University of Southern California Dissertations and Theses

The Retroactive Effect Of Strenuous And Exhaustive Exercise On Maze Task Learning

(USC Thesis Other)

The Retroactive Effect Of Strenuous And Exhaustive Exercise On Maze Task Learning

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content This dissertation has been . . . . . . , , 69-17,897 microfilmed exactiy as received HUTTON, Robert Stanley, 1939- THE RETROACTIVE EFFECT OF STRENUOUS AND EXHAUSTIVE EXERCISE ON MAZE TASK LEARNING. University of Southern California, Ph.D., 1969 Education, physical University Microfilms, Inc., Ann Arbor, Michigan THE RETROACTIVE EFFECT CF STRENUOUS AND EXHAUSTIVE EXERCISE ON HAZE TASK IEARKDJ3 by Robert Stanley Hatton A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR CF PHELCSCPHY (Physical Education) June 1969 UNIVERSITY PARK LOS ANGELES. CALIFORNIA 9 0 0 0 7 This dissertation, written by ________ Saksrfc..S£aifley-ifofcfcp.n_________ mmder the direction of h is — Dissertation Com mittee, and approved by all its members, has been presented to and accepted by The Gradu ate School, in partial fulfillm ent of require ments for the degree of D O C T O R O F P H IL O S O P H Y Deam Date June+.19$9 DISSERTATION COMMITTEE r \ /Chmtrmm* ~ ? ; ....... co - cAd-cU'/iui ACKNOWLEDGMENTS Prior to and during the course of this investigation the author was fortunate to receive assistance from several individuals ■who wnselfishly devoted their time and effort to help make tMs study possible. The author would, therefore, like to express Ms gratitude to: Mr. Robert Fuller for his ingenuity and assistance im assembling the maze task, Mr. Harry Horii for designing and constructing the ehectri- cal unit through which shock could be administered to the subjects tested. Miss Cynthia Gillespie and Mr. Harvey Wallace who had the rather tedious task of supervising the exercise task while the experimenter was testing subjects on the maze task, the Department of Physical Education, University <of California., Los Angeles, for financial assistance and use of their Performance Laboratory facilities, Ms committee members Mrs. Roxie Morris, Mr. Herbert dearies, with special thanks to his co-chairmen Dr s. Lockhart and lerstem. The author extends his greatest appreciation to his wife, Judy, who worked so hard in typing the rough drafts and the final manuscript of this dissertation. TABLE OF CONTENTS CHAPTER . I B I. INTRODUCTION................................. 1 Statement of the Purpose............ & Statement of the Problem.............. & Assumptions and Hypothesis........... & Significance of the Study............ 5 Theoretical Applications........... 5 Practical Applications............ & Delimitations ..... 7 Organization of Remaining Chapters............ f f i H. REVIEW OF LITERATURE......................... 3 Part I: Potential Neurophysiological Mechanisms of Learned Behavior....... 9 Neuroanatomical Changes.......... I® Neurochemical Changes............. M> Electrophysiological Changes...... 21 Part II: Theory of Consolidation %7 Related Investigations............. . Part III: Strenuous Muscular Exertion and Activity of the Central Nervous System........ iii CHAPTER P M Changes in Metabolic Functions............ 5& Behavioral Adjustments............ 59 Part IV: Maze Learning ...... 6>3 Procedural Considerations in Maze Learning.............................. 6k- Human Versus Animal Performance on a Maze Task............ 65 Summary....... 68 III. EXPERIMENTAL DESIGN AND PROCEDURE............. 68 Design and Construction of the Learning and Exercise Task.................... 68 Learning Task................. 68 Exercise Task................ 73 Pilot Study 1.............................75 Pilot Study II............................®2 Experimental Procedure....... 8k Treatment of the Data 85 Analysis of the Findings............. 8$) Summary and Conclusions........... 92 Experimental Design for the Principal Investigation.............. 9k i Procedure ............. 95 Treatment of the Data. 9® Summary....... 991 iv CHAPTER PAGE IV. RESULTS................................... 100 Error Scores... .........................100 Time Scores.............................. 105 Time Scores on Exercise Task........... 109 Additional Findings....................... 113 V. DISCUSSION................................. 115 Error Scores........... ............... 115 Time Scores ......................... 120 Exercise Task...... ......... 121 Limitations of the Findings.................122 Summary................................ * 12*f- VI. SUMMARY, CONCLUSIONS, IMPLICATIONS AND RECOMMENDATIONS............... 125 Summary....................... 125 Conclusions.............................. 129 Implications............................ 129 Recommendations ...................... 131 REFERENCES................. !33 APPENDIX.......................................... I**0 v LIST OF TABLES TABLE PAGE I. Group Means for Error Scores................ 79 II. Group Means for Time Scores...... 80 III. Group Means and Standard Deviations for Error Scores................................86 IV. Group Means and Standard Deviations Per Blocks of Trials for Time Scores.................87 V. One Tailed t Tests for Intra- and Inter group Comparisons Among Selected Trials...., 88 VI. Reliability Coefficients Between Selected Trials - Group A (Control Group)............... 88 VII. Individual Times Per Trial on the Exercise Task ..............................91 VIII. Group Means and Standard Deviations Per Blocks of Trials for Error Scores............ 101 IX. Analysis of Variance of Error Scores........... 102 X. Newman-Keuls Test for the Differences Between All Pairs of Ordered Means for Factor A - Exercise Treatment...... 10^ XI. Newman-Keuls Test for the Differences Between All Pairs of Ordered Means for Factor B - Blocks of Trials .... 10^- vi TABLE PAGE XII. Group Means and Standard Deviations Per Block of Trials for Time Scores (Expressed in Seconds)......................... 106 XIII. Rank Order and Sum of Ranks (R) for Delay Time Scores Out of the Starting Box and the Kruskal-Wallis One-Way Analysis of Variance (H)........... 110 XIV. Duration of Exercise for Animals in Group C......................... • HI XV. Group Means and Standard Deviations of Pre- and Post-Test Weights........................ ll^f vii n LIST OF i F T T I G iliMIR FIGURE PAGE 1. Six Unit, Multiple-U "Water Mas® mtiEn Guillotine-Type Gates........ ..... 69 2. Outside Dimensions (In Inches]) mff the Maze 72 3. Trip Mechanism........ 7^ h. Gate Raised......... 7^ 5. Gate Raised (Inside View])... ...... 7^ 6. Gate Lowered (Inside View]).................... 7^ 7. Gate Prior to Release.. ....... 7^ 8. Gate Released....... 7^ 9. Control Panel, Timer, and ©onrtbsr............. 76 10. Experimenter's View.. ....... 76 11. Exercise Apparatus.. ........ 76 12. Learning Curves for Group Mean g&nraar Scores.... 79 13. Performance Curves for Group ifegm Time Scores.. ......... 80 lh. Learning Curve for Error Scnaes................ 86 15. Performance Curve for Time Snares.............. 87 16. Learning Curves Based on (Group Meae m Error Scores......... 101 17. Performance Curves Based con T5ane SJcores........ 106 TQ33. CHflPSEE I IHTEOBICIIOT Basic to the learning process is the capacity of the nervous system to register, store, and recall experiential events. It appears that this process is subserved by a compiex; interaction between sensory and motor channels which operate more or less in a closed circuit fashion, i.e., sensory stimuli initiated by environmental factors are not introduced into the central nervous system via some distinct lin ear arrangement but rather are superimposed upon already existing in teractions between afferent and efferent neural components. Thus, the fate of any sensory stimulus depends, to a large extent, upon the state of the organism at the time the stimulus is introduced; there fore, it might be assumed that the met effect of external stimuli on behavior is based upon the interaction between the nature of the stim ulus and the state of the organism at the time the stimulus is intro duced. This resultant effect on the nervous system must in turn in fluence subsequent sensory inf ormation which is introduced into the system. This progressive interaction is well described by McGeogh: Inasmuch as living individuals can never escape the occurrence of progressively interpolated experience whereby each succeeding experience serves as interpolation to all preceding ones, the effect of each interpolation must be among the most ubiquitous phemnmmiom of mental life. (^i:^00) Investigators interested in learning behavior have been 1 especially concerned with the effects of seemingly unrelated inter- i polated experiences or sensory stimuli on the learning of a verbal or motor task. Out of this research have come, for example, such concepts as Hull's theory of reactive inhibition (IR) and conditioned inhibi tion ( sIr). Gibson's concept of generalization and differentiation, and Hebb's theory of consolidation (9) (^2) (8). A great deal of in sight into the problems associated with learning and memory processes has been gained. Compilation of experimental findings has provided a reference for new ideas regarding how learning and memory might be facilitated or inhibited by carefully controlled interpolated experi ences, i.e., stimuli introduced between learning trials. I ' S . t h respect to learning and retention of motor skills, the following generalizations have been made: 1. Motor skills seem more resistant to forgetting than verbal skills (1). 2. Motor skills that are continuous are better retained than those that are discrete (1) (6). 3. Rapid relearning of a gross motor skills seems to take place even after long intervals of time (65). k. Overlearning the task tends to minimize the amount of forgetting (3). It is interesting to note that little attention has been given to the effects of excessive exercise on the process of memory or forgetting, the reason being, perhaps, that motor skills have always been reported to be somewhat impervious to the forgetting process. It might easily be hypothesized that since strenuous ; exercise alters the organismac state of activity it may also effect in ■ some way the neural integration of subsequent stimuli. Much of the related wrarfe 3m tMs area has been concerned with the effects of fatigue on sahsBcimgmfc learning experiences or with the interpolated effects of local ffafcSgme i&s in a single limb) on the performance of a motor task. 3Bm either case the results have been negligible. Two experimental procedures, however, have received little at tention in the research literature. These procedures involve 1) de termining the effects of interpolated exercise of a strenuous and general nature on the learns mg of s . perceptual-motor task, and 2) determining the effects of sammlacr exercise conditions on the learning of verbal tasks. The rationale behind these procedures is two-fold. First, Hebb proposes that the cozmectamms or traces formed in learning are not completely laid down at the tame of a learned experience. Rather, some process goes on for narrates or, perhaps, hours afterward, during which time a raemoiy trace is 3m some way consolidated (*H) (11). If this be true, perhaps strenuous exercise introduced during the "period of consolidation" (e.g., of a motor task) might alter in some manner the memory trace. Secondly, am alternate hypothesis is that the exer cise task might interfere mtfia the neural processes intimately asso ciated with the task to be learned. Of prime concern am the present, study was the former of the two propositions. Since it fans been found that factors which tend to depress central nervous activities also tencL to disrupt or depress * J memory consolidation, providing the two factors overlap in time, would it not also be possible that exhaustive exercise or exercise of a strenuous nature could have the same effect, assuming that exhaustive I exercise exerts depressing effects on neural activities? With regard to this latter assumption, it is well known that among the many physio logical concomitants to an exhaustive work load are metabolic changes which tend to decrease blood pH. A decrease of blood pH is also be lieved to have depressing effects on central nervous activity. Other mechanisms could be cited to support the aforementioned assumption but these are discussed more thoroughly in Chapter II - Review of the Literature. Statement of the Purpose The purpose of this investigation, based upon theoretical con siderations related to a consolidation theory of memory, was to deter mine the effects of exercise on maze task learning. Statement of the Problem The specific problem of this investigation was to determine whether or not strenuous or exhaustive exercise, by producing a pos sible retroactive effect on the period of consolidation, influences the amount and rate of maze learning by rats. Assumptions and Hypothesis Two major assumptions were basic to this investigation: 1. It was assumed that high intensities of exercise result in acute metabolic changes which tend to decrease the 5 general level of central nervous activity. 2. It was also assumed that the memory theory of consolida tion represents a valid explanation of neural processes related to learned behavior. Evidence which “ tends to support or challenge these assumptions is presented in the Review of literature. The hypothesis tested was: strenuous or exhaustive exercise, introduced immediately after a learning experience, will have a detri mental effect upon the amount and rate of learning. The hypothesized deficit, in the amount and rate of learning was believed to occur as a function of overlap between induced organismic changes attributed to exercise and induced neural changes (period of consolidation) at tributed to the original experiential stimuli. Stated more precisely: Given three experimental conditions related to learning a maze task, where one group (group A) receives learning trials with no specified interpolated activity, a second group (group B) receives a mild bout of exercise immediately following each testing period, and a third group (group C) receives a strenuous and exhaustive bout of exercise immediately after each trial, it will be found that for the group means per error and time scores: group A group B< group C, i.e., learning for group B will be equal to or greater than group A and both group A and B will evidence a faster rate of learning than group C. Hq = Ma = Mg = Me, where H© = null hypothesis and M = mean of the respective group. Significance of the Study Of special concern in this investigation were the potential theoretical and practical applications of the findings. Theoretical Applications Through the years, many theories have been developed in which j it was postulated that changes occur in neuromechanisms associated with learned behavior. One concept which has been well received is that of ’ 'perseveration" or "reverberation", i.e., a memory trace may | be stored temporarily in a circular arrangement of continuous neural impulses. Existence of such a process leads to serious consideration of the kinds of experiential events which should follow a learning experience in order to assure maximum retention and improved rates of learning. It was hoped that findings from this study might provide further evidence related to the concept of perseveration. If an interpolated task of strenuous exercise significantly retards the rate of learning, as reflected by retarded intertrial improvements, whereas negligible results are obtained with moderate exercise, the "perseveration" concept might explain such results. In any case, additional investigations with additional controls would be necessary before more precise interpretations could be made concerning theo retical concepts. Practical Applications It is not uncommon for the athletic performer to be subjected to rather demanding and stressful training conditions which are specifically designed to induce acute metabolic disturbances. It is also not uncommon for the coach to teach complex movement patterns which the performer is expected to learn and execute perfectly. During the usual practice session both of these training procedures are used, without any particular attention being given to the 7 •» consequences of such interactions. With respect to the experimental hypothesis, it might be assumed that learning might be served best if there is a suitable period of delay before exhaustive exercise is introduced. This would hold true regardless of the type of material learned, e.g., a chalk talk covering strategy or practicing specific skills. Exercise of a more moderate nature, by virtue of its arousal effects on central nervous activity, may have a facilitative effect on the learning process. In either case, findings from this study could provide valuable information relative to the appropriate scheduling of experiential events where environmental factors which serve learning best could be, in part, accounted for. Delimitations This investigation was limited to the study of animal behavior. Although this restriction imposed limitations upon generalizations which might be derived from the findings, reasonably good relation ships have been demonstrated between findings obtained from animal and human research with respect to studies of memory consolidation. In order to provide a large enough sample size for statistical treatment, it was necessary to restrict the number of groups tested to three. The present study represents an initial investigation and additional studies must be considered. Other limitations, more specifically related to the experi mental procedure, are discussed in Chapters III and V. 8 gftygamization of Remaining Chapters A x&msmi off the experimental literature appears in Chapter H : Part 1 — JtofcmnifcSal Beurophysiological Mechanisms of Learned Behavior, Part H — TDfaBnny off Memory Consolidation, Part III - Strenuous Mus cular BxErtSnm ainfl Activity of the Central Nervous System, and Part IV - 35b®® lemming, ^-^twrMni g r i - t r gl design and procedure utilized in this investi gation are ( S a s u i f f i f f i e t i . in Chapter III, This material is presented in the fdULoraaqg sapience: of topics: the design and construction of the experimental apparatus used for testing, two preliminary pilot studies to determine ami refine the experimental technique, and a discussion of the - rterefjgrm arrtS procedures followed in the main investigation. Treatment < eF the dfeta. is also discussed in the latter section. In rtSfoapfoigr w the results are presented and interpreted with respect it® itfo® experimental hypothesis. A discussion of the findings appears in (Dfcapftear W, A summary af the study, conclusions derived from it, implica tions and rerammiasjfetions are found in the sixth and final chapter. CHAPTER II REVIEW OF LITERATURE Dae to the number of topics to be considered and the wealth of information available in the area of memory and learning, this review is presented in four main parts. Part I is concerned with a brief discussion of potential neurophysiological mechanisms associated with learned behavior. In Part II a theory of memory consolidation which has evolved, for the most part, from the theoretical concepts presented in Part I is con sidered. In order to show the gradual emergence of the memory con solidation theory, the material presented in Part II is organized in chronological order. Part III is primarily concerned with one major assumption underlying the present study, i.e., strenuous or exhaustive exercise causes acute changes in metabolic function which tend to depress general activity levels of the central nervous system. The last section. Part IV, is involved with a consideration of experi mental research relative to the effective use of maze tasks in small animal research. Procedural guides were abstracted from this review. A summary is provided at the end of the chapter. Part I: Potential Neurophysiological Mechanisms — of Learned Behavior In most theories, in which neuromechanisms underlying the learning process are postulated, particular attention is devoted to 9 10 structural and functional changes which occur in neural tissue as a result of repetitive stimulation. Of particular concern are the possible ways in which neuronal and synaptic mechanisms participate in learning, since it is at these latter junctions where the neural impulse is interrupted and susceptible to alteration. At this level the potential mechanisms most often considered are anatomical changes in the configuration of the synapse, neurochemical adaptations which influence the transmission of impulse, and electrochemical forces which are altered by the specific dynamic action of the neuron, i.e., whether the neuron is of the kind that inhibits or facilitates further transmission. The following discussion is concerned with research related to each of these respective viewpoints. It should be recog nized that the available research literature in this area is plenti ful and complex; therefore, only a review of selected material is presented here. Neuroanatomical Changes Bccles, in a conference on learning, remembering, and for getting, made reference to two theories of learning. One was a theoiy which "...postulates that memory is subserved by circulating impulses, tracking around in pathways with specific spatio-temporal patterning" (10:12). It was suggested that as the impulses circulate they give rise to specific behavioral responses. The second theory "...postulates a structured change and does not require continuing circulation of impulses. Some enduring change is built into the fine structure of the nervous system, making some particular 11 pathways more favorable for impulse transmission than others. This is the kind of memory tract that may persist throughout the whole of life, and is effective as long as we have a method of recall" (10:12). It was assumed that this structural change is synaptic. In his presentation, Eccles chose to concentrate on the latter theory. The following statements present interesting views which were brought out in Eccles1 presentation (10): 1. Memory and learning are preeminently functions of the brain in highly developed animals. 2. Qualitatively, synapses are the same at all levels; quantitatively there are important differences at the higher levels of the brain. 3. These differences involve: a. frequency-potentiation, that is, rapid repetitive activation of synapses at higher levels favors synaptic plasticity or flexibility. b. inhibitory synapses, which are more developed, more powerful, and more prolonged in action in the brain than at lower levels. c. immense dendritic structures and special kinds of synapses to be found on dendrites at higher levels, e.g., pyramidal cells, the most prominent feature in the cerebral cortex, have spines on their den drites. (On each spine is a synapse, which Eccles suggests is only excitatory.) 4-. Impulses are discharged only from a synapse if summation occurs. 5. At higher levels, most of the synapses of the cell body of the neuron are inhibitory and in certain places all of them are inhibitory. i 6. Pores of the subsynaptic membrane are quite fixed in size, being at least twice as large for excitatory than for inhibitory synapses. The holes for the excitatory synapse are big enough to let sodium go through freely, and they are small enough to stop sodium completely 12 for the inhibitory synapse.* In the case of an inhibi tory synapse the pores would allow the hydrated potassium and chloride ions to go through, but not sodium. 7. Chemical synaptic transmission introduces an amplifying mechanism, which may be at least 100 fold. The currents flowing in the presynaptic terminals are less than one per cent of what is required to generate the impulse in the muscle. 8. Cortical inhibitions are different from the spinal cord with a different pharmacological mechanism. The in hibitory potentials in the central cells are ten times larger and have ten to twenty times longer duration, e.g., 200 msec. - 60 msec, duration in central cells as opposed to 10 msec, duration in spinal cells. 9. Eccles suggests that if learning is due to better and bigger synapses, it is not enough to have a larger synaptic knob. The really significant change is in creased transmitter output, and this means more vesicles in position to be liberated by impulses. 10. Eccles thinks we (members of the conference) all would agree that we do not expect, under conditions of learning, to have all the synaptic structures in excess function, only those in specific areas. It might be only a very small faction of the available synapses. In summary, Eccles suggested that learning may be a phenome non associated with specific structural changes in synapses located principally at higher levels. These changes are assumed to be brought about by the effects of use and disuse of the synapse, i.e., synapses can be made more efficient by use, and less efficient by disuse (disuse in this case being the counterpart of forgetting). It should be noted that Eccles attributes great importance to the role of inhibitory synapses in memory or learning. ♦Eccles stated that the theory of the nerve impulse proposed by Hodgkin and his colleagues is established beyond all reasonable doubt. 13 Two suggested structural changes in synapses, thought to be directly related to learning by Eccles, are: 1, the possibility of collateral growth of terminal fibers, thereby increasing the number of synaptic connections, and 2. an increased size in the synaptic knob with a cor responding increased number of neurotransmitter cariying vesicles. The real significance of these structural changes is to increase neurotransmitter output. The possibility that glia may have a role in the transmission of impulses was thoroughly discounted by Eccles, One assumption made by Eccles is that a synapse may change in shape as a result of increased use. Russell (71)* in a presi dential address to the Royal Society of Medicine, had this to say about structural changes at synapses: The physiological mechanism whereby an active synapse becomes facilitated for future use is not fully understood, but there is some evidence to suggest that physical changes may play a part. Thus, the synaptic knobs on the surface of anterior horn cells may, in response to great activity, get larger (Eccles, 1953)* and recent work raises some interesting possibilities about a somewhat similar process in the inter fibre synapses of the cortex. For long there has been some thing of a mystery about inter-fibre transmission in the cor tex but studies described by Bok (195&) suggest a plausible explanation. He has found that 90% of the cerebral cortex consists of spaces which separate nerve cells, glial cells and blood vessels. In this space an immense number of nerve fibres cross each other in various directions, and this fibre density remains much the same in different layers of the cortex, and indeed as between man and ani mals. The distance which separates crossing nerve fibres is also remarkably constant, and this separation of one fibre from another is apparently maintained by a foamlike structure of spheres whose diameter is about 5 *1 • It is suggested therefore that fibre synapses depend partly 14 < just on the distance which separates fibres from each other. In other words close crossings are facilitated synapses, A further clue to this is probably provided by a study of the spikes (gemmuli) which appear on the den drites of cortical pyramidal-type cells. Bok has demon strated that the gemmuli on a dendrite are. on the whole, separated from each other by a regular distance ( 2. 5a. ) . This corresponds to the distance apart of the points where crossing fibres would come in close contact with the den drites on the above arrangement of foam particles. It seems therefore that every spike on the dendrites may become a close-crossing synapse. Further the development of the gemmuli greatly enhances the effectiveness of the close-crossing synapse by further approximating the crossing fibres. Indeed this increase in size of the fibre at the point of crossing almost (and perhaps en tirely) eliminates the space between the crossing fibres. In this simple way the effectiveness of the synapse re ceives the maximum enhancement. The development of these gemmuli is presumably a physical change which results from the activity of the synapses concerned, and plays an important part in en hancing their activity. Tremendous multiplication of these occurs during first two years of life. The crossing at the gemmuli suggests a very firmly established type of close-crossing synapse, but it may be assumed that simpler mechanisms of enhancement occur without the formation of so much physical change. (71:3) It is apparent that Russell concurs with Eccles' view that enhanced synaptic output may be due to changes in the size and shape of synaptic knobs. Unfortunately, the assumption that repeated use of a neuron increases the potency of normal excitatory processes re sponsible for transmission does not seem entirely clear. Much of the earlier evidence provided by Eccles for the "use-disuse" hy pothesis concerned changes in monosynaptic excitation of a moto neuron at the spinal level. Although experiments involving dorsal 15 root transections of peripheral neurons with maeroelectrode moni toring of monosynaptic reflexes provided gvSidamce < a £ increased neuronal efficiency (3*0 (35)» in other 5$nras4i@aMjmas using tenotomized animals just the opposite has found! to be truev i.e., the monosynaptic reflex recorded was greater from the tenoto mized muscle than from the normal control muscle; cm the other side (10), Although the hypothesis proposed by jtecles provides an in teresting basis for many learning theories, f'mrit&er evidence is needed before there can be widespread agreement cm this point by learning theorists. Young (18) has suggested that learning involves the inhibi tion of unwanted pathways and that this occurs suddenly and com pletely in each cell concerned, as in a single occasion of learning; it is assumed that only some mnemons are switched at each occasion of learning. Although no provision was made in Yeung*® theory for the erasure of memory, his emphasis on inMtoitory sjmaptic functions in learning is comparable to views proposed toy Steeles. Sharpies s (7*0, in an extensive review cm the reorganization of function in the nervous system— use and disnsse, reported that there are two diametrically opposed opinions regarding the effects of long term use on the synaptic junction: 1) repeated exercise increases the potency of normal excitatory processes of transmissions, and 2) repeated exercise decreases this potency of transmission. A dualistic position was suggested where some junctions are thought to increase in efficiency with use while others decrease. Many of 16 the findings reported by Sharpiess are difficult to interpret with reference to their possible implications for learning. Neurochemical Changes Prior to the relatively recent breakthrough in the chemistry of nucleic acids, Halstead and Katz proposed that individual neurons produce new and specific proteins to record a memory trace. It was suggested that a neuron could become involved in the permanent re cording of an experiential event only by making a new protein, in this way providing the means through which a memory eng ram is formed and after which the activity of the neuron would then become inde pendent of the original stimulus (h-9). Partial evidence to support the protein hypothesis was provided by Thompson and McConnell (77) in their study of planaria. Planaria were first trained by classical conditioning procedures to acquire a conditioned response (that of contraction) at the pre sentation of light. The worms were then cut in half and the front and tail sections were allowed to regenerate the missing half. Upon retesting, each of the regenerated organisms showed substantial re tention of the conditioned response. It was presumed that this regeneration could take place only through the DNA-RNA template (dioxyribonucleic acid-ribonucleic acid). In later studies, McConnell (33) found that naive planaria, who were fed the ground-up remains of previously conditioned pla naria, were able to perform the same conditioned response. It was hypothesized that RNA from the sacrificed planaria somehow stimulated the appropriate protein synthesis needed for the other planaria to leaxn the response. One of the leading proponents of a protein theory is Hyden (10) who has shown in a series of investigations that an increase of the neural stimulations, sensory or motor, causes an increase of the neuronal content of RNA, proteins, and enzyme activities. Although the end products mediated by this synthesis can be many, Hyden suggests that the neural frequency-modulation associated with experiential events, as in learning, may specifically alter the pro duction rate of HSA, thereby, making MA a likely means for carrying memory. To provide evidence that there are quantitative RNA. changes specific to learning, Hyden carried out a series of experiments on the vestibular apparatus of the rat. It was thought that by having rats learn to climb a wire to obtain food that this training would specifically stimulate the vestibular apparatus under conditions of learning. For controls, one set of rats were restricted to small cages and maintained on the same diet as the experimental group while another control group received passive rotatory stimulation on a specially designed apparatus, i.e., the animals did not perform any muscular activity which was purposeful yet the task still stimu lated the vestibular apparatus. A third control group was stimulated under ’ ’ stressed" conditions. This group was situated in a vertically placed wheel with a pathway of wire netting. The wheel was then moved back forth for thirty minutes through a 0° to 180° arc. The animals showed that which was considered to be clear symptoms 18 [ of stress. In a fourth type of control, cellular material was taken I r from another part of the brain, the reticular formation, in the : learning animals. A significant increase in the amount of RNA per Deiters' “nerve cell was found in the learning rats. With reference to nuclear RNA changes, a clear increase of adenine and a decrease in the uracil value giving a significant shift in the ratio of adenine to uracil, was found. There were no changes in the cytoplasmic RNA composition. The production of nuclear RNA fraction with an increased adenine to uracil ratio ceased within twenty-four hours after the experiment was terminated. In contrast, the four controls showed a significant increase in the amount of neuronal RNA in the vestibular nucleus, but no base ratio changes. A difference in glia was found in the adenine and guanine values. Glia RNA changes were similar but not identical with the nuclear RNA changes in the neuron. The adenine value was increased but the uracil value remained unchanged, thus giving a significant rise of the ratio. No significant base ratio changes were found in the glia RNA of controls. On the basis of these findings and others not reported here, Hyden proposed the following theory: Let us assume that the material background of memory consists of a spatial arrangement of molecules and intra molecular properties in the brain cells. Patterns of electrical currents are assumed to be part of a mechanism of their formation, but not the part which secures the stable, long-lasting memory formation, since memory can survive a stoppage of the electrical activity of the brain. 19 Can we visualize a molecular mechanism for learning and memory in brain cells that is linked to the mechanism for coding of genetic information that we have just dis cussed? I would like to bring up two aspects of such a memory mechanism. The theory is based on the results of analyses of the RNA and protein production and enzyme activities of nerve and glial cells, after neural stimu lation, and in the rat-learning experiments I have al ready discussed. The concept of the neuron, plus the surrounding glia, as the functional unit of the nervous system is important. Based on the data, the neuronal glia is considered as part of the unit which can regulate the induction of the neuronal RNA and protein production... I. The Selection Mechanism. ...What happens when an animal is placed in an acute learning situation for which there is no precedence in its life? I would suggest that the time pattern of frequencies set up in the neurons in volved leads to a release of repressed regions of chromo somal MIA. This leads to a production of highly specific DNA - copied RNA. In its turn, the RNA synthesis occurring on the demand of the situation gives, as an end product, specific proteins in the neuronal soma. The presence of these proteins or, at later stages, their rate of produc tion, lends to an activation of the transmitter substance. The next time the same modulated frequencies enter, these specific proteins answer with a rapid reaction, leading to an activation of the transmitter substance. In analogy to the mechanism of an antigen-antibody reaction, the specific proteins react to the same modulated frequencies which first lead to the release of the chromosomal activity, the synthesis of the DNA - dependent RNA, and their own formation. By such a mechanism, there could occur a chemical specification of neurons situated within the phylogeneti- cally given pathway where the stimulus entered, and also specification of millions of neurons situated outside of the first area. The question of whether a certain neuron may respond to a specific stimulus or not, then, depends on how similar in temporal patterns this stimulus is to the modulated frequency which specified the neuron the first time. It will also depend on the rate of production of the specific proteins in the neuron. If one of the two factors fails, this will lead to a weak reaction, or to no facilitation. ...A chemical specification of neurons in learning in volving synthesis of chromosomal RNA and specified proteins reacting on modulated frequencies in millions of neurons in different parts of the brain, stronger in some, weaker in others, would also fit the concept that a complicated task is learned and remembered, not as a series of bits, but in whole contexts. More easily specified areas of the „brain, for example the hippocampus in mammals, would get a more dominant position in learning, II. The Glia Function. An activation of chromosomal parts with synthesis of RNA and specific proteins is an induction process, involving feed-back systems. I would like to suggest that, in the learning and memory mechanism, the neuronal glia regulate the induction of RNA synthesis in the neuron, being an integrated part of the memory mechanism, also. The kinetic studies of enzymatic activities and RNA synthesis in the glia around the Deiter1s nerve cells have definitely shown that these two types of cells are ener getically linked... It seems that the glial RNA can react more rapidly than does the neuron, and this response prob ably precedes that of the neuron, also. In an acute learning situation, the modulated fre quencies set up by the neuron are also transferred to the glia. The glia are characterized by potentials of a 500 - 1,000-fold longer duration than those recorded from nerve cells. When the neural frequency is changed, a lock-in effect brings the slow frequency of the glia in synchrony, the difference being a multiple. This coupling of the frequencies of the neuron and the glia forms an informa tion system. The glial ionic equilibrium is disturbed and substrates in the form of nucleotides are transferred from the glia to the neuron to release the repressed chromosome region and induce the necessary enzyme syn thesis for the RNA production. This lock-in mechanism would, therefore, constitute the information system whereby the specific RNA was synthesized, triggered, or mediated by the glia as a regulator. (10:228-231) Two considerations in this theory that have received criti cal attention are the mechanism of modulated frequencies and the question of whether the RNA changes observed are merely associated with performance and not with learning. Burden's theory must await further supporting evidence from neurophysiological research before these issues can be fully resolved, Krech and his associates (31) have studied the biochemical correlates of learning using a slightly different approach. They have spent considerable time studying the quantitative changes of acetylcholinesterase (ACh£)~-an enzyme which deactivates the well known neurotransmitter acetycholine (ACh)--at the cortical level. Since reliable assay techniques for ACh were not available these investigators had to rely on an indirect measure of ACh activity, i.e., by measuring AChE. Initial findings supported the hypothesis. The basic assumption was that overall cortical levels of AChE activity are positively related to efficiency of synaptic trans mission and positively correlated with problem-solving ability. Findings from later studies (76), using a variety of learning tasks, tended not to support the hypothesis. Although correlations were quite low, there appeared to be a general tendency for the number of errors in maze learning to correlate positively with AChE concentrations. If their hypothesis were correct, the correlation should have been negative. As a result, these investi gators concluded that there is no necessary relationship between measured AChE levels and learning. Electrophysiological Changes It is fairly well accepted that the discharge rate of a neural impulse is regulated by two conflicting electrochemical forces— that of inhibition (hyperpolarization) and facilitation 22 (depolarization}. TPMs process has been well demonstrated in I | studies concerned mith. the tonic stretch reflex as exemplified by : single fiber preparation. As a muscle is stretched, muscle spindles i and their acconpanyimg neuron produce an "excitatory drive" in the form of a barrage off activating impulses per unit time. The outcome is a net depolarising current, across the cell membrane of a ventral horn cell, which is connected in series with the muscle spindle unit. This process is opposed by repolarizing forces (e.g., orthodromic inhibition) from affffererats over intemuncial cells and natural inhi bition initiated by the firing tonic ventral horn cells. The outcome is a net polarizing cnnrremt; therefore, the normal frequency of dis charge Fn, as proposed by Granit (5)» will be some function of the net depolarizing cmnnremfc, which is the algebraic sum of the two op posite forces of = fCP^p + Ppol) where f refers to strength of discharge of depolarisation; currents (or pressures) and polarization currents. Since Bedes has found that for some reason a constant polarization pressure does not deliver a constant inhibitive effect, a concealed factor that is not represented by this formula must still be present. Steeles suggests that temporal summation may be the answer. The important point to be made here is that regulation of current flow may he discussed in purely electrophysiological terms. It is partly on this basis that Hebb (8) formulated a most exquisite theory of behavior. Hebb*s theory of behavior has received widespread attention. This is possibly because it was proposed in such a global manner that the basic constituents of both S-R and "holistic” theories were par tially retained and integrated with neurophysiological concepts. Hebb's theory is also in accord with and influenced by the neuro physiological concepts proposed by Lashley. Hebb assumed that a growth process accompanying synaptic ac tivity makes the synapse more readily transversed. Hebb's "first stage of perception” was based on the following axiom: ]/flh em an axon of cell A is near enough to excite cell B and repeatedly or persistently takes part in firing it, some growth process or metabolic change takes place in one or both cells smch that A's efficiency, as one of the“Cells firing B, is increased. (8:62) Hebb has extended the concept of a linear association of conditioned learning so that it includes associations of afferent fibers of the same order— that is, sensori-sensori association; therefore, two different afferent fibers, which once acted inde pendently can now, by process of repeated stimulation and altered growth, perform the same function in a dependent relationship. Hebb has said that the "fundamental meaning of the assump tion of growth is the effect this would have on the timing of action by the efferent cell. A fiber of order N thus gains increased con trol over a fiber m + 1 making the fibers n + 1 more predictable" (8:73). Elaborating upon this concept, Hebb proposed that the first stage of perception involves a simple assembly (cycle) of cells where it is possible to demonstrate an "alternating" reverberation of neural impulses which is not extinguished as readily as a simple | closed circuit. This system inherently involves equipotentiality (a concept popularized by Lashley) so that brain damage might remove 1 some pathways and still not prevent the system from functioning. With perceptual development there would be a slow growth in the assembly. Hebb's next phase of perception is described as a phase sequence. Several activities may coexist and be aroused in any order by determinate effects of cell assemblies. "Activity in a superordinate structure is then best defined as being whatever de terminate organized activity results from repeated activity in the earlier developed or subordinate structure giving rise to it.1 1 (8:98) The phase sequence (phase cycles) in perception is an ideational series with its motor elements; i.e., the basis of per ceiving a distinctive total figure involves a sequence of cortical events with motor components. Perception of an actual object involves more than one phase cycle; i.e., it involves a heirarchy of phases, phase cycles, and series of cycles; therefore, two ideas or concepts to be associated might have one or more subsystems in common. Hebb "proposed that all learning tends to utilize and build on any earlier learning instead of replacing it so that much earlier learning tends to be permanent, therefore, learning may be facili tated or limited and canalized by it" (8:109). Perhaps the most essential evidence needed to substantiate Hebb's theory was further confirmation that a reverberatory circuit | does, Indeed, exist in neural tissue. Some evidence of this sort was provided by Lorente De No (29) in 1938, when he investigated the ! activity of intemuncial chains. In discussing multiple chains of intemuncial cells, Lorente De No suggested that intemuncial bom bardment can constitute a statistically constant stimulus for the motoneurons of the active group. The frequency of response was thought due to the strength of the bombardment and the rate of re covery of the neuron. Bums (25), in 1951 and 195^» provided evidence which tended to corroborate Hebb's theory by demonstrating a reverberating circuit (Bums referred to this phenomenon as "after bursts") within cortical tissue. It was shown that a few strong stimuli could produce a series of neural activity lasting many minutes after the initial stimulus. If ten stimuli were given to the cortical surface at three second intervals, responses of identical type continued to spread for as long as one hour. A linear response was found between the number of stimuli and after bursts and between the strength of stimuli and after bursts. In conclusion, Bums states that: The experimental results have been interpreted as indicating that after activity, whether driven or spon taneous, the deep somatic ends of neurons conducting the burst response repolarize more slowly than do the super ficial ends. It is supposed that current flow between superficial and deep ends of these structures, during recovery from activity, is responsible for their subsequent spontaneous discharge. (25{^5-^6) Gerard, in 1955, said "...the notion of neurons connected in a network, not merely in linear series, and of nerve impulses 26 passing about the connections in a circular, more or less continuing fashion, is now widely accepted" (41:229). Since recent findings also tend to fall into the same pattern, the concept of reverber ating circuits appears to have been firmly established. On the basis of present research findings Halstead and Rucker (49) have postulated a three phase model of memory: dynamic, inter mediate, and permanent. These investigators, as do many theorists, suggest that reverberatory circuits may account for memory that is relatively short lived. Long term memory, that which is relatively fixated and permanent, may be related to specific structural and biochemical changes at the cellular level. The intermediate phase would, perhaps, relate to the transitional period where neural im pulses are perpetuated in circular units for a period long enough to trigger specific structural and metabolic changes in the stimulated neurons. It appears that the views expressed in the foregoing review are not as diverse as might at first be suspected. In fact, the difference among many neurophysiological constructs which attempt to explain learning and memory phenomena, is, in part, a function of the experimenter's focus of study, e.g., Hyden studied biochemical adaptations at the neuronal level; Eccles has been primarily con- „ cemed with the electrophysiological changes associated with behavior; while Hebb has been more involved with the electrophysiological changes in neuronal activity associated with perception. In spite of this, the theories reviewed in this chapter exemplify the definite j progress which is being made toward determining the mechanisms i I which underlie and account for learned behavioral change. Part II; Theory of Consolidation Although relevant to any of the aforementioned theories, the consolidation theory has been often linked with the works of Hebb. In this theory, it is suggested that after a learning period, a continuing process of activity goes on in the central nervous sys tem and because of this a memory trace may be more firmly fixed. It is also suggested that the memory trace, during this period, is ex tremely vulnerable to disruption. The amount of disruption is, in part, a function of the time which elapses between the learning period and the period of disruption, i.e., the shorter the time in terval the more serious the consequences of the disruption. Over the years, the consolidation theory has been under in vestigation by several related fields of research, thus it has been subjected to a variety of experimental techniques. The problem is unique in that evidence provided by the neurophysiologist may be conveniently tested by using typical behavioral and psychophysio- logical experimental procedures. Because the consolidation theory has been subject to such widespread investigation, the following re view of literature about it represents only a selection of the more significant experiments, presented in chronological order. 1900.— It has been reported that Muller and Pilzecker (12), while studying the effects of interpolated activity on intertrial transfer of recently learned information, found that the earlier an activity was interpolated, the greater the degree of retroactive inhibition. It was concluded "that following original learning there was a gradual diminution of the physiological processes which served to intensify the established associations and further that these processes were inhibited by the interpolation of activity when given immediately after original learning" (57:25). Thus, Miiller and Pilzecker are given credit for having first proposed a "perseveration theory" of learning; therefore, the concept of con solidation commonly associated with Hebb's work was actually proposed at a much earlier date, 1901, — According to Glickman, McDougall called attention to the perseveration theory in attempting to explain the phenomenon of retrograde amnesia, i.e., the type of amnesia which is commonly ob served among individuals suffering from major head injuries (e.g., concussions), where memory of recent events is severely disrupted (1J4:219). 190^. — Glickman also reported that Burnham appears to be the first individual to discuss the relationship between retrograde amnesia and memory consolidation at length. It was speculated that: a) the time required for memory consolidation may vary among indi viduals and experimental conditions, b) shock may arrest the per severation process, c) such shock may be the result of fatigue, excitement, unconsciousness, or narcotics, d) the extent of retrograde amnesia is related to the elapsed time before inter ruption, and e) autonomic neural activity is an important factor in fixing impressions (44:219-220). 1913* — Ebbinghaus (4) apparently performed the very first experimental study of memory (10). He observed that forgetting is a function of time and maintained that all curves ■which demonstrate retention or memory are characterized by negative acceleration. 1938.— It has been reported that electroconvuLsiwe shock therapy (ECS) was introduced by Cerletti and Bini (10). Rrior to this time there was only indirect psychological evidence of a neural fixation process. ECS therapy provided a technical apparatus for laboratory study of retrograde amnesia. 1941. — Glickman gives Zubin and Banera credit for being; the first to systematically investigate the effects of ICS can retrograde amnesia; unfortunately, time relationships were not adequately de fined (44:220). 1949. --In most studies concerned with the retroactive effects of ECS on learning, Duncan (32) is given credit for providing the classic investigation which triggered subsequent experimentation. Duncan was concerned with the retroactive effect of ECS on learning as a function of the time elapsing between the tennimation of a single trial and the application of the ECS treatment. An avoidance task was used where the animal escaped from am electrically 30 charged grid to a safe cnmjgaigtmBnt. The assumption mas that the sooner the animals were give® ECS after each trial the greater the time mould be for the amoral, it® perform the avoidance response on the succeeding trial. Experimental grumps, given time intervals of twenty seconds, forty seconds, sixty seconds, and four minutes between each trial and ECS treatment, performed. significantly worse than the control group, which was not subjected to> ECS. These results were not true of groups given time intervals of one, four, and fourteen hours. To test the hypothesis that ECS is merely a nociceptive stimulus, Duncan repeated the tests by shocking the animals through the legs instead of the head. Fat only were no differences found among groups but the animals were noted to behave much differently. Those animals shocked ithrnragfe the legs struggled and squealed on successive trials in anticipation of the shock whereas the animals shocked through the head did mot. It was concluded that in those studies where disruption car disorganization of memory traces is emphasized the consolidaticm theory of memory is better illustrated than the competing theory off habit interference. Bussell, Braun, and fathom (72), that same year, provided evidence indicating that the effects of ECS on learning were, in part, a function of task difficulty. It was found that ECS caused decrements in learning and retention in water mazes with five-choice points, but not with cam. The number of eaqgeriments conducted during the ensuing 31 years was voluminous. Many were quite similar in technique, with differences being only in the experimental variable under investi gation. For this reason, the writer chose to advance ahead a few years to the work of Gerard. — Gerard (41), experimenting with hamsters, reported that shocks following trial runs by an interval of four hours or longer provided as good learning curves as when no shock was de livered. One hour produced some deficits, fifteen minutes produced serious deficits, and five minutes yielded no measurable learning whatsoever. it was also reported that hamsters kept at lower tempera tures during the interval between the learning experience and electroshock showed as great a disruption of learning at an inter val of one hour as warm ones did at an interval of fifteen minutes. Gerard therefore proposed that "if circuits enduring for minutes or hours may be required to fix experience in memory, perhaps circuits reverberative for seconds or fractions may be necessary even for the initial consciousness of an experience— or of some inner trig gered awareness" (41:229). 1957.— Leukel (52) explored the possibility that consolida tion may be affected by inactivating the central nervous system— for example, by low body temperature, ECS, general anesthetics— or by disrupting memory traces during periods when they are most vulner able, just after registration of the new experience. A fourteen 32 unit Stone multiple T water maze was used as itfce CTqperfimfflirtal ap paratus. ECS and a general anesthesia (sodium pmnltottftalj was found to decrease the rate of habit acquisition. ECS tobs riTfeetive up to two hours after the trials whereas sodium pentotMl produced both a decreased rate of error elimination and rate of Imganwamsit, when administered one minute after training but not tMrfy minutes after the training period. 1958. — Thompson et al. (76), in analyzing i f c f i n © differential effects of ECS on memory in young and adult rats, fgrnmcfi that thirty day old subjects made more errors than fifty nr siariiy day old sub jects. Forty day old subjects were more severely affected by ECS than fifty day old subjects but their error scores amply marginally exceeded sixty day old subjects. Deficits of sixSy day aid rats appeared to be of the same magnitude as that foanmd 3m 20© day old rats. No significant difference was noted foefcwBsm ECS treatment and either a) light or dark reared rats or, bD anmmals suffering from hypothyroidism as compared to normal animals. It appeared that differential metabolic activity of the brain did mot account for the age effects. The increased effects of ECS on memory impairment, of the young was thought to be related to the differential rate of matura tion of brain tissue; e.g., in rats myelinatton < n f f the brain is not completed until the age of fifty days. Thompson amd his colleagues assumed that the amount of myelinated neurons oEjrweBatedS with the 33 number of functioning neurons. As an additional experiment thirty- two animals (ages sixty-five to one hundred days) underwent brain surgery, where brain damage was surgically induced. ECS in this case caused an increase in memory loss which was comparable to those defi cits experienced in normal young rats. For this latter experiment, the authors’ concluded that since brain damage increased memory loss, and since myelination is complete at about fifty days, the greater deficit in young rats was related to the presence of fewer func tional neurons within the brain. Since ECS treatment was usually administered via alligator clips attached to the ears or outer layer of the skin surrounding the skull, Glickman (43) investigated the effects of direct electri cal stimulation of the brain on avoidance learning. The stimulus was applied to the ascending reticular formation (ARF). It was found that by stimulating the ARF, after presentation of an unconditioned stimulus in an avoidance conditioning situation, an increased re sistance to extinction was noted in the stimulated animals as com pared to non-stimulated controls. In explaining the results, a rival hypothesis to the memory consolidation theory was suggested, i.e., stimulating the ARF may have disturbed the process of atten tion thereby causing the observed deficits in avoidance learning. I960.— Coons and Miller (26) have seriously opposed the consolidation theory as an explanation for retrograde amnesia pro duced by ECS. The first part of their experiment was merely a replication of Duncan's earlier work. Although similar results ; were found, it was also noted that the animals receiving ECS exhibited l increased fear, this being determined by measuring changes in urina tion and defecation rates. The second part of the experiment involved training the rats so that they stopped performing an avoidance response, i.e., the electric shock in the starting box was turned off and a shock was introduced into the goal box. Under these conditions, it was found that learning the avoidance response was faster the sooner the ECS followed each trial. Generalizing from the findings, these investi gators suggested that: When any effects of either amnesia or conflict would hinder learning to make anticipatory runs, learning is poorer the shorter the interval between each training trial and the ECS. (26:527) When the effects of conflict would help and amnesia would hinder learning to stop the avoidance response, learning is better (as indicated by longer avoidance times) the shorter the interval between each training trial and the ECS. (26:529) As the authors pointed out, their findings did cast serious doubts on the conclusions drawn from previous studies with regards to the role of ECS in producing retrograde amnesia. 1961. — If the period of consolidation depends on a per sistent state of neural activity, perhaps enhancing impulse propa gation in the central nervous system might facilitate learning. Based on earlier evidence that low doses of the neural stimulant, strychnine sulphate, injected shortly before daily training trials, facilitated learning in rats, Breen and McGaugh (23) set out to 35 determine if mild injections of picrotoxin— known to have pronounced effects on brain cortex— would similarly facilitate learning if in jected immediately after a learning period. Picrotoxin was found to facilitate the learning of a maze task. It was hypothesized that any stimulant might have facilitating effects, varying directly with intensity. That same year, Pearlman et al. (62) took exception to the manner in which prior investigators administered anesthetic and con vulsive agents to induce retrograde amnesia. In this experiment a more precise control was exerted over the interval between the learning experience and the memory disrupting event. This was achieved by using a one-trial learning procedure and permanent in dwelling, intravenous catheters. Rats were conditioned to a bar pressing response for water 9 ten minutes a day until a stable rate was attained. After the cri terion was reached, an avoidance response was established by elec trifying the bar and reward nozzle. Three groups were anesthetized with ether ten seconds, five minutes, and ten minutes after shock; surgical anesthesia was produced in about thirty-five seconds and the animals recovered in ten minutes. Four groups were administered sodium pentobarbital at approximate3y twenty seconds, five minutes, ten minutes, and twenty minutes after shock; surgical anesthesia was produced within ten seconds after injection and persisted for one hour. In five groups, pentylenetetrazol was injected through cathe ters about twenty seconds after shock or intraperitoneally two hours, four hours, eight hours, and four days after shock. All injections of this drug produced seizures for about four minutes. The experimental groups were tested for avoidance response twenty-four hours after shock except the pentylenetetrazol group, which was tested one day after being convulsed. The following con clusions were drawn from the findings; 1. Surgical anesthesia severely impaired retention of the avoidance response if the anesthesia were induced within ten to fifteen minutes after the learning trial. 2. Intravenous pentobarbital was more effective than ether. 3. A single pentylenetetrazol convulsion abolished the avoidance response when as much as eight hours inter vened between learning and treatment and even to four days. 4- . It was concluded that the consolidation process, at least in initial stages, is incompatible with a state of surgical anesthesia. 5. Interference with memory by convulsions occurring hours or days after the learning trial probably depends upon a different mechanism, related to the clinically ob served sequelae of convulsive treatment, which are characterized by various transient derangements in cluding confusion and memory lacunae. (52: 111-112) To account for conflicting findings reported elsewhere, Pearlman et al. expressed the belief that "...it is probable that a general anesthetic can have two effects on memory. When given after consolidation is more or less complete, the anesthetic may reduce interference and retroactive inhibition and thus conserve the memory trace; when given .within a few minutes after the learning experience, it may block the consolidation process before the memory has attained stability and permanence" (62: 111). 37 Madsen and McGaugh (55) replicated the experiments (a dif ferent experimental task was used) reported previously by Coons and Miller. Animals were given one BCS after making a response which was punished by shock to the feet (stepping off a platform onto an electric grid). If SCS should affect performance by inducing fear then the animals given an ECS following a response punished by shock should not tend to make the punished response on a subsequent trial. The results indicated that there was a significant tendency for the animals which received the BCS after stepping off the platform to step off again twenty-four hours later; thus, amnesia was indicated making the results in conflict with those reported by Coons and Hiller. Glickman (*J4), during this period, provided an excellent historical review (referred to a great deal in the early portion of this review) entitled "Persevative Neural Processes and Consolidation of the Memory Trace," in the Psychological Bulletin. Deutsch (30), just one year later (1962), also published an excellent review en titled "Higher Nervous Function; The Physiological Bases of Memory" which appeared in the Annual Review of Physiology. These articles are highly recommended to those readers interested in gaining further insight concerning neural mechanisms of memory. 12*.- -As pointed out earlier, certain drug stimulants injected immediately after the learning experience facilitated the rate of learning and increased memory. Westbrook and McGaugh (83) decided to test whether the drugs merely acted as positive 38 reinforcers (reward stimulus} car w&efefier the drugs actually enhanced the consolidation process. The hypothesis was that if drugs rein force rats, given doses mithsmit other reinforcement should produce improvement in performance; if dungs only enhance memory storage, the animals should learn more tamt mat improve in performance until the reward is given; the animals receiving both the reward and the dose should be superior ibo ©nmittannills. The drug used was a strychnine like compound. A six-unit alley 10 anaa© was used as the learning task. The animals were given one trial per d^y for ten days. The groups were described as follows: Group I Reward {food} — Reward (drug) Experimental (RRE), Group II Reward ([food} — Reward (control solution) Control Group III First five trials no food reward (NRE - drug injections}, second five trials - RRE (control injections only}. Group IV First five trials no food reward (NRC - control injections)), second five trials - RRC (control injections only)). It was found that sgnmap I was significantly better than group II (compared trials two amffl ten)'; during the first five trials the mean error of the two rewarded groups decreased while those of the nonrewarded subjects remained fairly constant; for tests between trials two and five the mean errors of group I was lower than 39 | group II but means of groamp 331 and IV were not significantly dif ferent; for trial six and item comparisons there was a significant drug effect for group I and 33T as compared to groups II and IV. The findings favored the cqnsoliidation hypothesis, i.e., post-trial injections of a mild stimulant {175? I.S.) facilitated latent learning as well as ccmveratausmal maze learning. 1965. — The theory ©f consolidation, after achieving wide spread acceptance, became a rival hypothesis to reactive inhibition (Ijj) in explaining increased performance scores (reminiscence) fol lowing a rest interval. The reactive inhibition hypothesis follows the Hullian theory that t n wfoggiigwer- a reaction is evoked in an organism there is created as a result a primary negative drive; this has an innate capacity (I^) to imMMt the reaction potentiality (3%) to that response” (9:300). According; to this hypothesis reminiscence would result from the dissipation of 1^ during the rest period, Eysenck (36) has argued that reminiscence results from a combination of consolidatioaa and depressed Ijj, After memory is well ingrained, then I^ is believed to play a more important role. Rachman and GroSSi {&£>)) set up the following experimental design, using a pursuit rotor task (human subjects), in an attempt to analyze the factors of 3 j j j amd consolidation on reminiscence. All practice periods were analyzed in terms of ten second trials; therefore, five mmnmibes of practice was the equivalent of thirty trials. • Reminiscence was calculated by determining the dif ference in performance between the last ten seconds prior to the 4o ! rest period and the first trial of the post-rest period. More specifically, the eaqjerhmental design was as follows: 5 Min. 10 Min. 4 Hr s. 2 Min. Massed Practice Rest Rest Massed Practice " R.C.*/rest « — » « ' » w rest/R.C. / rest " " M " ” rest/R.C. 1 1 " " ♦Reversed cue test for three minutes. It was found that the mean score for the control group differed significantly front those of groups A, B, and C. The means for groups A and B differed significantly, while group C did not differ from A or B. In all, the control group exhibited normal reminiscence; group A showed no reminiscence; groups B and C ex hibited an intermediate amount of reminiscence. Since groups A, B, and C had equivalent amounts of rest, Ijj should have been factored out in comparisons made among these groups. Therefore, it was suggested by the findings that consolida tion is an additional factor to be considered in explaining remi niscence. It should also he noted that if a pure interference theory were to account for the differences among groups A, B, and C one would expect the outcome to be independent of the temporal point of interpolation, i.e., the scores among groups A, B, and C should not have differed. That same year Gottlieb (45), working with human patients, reported evidence of temporary disruption of verbal memory while Control Exp. A Exp. B Exp. C ! 41 ! | ! j localizing ECS treatment to the left or right cerebral hemisphere. t | Of major significance was the finding that ECS to the left parieto- i | ; temporal region retarded verbal recall more than ECS to the right ; corresponding region or to the frontal lobes. Thus, the notion i that the left parietal and temporal lobes are more intimately in volved in verbal memory functions than the other regions tested was supported. "Recall of right-sighting right-handed patients was swifter than non-right-sighting right-handed patients when both received ECS to the right hemisphere, extending the previous indi cation that contralateral cerebral aspect of sighting dominance is involved in the verbal process" (45:368). 1968.— Buresova et al. (24) studied the effects of ECS in two learning situations— original learning and its reversal. Several conditions were considered— namely, short and long intervals between original learning and reversal, with or without overtraining. ECS was found to impair retention of original learning. Impairment in retention of reversal learning was greater when overtraining was given in the original learning and when the learning-reversal inter val was increased from one to twenty-four hours. In summary, the authors give notice to other experimenters by stating: It is therefore concluded that the disruption of the consolidation process by ECS may always be present in animals but that experimenters have not always used situa tions and measurements appropriate to finding it. (24:159) Franchina and Moore (31) changed prior procedures a bit by investigating the effects of strychnine on the inhibition of a UrZ previous performance, After training animals to jump a hurdle on their way to a goal box, an inescapable foot shock was administered in the goal box to all animals after which (fifteen minutes later) the animals received an intraperitoneal injection of either various concentrations of strychnine (experimental groups) or ninety-five per cent saline (control group). All experimental groups except one, which received the lowest concentration of strychnine, exhibited significantly different delay and longer delay times in crossing the hurdle than the saline group. A follow-up study provided findings which allowed the authors to reject the hypothesis that strychnine merely acted as a punishing stimulus; therefore, the findings were consistent with a consolidation theory. Ray and Bivens (6?), experimenting with the magnitude of the reinforcement as well as the interval between ECS and each trial, found that in a passive avoidance task (foot shock with varying in tensities per group) the performance of those animals learning at low intensities of foot shock was impaired more from ECS than was that of those learning at high intensities. It was noted, however, that ECS given within ten seconds of the avoidance trial produced disruptions in learning regardless of the foot shock intensities. In a very recent investigation, Misanin, Miller, and Lewis (58) reported findings that challenge the role of ECS in amnesia: Rats had a memory loss of a fear response when they received an electroconvulsive shock twenty-four hours after the fear-conditioning trial and preceded by a brief presentation of the conditioned stimulus. No such * * 3 loss occurred when the conditioned stimulus was not presented. The memory loss in animals given electro- convulsive shock twenty-four hours after conditioning was, furthermore, as great as that displayed in animals given electroconvulsive shock immediately after con ditioning. This result throws doubt on the assertion that electroconvulsive shock exerts a selective amnesic effect on recently acquired memories and thus that elec troconvulsive shock produces amnesia solely through interference with memory trace consolidation. (58:55*0 It would have been interesting if these investigators had continued with repeated conditioning trials preceding ECS over a longer period of time. Perhaps they have uncovered differential ECS effects due to the presence of two different cortical states of activity associated with memory or learning— one being an elec trically "quiet" state where memory is associated with metabolic or structural changes at subthreshold levels and the other being a "triggered state" where impulse frequencies associated with the aforementioned are fired in a reverberatory fashion. The latter state may be susceptible to disruption by ECS whereas the other may not. Nielson (59)» in a series of experiments, tested the following assumptions: 1. ECS produces a massive inhibition, part of which be comes associated with the environmental stimuli (e.g., room where ECS was given) present at the time of ECS administration. 2. ECS produces amnesia by changing the levels of brain excitability levels. 3. Following ECS, there is a recovery of a passive avoidance response, corresponding to the changes in brain excitability produced by ECS, when the increased activity levels of convulsed animals are suppressed by- ear clips. U r . If amnesia produced by ECS is the result of differences in brain excitabilities existing between the learning state and the recall state, and if, by grid shock and ECS, the threshold is raised prior to step-down training (stepping off a platform onto an electric grid), ECS administered immediately after step-down training should fail to produce a performance decrement. Results of experiment one provided supporting evidence for the first assumption, i.e., activity of rats with attached ear clips was reduced and emotionality was increased. Since on experiments two and three a) ECS seemed to alter threshold levels of brain tissue, and b) recovery of a passive avoidance response following ECS (ninety-six hours after ECS treatment) was demonstrated, Itfielscm suggested that "...ECS does not interfere with memory consolidation but rather, teraporily interferes with memory retrieval mechanisms" (59:12). To test the fourth assumption, the following procedure was used: 1. active avoidance learning for the animal to learo to avoid shock, 2. ECS treatments to induce the various brain states, ^5 3. passive avoidance training at various brain excitability slates| k - m EES treatments, either to disrupt consolidation of memory of the grid shock associated with stepping off the platform, or to match brain excitability states induced by the first ECS treatments, and - 5- test for retention of the passive avoidance response (59:1*0. Based on the findings, it was apparent that ECS effects were changed through time and altered by prior experience. ECS was not found. to produce amnesia. Most similar response latencies (time spent can platform) were obtained from the same grid-shock ECS treat ment. iSielson concluded his experiment by offering the following hypothesis: The hypothesis is offered here that the neurological aspect of learning may involve changes in levels of brain excitability as reflected in the thresholds of functional memral systems, that retention implies a maintenance or reconstruction of these modifications of brain excitability, and that failure of retention occurs whenever brain ex citability is modified away from that established by the training: procedure. (59:3) ' Pirns, Nielson proposed that ECS alters memory retrieval rather than memory consolidation and that performance differences due to EfcfS treatment are related to changes in neural activity. Actually, there is an increasing amount of evidence which suggests that memory and/or learning may be highly specific to a particular level of neural activity. Perhaps a quick review of a "classic1 1 investigation in this area would help to explain this concept which i j is often referred to as "dissociated” learning. i | Overton (6l) in 19&J- investigated a "state dependent" phe nomenon in learning, i.e., in certain cases a response learned under a drug will not reappear until the animal is again under that same drug. The purpose of his experiments was to determine whether state dependent learning could be demonstrated with centrally acting drugs and to study the properties and mechanisms of such state dependence if it occurred. Using a T maze subjects were trained to escape from an unavoidable shock. Experiment 1.— The purpose was to determine whether dis sociation of learning would occur between the nondrug state and drug state using sodium pentobarbital. Results.— In every group, learning was slower during training while drugged than while nondrugged. Performance was also poorer. Learning was found to be state dependent, i.e., one group performed well in the nondrugged state (state of original learning) bait ran domly in drug state and another group performed well in the drug state (state of original learning) but not well in the nondragged state. Other groups tested under a slightly different procedure showed complete dissociation of learning as well. Generalization from one state to another would only take place if the subjects were under light drug conditions. Experiment 2.— The purpose was to evaluate the degree of state dependence by having subjects perform a response in the drug state (six consecutive maze trials— one per day) and then continue on with the same task in the nondrug state. Results. — Virtually no transfer of training occurred between performance under different states. Experiment 2,.— Training trials were alternated under two drug conditions. Subjects were trained to turn one way while 4? dpngged and the other way while nondrugged. These results were compared. with controls who were trained in one drug state only. The subjects were given two training trials per day--one in the mourning (nondrug) and one in the evening (drugged). Under these ctmditicms one would expect negative transfer of learning for the cnantroX and experimental groups. BesuLts. — The experimental groups acquired differential responses, i.e., there was no transfer of learning. Bgperiment b.— The purpose was to determine the relation- sMp between doses of pentobarbital and degree of resulting dis sociation. Results.— Amount of transfer of training increased regularly as the dose used to establish the drug state decreased. Experiment £•— The purpose was to compare the effectiveness off drug states, as agents controlling differential responses, with that off various exteroceptive stimuli, interoceptive stimuli, and states. Results.— It was found that controlled responses under pamtonfearbital were acquired much more rapidly than responses related to other sensory cues, suggesting the existance of a control mech anism wMch is different from one which allows enhanced discrimina- Overton summarized his findings by postulating three features off state dependent learning: 1) heavy doses of pentobarbital produce complete dissociation of learning between nondrugged and drugged states; 2) complete dissociation only occurs in extreme forms of conditions, i.e., the more similar the drug states, the more com plete the transfer of learning between states; and 3) learning ■ i f t m r i i m g; the drugged state does not appear to be based on changes in sensory cues. Thus, there is some evidence to suggest that learning is associated with specific changes in neural activation. Although findings by Nielson, Overton, Kisanin et al, do not fully discount a theory of consolidation, the usual interpretation of the exact role consolidation plays in memory has certainly been challenged. In summaiy, it appears that: 1, Memory is more vulnerable to disruption or retroactive amnesia the sooner a stimulus, which tends to depress or disrupt neural activity, is introduced after the learning period; 2, Memory can be facilitated by introducing stimulants, which tend to heighten neural activity, immediately after the learning experience; 3, Memory consolidation is more easily disrupted in the less matured nervous system; h. The amount of disruption in memory consolidation is, in part, a function of: a, the nature of the disrupting stimulus, b, the complexity of the task, c, the magnitude of the reinforcement stimulus, and d, the active state of the nervous system. Among the types of stimuli that have been reported (not necessarily in this review) to either directly or indirectly affect memoiy consolidation are: ECS, drugs (stimulants and depressants), temperature changes, RNA inhibitors, RNA facilitators, anoxia, brain stimulation (subconvulsive), changes in carbon dioxide concentrations, 49 fatigue, sleep, and interpolated tasks (e.g., reversal cues). Since the purpose of this review was to inform the reader of the merits and faults of the consolidation theory, certain investigations with experimental procedures more directly related to this study were temporarily by-passed. These will be discussed now. Belated Investigations Three studies were found that had implications for the inves tigation under consideration. They were as follows: Corey (27), using an elevated maze with eight cul de sacs, investigated the relationship between compulsory physical exercise and the ability of the white rat to learn and relearn the maze task. The exercise task involved running in a rotated wire cage at a speed of %5<D0 feet per hour (less than one mile per hour). The procedure called for five trials per day on the maze task until a criterion score of five consecutive errorless trials was obtained. Rate of learning was based on 1) the total number of errors before the criterion score was reached, 2) the total number of trials, 3) the total number of seconds spent in maze, and 4) the total amount of active time while running the maze, i.e., the time elapsing while an animal was not moving was not counted in the total time. The daily order of procedure for the exercise animal, was 1) testing on the maze task, 2) exercise in the rotating drum, and 3) feeding. The exercised groups received three weeks of exercise prior to the experiment. 50 It was indicated from the findings that, although the exercised animals made better scores, the difference was not sig- ! nificant at the .05 level. Among a number of statistical compari- i | sons of scores, no significant difference was found between the exercised animals and the control animals. The investigator con cluded that compulsory exercise did not have a consistent or no ticeably peculiar effect upon learning. Unfortunately, the exercise task in this study was poorly controlled. As Corey stated: "Regardless, however, of whether the animals were running or were clinging to the wire floor for a com plete revolution, they were exercised" (27:293). In this case it is obvious that the exercise task was extremely moderate and that animals could actually rest by clinging to the floor and riding the i wheel. Hutton (refer to pilot investigations— Chapter III) found : that some rats were capable of running well over three hours at one mile per hour with only moderate preliminary training. It should also be noted that the control groups were not well equated with the exercised groups, since the control groups did not receive the additional preliminary training. No direct reference was made to the amount of time elapsing between the exercise task and the maze task. Gray (^6), using a multiple-U maze with six choice points, studied the effects of forced activity on maze performance of white rats. The animals were exercised on a motor driven revolving drum at two different speeds— ten and twenty revolutions per minute (both 51 speeds were under one mile per hour). Shock reinforcement was used in the running task whereas shock or food reinforcement was used in the maze task. The regular procedure was to place the activity animals in the maze immediately after running. One maze trial was given daily for a minimal period of twenty days (for the maze task). Data were in the form of errors and elapsed time. No reliable difference among groups was foiwd, although the activity animals had average error scores slightly superior to con trol animals (found to be reliable when all groups were combined). Gray concluded that: 1. forced activity produced no deleterious effects on the health of white rats or on their maze performance, and 2. forced activity provided slightly beneficial effects on maze learning, as evidenced by error scores. Although the procedure used in Gray's study was not similar to this investigation tinder consideration, the findings were of some interest. Duffy and Freeman (31) (38) have long stressed the im portance of an intensity dimension in behavior, fhsey believe that there is a lawful relationship between the level of arousal and per formance, i.e., there is an optimal level of arousal for maximum performance. Findings by these authors, and others, have shown that pre-task muscular exertion can enhance the quality of performance. In the case of Gray's experiment, even though a fear animals ran for considerably long times (up to four hours), the pre—task exercise 52 may have been moderate enough in intensity (speed) to enhance the performance of the experimental groups by favorably altering their level of arousal. Hinami and Dallenback (57) investigated the effect of forced activity upon learning and retention. Cockroaches, oddly enough, were exposed to treadmill running for ten, twenty, or thirty minutes following a discriminatory (avoidance response) learning task. Shock reinforcement was used in both tasks. It was found that learning and retention were poorer in the exercised group regardless of the duration of exercise. By intro ducing longer recovery periods in an additional experiment, it was found that relearning did not vary directly with the length of the recovery period, i.e., exercise administered immediately after the learning experience (thus, providing a longer recovery period before retesting) also seemed to have an adverse effect on retention. The investigators concluded that the interpolated activity was a factor in determining the amount of retroactive inhibition. It was assumed that two factor's determine 1^— an "anticonsolidation factor" and an excitement or irritability factor. It was proposed that the latter factor becomes less effective as recovery time increases. Thus, few studies were found which considered the retro active effect of strenuous exercise on memory consolidation. In the references noted, the exercise task appeared either too moderate to produce the hypothesized depressant effects on central nervous system activity or the exercise task was not properly controlled. 53 A wealth of information related to the effects of fatigue on human performance is provided in journals concerned with motor learning. Surprisingly, no study was uncovered which was directly concerned with the retroactive effects of strenuous exercise or "fatigue" on learning. For example, investigations by Alderman (19), Benson (21), Geddes (*M)), Gutin (4?)* and Nuimey (60) were, for the most part, concerned with the pro-active effects of either local and/or general fatigue on performance or learning, i.e., either the subject was exposed to the learning task immediately after strenuous exercise or the exercise task was interpolated between trials with no provision for adequate recovery. There seems, therefore, to be a definite need for further research concerning additional conditions which may affect the rate of learning a perceptual-motor task. It should be noted that Gutin did find local fatigue, ad ministered by exercising the "test" arm to exhaustion immediately after a pursuit rotor task, had no effect on the learning of that task even though adequate time was provided for recovery from fatigue. The basic assumption underlying the hypothesis of this dissertation was that exercise tends to alter neural activity of the central nervous system (CNS). As mentioned earlier, there is overwhelming evidence to suggest that moderate exercise facilitates CNS activity (heightened arousal) and, in many cases, learning as well (^7). But, what about the effects of strenuous exercise on CNS activity? i i | Although not so prevalent, there are some investigations the i : results of which suggest that exhaustive exercise may have depressing i ! effects on CNS activity. If this be true, then exhaustive or stren uous exercise of a general nature may have a detrimental influence on memory consolidation if it is introduced immediately or within a short period after the learning experience, i.e., the rate of memory consolidation would be depressed. In the following review evidence that strenuous exercise may lead to physiological changes which tend to depress CNS activity is considered. Part III; Strenuous Muscular Exertion and Activity of the Central Nervous System It has long been known that a decrease in blood pH and cerebrospinal fluid will cause depression of CNS activity (7). Since exercise may cause a decrease in blood pH due to the produc tion of metabolic by-products, this may be one mechanism through which CNS activity might be depressed by exhaustive work loads. For this reason the first part of this review is concerned with metabolic changes relative to acute exercise. Changes in Metabolic Functions In 19^2 Turrell and Robinson (80) reported that as the con centration of base bound bicarbonate is decreased beyond a point by higher lactate concentration the C^ctat£~"^ becomes progressively smaller, therefore giving rise to a decrease in blood pH. As the concentration of lactate increases the plasma proteins and hemoglobin will account for a greater fraction of the base binding capacity. 55 Using varied intensities of exercise (treadmill running), the maximum blood lactate recorded was 22 mEq/l with serum pH of 6.97 (arterial blood samples). It was concluded that at higher lactate concentrations hemoglobin and plasma proteins accounted for an increase base fraction while at low concentrations it was bicarbonate. Hickham et al. (50) tested eleven untrained male subjects (ages twenty-two to thirty-five years) using a treadmill exercise task. Speed and grade of running was set at 5*1 miles per hour and 3*5 par cent, respectively. Per cent oxygen (Og) saturation and O2 tension were found to fall during exercise. For the group as a whole there was a considerable fall in pH and serum carbon dioxide (CO2) content. It was also found that resting pH and CO2 content were not restored for many minutes. Thus, the earlier findings reported by Turrell and Robinson were confirmed, Bouhuys et al. (22), using a heparinized capillary tube vasodilated fingertip method, found that lactic acid and pH changes in humans within the first minutes following cessation of work were small. Pyruvic acid and lactic acid changes were found to correlate ,818 together with lactic acid concentrations being the larger of the two. In all, their findings corroborated results from previous investigations. The pH changes did seem small as compared to others reported by Robinson (6?) where pH was found to decrease as much as 7.00 as compared to resting values of 7«^0. (It should be pointed out ! 5 6 I i i | ‘ | that Robinson used a treadmill as the exercise task whereas Bouhuys j et al. used the bicycle ergometer.) Robinson suggests that recovery j I of an athlete from such fatigue is rapid enough that he may perform well again in less than an hour, although complete recovery may re quire two hours or more. Scheinberg et al. (73) investigated the effects of vigorous exercise (walking at a 4,0 per cent grade for approximately forty minutes) on cerebral circulation and metabolism. The following statements reflect some of the findings relevant to this disserta tion: 1. Cerebral blood flow (CBF) varied and could not be attributed to any known factor. 2. Cerebral On consumption increased during exercise from a mean of 5*78 to h.65 ml. (^/min./lOO grams of brain. 3. There was no significant change in cerebral glucose consumption. 4. The mean cerebral arterial pressure decreased twenty- nine per cent during exercise. 5. The mean cerebral vascular resistance decreased thirty- three per cent. 6. There were small but consistent increases in arterial 0g capacity but no alteration in per cent of O2 satu ration. 7. There was a consistent and considerable decrease of pH and CO2 content in both arterial and cerebral venous blood during exercise with no significant alteration in arterial or venous CO2 tension. The investigators concluded that strenuous exercise is some what analagous to diabetic acidosis, i.e., in either case blood pH is decreased. With exercise, changes in blood pH are attributed to an accumulation of lactic acid, whereas in diabetic acidosis an in crease in keto acids plays a more prominent role. i ! Zobl et al. (85) also studied the effect of exercise on cere- I : bral circulation and metabolism. In contrast to the above mentioned i findings, these investigators found no significant rises in cerebral 02 consumption and blood flow. It was suggested that cerebral metab- , olism does not share increased body metabolism and that during exer cise the brain behaves as a steady-state organ. Van Vaerenberg et al. (81) were interested in lactate changes in cerebrospinal fluid (CSF) due to muscular exercise. CSF was studied in anesthetized animals in which muscular work was produced by electrical stimulation. An interesting finding was that CSF lactate levels increased slightly during exercise but when lactate concentrations in the blood 1 were artificially induced, no such changes in CSF were demonstrated as compared to control levels. These investigators suggest that, with exercise, factors other than simply an increase in blood lactate concentration must be responsible for the rise in CSF lactate con centrations. In general, blood lactate increased; alkaline reserve and C02 partial pressure decreased; as for pH, a nonsignificant in crease was found. In recent and extensive study of the effects of exercise on tissue, respiration of the brain, skeletal muscle, heart, liver, kidneys and spleen of rats, Romanowski and Strazyriski (69) reported the following findings: 58 Oxygen uptake by the brain tissues was reduced in our experimental animals by a statistically significant 17.2 per cent. As has been shown by a number of observations, physi cal exercise, strenuous or not, does not modify cerebral blood supply. Therefore, we should look for the cause of the reduced brain respiration in changes produced by marieed fatigue in the chemistry of the brain, Yakovlev believes that a causal relationship exists between the reduction of the blood sugar level during severe exercise and the de pression of oxidative processes in the brain. The mechanism of this effort-induced diminution of cerebral respiration has been indicated by Leszkiewicz, who found cytochrome oxidase activity and oxidative phosphorylation to be di minished in the brain of animals in a state of great fatigue after strenuous exercise. According to Missiuro *exhaustion of the functional capacity of nerve centers by severe exer cise depresses the efficiency of oxidative resynthesis * of such compounds as ATP, glycogen, and others. (69:332) These investigators suggested that inhibition, due to severe exercise, of oxygen uptake of brain may be caused among other sub stances, by the inhibitory effects of serotonin and, possibly, corti costeroids. It appears that cerebral blood pH may decrease significantly with strenuous exercise; however, no significant changes are found in cerebrospinal fluid. Although lactic acid concentrations may be found to increase in both CSF and arterial cerebral blood, O2 avail ability to nerve tissue does not appear to be a problem. Although Zobl believes that the brain behaves as a steady-state organ, Romanowski and Strazynski indicate that O2 uptake by brain tissue is significantly reduced; it must be remembered that Zobl was only concerned with circulatory changes. These findings are confounding, in part, due to differences in intensity of the exercise task and difficulties inherent in comparing parameters of animal and human metabolism. At this point, findings regarding physiological factors 59 which may cause neural depression or inhibition seen highly incon clusive. The aforementioned investigations represents only one ap proach to studying the relationships existing between exhaustive exercise and neural activity. Other studies which are more be- haviorally oriented mast also be considered. Behavioral Adiustroemts Simonson, 3texer, and Benton (75)» using fusiom flicker frequency as a criterion for the state of the central nervous sys tem, attempted to determine the amount of muscular work: which would produce depressing effects on the CNS. Using a variety of work loads, designed to induce local and general fatigue, it was found that strenuous static exercise increased fusion frequency whale dynamic exercise (running to exhaustion) decreased fusion frequency. The magnitude and duration of depressed CNS activity (i.e. , decreased fusion frequency) corresponded to the severity of the exercise. Pineda and AdSrisson (63) postulated that physical fatigue should alter resting encephalographic (EEG-) patterns. BEE- records were taken for a period of fifteen minutes, two to three minutes after stopping exercise (treadmill walking at a twelve per cent grade, increased one per cent every five minutes). In all subjects (N = 16) there was a significant change in the appearance of the TgRr? recording after fatigue. Same changes lasted as 1«ng as ten minutes. The most marked difference was an increase in the alpha index (per cent time alpha). Frontal to 60 central bipolar leads showed an increase of twenty-two to twenty- eight per cent more alpha activity. Central to occipital bipolar leads showed a seventeen to nineteen per cent increase of alpha activity. These resalts were similar to those reported by Beaussart et al. (20) in which SEE recordings of boxers before and after fights were compared. (In this latter experiment an increase in the amplitude and amount of alpha was found in a significant number of boxers after a contest.) It was suggested that: Assuming that alpha rhythm is a rhythm of rest, which may be blocked by attention, and that it is a rhythm which maintains the individual in an optimum state of prepared ness to receive incoming stimuli, it can be postulated that the alpha rhythm increase indicates that after physical fatigue the subject is less attentive to incoming sensory stimuli and the awareness of his surroundings is not so acute. (63:3&0) It was also suggested that the physiological mechanisms underlying these changes may have involved a) slight inhibition of the ascending reticular activating system and/or b) metabolic im balance. Fromroel et al. (39) undertook a neuropharmacological study of central reactions to a variety of stimuli, including the inter active effects of intense muscular exertion and drugs. It was pro posed that ’ brascular exertion produces an initial sympathetic phase, which is followed by vagal dominance combined with somnolence, in difference, passivity, and inattention" (39:10). In referring to works of Bugand (2), Frommel and his associates suggest that "auto matic work produces a peripheral fatigue, while voluntary work, involving nervous tension necessary for the coordination of movement, causes a more central fatigue. The degree of fatigue depends on the functional level; it occurs quickly in the cerebral cortex and the central grey nuclei, and only slowly at the periphery” (39:HD- Kuascular fatigue was induced in guinea pigs and nice by forced walking on a treadmill. The experimental technique was rather unusual in that the extent of depressed CNS activity due to muscular exertion was determined by the amount of increase in sleeping time after treating the animal with a hypnotic drug, e.g., phenobaaMtaZ, meprobamate, and chlordiazepoxide. Based on this criterion, muscular exertion showed a depressant influence on CNS centers concerned mth sleep since the aforementioned medication, given thirty minutes adRfcer exercise, significantly increased sleeping time as compared to ©con trol animals; however, sleeping time was not significantly influenced when morphine, chlorpromazine, ether, or hexobarbital was used. ICf the medication is given before exercise, chlorpromazine, chlor diaz epoxide, and pentobarbital were related with significant in creases in sleeping time. Evidence was also presented which indicated that intense muscular exertion diminishes the normal excitatory response of the cortical and subcortical structures caused by amphetamine. The authors concluded that muscular fatigue seems to have an extensive depressant effect on central structures, as on those concerned mth sleep, perception, and thermoregulation (the latter two were not discussed in this review). 62 Although this latter investigation should be seriously questioned in terms of experimental design and the conclusions drawn, it does exemplify the strong belief that some investigators have in the assumption tinder question: that is, strenuous exercise produces depressant effects on the CNS. Of course, if this assump tion were true, the magnitude of depression would more than likely depend upon a number of factors, e.g., intensity of the exercise, type of exercise, fitness of the performer, etc. Although it is not clear what mechanisms associated with exercise might cause CNS depression, the assumption that such mech anisms exist seems reasonable. In fact, Frommel's concept that muscular exertion produces initially a sympathetic pause which may eventually lead to a vagal dominance, if exercise is severe enough, is not unlike Duffy's inverted U hypothesis related to level of arousal and performance. In light of the evidence presented in Part I, II, and HI of this review, it seems reasonable to postulate that severe exer cise, if introduced immediately after the learning experience, might cause retroactive effects on memory consolidation, due to CNS de pression. It might also be hypothesized that moderate exercise introduced in the same manner may facilitate exercise by enhancing the consolidation process through CNS excitation. The type of exercise task selected should be similar in structure to the learning task if interference effects which might occur between tasks are to be avoided. Because a shallow water maze 63 was selected as the task to be learned, running on a treadmill was selected as the exercise task since the locomotor patterns (wading ; versus running) in each task would be similar, if not identical. This i topic and other problems related to animal research and maze learning are discussed in more detail in the following review. Part IV; Maze Learning A task, involving a multiple-U water maze, was selected for use in the present investigation after a careful review of the liter ature, The present review is ended by referring to a few articles which helped to guide the investigator in the selection, design, and construction of the maze apparatus. The assumption that water is an effective incentive (used in this case as a nociceptive stimulus) in encouraging animals to learn is also defended. Procedural Considerations in Maze Learning Wever (8*0, in 1932, investigated the effects of changes in water temperature as an incentive to swimming activity in the rat. It was found that incentive varies with water temperature. Animals exposed to low temperature performed better than those near body temperature. High temperatures seemed to have a disorienting effect upon performance, Wever suggested that, if special temperature con trols are not available, a value of around 20°C would be satisfactory. This study brings to light an important consideration re garding an experimental design. Because the level of performance in water was found to be highly related to water temperature, it is essential that water temperature remains constant throughout the experimental procedure, if intercomparisons are to be made between the rates at which animals learn. Hack (48) later demonstrated that at water temperatures of 15°» 37.5°» and 45°C the lowest time scores in a maze task were associated with the lowest temperature. The poorest time scores were associated with the temperature of 37.5°C. Using a shallow water T maze with a water depth of two ami one-half inches, Dunn (33) found that escape from water was comparable to food reinforcement as an incentive to learn. In fact, the water group was superior in the relearning tests. Dunn attributed this finding to a better control over the incentive when using water. It should be noted that in the later trials (thirteen and fourteen) of Dunn's experiment the water group increased in errors which he suggests indicated, perhaps, that the animals adapted to the water, therefore decreasing the strength of water an an incen tive. No such adaptation has been reported by others using deeper water. The most used reference concerning the experimental task in the present study was the Handbook of Psychological Research cm the Rat, prepared by Munn (13). Although reference is made to Nunn in Chapter HI there are a few important and helpful points related to the use and design of maze tasks that should be considered at this time. 1. Time is regarded as an ambiguous measure since it is considerably influenced by initial errors and because it correlates high with temperamental and physiological factors extraneous to learning, e.g., a cautious rat 65 may increase time but decrease errors, whereas a faster rat may decrease time but increase errors. (Hasbbum (82) reported a correlation between speed of maze run ning and learning ability to be .20 + .1*48.) 2. Entrances into blind alleys offer the best means of measuring and comparing learned behavior. 3. The number of trials to reach an arbitrary criterion of efficiency is not as advantageous a measure of the learning process as total number of scores. 4. Error scores are regarded by most investigators as the best single score to use as a basis for plotting learning curves and comparing performance. 5. Initial trials are affected by a forward-going tendency, i.e., there is a tendency for an animal to go by path ways that lead off the main pathway as opposed to those pathways which dead end, thus forcing the animal to make a turn one way or the other. 6. Initial trials are affected by centrifugal swing, i.e., when momentum carries the rat to the outside wall of an alley, the tendency is to follow that wall and make the turn continuous with it. This seems to be more of an influence than forward-going tendency. 7. Goal pointing blinds (cul de sacs) are entered about three times more frequently than those pointing away from the goal. 8. In general, maze learning curves are negatively ac celerating. Human Versus Animal Performance on a Maze Task One last article seems worthy of mention. In 192k, Tsai (19) tested the universal application of Ebbinghaus' formulation that all curves for retention of memory are negatively accelerated by com paring retention curves between humans and rats. Using a maze task for each group (stylus maze for humans and a normal maze for rats), measurements were compared between groups. - . 6 6 Overlooking the obvious limitations of such an amMftmmus experimental design, it was found that: 1. The general shape of the learning curves <nf rafts dif fered from human subjects, the learning caavnss fear rats being linear, whereas learning curves for i t namMms were: negatively accelerated. 2. Animals retained more effectively during ftft® fSrsfc three to five weeks but humans were better after seven weeks. 3. The range of variability differed between gwr n ugs; per formance variability for rats increased wraith the length of the rest interval, whereas humans mairftaramtwB a simi lar range throughout; animals were less wardaBdLs than humans when the intervals of rest were short. In summary, a great deal of technical information was gained by attending to suggestions and experimental procedures used by other investigators. Further considerations regarding the dtesSgm of the learning task and the exercise task, as well, are discsnssasdl in Chapter HI. Summary In Part I of this review, learning was discussed as a physiological function associated with changes at the ^ maipfcic: or _ _ neuronal level of the central nervous system. These msrn na rnal changes are theorized to include neuronatomieal, neurochemical, arad related electrophysiological adaptations associated with repstnftftw® stimuli 67 across synaptic junctions. More specifically, it has been proposed that learning and memory may be a function of 1) synaptic growth with a corresponding increase in neurotransmitter substance, 2) changes in nitrogenous base ratios of nuclear RKA resulting in the production of specific protein substances related to neurotrans- mitting functions, and 3) modulation of impulse frequencies such that a specific reverberatory circuit links a desired response with the incoming stimulus. Part II involved a discussion of the memory theory of con solidation as related to present theoretical concepts and experi mental findings. Rival hypotheses were also considered. Part III dealt with evidence related to the effects of exhaustive or strenuous exercise on CNS activity. Of prime concern was a consideration of the physiological effects of strenuous exer cise on metabolic functions and on associated behavioral adaptations. The review was concluded in Part IV with a brief discussion of maze learning of small animals. Prime attention was given to experiments concerned with the effectiveness of water mazes in learning studies. Guidelines for determining an experimental pro cedure were considered. A comparison between human and animal per formance in maze tasks was also discussed. CHAPTER HI EXPERD9BKTAL BESIGN AND PRDCEOT5E To determine the effects of strenuous eawrcise on learning, a two factor experimental design with fixed effects was employed in this investigation. The factors were identified as follows: 1. Factor A - the exercise treatment. 2. Factor B - the treatment of repeated measures, i.e., trials. The experimental design and procedure were determined after a careful review of the research literature and after a series of preliminary undertakings (described in the following pages) which included: 1. The design and construction of the learning and exercise task; 2. A pilot study to determine a suitable experimental technique and to test the adequacy of the learning task; and 3. A pilot study to refine the experimental technique. Design and Construction of the Learning; and Bxercise Task Learning Task A multiple-U water maze (see Figure 1) with six choice points (point at which a right or left turn must be made) was 68 FIGURE 1 SIX UNIT, MULTIPLE-U WATER MAZE WITH GUILLOTINE-TYPE GATES 7° constructed out of one-half inch thick marine plywood. Design specifications of the maze were adapted from an earlier design described by Gray (46). The type of maze selected was based on the satisfaction of criteria described by Munn (13) as "general features of mazes having high reliability." These criteria include the following: 1. The alleys should be of equal length. This aids in rendering equal difficulty throughout the task. 2. The animal should be confronted at various points in the maze by two similar paths. The advantage being that the animal is prevented from overshooting the points where turns are required. 3. The maze should be of sufficient complexity to be difficult for the animal to solve. Longer and more complex patterns tend to give higher reliabilities. 4. Doors or gates should be included to prevent retracing. 5. The performance should be objectively scored. 6. All of the conditions surrounding the maze should be constant. There were two desired conditions that were not satisfied in this maze design due to limited funds. There was no self- recording device which would provide ultimate objectivity in recording performance scores, and there was no automatic device to place or remove the animals from the maze thereby reducing handling; however, handling of the animals before or after testing was minimized. Specifications of the maze were as follows: The walls of the maze were twelve inches high and assembled on a continuous floor plan (specifications for the floor plan are given in Figure 2) such that a pathway four inches in width was formed. The correct route consisted of a series of right-left-right- right-left-left turns. All seams were sealed with a Butyl rubber calking compound before the entire interior was sprayed with a water sealant. The underside of the floor was reinforced by four two inch by two inch beams, two of which ran the entire length of the maze. Embedded at each end of the floor was a one-half inch drain plug which led into an "L" joint brass fitting; to this was attached polyethylene tubing which could be cut to length. A three-quarter inch pipe was located two inches above the floor in the anterior wall of the starting box; to this pipe was attached a standard garden hose through which the maze was filled to the appropriate water level— three inches. Two one inch overflow pipes were located three inches above the floor in the back wall of the goal box, enabling the maintenance of a constant water level as well as a slow circulation of water to pre vent temperature changes that might occur with long testing sessions. A gillotine type gate was located at each choice point. The gates were constructed of three parallel dowels one-half inch in diameter (each dowel was anchored three inches from the bottom and two inches from the top by a wood bridge), which passed freely through a guide block located above each choice point.* The length of each gate was such that there was a one-quarter to one-half inch clearance between the floor and the bottom of the gate after the gate was released. When a gate was dropped, rubber pads located on top of the guide block helped to muffle the sound created by the contact between the upper wood bridge and the guide block. The trip mechanism for each gate consisted of a heavy gauge wire (see Figure 3) which penetrated the walls at a point three inches from the top. The wire was allowed to rotate freely on its longi tudinal axis. Between the walls the wire was permanently bowed so that in a horizontal position a wire shelf was provided upon which the raised gate rested (see Figure *0. One end of the wire was ex tended beyond the wall in the form of a one inch lever, bent in the same plane as the bowed portion, which rested on another ’ trip" lever that was attached to the external side of the wall. The "trip" ♦Because young rats were found able to squeeze between the dowels, a solid metal plate was later attached to the goal side of the wood bridges. 72 T 6 1 gcftel gate 2 to next unit Starting Box and Unit 1 (dimensions same for all units) 16 I2i 1 from sixth unit Goal Box and Alleyway Leading from Unit 6 T 9 _ L FIGUH3 2 OUTSIDE DIMENSIONS (IN INCHES) OF THE MAZE lever could be pivoted freely in a horizontal plane (see Figures 5 and 6). When the • ’ trip” lever was pulled by the operator, via an attached cord, the wire shelf was allowed to fall, thereby releasing the gate (see Figures 7 and 8). A starting box and goal box were located at opposite ends of the maze. The starting box was covered with a hinged lid and had slightly smaller dimensions than the goal box. A platform was located within the goal box upon which the animal was allowed to escape from the water. The passageway out of the starting box was blocked by a trap door which could be operated manually by the experimenter. The controls to the gates and trap door were centrally located on the right-hand side of the maze (see Figure 9~the white blemishes on the seams are due to the calking compound; the timer and hand counter are seen in the foreground). The maze was elevated from the floor by two tables of standard height. This placed the experimenter, while in a sitting position, out of direct view of the animals being tested. Pro gression of each animal through the maze was indirectly viewed through two large overhanging mirrors (with dimensions of sixteen inches by forty-five inches and sixteen inches by sixty inches) which ran lengthwise above the maze (see Figure 10 for an example of the experimenters view). Exercise Task Treadmill running was used as the exercise task, A rec tangular unit with fifteen running compartments was placed over the belt of a Quinton Treadmill (Model B). Dimensions for each compartment were three inches by thirteen inches by four and one- half inches. Plastic lids were attached on top of the individual compartments by a posteriorly located hinge. Electric shock reinforcement was used as the incentive to run. Shock was provided by a Grass Stimulator, Model The shock was administered to the animal through means of five vertical bars, located one-quarter inch apart in the back of each running 7^ ! FIGURE 3 TRIP MECHANISM FIGURE 4 GATE RAISED FIGURE 5 GATE RAISED (INSIDE VIEW) FIGURE 6 GATE LOWERED (INSIDE VIEW) FIGURE 7 FIGURE 8 GATE PRIOR TO RELEASE GATE RELEASED 75 compartment. Ground and hot wires were alternated such that the walls of each compartment were connected to a ground wire. An illustration of the exercise apparatus is provided in Figure 11. Pilot Study I The purpose of this study was to determine a suitable experimental technique and to test the adequacy of the learning task. The study was conducted in the Performance Laboratory of the Department of Physical Education, University of California, Los Angeles. Twelve male albino rats of Sprague-Dawley strain were randomly divided into three groups— group A, group B, and group C. Group A received two trials through the maze task every other day with forty-five seconds of rest between trials; group B received one trial every other day; and group C received one trial every other day with performance scores recorded under a different marking system. Error scores for groups A and B were recorded when an amwal entered a blind alley. Error scores for group C were re corded when an animal made a wrong turn and when the animal entered a Kliwri alley. Elapsed time was measured with a stop watch. Time scores indicated the duration it took an animal to reach the escape platform after leaving the starting box. All animals were thirty-five days old at the onset of the experiment. Animals were housed one per cage, maintained on a normal laboratory diet, and allowed to eat ad libitum throughout TtH O T f f R 9 CONTROL PAEfflL, TIMER, AND COUNTER VIEW FTSHE 11 wanBMJSB APPARATUS the experiment. Three days were allowed for the animals to become acclimatized to the laboratory environment before testing. Testing began on the fourth day and was continued every other day over a period of fourteen days, thus each animal was tested seven times. Since the principal investigation included an exercise task, suf ficient time had to be allowed between test periods to avoid the possibility of carry-over effects associated with the exercise treatment. As Winer has suggested: "The utility of designs calling for repeated measures is limited where carry-over effects are likely to confound results" (17:300). The testing procedure was as follows: 1. Testing was initiated at the same time, 1:00 P.M., throughout the experiment. Animals were tested in an alternating order among groups. Once established, the order of testing remained unchanged throughout the experiment. 2. The maze was filled to a water level of three inches. Water temperature was maintained at 19°C (tap water temperature). 3. Testing was begun by placing the animal in the starting box. After a ten second delay, the trap door was raised which allowed the animal to traverse the maze. i | . . Measurement of elapsed time was begun when the animal moved out of the starting box. 5. Each time the animal entered a new unit, passing a 78 choice point, a guillotine type gate was released behind him, 6. Error scores were recorded with a hand counter. 7. When the animal reached the escape platform, the timer was stopped and elapsed time was recorded. 8. After each testing session, which consisted of one or two trials, the animals were dried and replaced in their cages. Three animals did not respond to the treatment--one from group B and two from group C. Since no effort was made by these animals to traverse the maze, they were eliminated from the study. Typical learning and performance curves were obtained for each group.* Kean error (see Figure 12 and Table I) and time (see Figure 13 and Table H) scores were plotted along the ordinate, and trials were plotted along the abscissa, i.e., each point of the graph indicates a group mean score. Although only cautious interpretations of the data were considered possible because of the small sample size, the following findings were thought applicable and were subsequently used in rendering improvements to the maze task and later experimental ♦Since it was indicated in the review of the literature that most investigators concerned with measuring learned animal behavior by using maze tasks considered error scores as the single measure most indicative of learning, curves based on error scores are herein referred to as learning curves. Since the value of time scores has been questioned as a valid measure of learning, curves based on time scores are herein referred to as performance curves. 79 Group A E R R 0 R S x Group B • Group C x— 12 11 10 \/ TRIALS FIGURE 12 IEHttUNG CURVES FOR GROUP MEAN ERROR SCORES. TABLE I GROUP MEANS FOR ERROR SCORES* TRIALS 1 2 3 4 5 . 6 7 Group A n =4 4.25 4.50 3.75 4.25 2.37 1.87 1.75 Group B n = 3. _ 4.00 3.0.0 3.67 4.67 _ 3.67 4.00 2.00 Group C n = 2 13.00 4.50 8.00 10.50 11.00 6.00 8.50 ♦Jn waking group comparisons, please note that a different scoring system for errors was used for group C. • * Group A * ----* Groagi 3 •— — • Group C T 330 I 300 H 2?0 E 240 210 I ISO 15 150 120 B TRIALS FIGURE 13 PERFORMANCE CURVES FOR GROUP MEAN TIME SCORES TABLE II GROUP MEANS FOR TIME SCORES TRIALS 1 2 . 3 4 5 " 5 7 Group A n - 4 155.25 88.10 51.00 54.60 43.75 28.38 31.13 Group B n = 3 148.67 90.00 48.67 52.00 92.00 92.00 43.00 Group C n = 2 322.00 99.00 81.50 117.00 63.00 76.50 Tfcp water temperature was found to increase from X9°te i® 23°C during the weekends. This was attributed to the normal lack of demand for water facilities at the University during the weekend. Since the maze task was constructed without a cooling device for lowering water temperatures, it was decided that future studies should be conducted with the water temperature main tained at 23°C. Since the time spent between leaving the starting base and the first choice point varied considerably among animals, it was decided to base future scores am the duration it takes the animal to reach the escape plat form. after passing the first gate. Since three animals refused to traverse the maze, it was thought that an increase in water level to four or five inches might increase the effectiveness of water as am incentive in subsequent studies. After inspection of the learning and performance curves it appeared that one trial, administered every other day, was insufficient to produce adequate improvement in scores. Since the manner of presenting trials as found in group A seemed to provide better evidence of improvement, this procedure was selected for use im future studies; however, it must be pointed ant that the selection of this procedure led to some disadvan tages in studying the effects of exercise on memory consolidation. If an immediate post-trial treatment of strenuous exercise should prove detrimental to memory consolidation, a second trial and the inclusion of a rest period between trials would provide the animal more time for consolidation, making an interpretation of the time course more difficult. 5. Testing sessions were found long and arduous. To facilitate this procedure, it was decided that pre test waiting stalls, located in closer proximity to the maze track, should be used. This procedure also provided an intermediary period before testing where the arousal level among animals could be more closely equated. Pilot Study II T6® purpose of this study was to refine the experimental ibadfemmcrae to be used in the final investigation. Sixteen male albino rats, thirty-five days old and of Spragnoa>-]IIawley strain, were randomly divided into two equal groups— gjromrgs A group B. Group A, the control group, was administered (cmOly tfe© maze task, i.e., the animals were returned immediately to if fl i e e d r cages after being tested. Group B, the experimental group, was exposed to strenuous exercise thirty seconds after performing it2ae mtjgg* task. Strenuous exercise consisted of running to exhaustion 83 on a motor driven treadmill. Two animals were housed per cage— one from each group. After one day for acclimatization, all animals received one screening trial through the maze task. It had been decided that any animal which did not traverse the maze in ten minutes or less' would be eliminated from the study— all animals met this criterion. On the third day all animals were numbered and marked according to their respective group. Animals in group 6 were marked with a red dye and numbered one through eight. Animals in group A were similarly numbered but not marked. All animals were shaved around the buttocks to facilitate electrical conduction when contact was made with the stimulating device used in the treadmill task; however, only the experimental group was administered the tread mill task. All animals in group B received four practice sessions on the treadmill. Practice sessions were distributed over four days and administered in the following order: l) third day— five minutes at 1 mi. per hr.; 2) fourth day— three minutes at 1.5 mi. per hr. (this speed was found to be too fast at this time); 3) fifth day (morning)— five minutes at 1.2 mi. per hr.; 4) fifth day (late afternoon)— five minutes at 1.2 mi. per hr. All animals were weighed prior to the first day of testing. This was done because it was reasoned that drastic weight differ ences between groups at the conclusion of the experiment might in dicate undesirable stress effects of exercise and shock reinforce ment on normal growth patterns. Experimental Procedure The experimental procedure was as follows: 1. All animals received two trials, every other day, in the maze task. Trials were separated by a forty-five second rest period. Each animal received a total of twelve trials. 2. Administration of the maze task was conducted in a similar manner as in Pilot Study I (procedure used with group A) with the following exceptions: a. the water level was maintained at 4.5 inches, b. water temperature was kept near room temperature— 23°C, c. there was no need for a constant flow of water, since the water temperature could be easily main tained at near room temperature, d. the timer was started when the animal passed gate one, e. all animals were placed in a waiting stall before testing and at the onset of each test day, f. after completing the maze task, animals in group B were dried and prepared for the exercise task. To facilitate conduction, EKG electrode paste was placed on the animal’s hind quarters. One minute was allowed to elapse between tasks, g. the shock stimulus in the exercise task was maintained at forty volts, direct current, except in conditions where an animal was nearing exhaus tion. Animals were believed to be exhausted when they could no longer run despite sudden and short pulsating increases in the shock stimulus (to sixty volts) or despite assistance from the experi menter. Animals approaching the exhaustive state generally made more contacts with the stimulating device, wobbled or deviated from their normal running pattern, and exhibited lesser control over their hind legs. Treatment of the Data All raw data scores on the maze task were transformed into mean scores per test session, i.e., error or time scores were represented by one score instead of two. These individual mean scores were used in further analyses as raw data test scores. A learning and performance curve, using group mean scores, are illustrated in Figures 14 and 15. Group mean scores and standard deviations are also found in Tables III and IV. One-tailed t tests were administered to the data to compare a) intra-group means between trial one and six, and b) inter-group means between trials one and trials six (see Table V). Inter-trial reliability coefficients for trials one and two, two and three, five and six were computed for the control group (see Table VI). 86 • ---— Group A * ----* Group B 5 3 2 1 0 BLOCKS OF TRIALS FIGURE 14 LEARNING CURVE FOR ERROR SCORES TABLE III GROUP MEANS AND STANDARD DEVIATIONS FOR ERROR SCORES BLOCKS OF TRiAtS 1 2 3. . 5.. . T" Group A Ms 3 M 34? 3.69 2.94 2.25 1.88 n = 8 SD: 1.02 M .97 .86 .94 .99 Group B M: 4.67 4.08 3.42 3.67 _2.92 2.67 n = 6 SD: .99 1.10 1.14 1.11 .6l 87 i Group A *--- —* Group B 130 105 BLOCKS OF TRIALS FIGURE 15 PERFORMANCE CURVE FOR TIME SCORES TABLE IV GROUP MEANS AND STANDARD DEVIATIONS PER BLOCKS OF TRIALS FOR TIME SCORES BLOCKS OF TRIALS 1 2 3 4 5 6 Grouu A M: 52.19 45.44 46.00 40.69 33.63 33.81 n = 8 SD: 16.76 14.9 5 12.33 9.82 7.68 14>5 Group B M: 85.42 57.17 53.50 47.00 44.83 57.75 n = 6 SD: 47.92 13.46 24.77 1P,J58 _ 7.19 25.19 TA ELE-V - ONE TAILED T TESTS FOR JMJBfr- ABE® INTERGROOP COMPARISONS AMONG SBLfflOBB TRIALS Within Group Comparisons Group A: Trial 1 to Trial 6 it 3XDD)3PSSl {terrors. ) > 3.<W* t scores (time) 2.91* Group B: Trial 1 to Trial 6 1.2b* Between Group Comparisons Trial 1 to Trial 1 2.1©* 1M Trial 6 to Trial 6 1.7® 1.88+ ♦Significant at the .05 level. table wl RELIABILITY COEFFICIENTS BETaEM SS3LECTED TRIALS- GROUP A (COME, QfflUP) TRIALS 1 - 2 2 - 3 ....■ Error Scores CM C T t . 1 .4a -.Oil* Time Scores .63* .3® .37 ♦Significant at the .05 level. For the exercise task, the length of tine each animal exercised per trial is provided in Table VII. Treadmill speeds corresponding to each trial are also provided. (Note that in group 6 one animal was eliminated due to an infected eye and one animal died during the course of the study.) Analysis of the Findings Of particular concern was the discrepancy noted between group mean scores for trial one. Although the animals were randomly j, placed in their respective groups, a significant difference at the .05 level was found between the error scores of groups at trial one. A similar situation appeared in time scores but the difference was not found to be significant. Under nomal circumstances, where experimental and control groups are not properly equated before introducing the experimental variable (as in this case), the data should be analyzed with covariant statistical procedures. Had the learning and performance curves yielded more promising results, an analysis of covariance for repeated measures would have been appro priate. As it was, group differences due to the exercise treatment appeared negligible. Although a comparison of time scores did indi cate that the groups differed significantly at trial six, the dif ference was attributed to one extreme score. Intra-group comparisons between trial one and six indicated that performance scores improved significantly (.05 level) in all cases but one-time scores for group B. Learning was assumed to be a major factor in this improvement. 90 Inter-trial reliability coefficients computed for group A (see Table VI) were negligible to moderate in strength. Only one correlation was found significant, a .63 correlation between trial one and two for time scores. Tsai has reported that individual variability of rats increases with the length of the rest variable. Since the rest interval between trials was close to two days, this would account, in part, for the degree of variation noted among test scores. Of course, the sample size was quite small for this kind of statistic. The correlation between error and time scores was sur prisingly high, r = .63, compared to those reported in the research literature, generally r = .20 or less. Many problems were encountered in administering the exercise task. On the first day only one animal was capable of running at l.h mi. per hr. Even when the speeds were reduced to 1.2 mi. per hr., starting with animal #4, (see Table VII), only one additional animal (#8) ran. This latter finding was unexpected since all animals seemed to run well in pre-training trials. It was inter esting to note that, after the maze task, the animal's demonstrated an acute inability to maneuver their hind legs. Extreme plantar flexion was common. It appeared that either the animals were some what fatigued from the previous maze task or, possibly, the water acting as a stressor (temperature, etc.) somehow affected the animal's capacity to respond. (This same problem was evidenced in earlier exploratory work using a different animal population. It was found that with a longer delay period— forty-five minutes or TABLE VII INDIVIDUAL TIMES PER TRIAL ON THE EXERCISE TASK TRIALS 1 2 3 4 5 SPEED 1.4 mi/hr 1.2 mi/hr 1.2 mi/hr 1.2 mi/hr 1.4 mi/hr SUBJECTS 1. -* 16.5 36.0 17.5 4.5 2. 31 72.5 124.5 160+ 104.0 3. ELIMINATED 4. 16.0 36.5 9 6.5 34.0 5. 35.5 168+ 128+ 120+ 6. DIED 7. 12.0 72.0 7 0.0 36.5 8. 52.5** 70.0 63.0 52.0 53.0 *A dash line indicates that the animal would not run during this trial— refer to the text for further in formation. **1.2 miles per hour. 92 more— between tasks animals previously experiencing this difficulty would run anywhere from twenty minutes to an hour at 1.2 mi. per hr.) It seems unlikely that these animals were physically stressed enough or fatigued in the maze task to cause this magnitude of weak ness. Since this was a shallow water maze, the animal's progression through the maze was frequently interrupted with stops. It is also difficult to explain why all animals on the second day of testing ran twelve minutes or longer at 1.2 mi. per hr. It was thought that in the future longer periods of pre-training may alleviate the difficulties encountered on the first day. Another problem with the exercise task involved the inability to exhaust some animals within reasonable lengths of time (two hours or less). There were four cases where the animals were taken off the mill after running two hours or more. Selection of an appropriate treadmill speed was difficult due to large individual differences exhibited in running abilities as noted in time scores for trial five. No weight differences were found to exist between groups before or after testing. Summary and Conclusions Since the purpose of these investigations was to refine the experimental technique to be used in the final investigation, the following remarks, based on the findings, were used as points to be considered in the subsequent investigation. 1. In order to avoid group differences at the onset of the experiment, a larger sample is needed to minimize the effects of extreme scores on group means. A maximum cut-off time should also be used for elapsed time scores. AH animals should be pre-trained on the exercise task before randomly placed in groups. This would provide equal treatment to all groups prior to the introduction of the experimental variable. The sample should be large enough to allow for the elimi nation of animals due to death, illness, infection, etc. Twelve trials presented in blocks of two seem adequate to produce a significant improvement in error and time scores. Learning is undoubtedly a major factor in this improvement. Animals appeared to perform just as well in water at 23°C as in water at 19°C. Reliability coefficients between trials should improve slightly with a larger sample; however, the training regime used in this study favored a high degree of score to score variability. Although it can safely be predicted that improvement in performance will occur within twelve trials, it is not possible to predict on an individual basis how much improvement can be ex pected Problems encountered with the exercise task might be alleviated by providing a longer pre-training period with gradual increases in speed. It would be undesirable to increase the time period between tasks Experimental Design for the Principal Investigation This was a two-factor, three by six, experimental design with fixed effects having repeated measures on one factor. The design is reported by Winer as follows: FACTOR B A ' C . R ’ . 9 A ap Each G represents a random sample size n from a common population of subjects. Each of the subjects in G^ is observed under q different treatment combinations, all of these treatment combinations involving Factor A at level a^, The actual observations on the subjects within group i may be presented as follows: Subject hi bj bq 1 X m ... 1 k Xilk ... ... xijk ... n hln The symbol denotes a measurement on subject k in Gj_ under treatment combinations ab^. (17:302) bi bi b, a ° 1 ......Gl ........^ Gi ......Gi ........Gi gp gp Gp 95 In this stwty there were three levels of factor A (two groups receiving the exercise treatment plus the control group), and six levels of factor B (blocks of two trials) having repeated mea sures, The experiment can be illustrated as follows: TRIALS (B) E __________ 1 2 3 ^ 5 6 X E Group A Gj. G]_ G]_ G^ G^ R C (a) Group B G2 Gg G2 G2 Gg G2 I S Group C G-j G3 G^ G^ Gj Gj E where group A was the control group which received treatment B but not A; group B and C were experimental groups which received both treatments. Procedure Sixty-six male albino rats were housed three per cage and maintained on a normal laboratory diet. All animals were thirty- five days old at the onset of the experiment and of Sprague-Dawley strain. After three days of acclimatizing to their new surroundings, all animals were provided temporary markings and one screening trial through the maze. Any animal that did not traverse the maze in ten minutes was eliminated from the study. Two animals failed to meet the criterion time. On the fourth day ail animals were shaved and given the first of i rine screening and training trials on the treadmill task. All trials were given in the morning at approximately the same time each day. IMs procedure involved five minutes of running a day starting at erne mile per hour. Speeds were increased in the following order: 5th day - 5 minutes at 1.2 mi. per hr. 6th day - 5 minutes at 1.2 mi. per hr. 7th day - 5 minutes at 1.3 mi. per hr. 8th day - 5 minutes at 1.3 mi. per hr. 9th day - 5 minutes at 1.4- mi. per hr. l©th day - 5 minutes at 1.4 mi. per hr. 11th day - 5 minutes at 1.5 mi. per hr. 12th day - 5 minutes at 1.5 mi. per hr. the twelfth day fifty-one animals had qualified as sub jects for the experiment. Most of the animals eliminated during this time either would not run or died during the interim. A few pTyimalfi suffered injuries on the treadmill during the early stages of training and were eliminated from the study, 'The s n f f i T i m « r i ! s , were now randomly assigned to one of three groups (n = 17), %sd numbered accordingly, and weighed. Animals in group A (emafcpoOL group) were dyed green and group B red. There was no need to <%e group C. The animals were then caged so that one animal from group shared the same cage, i.e., cage one contained animal number cm© from group A, B, and C. Tha thirteenth day marked the beginning of testing. Since the procedure called for testing every other day it was possible to 97 split the groups and test half each day, i.e., twenty-seven animals were tested on the first day and twenty-four on the second day. testing procedure was as follows: 1. Testing began at 9:00 A.M. every day. 2. The animals to be tested were placed in waiting stalls, located in close proximity to the maze, with one xmmal from each group per stall. 3. All animals were tested approximately at the same time each day. h. The maze task was administered as previously described— two trials, forty-five second rest between trials. 5. After the maze task animals in group A were returned it© their stalls until testing was concluded, and animals in group B and C were dried and prepared for the tread mill task (a description of this task was provided in the previous investigation). Group B received five minutes of exercise. Group C ran for a maximum of ninety minutes or until they were exhausted. A one minute interval was provided between tasks. 6. Elapsed time on the treadmill for group C was recorded to the nearest one-half minute. 7. On the last test day all animals in group C were run tqe- til Exhausted in order to estimate the amount of stress placed on those animals which had run on previous test days for ninety minutes. All animals were re-weighed 98 in order to provide pre-test and post-test comparisons. Treatment of the Data Group mean and standard deviations were obtained from indi vidual mean scores per blocks of trials. A typical learning and performance curve was obtained by plotting group mean scores against respective trials. To test the null hypothesis that the population mean for the test scores were equal, that is, M^ = Mg = Mq, an analysis of variance for an experiment with repeated measures on one factor was used. The data, however, %ust conform to a prescribed pattern of variances and covariances before the statistical test can be con sidered exact” (17:123). Violations of the assumptions, homogeneity of within-treatment population variance and homogeneity of covariance, will generally result in a positive bias in the usual F test. Winer has suggested that when these assumptions are questionable, an approximate test (conservative or negatively biased) can be used where the F ratio has one and n - 1 degrees of freedom (17:123). If a test for homogeneity of variance is computed and satisfied, and the experimenter has reason to question the homo geneity of covariance, the degree of freedom can be adjusted, there by, providing the experimenter a more conservative test for the F ratio (17:306). Fro® an analysis of variance of control group test scores reliability coefficients were estimated. Thus, the estimated re liability for the mean of k measurements and the reliability of a single measurement (X^j) were computed. An intra-group correlation coefficient between error and time scores was also computed. If a significant F at the .05 level was found, the Newman- Keuls method for testing planned comparisons between pairs of means was applied to the data. The critical value was again set at the .05 level. Other statistical treatments that were administered included a Kruskal-Wallis analysis of variance (non-parametric) of delay time scores out of the starting box (to be discussed later in greater de tail) and typical t tests for pre- and post-test weights. Summary The experimental design for and technique employed in this investigation were based on a review of the research literature and on the findings from preliminary investigations conducted by the in vestigator. These preliminary investigations were discussed in detail in this chapter. The experimental design was described as having two factors with fixed effects and repeated measures on one of the factors. Of principal concern was the effect of exercise (factor A) on the rate of learning as produced by the effects of factor B (re peated trials or measures). CHAPTER IV RESULTS It was the purpose of this study to investigate whether strenuous and exhaustive exercise influence the amount and rate of maze learning by exerting a possible retroactive effect on the period of consolidation. Learning and performance were studied in terms of error and time scores. In this chapter, the results of the statistical treatment of these scores are discussed. Error Scores Inspection of the learning curves (Figure 16), which were based on group mean error scores, indicates that all groups improved as a result of repeated trials. There appeared to be occasional re gressions in the mean scores of groups B and C. Of particular in terest is the discrepency in the slope of the curves. The slope of the curve for group C is noticeably shallower than that of group A or B. This difference is in accordance with the research hypothesis. A compilation of group mean error scores and standard deviations is provided in Table VIII. To test homogeneity of variance, F max statistics were com puted (SSslibj. within groups: F max = 1,44; SSg x subj. within groups: F max = 1.37) and were found not to exceed the critical val ues (SSsubj. within groups: F max .95 (3. 13) = 4.16; SSg x subj. 100 101 E R R 0 R S LEARNING CURVES BASED ON GROUP MEAN ERROR SCORES TABLE VIII GROUP MEANS AND STANDARD DEVIATIONS PER BLOCKS OF TRIALS FOR ERROR SCORES Blocks of Trials 1 2 3 _ 4 5 6 Group A M: 4.11 3.57 2.89 2,75 2.32 1.93 SD: •9? 1.32 1.15 l.oS 1.23 1.03 Group B M: 4.04 2.82 2.96 2.46 1.96 SD: .87 1.29 _ .79 l.l4 i.o4 1.01 Group C M: 4.25 4.00 3.36 _ 3.43 3.07 3.14 SD: .80 O' CO . .61 1.66 1.37 •7.7- • ■- • Group A * ---* Group B •— —• Group C 5 3 2 1 0 BLOCKS OF TRIALS FIGURE 16 102 within groups: F max .95 (3. 65) = 85). The assumption of homo geneity of within-treatment population variance was therefore not violated. Since the homogeneity of covariances might be questioned, the degrees of freedom used to find within-subject tests was adjusted upward, providing a more conservative test. Critical values were set at the .05 level of significance. The analysis of variance of error scores is found in Table IX. TABLE IX ANALYSIS OF VARIANCE OF ERROR SCORES Source of Variation df MS F Between Subjects A (exercise) 2 9.15 3.60s * Subjects within groups 39 2.54 Within Subjects 210 B (trials) 5 19.68 19. 48** AB 10 .80 .08 B x subjects within groups 195 1.01 ♦Significant at the .05 level. ♦♦Significant at the .01 level. A significant F ratio at the .05 level was noted for factors A (exercise) and B (repeated blocks of trials). A test for AB inter action was found not to be significant. To determine which levels of factor A accounted for the observed difference a Newman-Keuls procedure was used to test dif ferences between »11 possible pairs of ordered group means for factor A, i.e., Ai - A^ (see Table X). 103 The information in Table X may be summarized as follows: %> A^. Treatments underlined by a common line do not differ from each other; treatments not underlined by a common line do differ. Therefore, group C differed significantly from group A or B and groups A and B did not differ. This result was in agree ment with the research hypothesis. Thus, the null hypothesis (Hq : = Mg = Mq) was rejected. A significant F ratio was also noted for factor B. To determine the effects of trials on the rate of learning a Newman-Xeuls procedure was computed for factor B, i.e. Bj; - By (see Table X). The information in Table XI may be summarized as follows: Bg Bj Bg %. Thus, the only trials that did not differ from each other were 5 - 6, 3 - k, and 1 - 2. All other comparisons rendered a significant difference between mean scores at the .05 level. These findings bore out quite well what may be readily seen in the learning curve. The findings indicated that significant im provements in error scores occurred rapidly with blocks of trials. As evidenced by the learning curves, group A and B displayed a faster rate of improvement than did group C. Retjability of the Measure for Error Scores Using an analysis of variance of control group scores, reliability coefficients computed for the mean of k measurements (repeated trials) and the reliability of X^j (single measurement) were found to be r = .39* and .10, respectively. VSner has stated ♦Significant at the .05 level. 104 TABLE X NEWMAB-ISroLS TEST FOR THE DIFFERENCES BETHBSS AIL FAIRS OF ORDERED MEANS fcr factor a « bxhicise treatment Exercise Treatment (Grp. A) A2 (Grp. B) A^ (Grp. C) 6rdered Means 2.93 3.02 3.5^ Differences .09 .61* Between Ag - .52* Pairs Aj - ♦Significant at the .05 level TABLE XI NEWMAN-K0GGLS TEST FOR THE DIFFERENCES BETWEEN ALL PAIRS OF ORDERED MEANS FOR FACTOR B - SLOCKS OF TRIALS Blocks of Trials BS B5 B3 B4 B2 BI Ordered Means 2.35 2.62 3.02 3.05 3.87 4.18 B6 .2? .67* .70* 1.52^ 1.83^ B5 — AO* A3* 1.25^ 1.56^ B3 — .03 .85^ 1.16+ B4 - .82^ 1.13* B2 - .31 BI - ♦Significant at t&e .05 level 105 that "...the analysis of variance model cannot be used to estimate j i reliability when the true score changes irregularly fftom one measure ment to the next, as for example, when practice effects are present in some nonsystematic manner. If, however, changes in the underlying true score are systematic and constant for all subjects, then adjust ments for this change may be made by eliminating variation due to change from the within-subject variation’ 1 (80: 132). Assuming that the underlying true secures were systematic and constant, reliability coefficients were recomputed with; adjustments for change, due to practice effects, taken into consideration. Ad justed reliability coefficients for means of k measurements and the reliability of X^j were .75* and .20**, respectively. Time Scores Inspection of the performance curves (see Figure 17), which were based on group mean time scores, indicates that a peculiar trend existed between the control and experimental groups. Also, it shall be noted that the standard deviations, in the table provided, were, in some cases, considerably high when compared mth the corresponding mean scores. This resulted despite the fact that a maximum time score of five minutes was used to reduce the effects of extreme scores on the mean. A compilation of group mean time scores and standard devia tions is provided in Table XII. A test of homogeneity of variance ♦Significant at the .01 level. ♦♦Significant at the .05 level. 106 T 100 I M 80 E 60 I N 40 S 20 E C 0 "t! T2 T3 55 ~~T5 W BLOCKS OF TRIALS Group A» ■ ■ Group B « * — * Group C •— • FIGURE 17 PERFORMANCE CURVES BASED ON TffiE SCORES TABLE XII GROUP MEANS AND STANDARD DEVIATKMS PER BLOCK OF TRIALS FOR TIME SCORES (EXPRESSED IN SECONDS) TRIALS 1 2 3 ' 5 r ................ Group M: 79.82 71.46 62.32 45.53 39.18 34.68 A SD: 28. l6 34.90 57,05 16.96 16.39 X3te Group M: 71.68 83.54 66.11 97.64 105.71 93.39 B SD: 29.49 38.01 48.58 76.40 100.06 86.46 Group M: 104.36 76.57 51,32 _ 66.93 58.64 72.77 . . . . . . C SD: 29.64 14.92 6?.o? 34.43 7^57. _ _ 107 revealed that the assumption of homogeneity of witMn-treatment population variance was violated: therefore, an analysis of variance treatment of time scores was not considered appropriate for this measure. In a case like this "...a transformation on the scale of measurement may provide data which are amenable to the assumptions underlying the analysis model" (80:2^0). For example, rather than time in seconds the scale may be the logarithm of time in seconds; however, transformations "...which yield homogeneity of error terms which are pooled do not necessarily yield the required homogeneity conditions on the covariances." With these limitations in mind it was decided by the experimenter to confine statistical treatment of elapse time scores to measures of mean and standard deviation. It was interesting to note the similarity between the curves of the control group time and error scores; in comparison, no remote similarity of this type existed for group B and C. Support for this observation was provided by correlation coefficients between error and time scores for each group: r(control group) = .66; r(group B) = .25; r(group C) = .28 (although moderate, all correlations were significant at the .01 level). The findings of mild positive cor relations for groups B and C was surprising since casual inspection of the respective learning and perfoxmamce curves would seemingly lead to the assumption of a negative relationship. This apparent discrepancy can be partially explained by the large effect which extreme time scores had on the trial means (see the raw data scores provided in the Appendix). There was still a much stronger relation between error and time scores for the control group than for groups B and C. Another finding of interest was that there were fifteen group B cases in the maze task where the maximum elapse time score (five minutes) had to be assigned. There were also six cases in group C but not one case was recorded for group A. In fact ( not one animal in group A scored a time over three minutes. There were ten incidences, all of which occurred in groups B and C, where the animals tended to sit in one of the alleys without making an attempt to move toward the goal. After the criterion time period elapsed, these animals were prodded, whereupon they usually progressed with out interruption toward the escape platform. In addition, it was noticed that as the experiment progressed the animals in groups B and C tended to wait longer in the starting box (i.e., after the trap door was raised) before moving toward the first gate. Since the elapse time score was measured between the first gate and the goal box, this delay was not reflected in the time scores; therefore, on the last block of trials an additional stop watch was included which allowed the experimenter to record this delay time between the starting box and the first gate. Delay times for each trial were measured and added together for each animal. Five minutes was the maximum time recorded for each trial, i.e., the maximum time that would be assigned to any animal was 600 seconds for the two trials. A test on the delay time scores for homogeneity of variance showed that the assumption of hraimganieity was violated; therefore, the three groups of scores were compared by means of the Kruskal- Wallis one-way analysis of variance which is a nonparametric test. Each of the N observations were replaced by ranks and then the corresponding ranks were summed for each group. The Kruskal- Wallis test (Table XIII) was administered to> determine whether these sums of ranks were different at tine .©5 level, of significance. The computed value was found to be !©.£©. This value was significant at less than the .01 level; therefore, the null hypothesis = Rg = Rq was rejected. This indicated that there were differences in delay time scores among groups. It was apparent from the sums of ranks that group A most likely accounted for the difference, since the sum total of ranks for group A was considerably lower than that found in group B or C. The question as to why these discrepancies in time scores occurred are discussed in (Chapter V. Time Scores on Exercise Task Problems were again encountered in administering the exercise task (see Table XIV). On the first day six animals from group C exhibited an inability to run at X.25 mi. per hr., although no simi lar problems were encountered during the training period. It was decided that rather than not run the animals, a delay in testing would be allowed and the animals would be given one more attempt. This procedure was successful for all but one animal ($6). Approxi mate delay times were also recorded in Table XIV. Due to the 110 n n i m RANK ORDER AND SOM <EF M R S (R)) FOR DELAY TIME SCORES OUT <EF USE) STARTING BOX AND THE ONE—WAY analysis m w m m s m (h) Group A ( G i w n r o f f n ® Group C 7 3S.5 36.5 6 34 40 1 24 38 16 32 8 22 25 26 35 1® 40 27 2® 33 12.5 m 31 10 SB 40 21 12.5 5 11 2® 30 19 23 3 14 15 9 17 3 3 R. = 21S.5 A ... _______ Bg = 3M5 .® Rq = 342.5 H = > 10.68 has probability <mT occurrence under Hq of P<.01. Ill TAHLB XIV DURATION GF EXERCISE FOR ANIMALS US GBCUF C TRIALS 1 2 3 4 5 , .. SPEED IN MI./HR. 1.25 1.25 1.25 1.40 1.50 1.50 SUBJECTS 1. (37,0)*3** 60.0 83.0 89.0 48.0 46.5 2. 32.5 72.5 87.5 90.0 55.5 55.5 3. 86.0 56.0 44.0 74.5 81.0 61.5 4. 58.0 39.5 90.0 90.0 90.0 131.0 5. OTTO 6. 90.0 90.0 90.0 90.0 157.0 . . . 7. (11.5)1 37.0 55.0 25.0 41.0 8. OTTO 9. . _(10.5)1 36.0 43.0 41.0 43.5 53.0 10. 55.0 55.0 90.0 72.0 88.0 125.0 11. 77.0 _ (66.5)1t 74.0 60.0 (82.0)2 82.0 12. OTTO 13. 75.0 72.0 90.0 90.0 90.0 230.0^ 14. 54.0 60.0 90.0 88.0 90.0 152.5 15. 15.0 36.0 63.0 64.0 90.0 162.5 16. (21.5)? 68.0 81.0 90.0 90.0 217.5+« 17. (6l.P)?/4 40.0 60.0 49.0 32.5 33.0 ♦Parentheses indicate that a longer delay period intervened between tasks; ♦♦the numbers outside the parentheses indicate the duration of the delay period (to the nearest ^ hour); ♦♦♦the tread mill speed was increased for these animals after three hours of running. stringent time demands im testing; other animals on the maze task, it was difficult to administer a high degree of control over these | delay times. On the second day, 'only two animals demonstrated dif- ■ ficulty in running* whsrennpom one ran, after a delay period, while ; the other did not {#?]). Btotte particularly that the exercise time for animal #6 on the second trial was the maximum time allowable. This was surprising sine® the same animal did not run on the pre ceding day. No other difffflcaOLti.es were encountered with group C ■until the fifth trial. M. tMs time animal #11 did not run on the first attempt. By trial three, five animals were running up to the criterion time~90 minutes. The itreadnmll speed was increased to 1.40 mi. per hr. for trial four and agaim to 1.50 ®i» per hr. for trials five and : six. On trial six all anrnmrmills were run to exhaustion, i.e., no maxi mum cut-off time was used. Times were recorded for two animals that were more than twine as long as the criterion time. It is highly unlikely that these ware severely stressed by the exercise task in the later stages off the experiment. This is probably also true of the three other «wnmT«fT!s who ran one hour longer than the criterion time. In summary, out off eighty-four possible exercise periods, there were: 1, two incidences where the animals would not run with or without a longer delay period (three per cent of the cases}; 2. seven incidences where the animals ran after a longer delay period (eight per cent of the cases); 3. seventeen incidences where the animals did not run to complete exhaustionf i.e. ( ran to criterion time (twenty per cent of the cases); and b. fifty-eight incidences where the animals ran to exhaus tion one minute after completing the maze task (sixty- nine per cent of the cases). For trial six, omitting the two extreme cases, the mean time (n =12) for running to exhaustion was 91.67 minutes. The times ranged from 33 to 162.5 minutes with five animals running beyond the original criterion time. Although of less magnitude, similar problems were encountered with group B. On trial one, one animal did not run with or without a delay period. On trial three, this same animal ran five minutes after a short delay period. Additional Findings Pre- and post-test weights were taken of all animals. This was done because it was reasoned that drastic weight differences between groups at the conclusion of the experiment might indicate undesirable stress effects of exercise and shock reinforcement on normal growth patterns. The mean and standard deviation for each group is reported in Table XV. A t test on pre-test and post-test weights was administered to the two means which showed the greatest discrepancy, i.e., for 114 pre-test weights group C was compared to group B and for post-test weights group A was compared to group C. No significant difference among means was found in either case. TABLE XV GROUP MEANS AND STANDARD DEVIATIONS OF PRE- AND POST-TEST WEIGHTS Pre-test Post-test Group A M: 103.86 150.21 SD: 15.29 25.95 Group B M: . 105.14 15M3 SD: 14.06 23.9& Group C M: 102.57 .. SDt 15706 31719 CHAPTER V i l DISCUSSION The problem entertained in this dissertation was to determine whether the amount and rate of learning a perceptual-motor task might be altered by the hypothesized effect of exercise on memory consolida tion. This chapter presents a discussion and further clarification of the findings. Error Scores Error scores were considered to be the primary criteria for ' measuring learned animal behavior as related to maze task performance. Munn (13) has reported evidence which amply supports this contention. According to Munn, error scores on a maze task appear to be con siderably more reliable than time scores since time scores are in fluenced by many factors not necessarily related to learning, i.e., an animal may decrease the time it takes to traverse a maze by merely increasing his speed, but at the same time he may make considerably more errors due to his inability to choose or remember the correct pathway. For this reason, plotted curves based on error scores were arbitrarily designated in this study as learning curves. This does not mean that «U factors which contributed to the shapes of the learning curves were reflections of learned behavior nor were the error scores entirely a product of behavioral factors which were ; ................. 115 _______ indicative of learning. Error scores were merely considered to be the best criterion available for measuring learned behavior in this iparticular task. It was noted that all groups showed a significant improvement in error scores as a function of practice. Thus, learning was assumed to have occurred. According to the research hypothesis group A and B should have demonstrated greater improvement than group C. Based on the appearance of the learning curves and the finding of a significant F value (.05 level) for Factor A (exercise treatment) this hypothesis was supported with reference to error scores. However, there are additional variables to be considered. It is difficult to interpret Factor A as comprised solely of an exercise treatment. Shock treatment to the hind legs and buttocks was also administered to the animals as a reinforcing stimulus; therefore, it might be argued that the slower rate of learning evidenced by group C was somehow related to longer periods of shock reinforcement. Two considerations make this assumption doubtful. One, it should be re membered that Duncan, Coons, and Miller (32) (26) administered shock to the legs and feet of rats as an experimental control in analyzing the effects of ECS on memory consolidation. Although Coons and Miller did find that animals receiving shock to the hind quarters exhibited increased fear responses it was also found that only those animals receiving ECS through the head demonstrated memory impairment in making an active avoidance response. Two, it should also be mentioned that the exercise task was suspected to have been a more punishing 217 task for animals in group B, since these animals experienced more jostling while running and more contacts per unit time with the stka> lating device than did animals in group C. A possible explanation for this observed difference in performance between groups B and C might be that group C animals performed better because they received more practice at running during the course of the experiment than did group B. (It must be remembered that group B animals were limited to five minutes of running. ) Group B animals also demonstrated .greater difficulty in adjusting to changes in treadmill speeds as the experi ment progressed. In spite of this, group B still demonstrated a greater amount and faster rate of learning than did group C. Theoretical implications associated with the findings are, indeed, difficult to interpret at this time, A number of additional experimental controls are needed before the findings can take ora greater theoretical significance. In spite of the risk, the investi gator offers the following interpretation of error score findings. Since both tasks demanded similar movement patterns, an inter ference theory does not seem applicable in explaining the findings. If interference were operable group B error scores probably should, have been influenced by the exercise treatment in a manner different from those scores found in group A. The concept of reactive inhibition (Ir) does not appear ap plicable in explaining the findings, since plenty of time was allowed 9 between testing sessions for the hypothesized net amount of Ig to dissipate. Thus, Ir should not have been an important factor 118 Influencing an animals performance during his next exposure to the maze task. It would appear that among the many alternative hypotheses to explain the error score findings two theories seem highly applicable: 1) the theory of memory consolidation, and 2) a "state-dependent" theory of learning. In keeping with a consolidation theory, the findings might be accounted for by the following rationale: During and after each block of trials on the maze task, memory traces associated with the act performed were established. The sub sequent bout of exercise produced acute metabolic changes in the organismic state of the animals. When allowed to persist long enough the exercise treatment induced depressant effects on certain levels of central nervous activity— among which may have involved changes in cortical activity. Thus, the period of neural consolidation asso ciated with the maze performance and occurring immediately after cessation of a maze trial, may have been disrupted by overlapping in time with the depressant effects of the exercise treatment on neural activity. Thus, a retroactive effect of exercise on memory was evidenced. It was theorized that the only experimental condition which could lead to memory disruption, in this case, was satisfied by the treatment provided group C. In accordance with a "state-dependent" theory of learning as proposed by Overton (6l) and Nielson (39) • the findings might be ex plained as follows: 119 By introducing the exercise task immediately after the naze trials an altered state of neurological events associated vith learning may have occurred causing changes in the state of brain excitability. If it is true that memory of the learned event involves a maintenance or reconstruction of these modifications of brain excitability, then failure of retention would occur whenever brain excitability is modi fied away from that established by the training procedure. Thus, the exercise treatment may have modified the state of brain excitability related to the maze task in such a manner that the observed poorer maze performance may have been a problem of memory retrieval rather than one related to the disruptive effects of exercise on memory con solidation. Within the framework of this latter theory, it would be diffi cult to explain why there were differences evidenced between groups B and C. It should be recognized that the theoretical explanations presented here are highly tenuous and inconclusive. At the present time such speculation rests on relatively weak findings; additional experimental evidence is needed. It is believed, however, that one important function can be served by introducing the aforementioned concepts. Ey attempting to theoretically evaluate the findings perhaps, in the near future, some interested student or investigator might be prompted to test further the hypothesis considered in this study by pursuing an approach, direction, or goal which extends beyond the mere tabulation of 120 additional findings. Tine Scores Wien curves obtained from elapsed time scores were compared, a rather unsuspected result was noted. For the most part, time scores for groups B and C showed a tendency to increase after T2 for group B or after T3 for group C (see Figure 18 and Table XII). This was also evidenced by noting the correlation between error and time scores, i.e., = .66, rg - .25, and tq = .28, Wiereas time scores for group A correlated moderately well with error scores, the same corre lations for group B and C were found to be quite low. How can these differences be accounted for? It would seem that the use of two different incentives (water and electric shock) between tasks was not wise. If, as a result of the exercise task (a punishing task), a response that is characteristic of avoidance learning was conditioned, this could well explain the tendency for animals in group B and C to hesitate in completing the maze task. This explanation compares favorably with findings reported earlier by Killer and Coons. It should also be noted that l) not one arrimil in group A took longer than three minutes to traverse the maze nor longer than one and one-half minutes to leave the starting box while 2) several animals in group B and C were assigned the maximum elapse time scores and were found to have significantly longer delay times in the starting box than did group A. Si light of the hypothesis tested, it was conjectured that the time score findings seemed more indicative of a learned avoidance response which tended to influence maze performance but which ap peared to have little effect on the learning of the maze task. This I was evident when error and time score differences which existed be- / tween groups B and C were compared. Exercise Task Difficulties which were encountered in administering the exer cise task might have been avoided had electric shock been used in both tasks. This is suggested since water temperature was suspected as a possible factor related to the difficulty some animals experienced in running at reasonably slow speeds as compared to speeds encountered during the training period. After considering the variables involved it seems possible to assume that changes in body or muscular tempera ture may have been an important factor related to this problem, i Longer training periods on the treadmill task with practice runs after exposure to cold water would probably have been beneficial. The in vestigation of this problem in itself would make an interesting study. It was unfortunate that the results were somewhat contaminated by the variable time factor which existed between tasks for a few animals in group B and C. The findings must be interpreted with con sideration given to the percentage of animals that performed up to the specified criterion for the exercise task. It is difficult to estimate how the results might have dif fered had all the animals in group C run to exhaustion before the nine ty minute cut-off period. Had these circumstances prevailed, one might 122 hypothesize that the differences evidenced in the amount and rate of learning for group C as compared to group A and 5 would have been greater. The exercise task did not appear to have debilitating effects on normal growth patterns as shown by the nonsignificant pre- and post-weight comparisons,* limitations of the Findings Final reference should be made to additional limitations of the findings not previously discussed in this chapter. Since a few animals in group B and C (ten incidences in all out of a possible 252) exhibited tendencies to sit in an alleyway, they were encouraged by the investigator to complete the task, i.e., it was found that the animal after being nudged from a sitting posi tion usually progressed onward. Unfortunately, this procedure intro duced an undesirable variable since the number of priors that might otherwise have been scored was possibly affected. If an animal needed to be prodded more than twice in any one trial, he was eliminated from the study (as was the case with one animal). The delay period between tasks was not adequately controlled in eleven per cent of the cases; therefore, although a one minute delay period was thought to be a factor related to the effects of ♦Price (&+) found that forced exercise retarded the rate of weight increase of young rats. 123 exercise on the amount of learning, similar results might have been attributed to animals who received longer delay periods. This con sideration points out the need for an additional control group which should have been exposed to the exercise task several minutes after the maze task. In this latter situation, if the theory of memory consolidation is to be found applicable, one should find that the results would be influenced by introducing a longer time interval between tasks, i.e., a group experiencing the strenuous or exhaustive exercise one-half hour after performing the maze task should suffer less learning deficits than a group receiving the exercise task im mediately after the maze task. The age of the animals sampled must also be considered in interpreting the findings. As Thompson (76) has pointed out, when dealing with problems related to memory consolidation one may find differential effects in memory disruption relative to the ages of the animals used, i.e., the younger the animal the greater the chances of evidencing memory disruption. Exercise times recorded for trial six (a run to exhaustion) showed that many of the animals ran considerably beyond the original ninety minute criterion time. Just how strenuous the exercise may have been for these animals earlier in the experiment cannot be accurately determined. In reference to individual time scores on the exercise task, it is only possible to state that the amount of exercise experienced for group C placed severe demands on the majority of animals. 12h Summary The hypothesis tested was found tenable when error scores were considered. No definite conclusions could be made regarding the theoretical significance of the findings. It was suggested, however, that the findings might best be explained by a consolidation theory of memory or, perhaps, by a "state-dependent" theory of learning. CHAPTER IT! SUMMARY, CONCLUSIONS, IMPLICATIONS AND RECOMMENDATIONS In this chapter, the investigator has attempted to present a concise review of the experiment undertaken in this dissertation. On the basis of the hypothesis tested and the experimental findings, conclusions are presented. These are followed by a brief discussion of theoretical and practical implications which are derived from the findings. Specific recommendations have been made for further research; these may be found at the close of this chapter. Statement of the Purpose It was the purpose of this investigation, which was based upon a memory consolidation theory, to determine the effects of strenuous and exhaustive exercise on learning. Statement of the Problem The specific problem of this investigation was to determine whether or not strenuous or exhaustive exercise, by producing a pos sible retroactive effect on the period of consolidation, influences the amount and rate of maze learning by rats. Hypothesis Given three experimental conditions related to learning a maze ___________ . 125 _......... ! - 126 task, where one group (A) receives learning trials with no specified !interpolated activity, a second group (B) receives a mild bout of I |exercise immediately following each testing period, and a third group (C) receives a strenuous or exhaustive bout of exercise immediately after each block of trials, it will be found that for the group means per error and time scores: group A > group B < group C, i.e., learning for group B will be equal to or greater than group A and both group A and B will evidence a greater amount and faster rate of learning than group C. The null hypothesis tested was Hq = = Mg = Mq, where M = mean of each respective group for error or elapse time scores. Procedure After nine days of pre-training on a treadmill task, fifty- one albino rats, Sprague-Dawley strain, were randomly distributed into three groups (A, B, and C). All animals were maintained on a normal laboratory diet and allowed to eat ad libitum. At the onset of the experiment, the animals were thirty-five days of age. All animals were given two successive trials (block of trials), every other day, in walking through a multiple-U, six unit, shallow water maze. Each animal received a total of twelve trials. One minute after each block of trials on the maze task group B and C received five minutes of exercise and ninety minutes of exercise (or until exhausted), respectively. Group A did not receive the exercise treatment. All animals were restricted to cage activity between test sessions. Error scores based on entries of the animal into a blind alley 12? and time scores based on the time it took an animal to travel from the first choice point to the escape platform were the parameters measured for the maze task. An analysis of variance for a two-factor, three by six, experi- I ment with repeated measures was used in treating the data. The sta tistical treatment was administered to individual mean scores per block of two trials. Findings Error scores: A significant F at the .05 level for Factor A ; (exercise treatment) was noted. Besults from a Neuman-Keuls pro cedure administered to ordered differences on Factor A indicated that group C differed significantly from group A and B, whereas .group A and B did not differ from each other. Thus, the null hypothe sis, = Mg = Me, for error scores was rejected. A significant F at the .05 level for Factor B (blocks of trials) indicated that a significant improvement in performance was evidenced among groups. This improvement was believed indicative of learning. Inspection of the learning curves also indicated that groups A and B displayed a faster rate of improvement than did group C. It was proposed that error score findings reflected the hypothe sized retroactive effects of strenuous and exhaustive exercise on the period of memory consolidation of the maze task. It was suggested that a theory of memory consolidation or a "state-dependent” theory of learning explained the error score findings. 128 Elapse time: Time scores for groups B and C slowed markedly after approximately the third block of trials, whereas group C pro- jgressively improved. | It was hypothesized that since treadmill running was punishing I and because it followed immediately after the maze task, perhaps animals in group B and C demonstrated a response characteristic of avoidance learning by hesitating in the maze task. Although the null hypothesis was rejected, the time score findings were not found to be in accord with the research hypothesis. Some difficulties were encountered in administering the exer cise task. Out of eighty-four possible exercise periods for group C, there were two incidences where the animals would not run, seven incidences where the animals ran after a longer delay period, seven- ;teen incidences where the animals did not run to complete exhaustion ! and forty-eight incidences where the animals ran to exhaustion one minute after completing the maze task. It should be noted that during the course of the experiment nine animals were eliminated due to sundry reasons. To summarize, in light of the limitations specified, it was found that: 1. Using error scores as the learning criterion, animals exposed to strenuous or exhaustive exercise immediately after a maze task exhibited a lesser amount and a slower rate of learning than did animals receiving no exercise or moderate exercise. It was suggested that maze task I I 129 error scores were influenced by retroactive effects | of exercise on memory consolidation* 2. Using time scores as the learning criterion, the animals receiving the exercise treatment displayed, for the most part, poorer elapse time scores as the experi ment progressed beyond the third block, of trials, whereas the animals not receiving the exercise treatment demon strated a progressive improvement in time scores. Thus, the interpolated exercise task appeared to produce nega tive transfer effects on maze performance, when perfor mance was measured by time scores. Conclusions It was concluded that strenuous or exhaustive exercise, introduced immediately after a maze task, influences the amount and rate of maze learning by rats when the criterion for measuring learning is error score. It was hypothesized that the observed effect of exercise on learning was due to retroactive disruption of memory consolidation, via central nervous system depression, caused by strenuous or exhaustive exercise. Implications In generalizing from the findings, it might be postulated that strenuous or exhaustive exercise which is introduced immediately after structured learning experiences may produce an environmental situa tion which does not favor optimal transfer of learning from one j practice session to another. If future experimentation tends to I substantiate this, the findings have important implications for ; those concerned with the proper scheduling of teaching and/or prac- j ; tice regimes where learning and exercise are totally integrated as a means of meeting pre-established objectives. Situations of this type are most numerous in the field of sports as, for example, when the coach applies the principle of overload by running his athletes to exhaustion with the intent to promote cardiovascular endurance, ; while during the same practice session attempts are made to 1) teach complex motor skills associated with a given position or play or 2) teach highly cognitive tasks related to defensive or offensive strategy. In these circumstances the instructor or coach may do i well to space intensive physical work outs at a reasonable distance ; in time from highly structured learning experiences. This procedure i ; may have the effect of reducing or eliminating an environmental at mosphere which might otherwise provide less than optimal conditions for learning. If the experimental hypothesis were found valid, it seems reasonable to postulate that physical fitness of the performer might be a critical factor in determining the magnitude of the effects which exhaustive exercise may have on learning. Unfortunately, implications derived from this study are limited due to the lack of needed experimental' controls. It would appear that the problem studied here shows promise of providing definite theoretical implications for a consolidation theory of ! 131 i {memory or a ^'state-dependent" theory of learning. ! | The major implication derived from this investigation, however, is the obvious need for additional research. A series of investiga tions, where time (e.g., length of the inter-task period) and intensity (e.g., amount of exercise) variables may be adequately studied, are needed. Recommendations The following recommendations are suggested jnrSmarily for the reader interested in testing further the research hypothesis of the present investigation: 1. Future investigations should include learning; and exercise tasks which utilize the same stimulus as a reinforcement. 2. A similar research design should be administered to animals of various age groups in order to determine to what extent strenuous exercise may affect the learning of animals at different levels of maturation. 3. Several experiments must be carried out to determine the effects of: a. different intensities of exercise, b. different time periods between tasks, and c. different levels of pre-training on memory or learning impairment. 4. A matched-pair control should be incorporated into a design, where one animal receives the exercise treatment •plus shock, wMle the control animal receives equal amounts off shod* administered at the same time but does not receive the exercise treatment. In this manner the possible interaction effects between exercise and shock may toe factored out. The interaction effects of task complexity, maze con- soM'daticmi, and exercise should be investigated. A meuraf&ffiimacolagical study, similar to that proposed by FrammeGDL et al. (39). could be pursued where the effects of exhaustive exercise plus a neural depressant on memoiy ©nraealidation are studied, as opposed to the effects produced only by a neural depressant, lastly, amd most importantly, the experiment undertaken in this dissertation plus the recommended investigations shouM toe replicated, when possible, using human beings as subjects. REFERENCES Books Bilodeau, Edward. Acquisition of Skill. jSfew York: Academic Press, Inc., 19o(o. Bugard, P. Ia Fatigue Physiologie. PsychoLogi-e et. Medicine Sociale. Paris: Masson, i960. Cratty, Bryant J. Movement Behavior and Kotor Learning. Philadelphia: Lea and Febiger, 19^51 Ebbinghaus, Herman. Memory: A Contribgtiogii to Experimental Psychology. New York: Bureau of Publications, Teachers College, Columbia University, 1913. Florey, Ernst, editor. Nervous Inhibitican. Hear York: Pergamon Press, 19^1. Gagn6, R. M., and E. F. Fleishman. Psychology and Human Performance. New York: Henry Holt and Company, 1959. Guyton, Arthur. Textbook of Medical Physiology. Philadelphia W. B. Saunders and Company, 1966. Hebb, D. 0. The Organization of Behavior. Hear York: John Wiley and Sons, Incorporated, 19*4*9. Hull, Clark L. Principles of Behavior. j E f e a r York: Appleton- Century-Crofts, Incorporated, 19^3. Kimble, Daniel P., editor. The Anatomy of Memory. Palo Alto: Science and Behavior Books, Incorporated, 1967. Morgan, Clifford T. Physiological Psychology. New York: McGraw-Hill Book Company, 19&5. Muller, G. E., and A. Pilzecker. EbcperimeaifcalLle Beitrage Zur Lehre von GedAchtnis. Leipzig: J. A. Barth, 1900. Munn, Norman L. Handbook of Psychological Research on the . . Rat. Boston: Houghton Mifflin Company, 1950. Russell, Ritchie W. Brain: Memory learning. London: Oxford University Press, 1959. 133 1 3 4 15. Siegel, Sidney. Nonparametric Statistics for the Behavioral Sciences. New York: McGraw-Hill Book Company, Incorporated, 1956* 16. Thompson, Richard F. Foundations of Physiological Psychology. Uesr York: Harper and Row, Publishers, 1967. 17. Winer, B. J. Statistical Principles in Experimental Design. Uensr York: McGraw-Hill Book Company, 19^2. 18. Young, J. Z. The Memory System of the Brain. Berkeley and. Xos Angeles: University of California Press, 1966. Periodicals and Unpublished Materials 19. Alderman, Richard B. "Influence of Local. Fatigue on Speed and Accmracy in Motor Learning," Research Quarterly, 36:13X-X40, 1965. 20. Beaussarfc, M. G., et al. "The EEG of Boxers Examined Immediately After Combat. Comparative Study with the EES Recorded Before Combat in 52 Cases," Revue Ifegmtlogiorue, 101:422-425, 1959. 21. Benson, David W. "Influence of Imposed Fatigue on Learning A JmjdLng Task and A Juggling Task," Research Quarterly, 39:251-257. 1968. 22. Bouhuys, A., et al. "Metabolic Acidosis of Exercise in Healthy Males," Journal of Applied Physiology. 2X:XOWul046, 19^ 23. Breen, Richard A., and James L. McGaugh. "Facilitation of Maze learning With Post-trial Injections of Picrotoxin,1 " joiaamal of Comparative and Physiological Psychology. , 54:h9&-5Q!. 1961. 24. Buresova, o., J. Bures, and L. K. Gerbrandt. "The Effect of am KLectrcconvulsive Shock on Retention of a Spatial Msarimination and of Its Reversal," Physiology and Behavior, 3:155-159. 1968. 25. Burns, B. BelisXe. "The Production of After-bursts in Isolated Unanesthetized Cerebral Cortex," Journal of Physiology. 125:427-446, 195^* 135 i j 26. Coons, Edgar and Neal E. Hiller. "Conflict Versus Consolida- { tion of Memory Traces to Explain 'Retrograde Amnesia' Produced by ECS," Journal of Comparative and Physiological Psychology. 53:524-531, 19^0. 1 ! j 27. Corey, Stephen M. "The Relationship Between Compulsory | Physical Exercise and the Ability of the White Rat to Learn and Relearn an Elevated Skeleton Maze," Journal of Comparative Psychology. 11:291-318, 1931. 28. Courts, Frederick A. "Relations Between Muscular Tensions and Performance," Psychological Bulletin, 39:347-367, 1942. 29. De No, Lorente. "Analysis of the Activity of the Chains of Intemuncial Neurons," Journal of Neurophysiology, 1:207-244, 1938. 30. Deutsch, J. A. "Higher Nervous Function: The Physiological Bases of Memory," Annual Review of Physiology, 24:259-286, 1962. 31. Duffy, Elizabeth. "The Relation Between Muscular Tension and Quality of Performance," American Journal of Psychology, 44:535-546, 1932. 32. Duncan, Carl P. "The Retroactive Effects of Electroshock on Learning," Journal of Comparative and Physiological Psychology. 42:32-44, 19^9. 33. Dunn, Bernard M. "A Comparison of Food Reward and Escape from Water in Motivating Learning in the White Rat," Journal of Comparative Psychology. 19:107-112, 1935. 34. Eccles, R. M., and R. A. Westerman. "Enhanced Synaptic Function Due to Excess Use," Nature. 184:460-461, 1959. 35. Eccles, R. M., W. Kozak, and R. A. Westerman. "Enhancement of Spinal Monosynaptic Reflex Responses After Denervation of Synergic Hind-limb Muscles," Experimental Neurology. 6:451-469, 1962. 36. Eysenck, H. J. "A Three Factor Theory of Reminiscence," British Journal of Psychology. 56:163-181, 1965. 37. Franchina, Joseph J. and Mark H. Moore, "Strychnine and the Inhibition of Previous Performance," Science. 24:903-904, 1968. 38. Freeman, G. L. "The Optimal Locus of 'Anticipatory Tensions' in Muscular Work," Journal of Experimental Psychology. [ 21:554-564,1937. j 136 i 1 39. Frommel, E.» et al. "Neuropharmacological Study of Central Reactions to Intense Muscular Exertion, Sensorial Fatigue, Sexual Excitation, Hungar, and Thirst,” Medicina et Pharmacologia Experimental!s: Supplement, l4:l-^9, 196^ '40, Geddes, David D. The Effects of Rest and Exercise on Performance of Simple Mental Tasks, Doctoral Dissertatican, University of Southern California, 1959. 41. Gerard, R. W. "Biological Roots of Psychiatry," Science. 122:225-232, 1955. 42. Gibson, Eleanor J. "A Systematic Application of the Concepts of Generalization and Differentiation to Verbal Learning," Psychological Review, 47:196-229, 1940 43. Glickman, E. "Deficits in Avoidance Learning Produced by Stimulation of the Ascending Reticular Formation," Canadian Journal of Psychology, 12:97-102, 1958. 44. Glickman, Stephen E. "Perservative Neural Processes and Consolidation of the Memory Trace," Psychological Bulletin, 58:218-233, 1961. 45. Gottlieb, Gilbert. "Cerebral Dominance: Temporary Disruptifflaa of Verbal Memory by Unilateral Electroconvulsive Shock Treatment," Journal of Comparative and Physiological Psychology, 60:368-372, 1965. 46. Gray, Wendell L. "The Effect of Forced Activity on the Mas® Performance of White Rats," Journal of Comparative Psychology. 23:475-512, 1937. 47. Gutin, Bernard. "Effect of Local Fatigue on Rotary Pursuit Performance and Learning," Paper presented at the Conference on Sports Medicine, Sponsored by the American College of Sports Medicine, Pennsylvania, I96B- 48. Hoch, Eugene R. "Learning as a Function of Water Temperature,* Journal of Experimental Psychology. 16:442-445, 1933. 49. Halstead, Ward C., and William B. Rucker. "Memory: A Molecular Maze," Psychology Today, 1:38-41, 66-67, 1963. 50. Hickham, J. B., et al. "Respiratory Regulation During Exercise in Unconditioned Subjects," Journal of Clinical Investigation, 30:503-516, 1951. 137 51. Krech, D., M. R. Rosenzweig, C. L. Bennett, and B. Krueckel. 'Enzyme Concentrations in the Brain and Adjustive Behavior Patterns,1 1 Science, 120:994-996, 1954. 52. Leukel, Francis. "A Comparison of the Effects of ECS and Anesthesia on Acquisition of the Maze Habit,1 1 Journal of Comparative and Physiological Psychology. 50*300-306, 1957. 53. McConnell, J. V. "Memory Transfer Through Cannibalism in Planarium," Journal of Neuropsychiatry: Supplement 1, 3:542-548, I9Z2. 54. McGeoch, John A. "The Influence of Four Different Interpolated Activities Upon Retention," Journal of Experimental Psychology. 14:400-413, 1931. 55. Madsen, Millard C. and James L. McGaugh. "The Effect of ECS on One-Trial Avoidance Learning," Journal of Comparative and Physiological Psychology. 54:522-523, 1961. 56. Malmo, Robert B. "Activation: A Neurophysiological Dimension," Psychological Review. 66:367-386, 1959. 57. Minami, Hiroshi; and Karl M. Dallenbach. "The Effect of Activity Upon Learning and Retention in the Cockroach," American Journal of Psychology, 59:1-58, 1946. 58. Misanin, James R., Ralph R. Miller, Donald J. Lewis. "Retrograde Amnesia Produced by Electroconvulsive Shock After Reactivation of a Consolidated Memory Task," Science, 160:554-555* 1968. 59. Nielson, Harold C. "Evidence That Electroconvulsive Shock Alters Memory Retrieval Rather Than Memory Consolidation," Experimental Neurology. 20:3-20, 1968. 60. Nunney, Derek C. "Physical Activity, Fatigue and Motor Learning," Unpublished Doctoral Dissertation, University of California, Los Angeles, 1961. 61. Overton, D. A. "State-Dependent or 'Dissociated' Learning Produced with Pentobarbital," Journal of Comparative and Physiological Psychology, 57:3-12, 1964. 62. Pearlman, Chester A., Seth K. Sharpless, and Murray E. Jarvik. "Retrograde Amnesia Produced by Anesthetic and Convulsant Agents," Journal of Comparative and Physiological Psychology. 54:109-112, 1961. Pineda, Anselmo, and Hike Adkissan. "ELectroencephalographic Studies in Physical Fatigue," Texas Reports in Biology and Medicine, 19:332-3^2, 1961. Price, Jones C. "The Effect of Exercise on the Growth of White Rats," Quarter Jonnamal of Experimental Physiology. 16: 61-6?, 19261 Purdy, Bonnie J. and Aileene Lockhart. ' ’ Retention and Relearning of Gross Motor Skills After Long Periods of No Practice," Research Quarterly. 33:265-277, 1962. Rachman, S. and J. Grass!. "Reminiscence, Inhibition, and Consolidation," British Journal of Psychology. 56:157-152, 1965. Ray, 0. S., and L. Bivens. "Reinforcement Magnitude as a Determinant of Performance Decrement After Electro convulsive Shock," Science, 160:330-332, 1968. Robinson, Sid. "Physiology of Muscular Exercise," Medical Physiology. Edited by Philip Bard, St. Louis: C. V. Mosby Company, 1961. Romanowski, Wieslaw, and Mladystaw StrazyAski. "Effect of Exercise on the Tissue Respiration of the Brain, Skeletal Muscle, Reart, liver. Kidneys, and Spleen," Internationale Zeitschgift Angewandte Physiologic Einschleisslich Arbeitsufcysiology. 25:329-338, 1968. Rosenzweig, M. R., D. Kreeh, and E. L. Bennett. "A Search For Relations Between Brain Chemistry and Behavior," Psychological Bulletin. 57:^76-^92, I960. Russell, Ritchie W. "The Physiology of Memory," Proceedings of the Royal Society of Medicine. 9:1-7, 1958. Russell, Roger. "Effects of Electroshock Convulsions on Learning and Retention in Rats as Functions of Difficulty of the Task," Journal of Comparative and Physiological Psychology. 42:137-1^2, 19^9. Scheinberg, Peritz, et al. "Effects of Vigorous Physical Exercise on Cerebral Circulation and Metabolism" American Journal of Medicine. 16:5^9-55^, 195^, 139 74. Sharpless, Seth K. "Reorganization of Function in the Nervous System - Use and Disuse," Annual Review of Physiology. 26:357-388, 1964. 75. Sinonson, Ernst, Norbert Enzer, and Roy W. Benton. 'The Influence of Muscular Work and Fatigue on the State of the Central Nervous System," Journal of Laboratory and Clinical Medicine. 28:1555-15^7, 19*0. 76. Thompson, Robert, et al. "An Analysis of the Differential Effects of ECS on Memory in Young and Adult Rats," Canadian Journal of Psychology. 12:83-96, 1958. 77. Thompson, R. and J. V. McConnell. "Classical Conditioning in the Planarian, Dugesia Dorotecephala," Journal of Comparative Physiological Psychology. 48:65-68, 1955• 78. Teraslima, Pentti, and D. F. MacLeod. "The Effect of Exercise on Blood, pH and pC02* Serum Glucose, Cholesteral, UEFA and Ketones," Journal of Sports Medicine and Physical Fitness. 6:235-2^3, 1966. 79. Tsai, Chiao. "A Comparison Study of Retention Curves for Motor Habits," Comparative Psychology Monographs. 2:1-29, 1924. 80. Turreil, E. S. and S. Robinson. "The Acid Base Equilibrium of the. Blood in Exercise," American Journal of Physiology. 137:742-745, 19%2. 81. Van Vaerenbergh, P. J. J., et al. "Lactate in Cerebrospinal Fluid During Muscular Exercise," Archives Internationales de Physiologic et de Biochimie. 73:729-737, 1965. 82. Whshburn, M. F. "Hunger and Speed of Running as Factors on Maze Learning," Journal of Comparative Psychology. 6: 181- 187, 1926. 83. Westbrook, William H. and James L. McGaugh. "Drug Facili tation of Latent Learning," Psychopharmacologica. 5:440-446, 1964. 84. Wever, Ernest G. "Water Temperature as an Incentive to Swimming Activity in the Rat," Journal of Comparative Psychology. 14:219-224, 1933. 85. Zobl, ELdred G., et al. "Effect of Exercise on the Cerebral Circulation and Metabolism," Journal of Applied Physiology. 20:1289-1293, 1965. MEAN SCORES PER BLOCK OF TRIALS FOR GROUP A T1 T2 T3 T4 T5 T6 s Er ET Er ET Er ET Er ET Er ET Er ET 1 4*2 103.0 2-2 63.6 3.0 44.5 4.0 92*2 3*2 72.5 2.0 67.O 2 3.0 37.0 4.0 66.0 3-2 6O.5 0.0 38.2 0.5 20.0 1-2 36.5 3 2-2 150.0 2.0 53.5 3-2 3?. 2 4.0 41.0 4.0 52.0 4.0 32.5 4* 5 4-2 69.0 4.5 74.0 3.5 44.0 3.0 37.0 3-2 43.0 4.0 42.5 6 2*2 80.5 3-5 154.5 6.0 263.0 2.2 28* 2 4.0 71.0 2.2 51.0 7 3*2 48.5 2.5 31.0 1*2 40.5 2-2 52.0 1.0 40.0 0.0 17.2 8 104.0 6.5 134.5. 2.5 2?« 2 2-2 2^-2 1.0 28.0 2.0 4?. 2 9 4.0 48.0 2.o 44.0 2.0 28.5 2.0 24.5 3-2 26.0 1.0 17.0 10 2.0 81.5 2.0 47.0 4.0 67.0 2.0 35.0 1.0 22.5 1.2 32.0 11* 12 4.^ 83.5 2-2 76.0 2-2 48.0 4.2 24.2 1-2 38.0 1-2 43.0 13* 14 2.0 81.0 5.0 107.0 3.0 62.5 3-2 45.0 1.2 27.2 2.0 29.5 15 5.0 95.0. 2-2 44.0 2.0 25.0 3.0 34.0 2.0 21.0 1.0 19.0 16 2*2 51.0 2.° 57.0 1.2 31.2 3.° 42.0 3-2 *2-2 2.0 25.O 17 4-2 85.5 3-2 48.5 2.0 23-2 2.0 31.2 2.0 42.5 2.0 29.5 S = Subject Er = Mean Error Score ET = Mean Elapse Time Score in Seconds Ti = Block of Two Trials * = Animal Died or Removed from Study (see reading text) MEAN SCORES PER BLOCK OF TRIALS FOR GROUP B s Tl Er ET T2 Er ET T3 Er ET T4 Er ET T5 Er ET T6 Er ET 1 3.0 90.5 3-5 137.0 4.0 197.1 3.5 204.5 4.0 204.5 3.0 201.5 2 4.0 6 0 6.0 126.0 2.3 36.0 2-5 91.0 3.5 84.5 3.5 105.0 3 ^•5 69.0 3.0 90.5 2.5 55-° 3.0 73.5 1.0 71.5 2.0 84.0 4 2.3 35.5 O 61.5 3.0 49.5 4.0 63.0 2.5 71.5 2-5 53-0 5 4.0 51.5 2.5 38.0 2-5 34.5 0-5 17-5 1.0 19.0 0-5 20.5 6* 7* 8 2-5 45.5. 5.0 40.0 1-5 17-5 3.5 197.0 2.5 300.0 1.5 300.0 9 5.0 102.5 5-5 103.0 2.5 84.5 5.0 259.0 3.5 223.0 2.0 204.0 10 3.0 73.0 3.0 75.0 3.0 55-5 3-0 86.0 1.5 48.0 1.0 31.0 11 ^•5 151.0 3.0 150.5 3.5 152.5 4.0 186.5 2.5 300.0 2.0 165.5 12 4-3 70.5_ 2.5 39.5 3-5 46.5 2.0 26.5 .1,5 .28.5 1.0 19.0 13 3-5 52.5 O 60.5 1-5 47.0 1.0 40.0 1.5 27.0 0.0 16.5 14 ?.o 41.0 7.0 132.0 3.5 57.0 3.0 35-5 3,5 35,5 3.0 30.5 15 5.0 96,5.. 4.5 58.5. 2.0 40.5 3.0 42.0 2.0 33.0 2-5 52.0 16* 17 3-5 60.5 5-° 57-5 4.0 45.0 3-5 45.0 4.0 34.0 3.0 23.O Subject Mean Error Score Mean Elapse Time Score in Seconds Block of Two Trials Animal Died or Removed from Study (see reading text) S = Er = ET = Ti = 143 j MEAN SCORES PER BLOCK OF TRIALS FOR GROUP C i | T1 T2 T3 T4 T5 T6 i S Er ET Er ET Er ET Er ET Er ET Er ET 1 3-3 92.0 4.0 98.5. 4.0 55.0 5,5 115.5 3.3 72.0 2.5 160.0 2 4. ^ 60.0 3.0 39.0 3.° 40.5 6,5 75.0 5.0 88.0 3-3 62.0 3 3-5 33.5 3-5 _ 55.5 4.0 57.0 2.0 37.3 2.0 39.0 3.0 47.0 4 4.0 72.5 4.0 . 61.5 2.5 31.3 2.0 32.5 1.3 39.0 2.3 23.5 5* 6 5.0 209,5. 4.3 92.0 3.3 83.5 4.0 45.0 3.0 59.0 3.0 62.5 7 2.5 108.5 5.0 146.5_ 3.5 71.5 2.5 58.5 4.0 37-3 5.0 69.0 8* 9 4.0 216.5 2.5 127.0 4.3 70.0 3.3 61.0 3.0 67.5 3.o 29.0 10 3*3 101.0 4.0 55.0_ 2.3 58.5 4.5 300.0 1.5 166.5 2.5 300.0 11 4.0 46.0 5.0 89.5 4.0 42.5 3.3 |34.0 3.0 36.5 4.0 119.0 12* 13 4.0 51.0 3.0 36.5 3.3 33.3 4.3 40.0 6.0 56.0 4.0 37.0 14 4.0 59.0 4.0 53.0 3.0 45.O 1.0 23.5 2.5 31.0 2.3 38.0 15 5.0 95,5_ 4.0 62.5 3.0 51.0 1.0 23.O 2.3 35.0 3.0 28.3 16 4.5 88.5 3-3 71.0 2.5 37.5 2.3 75.0 1.0 24.5 2.0 23.3 17 3.3 2?0.3 6.0 64.5 3.3 43.3 3.0 38.5 4.5 49.5 3.3 43.0 S = Subject Er = Mean Error Score ET = Mean Elapse Time Score in Seconds Ti = Block of Two Trials * = Animal Died or Removed from Study (see reading text)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The Effect Of Body Position On The Development Of Isometric And Isotonic Strength

PDF

The Relationship Between Work Capacity And Motor Learning

PDF

A Comparison Of The Ability To Control Single Motor Units In Selected Human Skeletal Muscles

PDF

Effects Of Exercise Upon Joint Mobility Of Young And Old Men

PDF

The Effect Of Three Different Pace Plans On The Cardiac Cost Of 1320-Yardruns

PDF

The Effect Of Participation In A Girls Inter-University Athletic Program Upon Selected Physiological Variables

PDF

The Differential Effects Of Viewing Selected Moving Visual Figure Patterns On The Performance Of A Dynamic Balance Task

PDF

The Rate Of Learning Motor Tasks

PDF

A Comparison Of Interference And Decay During Short-Term Motor Memory

PDF

Concepts Derived From Observed Movement Patterns Represented By Visual Forms

PDF

An Experimental Investigation Of The Relationship Of Age To Various Parameters Of Muscle Strength

PDF

The Effect Of Exercise Training On Systolic Time Intervals In Elderly Men

PDF

Patterns Of Error In Learning Balance Tasks

PDF

Warm-Up Effects Of Relaxation And Stretching Techniques Upon Gross Motor Performance

PDF

Transfer And Retention Of Selected Balance Skills

PDF

Augmented Knowledge Of Results Related To Constant And Variable Errors And Its Effect Upon Acquisition And Retention Of A Gross Motor Skill

PDF

Effects Of Light And Heavy Equipment On The Acquisition Of Sports-Type Skills By Young Children

PDF

Flexibility Changes As A Result Of Isometric And Isotonic Exercise Over Limited Ranges Of Motion

PDF

Heart Rate Response To Stress: A Mathematical Model

PDF

Economy Of Learning At Beginning Levels Of Gross Motor Performance

Asset Metadata

Creator Hutton, Robert Stanley (author)

Core Title The Retroactive Effect Of Strenuous And Exhaustive Exercise On Maze Task Learning

Degree Doctor of Philosophy

Degree Program Physical Education

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Education, Physical,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Lersten, Kenneth C. (committee chair), Lockhart, Aileene (committee chair), deVries, Herbert A. (committee member), Morris, Royce (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-708659

Unique identifier UC11361009

Identifier 6917897.pdf (filename),usctheses-c18-708659 (legacy record id)

Legacy Identifier 6917897.pdf

Dmrecord 708659

Document Type Dissertation

Rights Hutton, Robert Stanley

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA