S task m several key respects. Then , the rat D N M S task was used in three experiments, which were designed to determine the comparabiUty of the rat D N M S task to the monkey D N M S task in terms of the (1) the rate at wh ich it is learned (Exper iment 1), (2) the asymptotic level at which it is performed (Exper iment 1), (3) the memory abiUties that it taps (Experiment 2), and (4) the brain structures that it engages (Experunent 3). The foUowing were the sbc ma in findings of those experiments: 1. Intact rats readily learned the D N M S task. 2. Once they learned the task, intact rats performed at high levels of accuracy at delays of up to 120 s. 3. The performance of intact rats was better at shorter delays. 4. The D N M S of rats was disrupted by distraction during the delays. 5. Separate or combined lesions of the hippocampus and amygdala produced only a m i l d D N M S deficit in rats. 6. Lesions of the lateral entorhinal cortex, perirhinal cortex, and temporal association cortex appeared to produced a severe D N M S defict in rats. Because each of these six results paraUels the results of experiments on monkeys, they suggest that converging evidence from rat and monkey D N M S models of brain-damage-produced amnesia may help to elucidate the nature of human amnesia and its underlying causes. Ongo ing experiments are providing addit ional support for this view. R E F E R E N C E S Aggle ton , J .P . (1985). One- t r ia l object recognition by rats. The Quarterly Journal of Experimental Psychology, 37B, 279-294. Aggle ton , J .P. , Bl indt , H .S . , & Rawlins , J . N . P . (1989). Effects of amygdaloid and amygdaloid-hippocampal lesions on object recognition and spatial working memory i n rats. Behavioral Neuroscience, 103, 962-974. Aggle ton , J .P. , Hun t , P . R . , & Rawlins , J . N . P . (1986). The effects of h ippocampal lesions upon spatial and non-spatial tests of working memory. Behavioural Brain Research, 19,133-146. Aggle ton , J .P. , Hunt , P . R . , & Shaw, C . (1990). The effects of mammil la ry body and combined amygdalar-fornix lesions on tests of delayed non-matching-to-sample i n the rat. Behavioural Brain Research, 40,145-157. Aggle ton , J .P. , & M i s h k i n , M . (1983a). M e m o r y impairments fol lowing restricted med ia l thalamic lesions i n monkeys. Experimental Brain Research, 52,199-209. Aggle ton , J .P. , & M i s h k i n , M . (1983b). V i s u a l recognition impairments fol lowing media l thalamic lesions in monkeys. Neuropsychologia, 21,189-197. Aggle ton , J .P. , & M i s h k i n , M . (1985). Mammil la ry -body lesions and visual recognit ion i n monkeys. Experimental Brain Research, 58,190-197. Aggle ton , J .P. , N i c o l , R . M . , Hus ton , A . E . , & Fairbairn , A . F . (1988). The performance of amnesic subjects on tests of experimental amnesia in animals: Delayed matching-to-sample and concurrent learning. Neuropsychologia, 26, 265-272. Alex insky , T. , & Chapouthier , G . (1978). A new behavioral mode l for studying delayed responses in rats. Behavioral Biology, 24, 442-456. Bachevalier , J . , & M i s h k i n , M . (1989). M n e m o n i c and neuropathological effects o f occluding the posterior cerebral artery i n Macaca mulatto. Neuropsychologia, 27, 83-105. Bachevalier , J . , Parkinson, J .K . , & M i s h k i n , M . (1985). V i s u a l recognit ion i n monkeys: Effects o f separate vs. combined transection of fornix and amygdalofugal pathways. Experimental Brain Research, 57, 554-561. Baddeley, A . D . (1990). Human Memory: Theory and Practice. Toronto: A l l y n and B a c o n . Baddeley, A . D . , & Warr ington , E . K . (1970). A m n e s i a and the distinction between long- and short-term memory. Journal of Verbal Learning and Verbal Behavior, 9,176-189. Bechterev, W . V . von (1900). Demonstrat ion eines gehirns mit zerstorung der voderen und i imeren theile der hirnrinde beider schlafenlappen. Neurologishces Centralblatt, 19, 990. Beracochea, D . , & Jaffard, R . (1985). M e m o r y deficits subsequent to chronic consumption of alcohol i n mice: A n analysis based on spontaneous alternation behavior. Behavioural Brain Research, 15, 15-25.S Brooks , D . N . , & Baddeley, A . (1976). What can amnesic patients learn? Neuropsychologia, 14,111-112. B r o w n , J . (1958). Some tests of decay theory of immediate memory. Quarterly Journal of Experimental Psychology, 10,12-21. But terworth, R . F . (1989). Effects of thiamine deficiency on bra in metabol ism: Implicat ions for the pathogenesis of the Wernicke-Korsakoff syndrome. Alcohol & Alcoholism, 24, 271-279. Cermak , L . S . , Butters, N . , & Moreines , J . (1974). Some analyses of the verbal encoding deficit of alcoholic Korsakof f patients. Brain and Language, 1,141-150. C k i l l o , R A . , H o r e l , J A . , & George, P . J . (1989). Lesions of the anterior temporal stem and the performance of delayed match-to-sample and visual discrimination. Behavioural Brain Research, 34, 55-69. Cohen , N . J . , & Squire, L . R . (1980). Preserved learning and retention of pattern-analyzing sk i l l i n amnesia: Dissociat ion of knowing how and knowing that. Science, 210, 207-210. C o r k i n , S. (1968). Acqu i s i t i on of motor ski l l after bilateral media l temporal-lobe excision. Neuropsychologia, 6, 255-265. C o r r e l l , R . E . , & Scovil le, W . B . (1965). Performance on delayed match fol lowing lesions of media l temporal lobe structures. Journal of Comparative and Physiological Psychology, 60, 360-367. Drachman , D A . , & A r b i t , J . (1966). M e m o r y and the hippocampal complex. II. Is memory a mult iple process? ^ \/ r \/ \u00ee \/ v e ^ of Neurology, 15, 52-61. Drachman , D . A . , & Ommaya , A . K . (1964). M e m o r y and the hippocampal complex. Archives of Neurology, 10, 411-425. Ennaceur , A . , & Delacour , J . (1988). A new one-trial test for neurobiological studies of memory i n rats. I: Behaviora l data. Behavioural Brain Research, 31, 47-59. E t k i n , M . (1972). Light produced interference in a delayed matching task with capuchin monkeys. Learning and Motivation, 3, 313-324. Gaffan, D . (1974). Recogni t ion impaired and association intact in the memory of monkeys after transection of the tormx. Journal of Comparative and Physiological Psychology, 86,1100-1109. Gaffan, D . , & Har r i son , S. (1989). Place memory and scene memory: Effects o f fornix transection i n the monkey. Experimental Brain Research, 74, 202-212. Glees , P . & Grif f i th , H . B . (1952). Bi la teral destruction of the hippocampus (comu Ammonis) m a case of dementia. Monatsschrift fur Psychiatrie und Neurotome, 123,193-204. G o l , A . , & Faibish , G . M . (1967). Effects of human hippocampal ablation. Journal of Neurosurgery, 26, 390-398. Graf , P . , & Schacter, D . L . (1985). ImpUcit and expUcit memory for new associations i n no rma l and amnesic subjects. Journal of Experimental Psychology: beaming Memory, and Cognition, 10, 164-178. Graf , P . , & Schacter, D . L . (1987). Selective effects of interference on impUcit and expUcit memory for new associations.\/owma\/ of Experimental Psychology: Learning Memory, and Coalition, 13, 45-53. H o n i g , W . K . (1978). Studies of working memory i n the pigeon. In S . H . Hulse , H . Fowler , & W . K . H o n i g (Eds.) Cognitive Processes in Animal Behavior, (pp. 211-248). HiUsdale , N . J . : E r l b a u m . H o r e l , J A . (1978). The neiuoanatomy of amnesia: Cr i t ique of the hippocampal memory hypothesis. Brain, 101,403-445. H o r e l , J A . , & Misantone, L . J . (1976). V i s u a l discrimination impaired by cutting temporal lobe connections. Science, 193, 336-338. H o r e l , J A . , Pytko-Joiner , D . E . , Voy tko , M . L . , & Salsbury, K . (1987). The performance of visual tasks while segments of the inferior temporal cortex are suppressed by cold. Behavioural Brain Research, 23,29-42. Isaacson, R . L . , Douglas , R . J . , & M o o r e , R . Y . (1961). T h e effect of radical h ippocampal ablation on acquisit ion of avoidance response. Journal of Comparative and Physiological Psychology, 54, 625-628. Iverson, S .D . (1976). D o hippocampal lesions produce amnesia i n animals? International Review of Neurobiology, 19,1-49. Jagielo, J A . , Nonneman , A . J . , Isaac, W . L . , & Jackson-Smith, P A . (1990). H i p p o c a m p a l lesions impair rats' performance of a nonspatial matching-to-sample task. Psychobiology, 18, 55-62. Kaada , B . R . , Rasmussen, E . W . , & K v e i m , O . (1961). Effects o f hippocampal lesions on maze learning and retention i n rats. Experimental Neurology, 3, 333-355. Kesner , R . P . (1991). The role of the hippocampus within an attribute mode l of memory. Hippocampus, 1, 279-282. K i m b l e , D . P . (1963). The effects of bilateral hippocampal lesions i n rats.Ioumal of Comparative and Physiological Psychology, 56, 273-283. K n o t h , R . L . , & M a i r , R . G . (1991). Response latency and accuracy on a pretrained nonmatching-to-sample task i n rats recovered from pyrithiamine-induced thiamine deficiency. Behavioral Neuroscience, 105, 375-385. K o l b , B . (1990). Poster ior parietal and temporal association cortex. In B . K o l b and R . C . Tees (Eds.) The Cerebral Cortex of the Rat, (pp. 459-471). Cambridge, Mass : T h e M I T Press. K o l b , B . , B u r h m a n , K . , & M c D o n a l d , R . (1989). Dissociat ion of prefrontal, posterior parietal, and temporal cort ical regions to spatial navigation and recognition memory i n the rat. Society for Neuroscience Abstracts, 15,607. Kornetsky, C . (1977). A n i m a l models: Promises and problems. In I. H a n i n and E U s d i n (Eds.) y l n i m c \/ Models in Psychiatry and Neurology, (pp. 1-7). Oxford: Pergamon Press. M a h u t , H . , M o s s , M . , & Z o l a - M o r g a n , S. (1981). Retent ion deficits after combined amygdalo-h ippocampal and selective hippocampal resections i n the monkey. Neuropsychologia, 19,201-225. M a h u t , H . , Z o l a - M o r g a n , S. & Moss , M . (1982). H ippocampa l resections impai r associative learning and recognit ion memory in the monkey. The Journal of Neuroscience, 9,1214-1229. M a i r , R . G . , A n d e r s o n , C D . , Langlais, P . J . , & M c E n t e e , W . J . (1988). Behaviora l impairments , b ra in lesions and monoaminergic activity in the rat fol lowing recovery f rom a bout of thiamine deficiency. Behavioural Brain Research, 27, 223-239. M e u n i e r , M . , M u r r a y , E A . , BachevaUer, J . , & M i s h k i n , M . (1990). Effects o f per i rh ina l cor t ical lesions on visual recognition memory i n rhesus monkeys. Society for Neuroscience Abstracts, 16, 616. M i l n e r , B . (1959). The memory defect i n bilateral hippocampal lesions. Psychiatric Research Reports, 11, 43-52. M i l n e r , B . (1962). Les troubles de la m \u00e9 m o i r e accompagnant des lesions hippocampiques bUaterales. In Physiologie de l'hippocampe, (pp. 257-252). Paris: Centre Nat iona l de la Recherche Scientifique. M i l n e r , B . (1968). Disorders of memory after brain lesions i n man. Neuropsychologia, 6,175-179. M i l n e r , B . (1971). Interhemispheric differences in the localizat ion of psychological proceses i n man. British Medical Bulletin, 27, 272-277. M i l n e r , B . (1972). Disorders of learning and memory after temporal lobe lesions i n man. Clinical Neurosurgery, 19,421-446. M i l n e r , B . , C o r k i n , S., & Teuber, H . - L . (1968). Further analysis of the hippocampal amnesic syndrome: 14-year foUow-up study of H . M . Neuropsychologia, 6, 215-234. M i s h k i n , M . (1954). V i s u a l discrimination performance foUowing partial ablation of the temporal lobe: II. V e n t r a l surface vs. hippocampus. Journal of Comparative and Physiological Psychology, 47, 187-193. M i s h k i n , M . (1978). M e m o r y in monkeys severely impaired by combined but not by separate removal of amygdala and hippocampus. Nature, 273, 297-298. M i s h k i n , M . (1982). A memory system i n the monkey. Philosophical Transactions of the Royal Society of London, 298, 85-95. M i s h k i n , M . , & Appenzel le r , T . (1987). The anatomy of memory. Scientific American, 256, 62-71. M i s h k i n , M . , & Delacour , J . (1975). A n analysis of short-term visual memory m the monkey. Journal of Experimental Psychology: Animal Behavior Processes, 1,326-334. M i s h k i n , M . & Pr ib ram, K . H . (1954). V i s u a l discrimination performance fol lowing part ia l ablations of the temporal lobe: I. Ven t r a l vs. lateral. Journal of Comparative and Physiological Psychology, 47,14-20. M u m b y , D . G . , P ine l , J .P .J . , & W o o d , E . R . (1990). A new paradigm for testing nonspatial work ing memory i n rats: Nonrecurring-i tems delayed nonmatching-to-sample. Psychobiology, 18, 321-326. M i u n b y , D . G . , W o o d , E . R . , & Pine l , J .P .J , ( in press). Object-recognition memory is only mi ld ly impai red i n rats wi th lesions of the hippocampus and amygdala. Psychobiology. M u r r a y , E A . (1990). Representational memory i n nonhuman primates. In: R . P . Kesne r and D . S . O l t o n (Eds.) Neurobiology of Comparative Cognition, (pp.127-155). Lawrence E r l b a u m Associates: Hi l l sda le , N . J . M u r r a y , E A . ( in press). M e d i a l temporal lobe structures contributing to recognit ion memory: The amydaloid complex versus the rhinal cortex. T o appear in J .P. Agg le ton ( E d . ) : Neurobiology of the Amygdala. Wi ley : N e w Y o r k . M u r r a y , E A . , BachevaHer, J . , & M i s h k i n , M . (1989). Effects of rh ina l cort ical lesions o n visual recognition memory h i rhesus monkeys. Society for Neuroscience Absti-acts, 15, 342. M u r r a y , E A . , & M i s h k i n , M . (1984). Severe tactual as we l l as visual memory deficits fol low combined removal o f the amygdala and hippocampus i n monkeys. The Journal of Neuroscience, 4, 2565-2580. M u r r a y , E A . , & M i s h k i n , M . (1986). V i s u a l recognition m monkeys fol lowing rh ina l cor t ical ablations combined with either amygdalectomy or hippocampectomy. The Journal of Neuroscience, 6, 1991-2003. Nissen, H . W . , Riesen , A . H . , & NowUs, V . (1938). Delayed response and discr iminat ion learning by chimpanzees. Journal of Comparative Psychology, 26, 361-386. O l ton , D .S . , Becker , J.S., & Handehnann, G . E . (1979). Hippocampus , space, and memory . The Behavioral and Brain Sciences, 2, 313-365. O l ton , D ,S . , & Feustle, W A . (1981). H ippocampa l function required for nonspatial work ing memory. Experimental Brain Research, 41, 380-389. Oscar -Berman , M . , & Z o l a - M o r g a n , S . M . (1980a). Comparat ive neuropsychology and Korsakoff 's syndrome. I. Spatial and visual reversal learning. Neuropsychologia, 18, 499-512. Oscar -Berman , M . , & Z o l a - M o r g a n , S . M . (1980b). Comparat ive neuropsychology and KorsakofPs syndrome. II. T w o choice visual discrimination learning. Neuropsychologia, 18, 513-525. Orbach , J . , M i l n e r , B . , & Rasmussen, T . (1960). Learn ing and retention i n monkeys after amygdala-hippocampus resection. y4\/ic\/\u00abVej of Neurology, 3, 230-251. Overman, W . H . , M c L a i n , C , Ormsby, G . E . , & Brooks , V . (1983). V i s u a l recognit ion memory i n squirrel monkeys. Animal Learning and Behavior, 11, 483-488. Pa rk in , A . J . (1982). Res idua l learning capability in organic amnesia. Cortex, 18, 417-440. Paxinos, G . , & Watson, C . (1982). The Rat Brain in Stereotaxic Coordinates. Toronto : A c a d e m i c Press. Peterson, L . R . , & Peterson, M J . (1959). Short-term retention of individual verbal i tems. Journal of Experimental Psychology, 58,193-198. Postman, L . , (1950). Cho ice behavior and the process of recognition. y4\/nencfln Journal of Psychology, 63, 576-583. P r i b r a m , K . H . , W i l s o n , W . , & Conners, J . (1962). Effects of lesions of media l forebrain on alternation behavior of rhesus monkeys. Experimental Neurology, 6, 36-47. Rid ley , R . M . , & Baker , H . F . (1991). A critical evaluation of monkey models of amnesia and dementia. Brain Research Reviews, 16,15-37. R i n d i , G . (1989). A l c o h o l and thiamine of the hrain. Alcohol & Alcoholism, 24, 493-495. Rothbla t , L A . , & Hayes, L . L . (1987). Short-term object recognition memory i n the rat: Nonmatch ing with trial-unique st imuli . Behavioral Neuroscience, 101, 587-590. Rothbla t , L A . , & K r o m e r , L . F . (1991). Object recognition memory i n the rat: the role o f the hippocampus. Behavioural Brain Research, 42, 25-32. Rupn iak , N . M . J . , & Gaffan, D . (1987). M o n k e y hippocampus and learning about spatially directed movements. The Journal of Neuroscience, 7, 2331-2337. Schacter, D . L . (1987a). ImpUcit expressions of memory in organic amnesia: Lea rn ing of new facts and associations. Human Neurobiology, 6, 107-118. Schacter, D . L . (1987b). ImpUcit memory: His tory and current status. Journal of Experimental Psychology: Learning Memory, and Cognition, 13, 501-518. Schacter, D . L . , & Tulving, E . (1982). M e m o r y , amnesia, and the semantic\/episodic dist inction. In R . L . Isaacson and N . E . Spear (Eds.) Expression of Knowledge, (pp. 33-65). N e w Y o r k : P l e n u m Press. Schneider, K . (1912). U b e r einige kUnisch-pathologische untersuchungsmethoden und irhe ergebnisse. Zug le i ch e in Bei t rag zur Psycholpathelogie der Korsakowchen Psychose. Zeitscrift fur Neurologische Psychiatrie, 8, 553-616. Scovil le , W . B . (1954). The l imbic lobe i n maa. Journal of Neurosurgery, 11, 64-66. Scovil le, W . B . , & M i l n e r , B . (1957). Loss of recent memory after bilateral h ippocampal les ions . \/ou\/na\/ of Neurology, Neurosurgery, and Psychiatry, 20,11-21. S loman , S A . , Hayman , C A . G . , Ohta , N . , Law, J . , & Tulving , E . (1988). Forget t ing i n p r imed fragment complet ion. Journal of Experimental Psychology: Learning Memory, and Cognition, 14, 223-239. Squire , L . R . (1986). Mechanisms of memory. Science, 232,1612-1619. Squire , L . R . (1987). Memory and Brain. N e w Y o r k : Oxford Univers i ty Press. Squire , L . R . , Z o l a - M o r g a n , S., & Chen , K . S . (1988). H u m a n amnesia and animal models of amnesia: Performance of amnesic patients on tests designed for the monkey. Behavioral Neuroscience, 102, 210-221. Sternberg, S. (1966). H i g h speed scanning in human memory. Science, 153, 652-654. Strong, E . K . , (1912). The effect of length of series upon recognition memory. Psychological Reviews, 19, 447-462. Teuber , H . - L . , M i l n e r , B . , & Vaughan, H . G . (1969). Persistent anterograde amnesia after stab wound of the basal bra in . Neuropsychologia, 6, 267-282. V i c t o r , M . , A d a m s , R . D . , & ColUns, G . H . (1971). TTie Wernicke-Korsakoff Syndrome. Phi ladelphia: F A . Davis . V i c t o r , M . , Angevine , J .B . , Manca l l , E . L . , & Fischer, C M . (1961). M e m o r y loss wi th lesions of hippocampal formation, ^ rc \/ i \/ve j of Neurology, 5, 26-45. V i c t o r , M . & Yakovlev , P.J . (1955). S.S. K o r s a k o f f s psychic disorder. Neurology, 5, 394-407. Z i l l e s , K . (1985). The Cortex of the Rat: A Stereotaxic Atlas. B e r l i n : Spr inger-Ver lag . Z o l a - M o r g a n , S., Squire, L . R . , & M i s h k i n , M . (1982) The neuroanatomy of amnesia: A m y g d a l a -hippocampus versus temporal stem. Science, 218,1337-1339. Z o l a - M o r g a n , S., & Squire, L . R . (1985a). M e d i a l temporal lesions in monkeys impai r memory on a variety of tasks sensitive to human amnesia. Behavioral Neuroscience, 99, 22-34. Z o l a - M o r g a n , S., & Squire, L . R . (1985b). A m n e s i a in monkeys fol lowing lesions of the mediodorsal nucleus of the thalamus. Annals of Neurology, 17, 558-564. Z o l a - M o r g a n , S., & Squire, L . R . (1986). M e m o r y impairment i n monkeys fol lowing lesions of the hippocampus. Behavioral Neuroscience, 100,165-170. Z o l a - M o r g a n , S., Squire, L . R . , & A m a r a l , D . G . (1986). H u m a n amnesia and the med ia l temporal region: E n d u r i n g memory impairment following a bilateral lesion l imi ted to the field C A l of the hippocampus. The Journal of Neuroscience, 6, 2950-2967. Z o l a - M o r g a n , S., Squire, L . R . , & A m a r a l , D . G . (1989a). Lesions of the amygdala that spare adjacent cort ical regions do not impair memory or exacerbate the mipairment fol lowing lesions of the hippocampal formation. The Journal of Neuroscience, 9,1922-1936. Z o l a - M o r g a n , S., Squire, L . R . , & A m a r a l , D . G . (1989b). Lesions of the h ippocampal formation but not lesions of the fornix or the mammil lary nuclei produce long-lasting memory impairment in monkeys. The Journal of Neuroscience, 9, 898-913. Z o l a - M o r g a n , S., Squire, L . R . , A m a r a l , D . G . , & Suzuki , W A . (1989c). Lesions of per i rhinal and parahippocampal cortex that spare the amygdala and hippocampal format ion produce severe memory impairment. The Journal of Neuroscience, 9, 4355-4370. ","attrs":{"lang":"en","ns":"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note","classmap":"oc:AnnotationContainer"},"iri":"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note","explain":"Simple Knowledge Organisation System; Notes are used to provide information relating to SKOS concepts. There is no restriction on the nature of this information, e.g., it could be plain text, hypertext, or an image; it could be a definition, information about the scope of a concept, editorial information, or any other type of information."}],"Genre":[{"label":"Genre","value":"Thesis\/Dissertation","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/hasType","classmap":"dpla:SourceResource","property":"edm:hasType"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/hasType","explain":"A Europeana Data Model Property; This property relates a resource with the concepts it belongs to in a suitable type system such as MIME or any thesaurus that captures categories of objects in a given field. It does NOT capture aboutness"}],"GraduationDate":[{"label":"Graduation Date","value":"1992-05","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#dateIssued","classmap":"vivo:DateTimeValue","property":"vivo:dateIssued"},"iri":"http:\/\/vivoweb.org\/ontology\/core#dateIssued","explain":"VIVO-ISF Ontology V1.6 Property; Date Optional Time Value, DateTime+Timezone Preferred "}],"IsShownAt":[{"label":"DOI","value":"10.14288\/1.0098895","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt","classmap":"edm:WebResource","property":"edm:isShownAt"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt","explain":"A Europeana Data Model Property; An unambiguous URL reference to the digital object on the provider\u2019s website in its full information context."}],"Language":[{"label":"Language","value":"eng","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/language","classmap":"dpla:SourceResource","property":"dcterms:language"},"iri":"http:\/\/purl.org\/dc\/terms\/language","explain":"A Dublin Core Terms Property; A language of the resource.; Recommended best practice is to use a controlled vocabulary such as RFC 4646 [RFC4646]."}],"Program":[{"label":"Program (Theses)","value":"Psychology","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline","classmap":"oc:ThesisDescription","property":"oc:degreeDiscipline"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the program for which the degree was granted."}],"Provider":[{"label":"Provider","value":"Vancouver : University of British Columbia Library","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/provider","classmap":"ore:Aggregation","property":"edm:provider"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/provider","explain":"A Europeana Data Model Property; The name or identifier of the organization who delivers data directly to an aggregation service (e.g. Europeana)"}],"Publisher":[{"label":"Publisher","value":"University of British Columbia","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/publisher","classmap":"dpla:SourceResource","property":"dcterms:publisher"},"iri":"http:\/\/purl.org\/dc\/terms\/publisher","explain":"A Dublin Core Terms Property; An entity responsible for making the resource available.; Examples of a Publisher include a person, an organization, or a service."}],"Rights":[{"label":"Rights","value":"For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https:\/\/open.library.ubc.ca\/terms_of_use.","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/rights","classmap":"edm:WebResource","property":"dcterms:rights"},"iri":"http:\/\/purl.org\/dc\/terms\/rights","explain":"A Dublin Core Terms Property; Information about rights held in and over the resource.; Typically, rights information includes a statement about various property rights associated with the resource, including intellectual property rights."}],"ScholarlyLevel":[{"label":"Scholarly Level","value":"Graduate","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#scholarLevel","classmap":"oc:PublicationDescription","property":"oc:scholarLevel"},"iri":"https:\/\/open.library.ubc.ca\/terms#scholarLevel","explain":"UBC Open Collections Metadata Components; Local Field; Identifies the scholarly level of the author(s)\/creator(s)."}],"Title":[{"label":"Title ","value":"The development of a rat model of brain-damage-produced amnesia","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/title","classmap":"dpla:SourceResource","property":"dcterms:title"},"iri":"http:\/\/purl.org\/dc\/terms\/title","explain":"A Dublin Core Terms Property; The name given to the resource."}],"Type":[{"label":"Type","value":"Text","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/type","classmap":"dpla:SourceResource","property":"dcterms:type"},"iri":"http:\/\/purl.org\/dc\/terms\/type","explain":"A Dublin Core Terms Property; The nature or genre of the resource.; Recommended best practice is to use a controlled vocabulary such as the DCMI Type Vocabulary [DCMITYPE]. To describe the file format, physical medium, or dimensions of the resource, use the Format element."}],"URI":[{"label":"URI","value":"http:\/\/hdl.handle.net\/2429\/3293","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#identifierURI","classmap":"oc:PublicationDescription","property":"oc:identifierURI"},"iri":"https:\/\/open.library.ubc.ca\/terms#identifierURI","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the handle for item record."}],"SortDate":[{"label":"Sort Date","value":"1992-12-31 AD","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/date","classmap":"oc:InternalResource","property":"dcterms:date"},"iri":"http:\/\/purl.org\/dc\/terms\/date","explain":"A Dublin Core Elements Property; A point or period of time associated with an event in the lifecycle of the resource.; Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF].; A point or period of time associated with an event in the lifecycle of the resource.; Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF]."}]}