- Email: [email protected]
Interocular transfer of color discrimination after tectal lesions in goldfish
EXPERIMENTAL
NEUROLOGY
Interocular
67,504-512
Transfer
(1980)
of Color Discrimination Lesions in Goldfish
after Tectal
KAREN F. GREIF AND MARGARET Y. SCOTT’ Di~~ision
of Biology.
Culifornirr
Institlrte Recei\,ed
of Technology. August
Pasndenu,
California
91125
27, 1979
Interocular transfer of a color discrimination task after lesion of trained tectal regions was studied in adult goldfish, using differential suppression of respiration to highly localized stimuli as the behavioral measure. Stimuli were confined to the posterior visual field outside the area of binocular overlap. Either the caudal half-tectum containing the visual projections involved in training or the entire tectum was ablated. Excellent interocular transfer across the entire visual field of the untrained eye was observed. The results suggest that engrams for monocularly acquired visual learning may be laid down bilaterally during acquisition.
INTRODUCTION Interocular transfer, the generalization of monocular learning from the trained to untrained eye, has the potential for providing insight into the mechanisms of learning and memory. By analysis of patterns of transfer, it may be possible to determine the functional roles of structures involved in the transfer of information and to determine whether memory storage involves one or both halves of the brain. Unfortunately, there is frequently disagreement concerning interocular transfer due to variations in procedure and interpretation of results. Such is the case in teleost fishes. Interocular transfer in teleosts has been studied since Sperry and Clark (24) reported partial transfer of active discrimination of lures which differed Abbreviations: HM-horizontal meridian; CS+. CS--reinforced, neutral conditioned stimulus. ’ This work was supported by National Institutes of Health grant GM 00086. U.S. Public Health Servicegrant MH 03372 to R. W. Sperry, and the McCallum Fellowship to M.Y.S. We thank R. Meyer and R. W. Sperry for helpful criticisms of the manuscript, and J. Macenka for histology, Dr. Greif s current address is Department of Physiology. School of Medicine, University of California, San Francisco, CA 94143. 504 0014-4886/80/030504-09$02.00/O Copyright All rights
0 1980 by Academic Press. Inc. of reproduction in any form reserved.
INTEROCULAR
TRANSFER
IN GOLDFISH
505
in color, size, and position, for food reward. Although transfer of learning occurred, it was by no means complete or instantaneous. Color discrimination has in general been reported to show transfer in goldfish (9, 12, 14). Interocular transfer of simple patterns using both active avoidance and cardiac conditioning was confirmed (10, 12, 14,23). However, Ingle (9) found that fish trained with stimuli which differed both in color and pattern discriminated only on the basis of color when tested with the untrained eye, using an active discrimination paradigm. In a later study ( I 1), easy patterns were found to transfer, but stimuli which were difficult to discriminate transferred poorly to the second eye. It is not necessarily reasonable to assume that transfer occurs equally well to all portions of the visual field of the untrained eye. If, for example, fish normally attach greater importance to stimuli which appear in the area of the visual field for which it has binocular vision (anterior visual field, 25” on each side of the midline along the horizontal meridian), then interocular transfer from the posterior visual field of the trained eye might be expected to be less effective than from portions which do interact binocularly. Previous studies did not restrict initial training to limited regions of the visual field. If innate differences between binocular and monocular portions of the visual field exist, careful testing of performance across the visual field of the untrained eye may reveal systematic differences in the effectiveness of interocular transfer. Using an apparatus which permits highly localized presentation of visual stimuli (19, 20), generalization of task performance across the visual field of the trained eye after lesion of trained tectal regions was reported (7), using conditioned respiratory suppression as the behavioral measure. Using the same procedure, which is adapted from eye-in-water recording techniques (16, 18). we investigated the interocular transfer of a color discrimination task after lesions of the trained posterior tectum. Preliminary reports have been published (6, 21). GENERAL
METHODS
Common goldfish (Carassius aurafus) were used throughout, about 6 to 9 cm from snout to base of tail. Fish were housed under diurnal lighting conditions. Differential suppression of respiration to red or green light was used as the behavioral method of discrimination, as previously described (7, 20). Briefly, fish were held underwater and presented with stimuli at specified loci along the horizontal axis of a water-filled hemisphere; stimuli were projected via a fiber-optic source mounted on a perimeter. Stimuli were 5 s of flashing red or green light subtending 1.5” visual angle and equated for brightness according to Yager (25).
506
GREIF
AND
SCOTT
The extraocular muscles of one eye were cut while the fish was anesthetized with 0.05% tricaine methanesulfonate (Finquel) and the optic disk centered at the apex of the hemisphere. After habituation to the stimuli, conditioned suppression was achieved by repeatedly pairing the red light (CS+) with simultaneous shock to the tail, with intermittent exposure to the green light. Stimuli were presented in all fish 60” behind the optic disk along the horizontal meridian (-60” HM). Fish usually learned the discrimination in three training sessions of approximately 4 h each, spaced 4 days apart. In some cases, difficulty in suppressing response to the green light (CS-) was observed; these fish were given as many as four 15-min periods of exposure to the green light alone in an effort to extinguish the unwanted responses. Three days after reaching criterion (~90% response to CS+, ~20% response to CS- for at least 40 trials), fish were anesthetized for surgery. Tectal lesions were made by cutting down to the ventricle with a sharpened tungsten knife and removing the tectal laminae by aspiration. Fish were revived by infusion of water through the gills. The day after surgery, each fish was reanesthetized and the extraocular muscles of the untrained eye were cut prior to testing for interocular transfer. The eye was aligned as previously described and a 30-min period for adaptation elapsed before testing. Stimuli were first presented in the posterior visual field and thereafter alternated between anterior and posterior fields along the horizontal meridian. Test loci were spaced 20 apart, with three to five presentations of each color at each locus. EXPERIMENT
I: HALF-TECTUM
ABLATION
Methods. Six fish were used; all trained at -60” HM. After the final training day, lesions of the caudal half-tectum were made. The size of the lesions was varied to ascertain the effect on performance. Fish recovered rapidly with no visible motor deficits. Approximately 1 week after initial testing for interocular transfer, generalization in the remaining portions of the visual field of the trained eye was examined. Stimuli were presented as described along the horizontal meridian, beginning in the anterior visual field. Results. Results are summarized in Fig. 1. Good transfer of the visual discrimination task was observed across the entire visual field of the untrained eye (Fig. IA). Variability was greater in the posterior visual field, with some fish performing well and others poorly in this region. No overall deficit of response was observed in the posterior field, contrary to the expected result. Generalization of response within the trained eye is shown in Fig. 1B. Good intraretinal transfer was observed across the remainder of the visual
INTEROCULAR
TRANSFER
507
IN GOLDFISH
\ .\
-40-60 POST. ECCENTRICIT
POST
ANT. Y (Degrees)
FIG. I, Interocular and intraretinal transfer after lesions of trained tectal regions. Six fish were trained at -60” along the horizontal meridian. Eccentricity is measured in degrees from the optic disk. Ant,-anterior visual field. Post.-posterior visual field. Differential response = percentage response to CS+ minus percentage response to CS-, averaged for all fish. Dotted lines-absolute ranges for all fish. A-responses in the untrained eye. B-responses in trained eye, approximately 1 week after interocular transfer testing.
field, as was reported (7). Data for loci immediately adjacent to the edge of the scotoma were discarded to eliminate bias in averaged values. No trends were observed relating lesion size and postsurgery performance, or final performance before surgery and transfer performance. No improvement in retention was observed with a longer recovery period before intraretinal generalization was tested, compared with fish tested I day after surgery (7). EXPERIMENT
II: COMPLETE
TECTAL
LESIONS
Methods. Three fish were used. Surgery was carried out through a slightly larger bone flap. Tectal tissue was removed by essentially peeling the tectal laminae off the ventricle with care taken to remove extreme rostra1 and lateral portions. The fish showed minor motor deficits, which disappeared by the time of testing the following day. After testing of the untrained eye, the fish were reversed in the training tank and tested for residual vision in the trained eye. If no responses to the stimuli were noted. attempts were made to elicit startle responses to larger objects. At the conclusion of testing, the fish were given an overdose of anesthesia and the brains removed for histology. Brains were fixed in Bodian’s fixative (ethanol, acetic acid, Formalin), embedded in paraffin, and sectioned at 10 pm in the sagittal plane.
508
GREIF AND SCOTT
+60+40+20 ANT. ECCENTRICITY
O-20
-40-60 POST (Degrees)
FIG. 2. Interocular transfer after complete tectal lesion, N = 3. Conventions as in Fig. 1. Fish trained at -60” along the horizontal meridian.
Resulrs. Results are shown in Fig. 2. Good interocular transfer was observed across the visual field of the untrained eye. No deficits were observed in the posterior visual field and the increased variability in the posterior field in fish with partial tectal ablation was not observed. Fish showed little residual response in the trained eye. One fish (CTL 2) made a few weak responses in the extreme anterior visual field (+80”), but did not respond elsewhere. The two remaining fish made no responses to the stimuli, to a waving hand, or to a looming body shadow. Ablation reconstructions of histological sections are shown in Fig. 3. Approximately 85 to 90% of the optic tectum was removed, with remnants mainly in the extreme ventrolateral portion. This tissue was visually isolated by lesion of the remainder of the tectum, but presumably retained efferents to other structures. In fish CTL 2, which showed slight residual response, the extreme anterior portion of the tectum was spared.
DISCUSSION The finding of consistent interocular transfer of color discrimination after lesions of trained tectal regions has several implications. The absence of a deficit in transfer from the posterior visual field of the trained eye to that of the untrained eye indicates that binocular interaction is not required for the transfer of color information. It also appears that color is recognized in a similar fashion throughout the visual field, in terms of content and importance. This conclusion is supported by previous findings of consistent transfer of color discrimination learning within one eye after highly localized training. These results are in marked contrast to the reported failure to transfer of pattern discrimination in pigeons using a modified Lashley jumping stand
INTEROCULAR CTL
2
TRANSFER CTL
3
509
IN GOLDFISH CTL4
FIG. 3. Ablation reconstructions (sagittal sections) of brain sections from fish with complete tectal lesions (CTL). Diagonal lines encompass tissue removed in each fish. Abbreviations: C-cerebellum. F-forebrain, ON-optic nerve, OTec-optic tectum. T-thalamus.
(4, 131. These researchers, as well as Catania (2), pointed out that the training apparatus may require the bird to view the stimuli in the far lateral visual field, because of myopia in more central regions. Tasks learned in the monocular portion of the visual field, in this case, failed to transfer to the untrained eye, suggesting that innate differences between monocular and binocular portions of the visual field of birds might exist. The results of the present study, as well as of the preceding (7) do not support this hypothesis in goldfish. The absence of transfer deficits suggests that access to the memory store for this task is bilateral in the goldfish brain. It is clear that the portion of the
510
GREIF
AND
SCOTT
tectum receiving direct visual input is not required for retention of the task; after lesions of directly “trained” tectal regions, good retention is still observed. A possibility which could not be eliminated was that engrams are stored in the remaining portion of the trained tectum, or in an association area, and that information was tapped by the naive hemisphere at the time of testing. Experiment II was initiated to examine this possibility by studying the effects of complete tectal ablation after localized training on the interocular transfer of the color discrimination task. Removal of the entire trained tectum did not prevent interocular transfer of a learned color discrimination. This suggests that memory for the task is established bilaterally in the brain, or unilaterally in a region which is accessible by the contralateral hemisphere. If storage is unilateral, it is reasonable to suggest that such storage is not in a highly localized region, because of the resulting complexity in neuronal circuitry. A highly discrete region would require visual connections from the entire visual field or a means to integrate visual information, and access to a pathway across the midline to the contralateral hemisphere. Although extratectal visual projection regions have been demonstrated [3,22], their roles have not yet been determined and further projections to the contralateral hemisphere have not been delineated. Further support for bilateral engrams was advanced by Ingle (9) who reported that fish which are presented with “distracting” stimuli to the untrained eye during monocular training failed to show transfer of the task during testing, and that fish were able to learn conflicting discriminations with each eye individually. Ingle proposed that presentation of irrelevant information might inhibit the formation of bilateral engrams and thus prevent interocular transfer, implying that such transfer occurred only when memory traces were bilateral. The pathways involved in the interocular transfer of information have yet to be unequivocally determined. However, most evidence suggests that the postoptic commissure is critical for the transfer of visual discrimination learning (12,26) and that the tectal commissure may serve to transfer some motor elements of incompletely learned tasks (1, 12). If the postoptic commissure does mediate transfer of visual discrimination, then it may well be that extratectal visual regions are involved in information transfer, as tectotectal efferent fibers do not project through the postoptic commissure. The suggestion of apparent bilateral memory for color discrimination in goldfish is in accord with most studies in mammals (8, 17) and birds (5,15) in which evidence for bilateral storage was found after monocular training and either section of commissural pathways or specific lesions. A parallel finding in a teleost fish suggests that a strategy for memory processing may exist which does not require specific brain structures. Despite considerable
INTEROCULAR
TRANSFER
IN GOLDFISH
511
evolutionary divergence and profound differences in brain anatomy. most animals appear to use redundancy of storage as a general rule. Rather than increase capacity by limiting traces to a single hemisphere, it appears that even after monocular training, engrams are laid down bilaterally. It may be that if an experience is worthy of retention, efforts are taken to assure that important information is not lost. REFERENCES I. BENGSTON, L. 0.. A. FRANCIS, AND M. S. GAZZANIGA. 1979 Tests for interocular transfer after tectal commissure transection in goldfish. Erp. Nelrrol. 64: 528-534. 2. CATANIA. A. C. 1964. On the visual acuity ofthe pige0n.J. Erp. Awl. Behav. 7: 361-366. 3. FINGER, T. E.. AND H. J. KARTEN. 1977. The accessory optic system in teleost fishes. Sot. Neurosci. Ahsrr. 3: 90. 4. GRAVES. J. A.. AND M. A. GOODALE. 1977. Failure of interocular transfer in the pigeon (Columba livia). Physiol. Brhuv. 19: 425-428. 5. GREIF, K. F. 1976. Bilateral memory for monocular one-trial passive avoidance in chicks. Behav. Biol. 16: 453-462. 6. GREIF. K. F., AND M. Y. SCOTT. 1977. Localization and generalization of visual memory after tectal lesions in goldfish. Sot. Neurosci. Absfr. 3: 234. 7. GREIF, K. F., AND M. Y. SCOTT. 1980. Intraretinal transfer ofa color discrimination task after tectal and telencephalic lesions in goldfish. E-VP. Nertrol. 67: 492-503. 8. HAMILTON, C. R. 1977. Investigations of perceptual and mnemonic lateralization in monkeys. Pages 45-62 in S. HARNAD, R. W. DOTY, L. GOLDSTEIN, J. JAYNES, AND G. KRAUTHAMER, Eds.. Lareralization in the Ner\wts System. Academic Press. New York. 9. INGLE, D. J. 1965. Interocular transfer in goldfish: color easier than pattern. Science 149: IOOC- 1002. IO. INGLE. D. J. 1967. Two visual systems underlying the behavior of fish. Ps~chol. Forsch. 31: 44-51. I I. INGLE. D. J. 1968. Interocular integration of visual learning by goldfish. Brain Behot,. Ebd. 1: 58-85. 12. INGLE, D. J., AND A. CAMPBELL. 1977. Interocular transfer of visual discriminations in goldfish after selective commissure lesions. J. Camp. Physiol. Psycho/. 91: 327-335. 13. LEVINE, J. 1945. Studies in the interrelations of central nervous structures in binocular vision. I. The lack of bilateral transfer of visual discriminative habits acquired monocularly in the pigeon. J. Getter. Psycho/. 67: lO5- 129. 14. MCCLEARY. R. A. 1960. Type of response as a factor in interocular transfer of pattern discrimination in cichlid fish. E.xp. Neun~l. 16: 3 I l-32 I. 15. MELLO, N. K. 1968. The effect of unilateral lesions of the optic tectum on interhemispheric transfer of monocularly trained color and pattern discrimination in pigeon. Physiol. Behart. 3: 725-734. 16. MEYER, R. L. 1977. Eye-in-water electrophysiological mapping of goldfish with and without tectal lesions. E.rp. Nell&. 56: 23-4 I. 17. MYERS, R. E. 1961. Corpus callosum and visual gnosis. Pages 481-505 in GERARD AND KONORSKI, Eds., Brain Mechanisms and Learning. Blackwell. Oxford. 18. SCHWASSMAN. H. 0.. AND L. KRUGER. 1965. Organization of the visual projection upon the optic tectum of some freshwater fish. J. Camp. Nertrol. 124: 113- 126. 19. SCOTT, M. Y. 1975. Functional capacity ofcompressed retinotectal projection ingoldfish. Anaf. Rec. 181: 474.
512
GREIF
AND
SCOTT
20. SCOTT, M. Y. 1977. Behavioral tests of compression of retinotectal projection after partial tectal ablation in goldfish. Exp. Neural. 54: 579-590. 21. SCOTT, M. Y., K. F. GREIF, AND R. W. SPERRY. 1977. Survival of visual discrimination learning after ablation of trained tectal regions in goldfish. Anat. Rec. 187: 778. 22. SHARMA, S. C. 1972. The retinal projections in thegoldfish: anexperimental study. Brain Res.
39: 213-223.
23. SHAPIRO, S. M. 1965. Interocular transfer of pattern discrimination in the goldfish. Am. J. Psychol. 78: 21-38. 24. SPERRY, R. W.. AND E. CLARK. 1949. Interocular transfer of visual discrimination in a teleost fish. Phvsiol. Zoo/. 22: 372-378. 25. YAGER, D. 1967. Behavioral methods and theoretical analysis of spectral sensitivity and spectral saturation in the goldfish. Crrrussius UN~U~US.Vision Res. 7: 707-727. 26. YEO, C. H., AND G. E. SAVAGE. 1976. Mesencephalic and diencephalic commissures and interocular transfer in the goldfish. E.rp. Neural. 53: 56-63.
NEUROLOGY
Interocular
67,504-512
Transfer
(1980)
of Color Discrimination Lesions in Goldfish
after Tectal
KAREN F. GREIF AND MARGARET Y. SCOTT’ Di~~ision
of Biology.
Culifornirr
Institlrte Recei\,ed
of Technology. August
Pasndenu,
California
91125
27, 1979
Interocular transfer of a color discrimination task after lesion of trained tectal regions was studied in adult goldfish, using differential suppression of respiration to highly localized stimuli as the behavioral measure. Stimuli were confined to the posterior visual field outside the area of binocular overlap. Either the caudal half-tectum containing the visual projections involved in training or the entire tectum was ablated. Excellent interocular transfer across the entire visual field of the untrained eye was observed. The results suggest that engrams for monocularly acquired visual learning may be laid down bilaterally during acquisition.
INTRODUCTION Interocular transfer, the generalization of monocular learning from the trained to untrained eye, has the potential for providing insight into the mechanisms of learning and memory. By analysis of patterns of transfer, it may be possible to determine the functional roles of structures involved in the transfer of information and to determine whether memory storage involves one or both halves of the brain. Unfortunately, there is frequently disagreement concerning interocular transfer due to variations in procedure and interpretation of results. Such is the case in teleost fishes. Interocular transfer in teleosts has been studied since Sperry and Clark (24) reported partial transfer of active discrimination of lures which differed Abbreviations: HM-horizontal meridian; CS+. CS--reinforced, neutral conditioned stimulus. ’ This work was supported by National Institutes of Health grant GM 00086. U.S. Public Health Servicegrant MH 03372 to R. W. Sperry, and the McCallum Fellowship to M.Y.S. We thank R. Meyer and R. W. Sperry for helpful criticisms of the manuscript, and J. Macenka for histology, Dr. Greif s current address is Department of Physiology. School of Medicine, University of California, San Francisco, CA 94143. 504 0014-4886/80/030504-09$02.00/O Copyright All rights
0 1980 by Academic Press. Inc. of reproduction in any form reserved.
INTEROCULAR
TRANSFER
IN GOLDFISH
505
in color, size, and position, for food reward. Although transfer of learning occurred, it was by no means complete or instantaneous. Color discrimination has in general been reported to show transfer in goldfish (9, 12, 14). Interocular transfer of simple patterns using both active avoidance and cardiac conditioning was confirmed (10, 12, 14,23). However, Ingle (9) found that fish trained with stimuli which differed both in color and pattern discriminated only on the basis of color when tested with the untrained eye, using an active discrimination paradigm. In a later study ( I 1), easy patterns were found to transfer, but stimuli which were difficult to discriminate transferred poorly to the second eye. It is not necessarily reasonable to assume that transfer occurs equally well to all portions of the visual field of the untrained eye. If, for example, fish normally attach greater importance to stimuli which appear in the area of the visual field for which it has binocular vision (anterior visual field, 25” on each side of the midline along the horizontal meridian), then interocular transfer from the posterior visual field of the trained eye might be expected to be less effective than from portions which do interact binocularly. Previous studies did not restrict initial training to limited regions of the visual field. If innate differences between binocular and monocular portions of the visual field exist, careful testing of performance across the visual field of the untrained eye may reveal systematic differences in the effectiveness of interocular transfer. Using an apparatus which permits highly localized presentation of visual stimuli (19, 20), generalization of task performance across the visual field of the trained eye after lesion of trained tectal regions was reported (7), using conditioned respiratory suppression as the behavioral measure. Using the same procedure, which is adapted from eye-in-water recording techniques (16, 18). we investigated the interocular transfer of a color discrimination task after lesions of the trained posterior tectum. Preliminary reports have been published (6, 21). GENERAL
METHODS
Common goldfish (Carassius aurafus) were used throughout, about 6 to 9 cm from snout to base of tail. Fish were housed under diurnal lighting conditions. Differential suppression of respiration to red or green light was used as the behavioral method of discrimination, as previously described (7, 20). Briefly, fish were held underwater and presented with stimuli at specified loci along the horizontal axis of a water-filled hemisphere; stimuli were projected via a fiber-optic source mounted on a perimeter. Stimuli were 5 s of flashing red or green light subtending 1.5” visual angle and equated for brightness according to Yager (25).
506
GREIF
AND
SCOTT
The extraocular muscles of one eye were cut while the fish was anesthetized with 0.05% tricaine methanesulfonate (Finquel) and the optic disk centered at the apex of the hemisphere. After habituation to the stimuli, conditioned suppression was achieved by repeatedly pairing the red light (CS+) with simultaneous shock to the tail, with intermittent exposure to the green light. Stimuli were presented in all fish 60” behind the optic disk along the horizontal meridian (-60” HM). Fish usually learned the discrimination in three training sessions of approximately 4 h each, spaced 4 days apart. In some cases, difficulty in suppressing response to the green light (CS-) was observed; these fish were given as many as four 15-min periods of exposure to the green light alone in an effort to extinguish the unwanted responses. Three days after reaching criterion (~90% response to CS+, ~20% response to CS- for at least 40 trials), fish were anesthetized for surgery. Tectal lesions were made by cutting down to the ventricle with a sharpened tungsten knife and removing the tectal laminae by aspiration. Fish were revived by infusion of water through the gills. The day after surgery, each fish was reanesthetized and the extraocular muscles of the untrained eye were cut prior to testing for interocular transfer. The eye was aligned as previously described and a 30-min period for adaptation elapsed before testing. Stimuli were first presented in the posterior visual field and thereafter alternated between anterior and posterior fields along the horizontal meridian. Test loci were spaced 20 apart, with three to five presentations of each color at each locus. EXPERIMENT
I: HALF-TECTUM
ABLATION
Methods. Six fish were used; all trained at -60” HM. After the final training day, lesions of the caudal half-tectum were made. The size of the lesions was varied to ascertain the effect on performance. Fish recovered rapidly with no visible motor deficits. Approximately 1 week after initial testing for interocular transfer, generalization in the remaining portions of the visual field of the trained eye was examined. Stimuli were presented as described along the horizontal meridian, beginning in the anterior visual field. Results. Results are summarized in Fig. 1. Good transfer of the visual discrimination task was observed across the entire visual field of the untrained eye (Fig. IA). Variability was greater in the posterior visual field, with some fish performing well and others poorly in this region. No overall deficit of response was observed in the posterior field, contrary to the expected result. Generalization of response within the trained eye is shown in Fig. 1B. Good intraretinal transfer was observed across the remainder of the visual
INTEROCULAR
TRANSFER
507
IN GOLDFISH
\ .\
-40-60 POST. ECCENTRICIT
POST
ANT. Y (Degrees)
FIG. I, Interocular and intraretinal transfer after lesions of trained tectal regions. Six fish were trained at -60” along the horizontal meridian. Eccentricity is measured in degrees from the optic disk. Ant,-anterior visual field. Post.-posterior visual field. Differential response = percentage response to CS+ minus percentage response to CS-, averaged for all fish. Dotted lines-absolute ranges for all fish. A-responses in the untrained eye. B-responses in trained eye, approximately 1 week after interocular transfer testing.
field, as was reported (7). Data for loci immediately adjacent to the edge of the scotoma were discarded to eliminate bias in averaged values. No trends were observed relating lesion size and postsurgery performance, or final performance before surgery and transfer performance. No improvement in retention was observed with a longer recovery period before intraretinal generalization was tested, compared with fish tested I day after surgery (7). EXPERIMENT
II: COMPLETE
TECTAL
LESIONS
Methods. Three fish were used. Surgery was carried out through a slightly larger bone flap. Tectal tissue was removed by essentially peeling the tectal laminae off the ventricle with care taken to remove extreme rostra1 and lateral portions. The fish showed minor motor deficits, which disappeared by the time of testing the following day. After testing of the untrained eye, the fish were reversed in the training tank and tested for residual vision in the trained eye. If no responses to the stimuli were noted. attempts were made to elicit startle responses to larger objects. At the conclusion of testing, the fish were given an overdose of anesthesia and the brains removed for histology. Brains were fixed in Bodian’s fixative (ethanol, acetic acid, Formalin), embedded in paraffin, and sectioned at 10 pm in the sagittal plane.
508
GREIF AND SCOTT
+60+40+20 ANT. ECCENTRICITY
O-20
-40-60 POST (Degrees)
FIG. 2. Interocular transfer after complete tectal lesion, N = 3. Conventions as in Fig. 1. Fish trained at -60” along the horizontal meridian.
Resulrs. Results are shown in Fig. 2. Good interocular transfer was observed across the visual field of the untrained eye. No deficits were observed in the posterior visual field and the increased variability in the posterior field in fish with partial tectal ablation was not observed. Fish showed little residual response in the trained eye. One fish (CTL 2) made a few weak responses in the extreme anterior visual field (+80”), but did not respond elsewhere. The two remaining fish made no responses to the stimuli, to a waving hand, or to a looming body shadow. Ablation reconstructions of histological sections are shown in Fig. 3. Approximately 85 to 90% of the optic tectum was removed, with remnants mainly in the extreme ventrolateral portion. This tissue was visually isolated by lesion of the remainder of the tectum, but presumably retained efferents to other structures. In fish CTL 2, which showed slight residual response, the extreme anterior portion of the tectum was spared.
DISCUSSION The finding of consistent interocular transfer of color discrimination after lesions of trained tectal regions has several implications. The absence of a deficit in transfer from the posterior visual field of the trained eye to that of the untrained eye indicates that binocular interaction is not required for the transfer of color information. It also appears that color is recognized in a similar fashion throughout the visual field, in terms of content and importance. This conclusion is supported by previous findings of consistent transfer of color discrimination learning within one eye after highly localized training. These results are in marked contrast to the reported failure to transfer of pattern discrimination in pigeons using a modified Lashley jumping stand
INTEROCULAR CTL
2
TRANSFER CTL
3
509
IN GOLDFISH CTL4
FIG. 3. Ablation reconstructions (sagittal sections) of brain sections from fish with complete tectal lesions (CTL). Diagonal lines encompass tissue removed in each fish. Abbreviations: C-cerebellum. F-forebrain, ON-optic nerve, OTec-optic tectum. T-thalamus.
(4, 131. These researchers, as well as Catania (2), pointed out that the training apparatus may require the bird to view the stimuli in the far lateral visual field, because of myopia in more central regions. Tasks learned in the monocular portion of the visual field, in this case, failed to transfer to the untrained eye, suggesting that innate differences between monocular and binocular portions of the visual field of birds might exist. The results of the present study, as well as of the preceding (7) do not support this hypothesis in goldfish. The absence of transfer deficits suggests that access to the memory store for this task is bilateral in the goldfish brain. It is clear that the portion of the
510
GREIF
AND
SCOTT
tectum receiving direct visual input is not required for retention of the task; after lesions of directly “trained” tectal regions, good retention is still observed. A possibility which could not be eliminated was that engrams are stored in the remaining portion of the trained tectum, or in an association area, and that information was tapped by the naive hemisphere at the time of testing. Experiment II was initiated to examine this possibility by studying the effects of complete tectal ablation after localized training on the interocular transfer of the color discrimination task. Removal of the entire trained tectum did not prevent interocular transfer of a learned color discrimination. This suggests that memory for the task is established bilaterally in the brain, or unilaterally in a region which is accessible by the contralateral hemisphere. If storage is unilateral, it is reasonable to suggest that such storage is not in a highly localized region, because of the resulting complexity in neuronal circuitry. A highly discrete region would require visual connections from the entire visual field or a means to integrate visual information, and access to a pathway across the midline to the contralateral hemisphere. Although extratectal visual projection regions have been demonstrated [3,22], their roles have not yet been determined and further projections to the contralateral hemisphere have not been delineated. Further support for bilateral engrams was advanced by Ingle (9) who reported that fish which are presented with “distracting” stimuli to the untrained eye during monocular training failed to show transfer of the task during testing, and that fish were able to learn conflicting discriminations with each eye individually. Ingle proposed that presentation of irrelevant information might inhibit the formation of bilateral engrams and thus prevent interocular transfer, implying that such transfer occurred only when memory traces were bilateral. The pathways involved in the interocular transfer of information have yet to be unequivocally determined. However, most evidence suggests that the postoptic commissure is critical for the transfer of visual discrimination learning (12,26) and that the tectal commissure may serve to transfer some motor elements of incompletely learned tasks (1, 12). If the postoptic commissure does mediate transfer of visual discrimination, then it may well be that extratectal visual regions are involved in information transfer, as tectotectal efferent fibers do not project through the postoptic commissure. The suggestion of apparent bilateral memory for color discrimination in goldfish is in accord with most studies in mammals (8, 17) and birds (5,15) in which evidence for bilateral storage was found after monocular training and either section of commissural pathways or specific lesions. A parallel finding in a teleost fish suggests that a strategy for memory processing may exist which does not require specific brain structures. Despite considerable
INTEROCULAR
TRANSFER
IN GOLDFISH
511
evolutionary divergence and profound differences in brain anatomy. most animals appear to use redundancy of storage as a general rule. Rather than increase capacity by limiting traces to a single hemisphere, it appears that even after monocular training, engrams are laid down bilaterally. It may be that if an experience is worthy of retention, efforts are taken to assure that important information is not lost. REFERENCES I. BENGSTON, L. 0.. A. FRANCIS, AND M. S. GAZZANIGA. 1979 Tests for interocular transfer after tectal commissure transection in goldfish. Erp. Nelrrol. 64: 528-534. 2. CATANIA. A. C. 1964. On the visual acuity ofthe pige0n.J. Erp. Awl. Behav. 7: 361-366. 3. FINGER, T. E.. AND H. J. KARTEN. 1977. The accessory optic system in teleost fishes. Sot. Neurosci. Ahsrr. 3: 90. 4. GRAVES. J. A.. AND M. A. GOODALE. 1977. Failure of interocular transfer in the pigeon (Columba livia). Physiol. Brhuv. 19: 425-428. 5. GREIF, K. F. 1976. Bilateral memory for monocular one-trial passive avoidance in chicks. Behav. Biol. 16: 453-462. 6. GREIF. K. F., AND M. Y. SCOTT. 1977. Localization and generalization of visual memory after tectal lesions in goldfish. Sot. Neurosci. Absfr. 3: 234. 7. GREIF, K. F., AND M. Y. SCOTT. 1980. Intraretinal transfer ofa color discrimination task after tectal and telencephalic lesions in goldfish. E-VP. Nertrol. 67: 492-503. 8. HAMILTON, C. R. 1977. Investigations of perceptual and mnemonic lateralization in monkeys. Pages 45-62 in S. HARNAD, R. W. DOTY, L. GOLDSTEIN, J. JAYNES, AND G. KRAUTHAMER, Eds.. Lareralization in the Ner\wts System. Academic Press. New York. 9. INGLE, D. J. 1965. Interocular transfer in goldfish: color easier than pattern. Science 149: IOOC- 1002. IO. INGLE. D. J. 1967. Two visual systems underlying the behavior of fish. Ps~chol. Forsch. 31: 44-51. I I. INGLE. D. J. 1968. Interocular integration of visual learning by goldfish. Brain Behot,. Ebd. 1: 58-85. 12. INGLE, D. J., AND A. CAMPBELL. 1977. Interocular transfer of visual discriminations in goldfish after selective commissure lesions. J. Camp. Physiol. Psycho/. 91: 327-335. 13. LEVINE, J. 1945. Studies in the interrelations of central nervous structures in binocular vision. I. The lack of bilateral transfer of visual discriminative habits acquired monocularly in the pigeon. J. Getter. Psycho/. 67: lO5- 129. 14. MCCLEARY. R. A. 1960. Type of response as a factor in interocular transfer of pattern discrimination in cichlid fish. E.xp. Neun~l. 16: 3 I l-32 I. 15. MELLO, N. K. 1968. The effect of unilateral lesions of the optic tectum on interhemispheric transfer of monocularly trained color and pattern discrimination in pigeon. Physiol. Behart. 3: 725-734. 16. MEYER, R. L. 1977. Eye-in-water electrophysiological mapping of goldfish with and without tectal lesions. E.rp. Nell&. 56: 23-4 I. 17. MYERS, R. E. 1961. Corpus callosum and visual gnosis. Pages 481-505 in GERARD AND KONORSKI, Eds., Brain Mechanisms and Learning. Blackwell. Oxford. 18. SCHWASSMAN. H. 0.. AND L. KRUGER. 1965. Organization of the visual projection upon the optic tectum of some freshwater fish. J. Camp. Nertrol. 124: 113- 126. 19. SCOTT, M. Y. 1975. Functional capacity ofcompressed retinotectal projection ingoldfish. Anaf. Rec. 181: 474.
512
GREIF
AND
SCOTT
20. SCOTT, M. Y. 1977. Behavioral tests of compression of retinotectal projection after partial tectal ablation in goldfish. Exp. Neural. 54: 579-590. 21. SCOTT, M. Y., K. F. GREIF, AND R. W. SPERRY. 1977. Survival of visual discrimination learning after ablation of trained tectal regions in goldfish. Anat. Rec. 187: 778. 22. SHARMA, S. C. 1972. The retinal projections in thegoldfish: anexperimental study. Brain Res.
39: 213-223.
23. SHAPIRO, S. M. 1965. Interocular transfer of pattern discrimination in the goldfish. Am. J. Psychol. 78: 21-38. 24. SPERRY, R. W.. AND E. CLARK. 1949. Interocular transfer of visual discrimination in a teleost fish. Phvsiol. Zoo/. 22: 372-378. 25. YAGER, D. 1967. Behavioral methods and theoretical analysis of spectral sensitivity and spectral saturation in the goldfish. Crrrussius UN~U~US.Vision Res. 7: 707-727. 26. YEO, C. H., AND G. E. SAVAGE. 1976. Mesencephalic and diencephalic commissures and interocular transfer in the goldfish. E.rp. Neural. 53: 56-63.