Visuomotor Adaptation to Displacing Prisms by Adult and Baby Barn Owls - PDF

Please download to get full document.

View again

of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Government Documents

Published:

Views: 2 | Pages: 9

Extension: PDF | Download: 0

Share
Related documents
Description
The Journal of Neuroscience, September 1989, g(9): Visuomotor daptation to Displacing Prisms by dult and Baby Barn Owls Eric I. Knudsen and Phyllis F. Knudsen Department of Neurobiology, Stanford
Transcript
The Journal of Neuroscience, September 1989, g(9): Visuomotor daptation to Displacing Prisms by dult and Baby Barn Owls Eric I. Knudsen and Phyllis F. Knudsen Department of Neurobiology, Stanford University School of Medicine, Stanford, California 9435 The capacity of barn owls to adapt visuomotor behavior in response to prism-induced displacement of the,visual field was tested in babies and adults. Matched, binocular Fresnel prisms, which displaced the visual field 1 lo, 23, or 34 to the right, were placed on owls for periods of up to 99 d. Seven baby owls wore the prisms from the day the eyelids first opened; 2 owls wore them as adults. Prism adaptation was measured by the accuracy with which a target was approached and struck with the talons, a behavior similar to pointing behavior used commonly to assess prism adaptation in primates. Baby and adult owls exhibited a limited capacity to adapt this visuomotor behavior. cquisition of adapted behavior was slow, taking place over a period of weeks, and was never complete even for owls that were raised viewing the world through relatively weak (11 ) displacing prisms. When the prisms were removed from adapted owls, they struck to the opposite side of the target. The recovery of strike accuracy following prism removal was rapid; 7 of 9 owls recovered normal accuracy within 3 min of prism removal, despite having worn the prisms for months. This limited capacity for adaptation contrasts dramatically with the extensive and rapid adaptation exhibited by adult primates exposed to comparable prismatic displacements. The mechanism of adaptation used by the owls was to alter the movements employed for approaching targets. Instead of moving straight ahead, the head and body moved diagonally relative to the orientation of the head. Thus, in contrast to prism adaptation by humans that can involve reinterpretation of eye, head, and limb position, prism adaptation by owls is based on changes in the motor commands that underlie approach behavior. pproaching an object is a complex visuomotor task requiring the integration of information from a variety of sources. First, the position of the object s image on the retinae must be interpreted. Second, the direction of gaze, i.e., the position of the eyes in the head and of the head on the body, must be accounted for so that the perception of the object s location remains stationary as the eyes and head move. Third, the positions of the Received Feb. 6, 1989; revised Mar. 2, 1989; accepted Mar. 21, We thank S. du Lac, J. R. Lackner, and J. Nachmais for their comments on the manuscript. This work was supported by a neuroscience development award to E.I.K. from the M&night Foundation and a grant from the National Institutes of Health (ROl NS ). Correspondence should be addressed to Dr Eric I. Knudsen, Department of Neurobiology, Stanford University School of Medicine, Sherman Fairchild Science Building, Room D2, Stanford, C Copyright 1989 Society for Neuroscience /89/ $2./O limbs must be evaluated in order to determine what movement would be appropriate. If this information is analyzed correctly, motor commands can generate movements that bring the animal to the desired object, but only if the motor commands themselves are properly calibrated. Error in any one of these processes will cause an animal to approach objects inaccurately, interfering with the animal s vital abilities to acquire food and interact with the environment. feedback signal that the brain could use to evaluate the performance of these processes is the stability of an object s image on the retinae as the object is approached. n inaccurate approach causes the image of the object to move across the retinae in a direction opposite to the direction of error. This feedback signal, which is analogous to the retinal slip signals used to calibrate the vestibulo-ocular reflex for stabilizing the eyes during head movements (for reviews see Melvill-Jones, 1977; Miles and Lisberger, 1981) and to calibrate smooth pursuit eye movements for tracking moving objects (Heywood and Churcher, 1971; Pola and Wyatt, 198) could be used to adjust the perception of object location and the movements necessary to approach desired locations. Does the brain use visual feedback to adjust these processes? If so, are adjustments made at one or at several stages in the computation of approach behavior? re adjustments restricted in range or to a particular period of brain development? These questions can be addressed by subjecting animals to binocular prisms that shift the visual field to one side. Such prisms alter the relationship between the projection of an object s image onto the retinae and the movement required to approach the object. dult primates, particularly humans, exhibit extremely rapid and large-scale adaptations to displacing prisms; substantial adaptations can occur within minutes or hours under optimal conditions (Held and Bossom, 1961; Harris, 1965, 198). The mechanisms of adaptation vary with the nature of the experience. The usual mechanisms include alterations in the interpretation of (1) gaze direction (eye position relative to head, and head position relative to body), which is used to translate retinal location into a body-centered coordinate space, and (2) the positions of the limbs (Harris, 1963, 198; Lackner, 1973, 1981). lthough baby primates have not been tested quantitatively with displacing prisms, it seems unlikely that prism adaptation could be significantly faster or more extensive than it is in adults. In contrast, prism adaptation in chickens is apparently much more restricted, even when prisms of moderate strength (compared to those used on primates) are placed on newly hatched chicks to maximize the likelihood of adaptation (Hess, 1956; Rossi, 1968, 1969). fter days of wearing binocular displacing 3299 Knudsen and Knudsen l Visuomotor daptation by Owls Table 1. Period of prism experience Owl (number) (identity ) Prism strength (degree rkw Prisms on (age in days) Prisms off (age in days) 1 Bf Mb It W Fj Wa Fu Si a Other data from some of these owls appear in other papers. The number of the owl is adapted for each report so that the order represents a logical progression for that study. The identify of the owl is maintained across studies to enable comparison of data from the same animal presented in different reports. prisms, chicks continue to peck to one side of a target. The only evidence that adaptation has taken place is that pecking errors are made to the opposite side of the target immediately after prism removal. lthough these data suggest a tremendous disparity in the adaptation capacities of chickens and primates, the task required of the chicks was different from that required of primates. Pecking is a ballistic movement made on the basis of an initial judgment of target location relative to head position. The chick loses sight of the target during the strike, and information about the accuracy of the movement is not available until after the movement is finished. On the other hand, the accuracy of orientation, approach, or pointing movements, typically used to assess prism adaptation in primates, can be evaluated and calibrated by the animal as the movements are being made. If this difference in the nature of the movements is important, it might overestimate a discrepancy in the capacities for adaptation in these different classes of animals. The experiments reported here assess the capacity for visuomotor adaptation in another species of bird, the barn owl. The behavior used to evaluate adaptation was the accuracy with which an owl struck at a target with its talons. This behavior was chosen because it is the closest within the owl s repertoire to pointing behavior by primates. In contrast to pecking behavior, the approach of the talons during a strike is monitored visually up to the moment of impact (Fig. 1). Moreover, this behavior is included in the owl s daily experience. Nevertheless, owls that viewed the world through prisms from the day the eyelids opened exhibited little adaptation of this behavior, even though the same individuals exhibited major alterations in sound localization behavior (Knudsen and Knudsen, 1989). The visuomotor adaptation that did take place occurred in adults and in babies and resulted from modifications in motor commands rather than from changes in the perception of gaze direction or of limb positions, as in primates. The data support the contention that visuomotor plasticity is considerably more restricted in birds than it is in primates (Taub, 1968). Materials and Methods Subjects. Eleven barn owls (Tyto alba) served as subjects, 9 were used to study the effect of long-term experience with binocular displacing prisms on strike accuracy, and 2 were used only to measure normal strike accuracy (Tables 1, 2). Owls l-7 had prisms mounted over the Figure I. strike by owl 3 after 6 d of prism experience. The owl had worn prisms from the day the eyelids first opened. The prisms displaced the visual field 23 to the right. Visuomotor adaptation is far from complete. The bull s-eye target was used for photography only. Normally, the target was placed on a uniformly white piece of paper. eyes as babies, just as the eyelids began to open and while the ocular medium was still cloudy (Knudsen, 1989). Owl 8 was fully grown when the prisms were mounted and reached sexual maturity while wearing the prisms. Owl 9 was a sexually mature adult when the prisms were mounted. Owls 1 and 11 were juveniles between the ages of 8 and 11 d when tested. Prisms. Owls l-9 wore matched binocular Fresnel prisms (Vision Care/3M) of one of 3 strengths (Table 1): 2 diopter (11 displacement), 4 diopter (23 discplacement), or 6 diopter (34 displacement). The prisms were oriented to displace the visual field to the right and were mutually aligned such that a collimated beam passing through either lens diffracted identically. The field of view afforded by each of the prism spectacles was measured ophthalmoscopically; detailed descriptions are given in the companion paper (Knudsen and Knudsen, 1989). In brief, the visual field decreased with increasing prism strength. However, even in the worst case (34 prisms), the horizontal extent of the visual field was 6 for the right eye and 8 for the left eye, with 5 (normal) binocular overlap. Blinders on the sides of the frames blocked peripheral vision. The prisms were held in lightweight aluminum frames that were secured to the head with bolts. The bolts were cemented to the skull while the animal was anesthetized with halothane and nitrous oxide. Because the skull of baby owls is soft, the prisms on owls l-7 were held in place initially by stitching the frames to an elastic helmet that was secured with chin straps to the head. When these owls reached 3 d of age, the frames were bolted to the skull as described above. Experience. The baby owls (l-7) were raised in a single large cage where they interacted vigorously with each other and with objects in the cage. The prisms were cleaned at least once per day, and each bird was handled frequently. When the owls began to fly (about 5 d old), The Journal of Neuroscience, September 1989, 9(9) Prisms on center - 1 cm - talon Figure 2. The pattern of talon marks that resulted from a strike. Strike center was the geometric center of the medial 2 talon marks from both feet. training for the strike test began. t this time, the owls were placed in individual cages so that food intake could be controlled precisely. The owls continued to be handled daily as they were trained and tested. The adult owls (8 and 9) were housed in separate cages throughout the experiment. They were much less active than the baby owls. Therefore, to encourage adaptation, we released the birds for approximately 3 hr each day in a large room where they flew among obstacles and from one perch to another. dditional experience with those behaviors required by the experimental paradigm was gained by each bird as it was trained and tested. Nevertheless, throughout the period of prism exposure, all of the owls exhibited difficulty with landing accurately on small perches, flying through small passages, and striking at dead mice with the beak. It was apparent from these qualitative observations that the birds had not adapted completely to the prisms. Strike test. Owls were trained to strike at a 1 cm3 piece of mouse (target) from a perch 7 cm above the floor and 6 cm from the base of the perch to the target (Fig. 1). The target was placed on a sheet of butcher paper (45 x 45 cm). Under the butcher paper was a thin pad of foam rubber. When the owl struck at the target, the talons penetrated the paper, leaving a pattern of holes (Fig. 2). new piece of butcher paper was used for each trial. The position of the butcher paper and of the target on the butcher paper were varied from one trial to the next, although the distance of the target from the base of the perch was maintained at 6 cm. Training the owls to perform this paradigm took about 2 weeks. Throughout the training period, the owls received food only in the test room. During the first few days of training, the birds stood on the perch and became familiar with the environment. Owls wearing prisms had more difficulty with the paradigm and, therefore, were trained on a low perch (3 cm from the ground) to begin with. When the owls became sufficiently hungry, usually by the fifth day, they would jump from the perch for food placed on the ground. Typically, they would land nearby and walk to the food. Once accustomed to this manner of feeding, they were not permitted to eat the food unless the initial approach was an mark ,.,.I,,.I. I.,.....,2 1 SO Prisms off I 1' ; -5, ,.,.,,,.,,,,,., left right zimuth (mm) Figure 3. Strike centers of owl 4 recorded before and after prism removal. Strike centers are shown in raw form. The location of the target defined the origin of the coordinate space. Strikes with prisms on (top) were measured after 82 d of continuous experience with 23 right-displacing prisms. Strikes with prisms off (bottom) occurred in the order indicated by the numbers, immediately following prism removal. aggressive strike. When the owls wearing prisms attained this level of training, they were placed on the 7 cm test perch to complete the training. These owls rarely hit the target with the talons, but as they landed the head moved quickly to the food. Data analysis. fter each strike, the owl s direction of approach was marked on the butcher paper. Owls either wearing prisms or with the prisms recently removed sometimes turned in the air as they descended. For these strikes, the direction of movement just before impact was recorded as direction of approach. Strike center was quantified as the geometric center of the medial 2 talon marks from each foot (Fig. 2). Strike error was the distance from 14- g 12..o, L E 1. 8 SO. ' 6. E ; 2. I Table 2. Strike accuracy of normal owls O- Owl zimuthal Magnitude of accuracy Samnle size error (mm, mean -1- SD) (mm, mean I SD) It 9 Left Left6-t Left 2 * 1 Figure 4. Mean strike error measured just before prism removal for the 7 owls raised with prisms. The azimuthal component of the strike error was defined as the shortest distance from the strike center to a line oriented in the direction of the owl s approach and passing through the center of the target. Mean azimuthal error is represented by a closed circle. The SD of azimuthal errors is represented by the bars. Sample sizes ranged from 2 to 37 strikes. 34 33 Knudsen and Knudsen l Visuomotor daptation by Owls prisms off I w q ; Li_li::: R Figure 5. Recovery of strike accuracy immediately after prism removal for owls raised wearing prisms (l-7). Each symbol type represents data from one bird. Each symbol indicates the azimuthal error of a single strike plotted according to the order in which it occurred (trial number). The test sessions lasted about 3 min. The data are grouped according to the strength of the prisms worn by the owl as indicated L2C R2 x x = fi i 4. x x x 34 prisms 6 6 Ii. li. Trial Number off 1 4 the strike center to the center of the target. zimuth error was the shortest distance from the strike center to the line passing through the center of the target along the owl s direction of approach. Strikes by several owls were photographed with a 35 mm Nikon camera. Strikes by owls 7 and 9 during and after prism adaptation were videorecorded (Magnavox, newvicon) and were analyzed frame-byframe (3 msec per frame) to determine speed of descent and details of adapted behavior. Results Normal strike behavior Strikes carried out by 3 normal owls (owls 9-l 1) were stereotyped, rapid, accurate movements (Table 2). Before an owl sighted the target, it surveyed its surroundings using characteristic, bobbing movements of the head: forward and back, up and down, and side to side. When the target was sighted, head movements ceased with the head directed at the target, and the body turned to align with the head. The head lowered slowly in the direction of the target until it was at the level of the talons. t this point, owls often paused momentarily. Then, the wings raised and the owl pushed forward, propelling the face at the target. The descentook approximately 5 msec, during which time the wings and tail made small amplitude braking movements. Just before impact, the talons swung forward in front of the face and extended toward the target (Fig. 1). The spread of the talons always encoinpassed the target and was typically centered on the target. In 6 strikes by the 3 birds, The Journal of Neuroscience, September 1989, 9(9) I 2 8: k 6. l ts t LLl E L2. l R a..* l a** l R2.), 1 5 IO i 5 1' ge (days) Figure 6. Recovery of strike accuracy by owl 5 following prism removal after 74 d of experience with 34 right-displacing prisms. Strike errors recorded on 3 consecutive days following prism removal (at age 86) are plotted in the order they occurred. The age line at the bottom represents the history of the bird. Hatched bar indicates the period of prism experience. the average distance from the center of the talon spread to the center of the target was 13 mm (Table 2); the largest single error was 32 mm. For comparison, dimensions of the median talon spreads of each of these 3 owls were 8 x 97 mm, 78 x 89 mm and 66 x 113 mm, respectively. Strike accuracy in azimuth was quantified for each owl as the transverse distance from the center of the distribution of all strikes to the location of the target (see Materials and Methods). Strike accuracy in azimuth for owls 9-l 1 ranged from left 2 mm to left 11 mm (Table 2). Strike accuracy of prism-reared owls Binocular prisms were placed on 7 baby owls at an age when the eyelids were just beginning to open (Table 1). The owls grew to full size and learned to fly while wearing the prisms. Strike accuracy was tested when the owls were between 65 and 111 d old, after having worn prisms for from 55 to 99 d. The alignment of the eyes in the head of each bird was measured ophthalmoscopically and was found to be normal. (Details of the methods of measurement and of the results have been reported previously; Knudsen, 1989.) ll of the prism-reared owls struck consistently and substantially to the right of the target (Fig. 3). Even the 2 birds wearing relatively weak prisms (11 ) struck to the right by an average of 29 and 43 mm, respectively. The magnitude of mean errors increased with the dioptric strength of the prisms (Fig. 4). The data demonstrate that adaptation to the prisms was far from complete even though these animals had always viewed an optically shifted world. Evidence that some degree of adaptation had taken place was found in the strikes carried out immediately after the prisms were removed (Fig. 3). On the first trial, carried out within minutes of prism removal, all of the birds struck to the left of the target by from 36 to 117
Recommended
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks