- Email: [email protected]
Deficits in luminous flux discrimination by cats with lesions of the superior colliculus-pretectum
Brain Research, 173 (1979) 217-224 © Elsevier/North-Holland Biomedical Press
217
D E F I C I T S I N L U M I N O U S F L U X D I S C R I M I N A T I O N BY CATS W I T H LESIONS OF T H E S U P E R I O R C O L L I C U L U S - P R E T E C T U M
BARRY S. P1NCHOFF and JACQUELINE M. S. WINTERKORN* Department of Anatomy, Cornell University Medical College, New York, N. Y. 10021 (U.S.A.)
(Accepted January 18th, 1979)
SUMMARY (1) Intact cats and cats relearning or initially learning after bilateral ablation of the superior colliculus-pretectum were trained on a simple flux discrimination in a straight maze where the stimuli were either immediately side by side at the end of the runway or were separated by an opaque barrier. (2) After ablation of the superior colliculus-pretectum, each preoperatively trained cat had a significant deficit in retention of the luminous flux discrimination regardless of whether or not the stimuli were separated by a barrier or whether the correct stimulus was darker or lighter than the incorrect stimulus. (3) Under each of the different training conditions, every cat learning the flux discrimination initially after ablation of the superior colliculus-pretectum was retarded in comparison with intact cats. (4) The severity of the postoperative learning deficit correlated highly with the total extent of the lesion in the superior colliculus and pretectum.
INTRODUCTION Considerable experimental effort during the past 10 years has been directed toward attempting to define the function of the superior colliculus, especially in visual learning. Studies of the abilities of cats with lesions of the superior colliculus to perform luminous flux discriminations generally have shown no deficit in retention of a preoperatively learned discrimination 1-4. However, in these studies occasional animals have had incomplete retention, and in other studiesS, n cats with lesions of the superior colliculus trained preoperatively on successive flux discriminations or on pattern as well as flux discriminations, have been found to require retraining in order to perform the luminous flux discrimination postoperatively. * To whom reprint requests should be sent.
218 The present study was designed to resolve the inconsistencies among prior studies and to determine whether simple flux discriminations learned preoperatively are retained by cats after bilateral ablation of the superior colliculus-pretectum, and whether initial learning of flux discrimination is retarded in cats after bilateral ablation of the superior colliculus-pretectum. METHODS
Subjects Twelve experimentally naive cats were used in this study. Eight cats were trained both preoperatively and following bilateral ablation of the superior colliculus-pretectum; 4 cats received only postoperative training.
Surgery Reflecting the view that the superior colliculus and the pretectum comprise a functional unit I, the lesions in this study were intended to remove variable amounts of the pretectum in addition to the superior colliculus. Surgery was performed on cats anesthetized with sodium pentobarbital. After posterior craniotomy and removal of the bony tentorium, lesions were made by subpial aspiration under direct visual control through a surgical microscope. All cats were allowed a recovery period of one month before postoperative testing or training.
Apparatus Training was carried out in a previously described maze 5, which consisted of an enclosed box at one end opening into a runway 91.4 cm in length. At the other end of the maze, were two plexiglas stimulus panel doors, upon which the visual stimuli were rear-projected by 35 mm slide projectors. The runway could be divided into two parallel alleys leading to the stimulus panel doors by inserting an opaque barrier 50.8 cm in length.
Stimuli The stimuli consisted of a transilluminated door 3000 cd/sq.m in luminance and an unilluminated door. At the choicepoint 50.8 cm from the stimuli, the luminances measured 1500 cd/sq.m for the illuminated door and 1.5 cd/sq.m for the unilluminated door.
Procedure Since preliminary work had suggested that initial learning and postoperative retention might differ depending on whether the correct stimulus was the lighter or the darker of the pair, the cats were studied in 2 groups: one group consisted of 7 cats for which the correct stimulus was the unilluminated door ( D + ) and the other group consisted of 5 cats for which the correct stimulus was the illuminated door ( L + ) . In addition, because previous work 5 has indicated subtle differences in the patterns of learning and relearning when an animal must discriminate from a choice-point at a
219 distance from the stimuli in a divided maze, each of these two groups of cats was further subdivided into one subgroup in which cats were trained with an interstimulus runway divider and another subgroup in which cats were trained in the maze without a divider. All cats were shaped to open doors for a food reward and then were assigned arbitrarily to one of the four training groups: (1) Dq- in the divided maze (D2, D4, D5, D51); (2) D ÷ in the undivided maze (D21, D52, DH6); (3) L-I- in the divided maze (D61, DH5, DH2); and (4) L + in the undivided maze (D62, DH3). Cats were trained 40 trials per day for food reward using an error correction procedure. In the undivided maze, an error was scored whenever the cat pushed open the incorrect door. In the divided maze, a door-push error was scored whenever the cat pushed open the incorrect door and an alley-entrance error was recorded whenever the cat entered the alley leading to the incorrect stimulus, but did not push open the incorrect door at the end of the alley. Discrimination training was continued until a performance criterion was achieved of 90 ~o or more correct responses for 3 successive days of training, by both scoring methods. After criterional performance was attained, preoperatively trained cats were subjected to bilateral ablation of the superior colliculus-pretectum. Prior to brain surgery, one cat, D2, was sham-operated and tested for retention after a month. For all cats, postoperative training or retraining commenced one month after surgery.
Histology At the end of all training, cats were perfused, the brains embedded in celloidin, serially sectioned at 40/am, and every tenth and eleventh section stained respectively with thionin and by the Weil technique. TABLE I
Extent o f lesions Cat
D2 D4 D5 D51 D21 DH6 D52 D61 DH5 DH2 D62 DH3
% Superior colliculus
Depth*
% Pretectum
Thalamic damage * *
L*** R
L
R
L
R
L
R
90 80 60 80 100 100 98 70 95 100 75 90
3 3 2 3 4 4 3 3 3 4 3 4
4 2 4 4 4 3 3 3 4 4 4 4
0 75 60 60 100 100 60 10 90 100 60 90
10 25 80 90 90 100 80 70 80 100 70 100
0 2 3 0 1 3 0 0 3 3 0 1
0 0 0 0 2 3 0 1 1 0 2 0
95 60 90 100 100 95 100 50 80 95 100 100
* D e p t h : 1 = superficial layers; 2 -- dorsal to sulcus limitans; 3 = to sulcus limitans; 4 = deeper than sulcus limitans. ** Thalamic damage: 0 = no thalamic damage; 1 = damage to lateralis posterior; 2 = damage to habenula and lateralis posterior; 3 = damage to medialis dorsalis, habenula and lateralis posterior. *** L = left side; R = right side.
220
II
31 I
lll
141
D 21
D62
Fig. 1. Reconstruction of lesions i n cat D21. Stippling indicates extent of complete neuronal degeneration. Fig. 2. Reconstruction of lesion in cat D62. Stippling indicates extent of complete neuronal degeneration.
RESULTS AND DISCUSSION
Table I summarizes histological confirmation of the extent of the lesion on each side of the brain for each cat. Figs. 1 and 2 illustrate the extent of representative lesions in two cats, D21 in group D + / n o divider and D62 in group L + / n o divider. All cats in this study achieved a high performance criterion on the L-- or the D-tdiscrimination in either the undivided or the divided maze preoperatively and/or after bilateral ablation of the superior colliculus-pretectum. As can be seen from Table II, unoperated cats trained on D + in the divided maze required an average of 480 trials to achieve criterion, compared with an average of only 280 trials under each of the other three training conditions. However, unoperated cats committed more errors learning the D + discrimination than the L ÷ discrimination in the undivided maze as well as the divided maze. In addition, as can be seen in examples of learning curves presented in Figs. 3, 4, 5 and 6, performance on the initial day of training on the D + discrimination was at or below 50 ~ correct responses for every cat whereas performance on the
221 TABLE I1 Trials and errors to criterion Group
Cat
Postoperative
Trials to criterion Errors
Trials to criterion Errors
D*
A**
D
A
D
D2 D4 D5 Mean D51
400 400 640 480 . .
400 440 640 493 .
145 131 289 188
23 32 34 30
240 320 520 360 640
D21 DH6 Mean D52
320 240 280
D61 DH5 Mean DH2
320 240 280 --
D62 DH3
280 --
D÷ Divider
D-F No Divider
L÷ Divider
L+ No Divider
Preoperative
-
. 96 39 63
-
-
320 240 280 --
280 360 1040 560 840
720 520 620 560
-
44 34 39 --
A
1 12 7 -
-
55 --
320 400 360 640
D
A
50 59 257 122 283
15 14 157 63 118
296 162 229 173 320 400 360 640
360 530
44 67 56 153
11 9 10 22
71 113
* D = to door-push criterion. ** A = to alley-entrance criterion. D 21 PRE OP ~c w ir
POST OP
1°°1 75-
75
D+
50-
N
25-
25 0
i
0
l 18
8 DAYS
OF
TRAINING
Fig. 3. Learning curves of cat D21 on D ÷ discrimination in the undivided maze preoperatively and after bilateral ablation of the superior colliculus-pretectum. Each circle represents percent correct responses during 40 trials on one day of training. first d a y o f t r a i n i n g o n t h e L + d i s c r i m i n a t i o n w a s a l w a y s a b o v e 50 % c o r r e c t r e s p o n s e s , reflecting the normal phototropism After ablation
of the subjects.
of the superior colliculus-pretectum,
every cat trained pre-
o p e r a t i v e l y i n e a c h o f t h e f o u r g r o u p s e x h i b i t e d a s i g n i f i c a n t deficit i n r e t e n t i o n o f t h e
222
D-4 PRE OP
POST OP
IOO-
"~-
IO0-
...~
75-
D-I50-
div o
50-
25-
25-
DAYS
w I0 OF
0
TRAINING
Fig. 4. Learning curves of cat D4 on D + oiscrimination in the divided maze preo!ceratively and after bilateral ablation of the superior colliculus-pretectum. Circles connected by lines in upl:er curves indicate percent correct responses by door-push scoring, considering only door-push errors; diamonds connected by dotted lines in lower curve indicate percent correct responses by alley-entrance scoring, considering both alley-entrance and door-push errors.
preoperatively learned discrimination, whereas the cat which was tested after a sham operation and a month's rest performed at criterional levels immediately. The cats relearning the D-k discrimination in the divided maze relearned with an average savings of 25 ~ to door-push criterion; however, each of the cats in the other three groups required more trials and committed more errors during postoperative relearning than during preoperative learning. Since the cats learning the D ÷ discrimination in the divided maze had required the greatest number of trials and committed the greatest number of errors of all the groups preoperatively, their more extensive preoperative training may have contributed to their better postoperative performance. In addition, this result confirms prior work ~ showing that cats with lesions of the superior colliculus-pretectum perform well on visual discriminations when the stimuli are se-
D 62 POST OP
PRE OP
L+
o
I00-
I00-
75-
75-
50-
o
.~
25-
z5I
O
7' DAYS
t
9 OF
TRAiNiNG
Fig. 5. Learning curves of cat D62 on L ÷ discrimination in the undivided maze preoperatively and after bilateral ablation of the superior colliculus-pretectum. Each circle represents per cent correct responses during 40 trials on one day of training.
223 DH-5 PRE OP
LT div
POST OP
IO0-
I00 -
ta
75-
75-
";
50-
50-
= o¢j 250
25i
0
6 DAYS
I0 OF
TRAINING
Fig. 6. Learning curves of cat DH5 on L + discrimination in the divided maze preoperatively and after bilateral ablation of the superior colliculus-pretectum. Circles connected by lines in upper curves indicate percent correct responses by door-push scoring, considering only door-push errors; diamonds connected by dotted lines in lower curves indicate I=ercent correct responses by alley-entrance scoring, considering both alley-entrance and door-push errors.
parated in space or by a divider. This improved performance by cats with bilateral lesions of the superior colliculus when stimuli are widely separated may explain why Urbaitis and Meikle 4 reported minimal or no deficits in cats with collicular lesions which were relearning simple flux discriminations in a Y-maze. The cats relearning the D + discrimination without the divider required more trials and committed more errors than cats relearning the L 4 discrimination either with or without the divider. The data indicate no difference between postoperative performance on the L + discrimination with or without the divider. In addition, all the cats were somewhat phototropic postoperatively regardless of preoperative training. Each of the 4 training groups included one cat which was trained initially after ablation of the superior colliculus-pretectum. All four of these cats required more trials and committed more errors to achieve criterional performance than unoperated cats. Similarly, compared with the preoperatively trained cats in its group which were relearning after ablation of the superior colliculus-pretectum, each of the 4 cats initially learning required more trials and committed more errors to reach criterional performance. The two cats trained in the divided maze required somewhat more trials and even committed more door-push errors than the two cats trained without the divider. However, with the~e few cats no significant difference could be shown between the postoperative learning of the L + discrimination and of the D + discrimination. All cats trained both preoperatively and postoperatively in the divided maze on the L ÷ or D + discrimination, committed some alley-entrance errors during learning. Nevertheless, all cats, including those with large bilateral lesions of the superior colliculus-pretectum achieved criterional performance, as measured by alley-entrance errors, confirming prior work from this laboratory 5. During postoperative relearning, alleyentrance errors represented a higher percentage of total errors than during preoperative learning, although alley-entrance errors were not always committed in large numbers. Cats initially learning in the divided maze after ablation of the superior colliculus-
224 pretectum committed both an increased number and percentage of alley-entrance errors before achieving criterion. Finally, confirming previous studies of alley-entrance errors 5, all cats committed more alley-entrance errors and higher percentages of alley-entrance errors on the more difficult discrimination, D ÷ , than on an easier discrimination, L + . The one cat in this study with a lesion confined almost exclusively to the superior colliculus, cat D2, had a deficit in retention of the preoperatively learned Dq- discrimination and required retraining to achieve criterion postoperatively. For other cats, in which the lesions extended beyond the superior colliculus into the pretectum, the postoperative deficits were even greater. Within each training group, the postoperative deficits in retention were not highly correlated with the percentages of superior colliculus alone ablated; however, the retention deficits were correlated perfectly with extension of the lesion beyond the superior colliculus into the pretectum. Furthermore, for all cats trained preoperatively and post-operatively the relative number of trials required to achieve criterion postoperatively compared with preoperatively was highly correlated (r ----0.93) with the total extent of the lesion in the superior colliculus and the pretectum. These data are supported by results of previous studies in which no retention deficits were reported in animals with lesions limited to the superior colliculus 1-4, whereas postoperative deficits in retention were reported for animals with collicular lesions which included at least part of the pretectum 1,4-6. In summary, although cats with lesions of the superior colliculus-pretectum do not retain a preoperatively learned luminous flux discrimination, they can be retrained to a high performance criterion postoperatively. The amount of retraining required is related to the total amount of damage in the superior colliculus-pretectum. Furthermore, cats with lesions of the superior colliculus-pretectum are retarded in initially learning a brightness discrimination post-operatively. Thus, even though the geniculostriate visual pathway is intact, ablation of the superior colliculus-pretectum produces definitive deficits in learning or relearning simple luminous flux discriminations. ACKNOWLEDGEMENTS We would like to thank Mr. William Giddings for his care of the experimental animals and Ms. Heidi Markowitz-Patzelt for assistance in training several cats.
REFERENCES 1 Berlucchi, G., Sprague, J. M., Levy, J. and DiBerardino, A. C., Pretectum and superior colliculus in visually guided behavior and in flux and form discrimination in the cat, J. comp. physiol. Psychol., 78 (1972) 123-172. 2 Blake, L., The effect of lesions of the superior colliculus on brightness and pattern discriminations in the cat, J. comp. physiol. Psychol., 52 (1959) 272-278. 3 Fischman, M. W. and Meikle, T. H., Jr., Visual intensity discrimination in cats after serial tectal and cortical lesions, J. comp. physiol. Psychol., 59 (1965) 193-201. 4 Urbaitis, J. C. and Meikle, T. H., Jr., Relearning a dark-light discrimination by cats after cortical and collicular lesions, Exp. Neurol., 20 (1968) 295-311. 5 Winterkorn, J. M. S., Visual discrimination between spatially separated stimuli by cats with lesions of the superior colliculus-pretectum, Brain Research, 100 (1975) 523-541. 6 Wood, B. S., Monocular relearning of a dark-light discrimination by cats after unilateral cortical and collicular lesions, Brain Research, 83 (1975) 156-162.
217
D E F I C I T S I N L U M I N O U S F L U X D I S C R I M I N A T I O N BY CATS W I T H LESIONS OF T H E S U P E R I O R C O L L I C U L U S - P R E T E C T U M
BARRY S. P1NCHOFF and JACQUELINE M. S. WINTERKORN* Department of Anatomy, Cornell University Medical College, New York, N. Y. 10021 (U.S.A.)
(Accepted January 18th, 1979)
SUMMARY (1) Intact cats and cats relearning or initially learning after bilateral ablation of the superior colliculus-pretectum were trained on a simple flux discrimination in a straight maze where the stimuli were either immediately side by side at the end of the runway or were separated by an opaque barrier. (2) After ablation of the superior colliculus-pretectum, each preoperatively trained cat had a significant deficit in retention of the luminous flux discrimination regardless of whether or not the stimuli were separated by a barrier or whether the correct stimulus was darker or lighter than the incorrect stimulus. (3) Under each of the different training conditions, every cat learning the flux discrimination initially after ablation of the superior colliculus-pretectum was retarded in comparison with intact cats. (4) The severity of the postoperative learning deficit correlated highly with the total extent of the lesion in the superior colliculus and pretectum.
INTRODUCTION Considerable experimental effort during the past 10 years has been directed toward attempting to define the function of the superior colliculus, especially in visual learning. Studies of the abilities of cats with lesions of the superior colliculus to perform luminous flux discriminations generally have shown no deficit in retention of a preoperatively learned discrimination 1-4. However, in these studies occasional animals have had incomplete retention, and in other studiesS, n cats with lesions of the superior colliculus trained preoperatively on successive flux discriminations or on pattern as well as flux discriminations, have been found to require retraining in order to perform the luminous flux discrimination postoperatively. * To whom reprint requests should be sent.
218 The present study was designed to resolve the inconsistencies among prior studies and to determine whether simple flux discriminations learned preoperatively are retained by cats after bilateral ablation of the superior colliculus-pretectum, and whether initial learning of flux discrimination is retarded in cats after bilateral ablation of the superior colliculus-pretectum. METHODS
Subjects Twelve experimentally naive cats were used in this study. Eight cats were trained both preoperatively and following bilateral ablation of the superior colliculus-pretectum; 4 cats received only postoperative training.
Surgery Reflecting the view that the superior colliculus and the pretectum comprise a functional unit I, the lesions in this study were intended to remove variable amounts of the pretectum in addition to the superior colliculus. Surgery was performed on cats anesthetized with sodium pentobarbital. After posterior craniotomy and removal of the bony tentorium, lesions were made by subpial aspiration under direct visual control through a surgical microscope. All cats were allowed a recovery period of one month before postoperative testing or training.
Apparatus Training was carried out in a previously described maze 5, which consisted of an enclosed box at one end opening into a runway 91.4 cm in length. At the other end of the maze, were two plexiglas stimulus panel doors, upon which the visual stimuli were rear-projected by 35 mm slide projectors. The runway could be divided into two parallel alleys leading to the stimulus panel doors by inserting an opaque barrier 50.8 cm in length.
Stimuli The stimuli consisted of a transilluminated door 3000 cd/sq.m in luminance and an unilluminated door. At the choicepoint 50.8 cm from the stimuli, the luminances measured 1500 cd/sq.m for the illuminated door and 1.5 cd/sq.m for the unilluminated door.
Procedure Since preliminary work had suggested that initial learning and postoperative retention might differ depending on whether the correct stimulus was the lighter or the darker of the pair, the cats were studied in 2 groups: one group consisted of 7 cats for which the correct stimulus was the unilluminated door ( D + ) and the other group consisted of 5 cats for which the correct stimulus was the illuminated door ( L + ) . In addition, because previous work 5 has indicated subtle differences in the patterns of learning and relearning when an animal must discriminate from a choice-point at a
219 distance from the stimuli in a divided maze, each of these two groups of cats was further subdivided into one subgroup in which cats were trained with an interstimulus runway divider and another subgroup in which cats were trained in the maze without a divider. All cats were shaped to open doors for a food reward and then were assigned arbitrarily to one of the four training groups: (1) Dq- in the divided maze (D2, D4, D5, D51); (2) D ÷ in the undivided maze (D21, D52, DH6); (3) L-I- in the divided maze (D61, DH5, DH2); and (4) L + in the undivided maze (D62, DH3). Cats were trained 40 trials per day for food reward using an error correction procedure. In the undivided maze, an error was scored whenever the cat pushed open the incorrect door. In the divided maze, a door-push error was scored whenever the cat pushed open the incorrect door and an alley-entrance error was recorded whenever the cat entered the alley leading to the incorrect stimulus, but did not push open the incorrect door at the end of the alley. Discrimination training was continued until a performance criterion was achieved of 90 ~o or more correct responses for 3 successive days of training, by both scoring methods. After criterional performance was attained, preoperatively trained cats were subjected to bilateral ablation of the superior colliculus-pretectum. Prior to brain surgery, one cat, D2, was sham-operated and tested for retention after a month. For all cats, postoperative training or retraining commenced one month after surgery.
Histology At the end of all training, cats were perfused, the brains embedded in celloidin, serially sectioned at 40/am, and every tenth and eleventh section stained respectively with thionin and by the Weil technique. TABLE I
Extent o f lesions Cat
D2 D4 D5 D51 D21 DH6 D52 D61 DH5 DH2 D62 DH3
% Superior colliculus
Depth*
% Pretectum
Thalamic damage * *
L*** R
L
R
L
R
L
R
90 80 60 80 100 100 98 70 95 100 75 90
3 3 2 3 4 4 3 3 3 4 3 4
4 2 4 4 4 3 3 3 4 4 4 4
0 75 60 60 100 100 60 10 90 100 60 90
10 25 80 90 90 100 80 70 80 100 70 100
0 2 3 0 1 3 0 0 3 3 0 1
0 0 0 0 2 3 0 1 1 0 2 0
95 60 90 100 100 95 100 50 80 95 100 100
* D e p t h : 1 = superficial layers; 2 -- dorsal to sulcus limitans; 3 = to sulcus limitans; 4 = deeper than sulcus limitans. ** Thalamic damage: 0 = no thalamic damage; 1 = damage to lateralis posterior; 2 = damage to habenula and lateralis posterior; 3 = damage to medialis dorsalis, habenula and lateralis posterior. *** L = left side; R = right side.
220
II
31 I
lll
141
D 21
D62
Fig. 1. Reconstruction of lesions i n cat D21. Stippling indicates extent of complete neuronal degeneration. Fig. 2. Reconstruction of lesion in cat D62. Stippling indicates extent of complete neuronal degeneration.
RESULTS AND DISCUSSION
Table I summarizes histological confirmation of the extent of the lesion on each side of the brain for each cat. Figs. 1 and 2 illustrate the extent of representative lesions in two cats, D21 in group D + / n o divider and D62 in group L + / n o divider. All cats in this study achieved a high performance criterion on the L-- or the D-tdiscrimination in either the undivided or the divided maze preoperatively and/or after bilateral ablation of the superior colliculus-pretectum. As can be seen from Table II, unoperated cats trained on D + in the divided maze required an average of 480 trials to achieve criterion, compared with an average of only 280 trials under each of the other three training conditions. However, unoperated cats committed more errors learning the D + discrimination than the L ÷ discrimination in the undivided maze as well as the divided maze. In addition, as can be seen in examples of learning curves presented in Figs. 3, 4, 5 and 6, performance on the initial day of training on the D + discrimination was at or below 50 ~ correct responses for every cat whereas performance on the
221 TABLE I1 Trials and errors to criterion Group
Cat
Postoperative
Trials to criterion Errors
Trials to criterion Errors
D*
A**
D
A
D
D2 D4 D5 Mean D51
400 400 640 480 . .
400 440 640 493 .
145 131 289 188
23 32 34 30
240 320 520 360 640
D21 DH6 Mean D52
320 240 280
D61 DH5 Mean DH2
320 240 280 --
D62 DH3
280 --
D÷ Divider
D-F No Divider
L÷ Divider
L+ No Divider
Preoperative
-
. 96 39 63
-
-
320 240 280 --
280 360 1040 560 840
720 520 620 560
-
44 34 39 --
A
1 12 7 -
-
55 --
320 400 360 640
D
A
50 59 257 122 283
15 14 157 63 118
296 162 229 173 320 400 360 640
360 530
44 67 56 153
11 9 10 22
71 113
* D = to door-push criterion. ** A = to alley-entrance criterion. D 21 PRE OP ~c w ir
POST OP
1°°1 75-
75
D+
50-
N
25-
25 0
i
0
l 18
8 DAYS
OF
TRAINING
Fig. 3. Learning curves of cat D21 on D ÷ discrimination in the undivided maze preoperatively and after bilateral ablation of the superior colliculus-pretectum. Each circle represents percent correct responses during 40 trials on one day of training. first d a y o f t r a i n i n g o n t h e L + d i s c r i m i n a t i o n w a s a l w a y s a b o v e 50 % c o r r e c t r e s p o n s e s , reflecting the normal phototropism After ablation
of the subjects.
of the superior colliculus-pretectum,
every cat trained pre-
o p e r a t i v e l y i n e a c h o f t h e f o u r g r o u p s e x h i b i t e d a s i g n i f i c a n t deficit i n r e t e n t i o n o f t h e
222
D-4 PRE OP
POST OP
IOO-
"~-
IO0-
...~
75-
D-I50-
div o
50-
25-
25-
DAYS
w I0 OF
0
TRAINING
Fig. 4. Learning curves of cat D4 on D + oiscrimination in the divided maze preo!ceratively and after bilateral ablation of the superior colliculus-pretectum. Circles connected by lines in upl:er curves indicate percent correct responses by door-push scoring, considering only door-push errors; diamonds connected by dotted lines in lower curve indicate percent correct responses by alley-entrance scoring, considering both alley-entrance and door-push errors.
preoperatively learned discrimination, whereas the cat which was tested after a sham operation and a month's rest performed at criterional levels immediately. The cats relearning the D-k discrimination in the divided maze relearned with an average savings of 25 ~ to door-push criterion; however, each of the cats in the other three groups required more trials and committed more errors during postoperative relearning than during preoperative learning. Since the cats learning the D ÷ discrimination in the divided maze had required the greatest number of trials and committed the greatest number of errors of all the groups preoperatively, their more extensive preoperative training may have contributed to their better postoperative performance. In addition, this result confirms prior work ~ showing that cats with lesions of the superior colliculus-pretectum perform well on visual discriminations when the stimuli are se-
D 62 POST OP
PRE OP
L+
o
I00-
I00-
75-
75-
50-
o
.~
25-
z5I
O
7' DAYS
t
9 OF
TRAiNiNG
Fig. 5. Learning curves of cat D62 on L ÷ discrimination in the undivided maze preoperatively and after bilateral ablation of the superior colliculus-pretectum. Each circle represents per cent correct responses during 40 trials on one day of training.
223 DH-5 PRE OP
LT div
POST OP
IO0-
I00 -
ta
75-
75-
";
50-
50-
= o¢j 250
25i
0
6 DAYS
I0 OF
TRAINING
Fig. 6. Learning curves of cat DH5 on L + discrimination in the divided maze preoperatively and after bilateral ablation of the superior colliculus-pretectum. Circles connected by lines in upper curves indicate percent correct responses by door-push scoring, considering only door-push errors; diamonds connected by dotted lines in lower curves indicate I=ercent correct responses by alley-entrance scoring, considering both alley-entrance and door-push errors.
parated in space or by a divider. This improved performance by cats with bilateral lesions of the superior colliculus when stimuli are widely separated may explain why Urbaitis and Meikle 4 reported minimal or no deficits in cats with collicular lesions which were relearning simple flux discriminations in a Y-maze. The cats relearning the D + discrimination without the divider required more trials and committed more errors than cats relearning the L 4 discrimination either with or without the divider. The data indicate no difference between postoperative performance on the L + discrimination with or without the divider. In addition, all the cats were somewhat phototropic postoperatively regardless of preoperative training. Each of the 4 training groups included one cat which was trained initially after ablation of the superior colliculus-pretectum. All four of these cats required more trials and committed more errors to achieve criterional performance than unoperated cats. Similarly, compared with the preoperatively trained cats in its group which were relearning after ablation of the superior colliculus-pretectum, each of the 4 cats initially learning required more trials and committed more errors to reach criterional performance. The two cats trained in the divided maze required somewhat more trials and even committed more door-push errors than the two cats trained without the divider. However, with the~e few cats no significant difference could be shown between the postoperative learning of the L + discrimination and of the D + discrimination. All cats trained both preoperatively and postoperatively in the divided maze on the L ÷ or D + discrimination, committed some alley-entrance errors during learning. Nevertheless, all cats, including those with large bilateral lesions of the superior colliculus-pretectum achieved criterional performance, as measured by alley-entrance errors, confirming prior work from this laboratory 5. During postoperative relearning, alleyentrance errors represented a higher percentage of total errors than during preoperative learning, although alley-entrance errors were not always committed in large numbers. Cats initially learning in the divided maze after ablation of the superior colliculus-
224 pretectum committed both an increased number and percentage of alley-entrance errors before achieving criterion. Finally, confirming previous studies of alley-entrance errors 5, all cats committed more alley-entrance errors and higher percentages of alley-entrance errors on the more difficult discrimination, D ÷ , than on an easier discrimination, L + . The one cat in this study with a lesion confined almost exclusively to the superior colliculus, cat D2, had a deficit in retention of the preoperatively learned Dq- discrimination and required retraining to achieve criterion postoperatively. For other cats, in which the lesions extended beyond the superior colliculus into the pretectum, the postoperative deficits were even greater. Within each training group, the postoperative deficits in retention were not highly correlated with the percentages of superior colliculus alone ablated; however, the retention deficits were correlated perfectly with extension of the lesion beyond the superior colliculus into the pretectum. Furthermore, for all cats trained preoperatively and post-operatively the relative number of trials required to achieve criterion postoperatively compared with preoperatively was highly correlated (r ----0.93) with the total extent of the lesion in the superior colliculus and the pretectum. These data are supported by results of previous studies in which no retention deficits were reported in animals with lesions limited to the superior colliculus 1-4, whereas postoperative deficits in retention were reported for animals with collicular lesions which included at least part of the pretectum 1,4-6. In summary, although cats with lesions of the superior colliculus-pretectum do not retain a preoperatively learned luminous flux discrimination, they can be retrained to a high performance criterion postoperatively. The amount of retraining required is related to the total amount of damage in the superior colliculus-pretectum. Furthermore, cats with lesions of the superior colliculus-pretectum are retarded in initially learning a brightness discrimination post-operatively. Thus, even though the geniculostriate visual pathway is intact, ablation of the superior colliculus-pretectum produces definitive deficits in learning or relearning simple luminous flux discriminations. ACKNOWLEDGEMENTS We would like to thank Mr. William Giddings for his care of the experimental animals and Ms. Heidi Markowitz-Patzelt for assistance in training several cats.
REFERENCES 1 Berlucchi, G., Sprague, J. M., Levy, J. and DiBerardino, A. C., Pretectum and superior colliculus in visually guided behavior and in flux and form discrimination in the cat, J. comp. physiol. Psychol., 78 (1972) 123-172. 2 Blake, L., The effect of lesions of the superior colliculus on brightness and pattern discriminations in the cat, J. comp. physiol. Psychol., 52 (1959) 272-278. 3 Fischman, M. W. and Meikle, T. H., Jr., Visual intensity discrimination in cats after serial tectal and cortical lesions, J. comp. physiol. Psychol., 59 (1965) 193-201. 4 Urbaitis, J. C. and Meikle, T. H., Jr., Relearning a dark-light discrimination by cats after cortical and collicular lesions, Exp. Neurol., 20 (1968) 295-311. 5 Winterkorn, J. M. S., Visual discrimination between spatially separated stimuli by cats with lesions of the superior colliculus-pretectum, Brain Research, 100 (1975) 523-541. 6 Wood, B. S., Monocular relearning of a dark-light discrimination by cats after unilateral cortical and collicular lesions, Brain Research, 83 (1975) 156-162.