t
VISUAL
MEMORY
LATERALIZATION
LORENZO VON FERSEN*
0028-3932 90 s3.00+0.00 1994 Pergamon Presr plc
IN PJGEONS
and ONUR GtiNTi_kKtiNt
Allgemeine Psychologie. uhiversity of Konstanz. D-7750 Konstanz. F.R.G. (Received 4 April 1989; accepted 13 June 1989) Abstract-Previous experiments employing simple visual discrimination tasks have revealed a cerebral lateralization in the visual system of pigeons with a dominance of the left hemisphere. Until now, visual memory lateralization in birds has not been investigated. To study possible asymmetries of visual memory functions, a simultaneous instrumental discrimination procedure was used. The animals were trained to discriminate 100 different visual patterns from a further 625 similar stimuli. Retention tests were conducted under binocular and monocular conditions. When the subjects looked monocularly, retention performance was significantly higher with the right eye (left hemisphere) than with the left eye (right hemisphere). The results suggest that the lateralization of the pigeon’s visual system depends at least partly on an asymmetry in visual memory capacity.
INTRODUCTION ALTHOUGH cerebral dominance was discovered in the 186Os, the neuronal foundations of most lateralized functions in humans remains largely obscure. Probably the only exception is the neuronal system controlling language for which detailed anatomical and physiological data of its neuronal organization has accumulated in recent years [7]. Until recently one of the reasons for the limited success in elucidating the biological foundations of hemispheric asymmetries was the lack of animal models. Language and handedness were the first systems which were recognized to be asymmetric and both were assumed to be unique to humans. As a consequence, cerebral asymmetries were treated as a peculiarity of the human brain. The last two decades have seen a dramatic change in this long-held opinion. Several functional systems were discovered to be lateralized in a number of species besides humans [30]. Experiments on these species now provide partial insights into the neuronal organization of cerebral asymmetries up to the synaptic level. One of the most influential series of studies are those by NOTTEBOHM et al., e.g. [21], in canaries. These experiments rely on a left hemisphere dominance for song production as a model for brain lateralization. Since these pioneering studies several workers have demonstrated that the left brain hemisphere in birds is not only dominant for vocalization but also for a variety of visual tasks [ 1, 8, 11,22, 331. Due to the virtually complete decussation of the optic nerves of birds, the avian visual system is particularly suited for research on lateralization [31]. Temporary occlusion of one eye ensures that information entering through the unobstructed eye is processed mainly by the contralateral hemisphere. Pigeons performing a pattern discrimination task under these conditions shdw better performance when seeing with the right eye, which projects to the dominant left hemisphere [lo]. *Present address: Department of Psychology. University tTo whom correspondence should be addressed.
of Exeter, Exeter EX4 4~33, U.K.
2
LOKENZO VOXFERSEN and ONUH GCNTORKCN
So far none of the studies on visual lateralization in birds examined possible asymmetries of visual memory functions. Several species of birds are known to be able to memorize large amounts of visual information over long time periods. Nutcrackers remember the location of thousands of food-hoards over several months [26]. VAUGHAN and GREENE [28] showed that even after 2 years pigeons exhibited an above chance discrimination of 160 pairs of complex stimuli. In another experiment [6], pigeons were trained to distinguish up to 100 repeatedly presented patterns from another 640 patterns that were shown at most twice. After a pause of 3 months the birds recognized the familiar stimuli almost as well as after intervals of a few days. These latter studies suggest that the visual memory capacity of pigeons almost matches that of humans [25]. Cerebral dominance in humans is not only evident in perceptual recognition tasks, but also in experiments which require memorization of stimulus information over some periods of time [4, 5, 241. It has been suggested on the basis ofsuch findings that lateralization effects are not only a function of immediate perceptual processes, but may actually also reflect differential storage capacities. We were therefore interested in whether it is possible to demonstrate a hemispheric asymmetry of visual memory in pigeons. The experiments were performed in the course of a more extensive study on the visual memory capacity of pigeons [29]. Only the results concerning lateralization of memory functions will be reported here. METHOD Subjects
Subjects were four adult homing pigeons (Columba Ma) oflocal stock. They were kept in individual indoor cages. The pigeons were normally maintained at 80% of their free feeding weight. At a certain time point of the experiment (detailed in the next section) they spent 6 months in an outside aviary and were allowed to regain their full weight. Afterwards the animals were deprived again. Apparatus
A modified two-key Skinner box was used (Fig. I ). The chamber measured 33 x 34 x 33 cm. The front wall had an opening of I I x 12.5 cm above the floor. A houselight was located 5 cm above this opening. A horizontal working surface of I2 x 9 cm was attached outside of the opening and level with its lower edge. It was surrounded by an enclosure measuring I6 x I4 x 10 cm. Two translucent perspex keys of 2.2 cm dia. were positioned side by side on the working surface. with their centres separated by 9 cm. Immediately in front of the keys were two reward receptacles. Two automatic feeders on top of the enclosure could be activated to deliver a few grains of millet to the receptacles through lengths of tubing. A reward light was attached to the ceiling of the enclosure. A mirror placed at an angle of 45 below the keys deflected the beam of an automatic projector onto the back of the keys. Sttmulus projection was controlled by two electromagnetic shutters. The stimuli (Fig. 2) were black and white photographic negatives which were affixed. two at a time, onto specially made slide frames. The patterns on the keys appeared as white shapes of about I cm dia. on a dark background. Sessions were controlled on-line with a microcomputer that also recorded and printed out the results on a trial by trial basis. Traininy The animals were trained to peck the keys with a standard autoshaping procedure. When shaping was completed. the discrimination procedure started. Two daily sessions of 40 trials each were conducted. Within a given trial the left-right position of posittve and negative stimuli was determined by a quasi-random sequence [9]. A total of 725 stimuli was used during this phase. 100 of them randomly allocated to the class “few” (positive). and the remaining 625 to the class “many” (negative). Stimulus pairs consisting of a positive and a negative stimulus were assembled randomly. When subjects responded to a positive shape, both stimuli disappeared and the pigeon was rewarded with a few grains of millet accompanied by illumination of the reward light. After an intertrial interval of 3 set the next trial began with the presentation of a new pair of shapes. Responses to the negative stimulus led to a 5 set period of total darkness (time out). After such an error the same patterns were presented again (correction trial). until the pigeon made a correct choice. As the subjects became experienced the response requirements within trials were gradually increased. Towards the end of training the animals had to issue within a particular trial I6 consecutive correct pecks to obtain a reward whereas I6 incorrect responses, not necessarily performed consecutively. led to time-out. Training began with a random selection of 30 positive and all 625 negative stimuli. Each time the animals reached
VISUAL
MEMORY
Fig. 1. Schema
r
LATERALIZATION
of the experimental
IN PIGEONS
chambers.
posrtlve llool 1
n 597
c 603
P 615 w 618
rlc 620 Fig. 2. Examples
of the rewarded
(positive)
and non-rewarded
(negative)
stimuli.
criterion (at least 80% correct chorces over 10 consecutive sessrons). the number of positive stimuli was increased by 10 until the pigeons dealt with all 100 positive stimuli. Additional sessions were conducted to prepare the animals lor special tests. These tests are reported m detail elsewhere [29] and will not be described here. Suffice it to mention that the animals were habituated to occasional unrewarded/unpunished trials and were then exposed to variously modified and also totally novel stimuli. After each test there were retraining sessions. An experimental pause followed where the animals were kept in an outside aviary. A retention test followed. Its most salient result (reported in detail in [29]) was that the animals still could discriminate 40 positive and 40 negative stimuli which they had not seen for 222 days. Further tests followed with modified/novel stimuli as sketched above [29]. When these were completed, the present tests for lateralization of visual memory functions were started. Lareralization tests The animals were lightly anaesthetized [17]. Rings of velcro were fixed around
with a mixture of choral hydrate, pentobarbital and magnesium sulphate each eye with histoacryl. Observations after recovery from anaesthesia
4
L~RENZCIVON FERSEN and ONUR G~JNTORRCN
revealed that the pigeons were not disturbed by the rings. For monocular tests opaque hemispherical goggles could be stuck onto these velcro rings. The animals were first run for ten sessions with 40 trials each under binocular conditions according to the procedure described above. Then monocular tests began. The pigeons discriminated on alternate sessions with sight restricted to the left or right eye by means of eyecaps. The sequence of right and left monocular conditions was balanced amongst animals. Each pigeon completed 10 sessions under each condition.
RESULTS Training performance of each subject was evaluated by calculating the percentage of correct responses per session (excluding correction trials). One of the four pigeons exhibited a very low response rate and was excluded from the experiment after 50 training sessions. The remaining three pigeons completed 224 sessions before all of them had learned to discriminate the 100 positive from the 625 negative stimuli. Additional 205 sessions were run during eight test phases before the animals were released into the aviary for an interval of 6 months. A further 84 sessions were taken up by the retenton tests. Thus the animals had completed 513 sessions before the lateralization tests began. Results of the lateralization tests are shown in Fig. 3. An overall ANOVA over the three conditions (binocular, monocular left, monocular right) with three animals demonstrated a significant effect of the eyecap conditions (F(2,58)= 128.049, P~0.01). There was no significant session effect (F (9, 58) = 0.8 15, P> 0.05) and no significant interaction between treatment conditions and sessions (F (18, 58) = 1.427, P>O.O5). Binocular performance was significantly higher than the performances in each monocular condition (both F (1, 38) > 175.698, P-c0.01). Additionally there were significant differences between the two monocular conditions: with the right eye seeing, the animals obtained significantly higher percent correct responses than with the left eye seeing (F(1,38)= 16.735, P~0.01). 100
right
left
9u
60
__ 50 I I
.
I .
.
I 1
Fig. 3. Percent correct results of the pigeons in the monocular left. monocular right. and binocular conditions. SEMs are depicted on the histograms. The horizontal line at 53% indicates the 5% confidence interval for chance performance. Significant P values (P ~0.01) are indicated at the bottom of the figure.
VlSiiAL
MEMORY
LATERALIZATION
Ih’ PIGEONS
5
As depicted in Fig. 3 the values for the monocular left condition were close to the 50% chance level. To test whether the performance of the animals was significantly above this level, the confidence interval at the 5% level was calculated using the SEMs of the responses in all three treatment conditions. As can be seen in Fig. 3, the values for all three conditions are well beyond this confidence interval. Thus the pigeons performed significantly above chance level in the binocular and in both monocular conditions. DISCUSSION The results of this experiment clearly demonstrate a lateralization of visual memory function in pigeons. Thus visual lateralization in birds seems not only to depend on asymmetries in immediate perceptive-cognitive processes [!, 10, 12, 221, but also seems to reflect differences in storage locations. According to the present study these visual memory processes are mainly located in the left hemisphere. This is reflected by the significantly higher performance of the animals in the right eye seeing condition over the left eye seeing one. On the other hand the binocular performance of the pigeons was superior to the results of either monocular condition. That indicates that both hemispheres contribute to the performance of the task when both eyes are unoccluded. This interpretation is additionally supported by the fact that with the left eye seeing the performance of the animals was weak, but still significantly above chance level. Thus we are inclined to believe that visual memory functions are mainly but not exclusively located in the left hemisphere in pigeons. To our knowledge this is the first demonstration of a lateralization of memory functions within the visual system of birds. Concerning other systems of the avian brain previous experiments were successful in revealing memory asymmetries. For example Horn et al., showed in a series of lesion experiments that two distinguishable and asymmetric memory systems control the process of imprinting in chicks. There is an initial store which involves the left medial hyperstriatum ventrale (MHV), a part of the forebrain roof, and a second store which appears later and is consolidated by the right MHV [15, 181. Additionally asymmetries in cholinergic receptor sites [2] and in the extention of the synaptic apposition zones could be demonstrated in the MHV of chicks as a consequence of imprinting [3]. Similar results were obtained by Rose et al., studying passive avoidance training in chicks. Chicks which pecked a bead coated with a bitter tasting substance, developed a long-lasting avoidance of pecking onto this bead. ROLEand CSILLAG[23] showed that this avoidance learning results in lasting changes in deoxyglucose metabolism in the left hemisphere MHV of the chick forebrain. Subsequently STEWARTet al. [27] could demonstrate asymmetric post-training effects in a stereological analysis of the MHV neuropil. According to these results passive avoidance training results in a significantly higher number of vesicles per synapse in the left MHV. Only a small number of studies examined possible asymmetries of memory functions in humans. This is possibly due to the implicit assumption that stimulus information selectively projected to one hemisphere is quickly disseminated throughout the brain. Therefore tests applied beyond a short time span should not yield lateral asymmetries. Contrary to this assumption several studies demonstrated even increasing asymmetries when subjects were asked to memorize stimuli which were presented to one of the hemispheres. MANNHAUPT [17] demonstrated that the right visual field (left hemisphere) advantage for concrete nouns increased with an increase in the number of items in a memory set which had to be matched with the projected stimuli. Similarly CONEYand MACDONALD[4] could only demonstrate a
6
L~RENZ~VON FERSEN and ONUR Gih~ii~~ii~
right visual field advantage for concrete nouns when long retention intervals were used. HISCOCKet al. [14] showed that memorizing numbers decreased right hand tapping more than left hand tapping and WICKENSand SANDRY[32] inferred from their results that memory of abstract words seems to be localized in the left hemisphere. Similar results were obtained with stimuli which generally yield a left visual field (right hemisphere) advantage. DEE and FONTENOT[S] demonstrated that longer delays in a delayed matching-to-sample task with complex visual patterns increased the advantage of the right hemisphere. MILNER and TAYLOR[ 193 showed that split-brain patients are able to match complex tactile stimuli after several minutes using their left hand (right hemisphere). Using their right hand (left hemisphere) they generally failed to achieve correct responses even after short delays of a few seconds. These results with human subjects suppose that laterality effects depend at least partly on asymmetries in memory mechanisms. HARDYKet al. [ 133 and MOSCOVITCH [20] postulate that memory is a lateralized function which is conceptually distinguishable from other asymmetric cognitive processes. This would imply for the present study that visual memory is one of the lateralized components determining visually guided behaviour. It would be distinct from other lateralized functions such as visual discrimination [ 121 or visual control of pecking [lo]. An alternative point of view is that all these different visual subfunctions are controlled by a single component which is functionally asymmetric. For example previous studies demonstrated that the lateralized performance of pigeons in a visual discrimination task is mainly determined by an asymmetric interaction of the left and the right optic tecta [l 11. The optic tecta in the midbrain of birds are the main projection areas of the retinal axons. Further studies are needed to reveal whether visual memory is an independent lateralized component of its own, or whether it is under control of other asymmetric structures. Acknowledgements-Lorenzo von Fersen was supported by a research grant and Onur Giintiirkiin by a habilitation grant (Gu 227/1-l ) from the German Research Council. We are grateful to Daniela Musumeci for providing us with the velcro eyecaps. Additionally we thank Ralf Jlger, Clive Wynne and Juan D. Delius for critically reading the manuscript.
REFERENCES 1. ANDREW, R. J. The development of visual lateralization in the domestic chick. Behac. Brain Res. 29.201-209, 1988. 2. BRADLEY, P. M. and HORS. G. Imprinting: a study ofcholinergic receptor sites in parts of the chick brain. Exp. Brain Res. 41, 121-123. 1981. 3. BRADLEY. P. M., HORN.G. and BATESON,P. Imprinting: an electron microscopic study ofchick hyperstriatum ventrale. Exp. Brain Res. 41, 115-120, 1981. 4. CONEY. J. and MACDONALD. S. The effect of rentention interval upon hemispheric processes in recognition memory. Neuropq~choloqia 26. 287-295. 1988. 5. DEE, H. L. and FONTENOT, D. J. Cerebral dominance and lateral diflerences in perception and memory. Neurops!choloqia 11, 167-173. 1973. 6. DELIUS. J. D. In Cognifion. Information Processing and Motivation, G. D’YDEVALLE (Editor), pp. 3--18. Elsevier. Amsterdam, 1985. 7. GALABURL)A. A. M. In Cerebral Dominnnce, N. GESCHWIND and A. M. GALABURDA (Editors). pp. 11-25. Harvard University Press. Cambridge, London, 1984. training: left hemisphere 8. GASTON.K. E. and GASTON.M. G. Unilateral memorv after binocular discrimination dominance in the chick? Brain Res. 303, 19c-193, 1984. stimuli in visual discrimination experiments. J. gmer. 9. GELLERMANS. L. W. Chance orders of alternating Ps$~ol. 42, 206208. 1933. of visually controlled behaviour in pigeons. Physiol. Behao. 34,575-577. 1985. 10. GCNWRKCN, 0. Lateralization
VISL’AL
MEMORYLATERALIZATIOXIN
11. GtX%~ti~~tih-. 0. and B~HRINGER. P. G. L.ateralization
PIGEONS
7
reversal after intertectal commissurotomy in the pigeon. Brain Res. 408, 1-5. 1987. 12. GONT~~RK~~N.0. and KESCH, S. Visual lateralization during feeding in pigeons. Be/tar. Neurosci. 101.433435. 1987. 13. HARDYCK. C.. TZENG. 0. J. L. and WANG. W. S. Y. Cerebral lateralization of function and bilingual decision processes: is thinking lateralized? Brain Larry. 5, 5671. 1978. 14. HISCOCK. M., KINSBOURNE,M., SAMUELS.M. and KRAUSE, A. E. Dual task performance in children: generalized and lateralized effects of memory encoding upon the rate and variability of concurrent finger tapping. Brarn Cognir. 6, 2440, 1987. 15. HORN.G. Neural mechanisms of learning: an analysis of imprinting in the domestic chick. Proc. R. Sot. Land. B. 213, 101-137, 1981. 16. MALLIN. H. D. and DELIUS. J. D. Inter- and intraocular transfer ofcolour disciminations with mandibulation as an operant in the head-fixed pigeon. Behor. Artal. Letr. 3. 297-309, 1983. 17. MANNHAUPT, H. R. Processing of abstract and concrete nouns in a lateralized memory search-task. P.srcltol. Res. 45, 91-105, 1983. 18. MCCABE, B. J., CIPOLLA-NETO. J., HORN, G. and BATESON. P. Amnesic effects of bilateral lesions placed in the hyperstriatum ventrale of the chick after imprinting. Erp. Brain Res. 48. 13-21, 1982. 19. MILNER, B. and TAYLOR, L. Right hemisphere superiority in tactile pattern-recognition after cerebral commissurotomy: evidence for nonverbal memory. Neuropspchologia 10, l-15, 1972. 20. MOSCOVITCH. M. Afferent and efferent models of visual perceptual asymmetries: theoretical and empirical implications. Neuropsychologia 24, 91-l 14, 1986. 21. NOTTEBOHM,F. In Cerebral Dominance, N. GESCHWIND and A. M. GALABLJRDA(Editors), pp. 93-l 13. Harvard University Press, Cambridge, London, 1984. 22. ROGERS, L. J. and ANSON, J. M. Laterahzation of function in the chicken fore-brain. Pharmocol. Biochem. Behar. 19,679486, 1979. 23. ROSE, S. P. R. and CSILLAG. A. Passive avoidance training results in lasting changes in deoxyglucose metabolism in left hemisphere regions of chick brain. Behar. Neur. Viol. 44, 315-324, 1985. 24. ROTHENBERGER,A., HEESCBEN, C. and KEMMERLING,S. Language development: lateralization ofsemantics and the short-term-verbal-memory (STVM) of the right hemisphere. Acm paedopsychiat. 50, 119-130. 1984. 25. SHEPARD, D. F. Recognition memory for words, sentences, and pictures. J. Verb. Learn. Verb. Behur. 6. 156-163, 1967. 26. SHETTLEWORTH, S. J. Memory in food-hoarding birds, Scient. Am. 3, 102-l 10, 1983. 27. STEWART, M. G., ROSE, S. P. R.. KING. T. S.. GABBOTT, P. L. A. and BOURNE, R. Hemispheric asymmetry of synapses in chick medial hyperstriatum ventrale following passive avoidance training: a stereological investigation. Der. Brain Res. 12, 261-269, 1984. 28. VAUGHAN, W. JR and GREENE. S. L. Pigeon visual memory capacity. J. exp. Psvchol.: Anim. Eehac. Proc. IO, 256-271, 1984. 29. VON FERSEN, L. and DELIUS. J. D. Long-term retention of many visual patterns by pigeons. Erholoq~ 82, 141-155. 1989. 30. WALKER, S. F. Lateralization of functions in the vertebrate brain: a review. Br. J. Psycho/. 71, 329-367, 1980. 31. WEIDNER, C.. REPERAST. J.. MICELI. D.. HABY. M. and RIO. J. P. An anatomical study of ipsilateral retinal projections in the quail using radioautographic. horseradish peroxidase, fluorescence and degeneration techniques. Brain Res. 340, 99-108. 1985. 32. WICKENS. C. D. and SANDRY. D. Task-hemispheric integrity in dual task performance. Acta P.~yh/. 52, 227-247, 1982. 33. WORKMAN, L. and ANDREW. R. J. Asymmetrres of eye use in birds. Anim. Behac. 34, 1582-1584. 1986.
VISUAL
MEMORY
LATERALIZATION
LORENZO VON FERSEN*
0028-3932 90 s3.00+0.00 1994 Pergamon Presr plc
IN PJGEONS
and ONUR GtiNTi_kKtiNt
Allgemeine Psychologie. uhiversity of Konstanz. D-7750 Konstanz. F.R.G. (Received 4 April 1989; accepted 13 June 1989) Abstract-Previous experiments employing simple visual discrimination tasks have revealed a cerebral lateralization in the visual system of pigeons with a dominance of the left hemisphere. Until now, visual memory lateralization in birds has not been investigated. To study possible asymmetries of visual memory functions, a simultaneous instrumental discrimination procedure was used. The animals were trained to discriminate 100 different visual patterns from a further 625 similar stimuli. Retention tests were conducted under binocular and monocular conditions. When the subjects looked monocularly, retention performance was significantly higher with the right eye (left hemisphere) than with the left eye (right hemisphere). The results suggest that the lateralization of the pigeon’s visual system depends at least partly on an asymmetry in visual memory capacity.
INTRODUCTION ALTHOUGH cerebral dominance was discovered in the 186Os, the neuronal foundations of most lateralized functions in humans remains largely obscure. Probably the only exception is the neuronal system controlling language for which detailed anatomical and physiological data of its neuronal organization has accumulated in recent years [7]. Until recently one of the reasons for the limited success in elucidating the biological foundations of hemispheric asymmetries was the lack of animal models. Language and handedness were the first systems which were recognized to be asymmetric and both were assumed to be unique to humans. As a consequence, cerebral asymmetries were treated as a peculiarity of the human brain. The last two decades have seen a dramatic change in this long-held opinion. Several functional systems were discovered to be lateralized in a number of species besides humans [30]. Experiments on these species now provide partial insights into the neuronal organization of cerebral asymmetries up to the synaptic level. One of the most influential series of studies are those by NOTTEBOHM et al., e.g. [21], in canaries. These experiments rely on a left hemisphere dominance for song production as a model for brain lateralization. Since these pioneering studies several workers have demonstrated that the left brain hemisphere in birds is not only dominant for vocalization but also for a variety of visual tasks [ 1, 8, 11,22, 331. Due to the virtually complete decussation of the optic nerves of birds, the avian visual system is particularly suited for research on lateralization [31]. Temporary occlusion of one eye ensures that information entering through the unobstructed eye is processed mainly by the contralateral hemisphere. Pigeons performing a pattern discrimination task under these conditions shdw better performance when seeing with the right eye, which projects to the dominant left hemisphere [lo]. *Present address: Department of Psychology. University tTo whom correspondence should be addressed.
of Exeter, Exeter EX4 4~33, U.K.
2
LOKENZO VOXFERSEN and ONUH GCNTORKCN
So far none of the studies on visual lateralization in birds examined possible asymmetries of visual memory functions. Several species of birds are known to be able to memorize large amounts of visual information over long time periods. Nutcrackers remember the location of thousands of food-hoards over several months [26]. VAUGHAN and GREENE [28] showed that even after 2 years pigeons exhibited an above chance discrimination of 160 pairs of complex stimuli. In another experiment [6], pigeons were trained to distinguish up to 100 repeatedly presented patterns from another 640 patterns that were shown at most twice. After a pause of 3 months the birds recognized the familiar stimuli almost as well as after intervals of a few days. These latter studies suggest that the visual memory capacity of pigeons almost matches that of humans [25]. Cerebral dominance in humans is not only evident in perceptual recognition tasks, but also in experiments which require memorization of stimulus information over some periods of time [4, 5, 241. It has been suggested on the basis ofsuch findings that lateralization effects are not only a function of immediate perceptual processes, but may actually also reflect differential storage capacities. We were therefore interested in whether it is possible to demonstrate a hemispheric asymmetry of visual memory in pigeons. The experiments were performed in the course of a more extensive study on the visual memory capacity of pigeons [29]. Only the results concerning lateralization of memory functions will be reported here. METHOD Subjects
Subjects were four adult homing pigeons (Columba Ma) oflocal stock. They were kept in individual indoor cages. The pigeons were normally maintained at 80% of their free feeding weight. At a certain time point of the experiment (detailed in the next section) they spent 6 months in an outside aviary and were allowed to regain their full weight. Afterwards the animals were deprived again. Apparatus
A modified two-key Skinner box was used (Fig. I ). The chamber measured 33 x 34 x 33 cm. The front wall had an opening of I I x 12.5 cm above the floor. A houselight was located 5 cm above this opening. A horizontal working surface of I2 x 9 cm was attached outside of the opening and level with its lower edge. It was surrounded by an enclosure measuring I6 x I4 x 10 cm. Two translucent perspex keys of 2.2 cm dia. were positioned side by side on the working surface. with their centres separated by 9 cm. Immediately in front of the keys were two reward receptacles. Two automatic feeders on top of the enclosure could be activated to deliver a few grains of millet to the receptacles through lengths of tubing. A reward light was attached to the ceiling of the enclosure. A mirror placed at an angle of 45 below the keys deflected the beam of an automatic projector onto the back of the keys. Sttmulus projection was controlled by two electromagnetic shutters. The stimuli (Fig. 2) were black and white photographic negatives which were affixed. two at a time, onto specially made slide frames. The patterns on the keys appeared as white shapes of about I cm dia. on a dark background. Sessions were controlled on-line with a microcomputer that also recorded and printed out the results on a trial by trial basis. Traininy The animals were trained to peck the keys with a standard autoshaping procedure. When shaping was completed. the discrimination procedure started. Two daily sessions of 40 trials each were conducted. Within a given trial the left-right position of posittve and negative stimuli was determined by a quasi-random sequence [9]. A total of 725 stimuli was used during this phase. 100 of them randomly allocated to the class “few” (positive). and the remaining 625 to the class “many” (negative). Stimulus pairs consisting of a positive and a negative stimulus were assembled randomly. When subjects responded to a positive shape, both stimuli disappeared and the pigeon was rewarded with a few grains of millet accompanied by illumination of the reward light. After an intertrial interval of 3 set the next trial began with the presentation of a new pair of shapes. Responses to the negative stimulus led to a 5 set period of total darkness (time out). After such an error the same patterns were presented again (correction trial). until the pigeon made a correct choice. As the subjects became experienced the response requirements within trials were gradually increased. Towards the end of training the animals had to issue within a particular trial I6 consecutive correct pecks to obtain a reward whereas I6 incorrect responses, not necessarily performed consecutively. led to time-out. Training began with a random selection of 30 positive and all 625 negative stimuli. Each time the animals reached
VISUAL
MEMORY
Fig. 1. Schema
r
LATERALIZATION
of the experimental
IN PIGEONS
chambers.
posrtlve llool 1
n 597
c 603
P 615 w 618
rlc 620 Fig. 2. Examples
of the rewarded
(positive)
and non-rewarded
(negative)
stimuli.
criterion (at least 80% correct chorces over 10 consecutive sessrons). the number of positive stimuli was increased by 10 until the pigeons dealt with all 100 positive stimuli. Additional sessions were conducted to prepare the animals lor special tests. These tests are reported m detail elsewhere [29] and will not be described here. Suffice it to mention that the animals were habituated to occasional unrewarded/unpunished trials and were then exposed to variously modified and also totally novel stimuli. After each test there were retraining sessions. An experimental pause followed where the animals were kept in an outside aviary. A retention test followed. Its most salient result (reported in detail in [29]) was that the animals still could discriminate 40 positive and 40 negative stimuli which they had not seen for 222 days. Further tests followed with modified/novel stimuli as sketched above [29]. When these were completed, the present tests for lateralization of visual memory functions were started. Lareralization tests The animals were lightly anaesthetized [17]. Rings of velcro were fixed around
with a mixture of choral hydrate, pentobarbital and magnesium sulphate each eye with histoacryl. Observations after recovery from anaesthesia
4
L~RENZCIVON FERSEN and ONUR G~JNTORRCN
revealed that the pigeons were not disturbed by the rings. For monocular tests opaque hemispherical goggles could be stuck onto these velcro rings. The animals were first run for ten sessions with 40 trials each under binocular conditions according to the procedure described above. Then monocular tests began. The pigeons discriminated on alternate sessions with sight restricted to the left or right eye by means of eyecaps. The sequence of right and left monocular conditions was balanced amongst animals. Each pigeon completed 10 sessions under each condition.
RESULTS Training performance of each subject was evaluated by calculating the percentage of correct responses per session (excluding correction trials). One of the four pigeons exhibited a very low response rate and was excluded from the experiment after 50 training sessions. The remaining three pigeons completed 224 sessions before all of them had learned to discriminate the 100 positive from the 625 negative stimuli. Additional 205 sessions were run during eight test phases before the animals were released into the aviary for an interval of 6 months. A further 84 sessions were taken up by the retenton tests. Thus the animals had completed 513 sessions before the lateralization tests began. Results of the lateralization tests are shown in Fig. 3. An overall ANOVA over the three conditions (binocular, monocular left, monocular right) with three animals demonstrated a significant effect of the eyecap conditions (F(2,58)= 128.049, P~0.01). There was no significant session effect (F (9, 58) = 0.8 15, P> 0.05) and no significant interaction between treatment conditions and sessions (F (18, 58) = 1.427, P>O.O5). Binocular performance was significantly higher than the performances in each monocular condition (both F (1, 38) > 175.698, P-c0.01). Additionally there were significant differences between the two monocular conditions: with the right eye seeing, the animals obtained significantly higher percent correct responses than with the left eye seeing (F(1,38)= 16.735, P~0.01). 100
right
left
9u
60
__ 50 I I
.
I .
.
I 1
Fig. 3. Percent correct results of the pigeons in the monocular left. monocular right. and binocular conditions. SEMs are depicted on the histograms. The horizontal line at 53% indicates the 5% confidence interval for chance performance. Significant P values (P ~0.01) are indicated at the bottom of the figure.
VlSiiAL
MEMORY
LATERALIZATION
Ih’ PIGEONS
5
As depicted in Fig. 3 the values for the monocular left condition were close to the 50% chance level. To test whether the performance of the animals was significantly above this level, the confidence interval at the 5% level was calculated using the SEMs of the responses in all three treatment conditions. As can be seen in Fig. 3, the values for all three conditions are well beyond this confidence interval. Thus the pigeons performed significantly above chance level in the binocular and in both monocular conditions. DISCUSSION The results of this experiment clearly demonstrate a lateralization of visual memory function in pigeons. Thus visual lateralization in birds seems not only to depend on asymmetries in immediate perceptive-cognitive processes [!, 10, 12, 221, but also seems to reflect differences in storage locations. According to the present study these visual memory processes are mainly located in the left hemisphere. This is reflected by the significantly higher performance of the animals in the right eye seeing condition over the left eye seeing one. On the other hand the binocular performance of the pigeons was superior to the results of either monocular condition. That indicates that both hemispheres contribute to the performance of the task when both eyes are unoccluded. This interpretation is additionally supported by the fact that with the left eye seeing the performance of the animals was weak, but still significantly above chance level. Thus we are inclined to believe that visual memory functions are mainly but not exclusively located in the left hemisphere in pigeons. To our knowledge this is the first demonstration of a lateralization of memory functions within the visual system of birds. Concerning other systems of the avian brain previous experiments were successful in revealing memory asymmetries. For example Horn et al., showed in a series of lesion experiments that two distinguishable and asymmetric memory systems control the process of imprinting in chicks. There is an initial store which involves the left medial hyperstriatum ventrale (MHV), a part of the forebrain roof, and a second store which appears later and is consolidated by the right MHV [15, 181. Additionally asymmetries in cholinergic receptor sites [2] and in the extention of the synaptic apposition zones could be demonstrated in the MHV of chicks as a consequence of imprinting [3]. Similar results were obtained by Rose et al., studying passive avoidance training in chicks. Chicks which pecked a bead coated with a bitter tasting substance, developed a long-lasting avoidance of pecking onto this bead. ROLEand CSILLAG[23] showed that this avoidance learning results in lasting changes in deoxyglucose metabolism in the left hemisphere MHV of the chick forebrain. Subsequently STEWARTet al. [27] could demonstrate asymmetric post-training effects in a stereological analysis of the MHV neuropil. According to these results passive avoidance training results in a significantly higher number of vesicles per synapse in the left MHV. Only a small number of studies examined possible asymmetries of memory functions in humans. This is possibly due to the implicit assumption that stimulus information selectively projected to one hemisphere is quickly disseminated throughout the brain. Therefore tests applied beyond a short time span should not yield lateral asymmetries. Contrary to this assumption several studies demonstrated even increasing asymmetries when subjects were asked to memorize stimuli which were presented to one of the hemispheres. MANNHAUPT [17] demonstrated that the right visual field (left hemisphere) advantage for concrete nouns increased with an increase in the number of items in a memory set which had to be matched with the projected stimuli. Similarly CONEYand MACDONALD[4] could only demonstrate a
6
L~RENZ~VON FERSEN and ONUR Gih~ii~~ii~
right visual field advantage for concrete nouns when long retention intervals were used. HISCOCKet al. [14] showed that memorizing numbers decreased right hand tapping more than left hand tapping and WICKENSand SANDRY[32] inferred from their results that memory of abstract words seems to be localized in the left hemisphere. Similar results were obtained with stimuli which generally yield a left visual field (right hemisphere) advantage. DEE and FONTENOT[S] demonstrated that longer delays in a delayed matching-to-sample task with complex visual patterns increased the advantage of the right hemisphere. MILNER and TAYLOR[ 193 showed that split-brain patients are able to match complex tactile stimuli after several minutes using their left hand (right hemisphere). Using their right hand (left hemisphere) they generally failed to achieve correct responses even after short delays of a few seconds. These results with human subjects suppose that laterality effects depend at least partly on asymmetries in memory mechanisms. HARDYKet al. [ 133 and MOSCOVITCH [20] postulate that memory is a lateralized function which is conceptually distinguishable from other asymmetric cognitive processes. This would imply for the present study that visual memory is one of the lateralized components determining visually guided behaviour. It would be distinct from other lateralized functions such as visual discrimination [ 121 or visual control of pecking [lo]. An alternative point of view is that all these different visual subfunctions are controlled by a single component which is functionally asymmetric. For example previous studies demonstrated that the lateralized performance of pigeons in a visual discrimination task is mainly determined by an asymmetric interaction of the left and the right optic tecta [l 11. The optic tecta in the midbrain of birds are the main projection areas of the retinal axons. Further studies are needed to reveal whether visual memory is an independent lateralized component of its own, or whether it is under control of other asymmetric structures. Acknowledgements-Lorenzo von Fersen was supported by a research grant and Onur Giintiirkiin by a habilitation grant (Gu 227/1-l ) from the German Research Council. We are grateful to Daniela Musumeci for providing us with the velcro eyecaps. Additionally we thank Ralf Jlger, Clive Wynne and Juan D. Delius for critically reading the manuscript.
REFERENCES 1. ANDREW, R. J. The development of visual lateralization in the domestic chick. Behac. Brain Res. 29.201-209, 1988. 2. BRADLEY, P. M. and HORS. G. Imprinting: a study ofcholinergic receptor sites in parts of the chick brain. Exp. Brain Res. 41, 121-123. 1981. 3. BRADLEY. P. M., HORN.G. and BATESON,P. Imprinting: an electron microscopic study ofchick hyperstriatum ventrale. Exp. Brain Res. 41, 115-120, 1981. 4. CONEY. J. and MACDONALD. S. The effect of rentention interval upon hemispheric processes in recognition memory. Neuropq~choloqia 26. 287-295. 1988. 5. DEE, H. L. and FONTENOT, D. J. Cerebral dominance and lateral diflerences in perception and memory. Neurops!choloqia 11, 167-173. 1973. 6. DELIUS. J. D. In Cognifion. Information Processing and Motivation, G. D’YDEVALLE (Editor), pp. 3--18. Elsevier. Amsterdam, 1985. 7. GALABURL)A. A. M. In Cerebral Dominnnce, N. GESCHWIND and A. M. GALABURDA (Editors). pp. 11-25. Harvard University Press. Cambridge, London, 1984. training: left hemisphere 8. GASTON.K. E. and GASTON.M. G. Unilateral memorv after binocular discrimination dominance in the chick? Brain Res. 303, 19c-193, 1984. stimuli in visual discrimination experiments. J. gmer. 9. GELLERMANS. L. W. Chance orders of alternating Ps$~ol. 42, 206208. 1933. of visually controlled behaviour in pigeons. Physiol. Behao. 34,575-577. 1985. 10. GCNWRKCN, 0. Lateralization
VISL’AL
MEMORYLATERALIZATIOXIN
11. GtX%~ti~~tih-. 0. and B~HRINGER. P. G. L.ateralization
PIGEONS
7
reversal after intertectal commissurotomy in the pigeon. Brain Res. 408, 1-5. 1987. 12. GONT~~RK~~N.0. and KESCH, S. Visual lateralization during feeding in pigeons. Be/tar. Neurosci. 101.433435. 1987. 13. HARDYCK. C.. TZENG. 0. J. L. and WANG. W. S. Y. Cerebral lateralization of function and bilingual decision processes: is thinking lateralized? Brain Larry. 5, 5671. 1978. 14. HISCOCK. M., KINSBOURNE,M., SAMUELS.M. and KRAUSE, A. E. Dual task performance in children: generalized and lateralized effects of memory encoding upon the rate and variability of concurrent finger tapping. Brarn Cognir. 6, 2440, 1987. 15. HORN.G. Neural mechanisms of learning: an analysis of imprinting in the domestic chick. Proc. R. Sot. Land. B. 213, 101-137, 1981. 16. MALLIN. H. D. and DELIUS. J. D. Inter- and intraocular transfer ofcolour disciminations with mandibulation as an operant in the head-fixed pigeon. Behor. Artal. Letr. 3. 297-309, 1983. 17. MANNHAUPT, H. R. Processing of abstract and concrete nouns in a lateralized memory search-task. P.srcltol. Res. 45, 91-105, 1983. 18. MCCABE, B. J., CIPOLLA-NETO. J., HORN, G. and BATESON. P. Amnesic effects of bilateral lesions placed in the hyperstriatum ventrale of the chick after imprinting. Erp. Brain Res. 48. 13-21, 1982. 19. MILNER, B. and TAYLOR, L. Right hemisphere superiority in tactile pattern-recognition after cerebral commissurotomy: evidence for nonverbal memory. Neuropspchologia 10, l-15, 1972. 20. MOSCOVITCH. M. Afferent and efferent models of visual perceptual asymmetries: theoretical and empirical implications. Neuropsychologia 24, 91-l 14, 1986. 21. NOTTEBOHM,F. In Cerebral Dominance, N. GESCHWIND and A. M. GALABLJRDA(Editors), pp. 93-l 13. Harvard University Press, Cambridge, London, 1984. 22. ROGERS, L. J. and ANSON, J. M. Laterahzation of function in the chicken fore-brain. Pharmocol. Biochem. Behar. 19,679486, 1979. 23. ROSE, S. P. R. and CSILLAG. A. Passive avoidance training results in lasting changes in deoxyglucose metabolism in left hemisphere regions of chick brain. Behar. Neur. Viol. 44, 315-324, 1985. 24. ROTHENBERGER,A., HEESCBEN, C. and KEMMERLING,S. Language development: lateralization ofsemantics and the short-term-verbal-memory (STVM) of the right hemisphere. Acm paedopsychiat. 50, 119-130. 1984. 25. SHEPARD, D. F. Recognition memory for words, sentences, and pictures. J. Verb. Learn. Verb. Behur. 6. 156-163, 1967. 26. SHETTLEWORTH, S. J. Memory in food-hoarding birds, Scient. Am. 3, 102-l 10, 1983. 27. STEWART, M. G., ROSE, S. P. R.. KING. T. S.. GABBOTT, P. L. A. and BOURNE, R. Hemispheric asymmetry of synapses in chick medial hyperstriatum ventrale following passive avoidance training: a stereological investigation. Der. Brain Res. 12, 261-269, 1984. 28. VAUGHAN, W. JR and GREENE. S. L. Pigeon visual memory capacity. J. exp. Psvchol.: Anim. Eehac. Proc. IO, 256-271, 1984. 29. VON FERSEN, L. and DELIUS. J. D. Long-term retention of many visual patterns by pigeons. Erholoq~ 82, 141-155. 1989. 30. WALKER, S. F. Lateralization of functions in the vertebrate brain: a review. Br. J. Psycho/. 71, 329-367, 1980. 31. WEIDNER, C.. REPERAST. J.. MICELI. D.. HABY. M. and RIO. J. P. An anatomical study of ipsilateral retinal projections in the quail using radioautographic. horseradish peroxidase, fluorescence and degeneration techniques. Brain Res. 340, 99-108. 1985. 32. WICKENS. C. D. and SANDRY. D. Task-hemispheric integrity in dual task performance. Acta P.~yh/. 52, 227-247, 1982. 33. WORKMAN, L. and ANDREW. R. J. Asymmetrres of eye use in birds. Anim. Behac. 34, 1582-1584. 1986.
