FEAR OF SNAKES IN FERAL AND LAB REARED SQUIRREL MONKEYS. 'Sarah Gardiner Murray. A Thesis Submitted to the Faculty of the DEPARTMENT OF PSYCHOLOGY

FEAR OF SNAKES IN FERAL AND LAB REARED SQUIRREL MONKEYS by 'Sarah Gardiner Murray A Thesis Submitted to the Faculty of the DEPARTMENT OF PSYCHOLOGY In Partial Fulfillment Of the Requirements For the Degree of MASTER OF ARTS In the Graduate College. THE UNIVERSITY OF ARIZONA 1 9 7 2

STATEMENT BY AUTHOR This thesis has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University. Library to be made available to borrowers under rules of the Library. Brief quotations from this thesis are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author. SIGNED: 1 APPROVAL BY THESIS DIRECTOR This thesis has been approved on the date shown below: Dennis L. Clark \ Date Assistant Professor of Psychology

ACKNOWLEDGMENTS The author expresses her sincere appreciation to Dr. James E.' King for his guidance in designing and writing this study, to Dr. Dennis L. Clark for his assistance with the data analysis and constructive criticism of this manuscript, and to Dr. Sigmund Hsiao for his constructive criticism of this manuscript.

TABLE OF.CONTENTS Page LIST OF TABLES.. LIST OF ILLUSTRATIONS...... V vi ABSTRACT....».................. vii INTRODUCTION..................... 1 METHOD........................ 7 Subjects................... 7 Apparatus... 8 Desrgu.................... 9 Procedure............. 11 RESULTS....................... 13 DISCUSSION...................... 27 APPENDIX; FEAR BEHAVIORS; GROUP MEAN DURATION PER STIMULUS FOR SELECTED D A Y S... 32 REFERENCES...................... 38

LIST.OF TABLES Table Page 1. Analysis of Variance Summary Table for Latency to Grab Food............ 16 2. Fourteen-day Group Means per Stimulus Condition................ 17 3. Analysis of Variance Summary Table for Day 1 Latency to Grab Food.......... 21 4. Mean Group Latencies to Grab Food per Stimulus on Day 1.... 23 5. Analysis of Variance Summary Table for Latency to Grab Food on Day 1 4..... 24 6. Standing Erect: Group Mean Duration per Stimulus................. 33 7. Locomotion: Group Mean Duration per Stimulus................... 34 8. Vocalization: Group Mean Duration per Stimulus................. 35 9. Withdrawal: Group Mean Duration per Stimulus................. 36 10. Turning Away from the Stimulus: Group Mean Duration per Stimulus....... 37 v

LIST OF ILLUSTRATIONS Figure Page 1, Fourteen-day Mean Group Latency to Grab Food per Stimulus Condition 14 2. Changes in Group Mean Food-taking Latencies as a Function of Days for Live Snake, ' Painted and Plain Rubber Snakes, and Painted Coiled Tube.... 20 3. Visual Scanning: Group Mean Duration per Stimulus for Selected Days... 26

ABSTRACT Fourteen squirrel monkeys divided into three groups (feral adult, n=6; older non-feral, n=3; and laboratory infants raised in an enriched social environment, n=5) received 42 trials per condition of the following stimuli, ordered according to their approximation to a natural predator: live boa constrictor, rubber snake painted with a boa's colors, plain rubber snake, painted coiled tube, plain coiled tube, straight tube, block, and the container for the stimuli. Fear was measured by latency to grab a preferred food in front of the stimulus. The duration of the following fear behaviors was recorded: visual scanning, standing erect, withdrawal, locomotion, vocalization, and turning away from the stimulus. Overall, the feral group showed the following foodtaking latency pattern from longest to shortest: snake (2) rubber snakes (3) all other stimuli. (1) live The non-feral groups did not differ significantly from each other and showed no differences when stimulus conditions were compared. The feral group had longer latencies than the non-feral groups in the live and rubber snakes conditions. By the last day of testing, there were no differences between groups or stimuli. The results support the thesis that fear of snakes is learned. v n

INTRODUCTION Many species of primates exhibit fear of snakes. Since the nineteenth century, naturalists and experimental psychologists have observed or demonstrated this phenomenon. The earliest experiments were informal and lacked con trol. Warwick (1832) reported that a female orangutan exhibited fear of a live boa constrictor and a toy shake. Brown (1878a, 1878b) introduced a coiled snake, a tortoise, and a dead alligator into a cage of macaques at the Philadelphia Zoological Gardens and observed that the macaques exhibited persistent fear of the coiled snake. Mitchell (1912) presented a snake to a variety of primates at the London Zoologi cal Gardens and noted that lemurs exhibited no fear and marmo sets acted indifferent, while capuchins, howlers, spider monkeys, wooly monkeys, macaques, guenons, baboons, mandrills gibbons, chimpanzees, and orangutans showed fear. It is difficult to draw definite conclusions about these primates' fear of snakes because all animals were tested in a social situation, and several primate species were housed in one cage. Mitchell also tested a baby lemur, a baby chimpanzee, a young cebus monkey, and a young gibbon. Of these, only the cebus monkey exhibited fear behavior. 1

' : '2 In the field, Antonins (1939) observed that chimpanzees showed caution in the presence of large snakes (Python sp sebae) and occasional aggression toward smaller snakes, both poisonous and npnpoisonous. Hall (1963) reported that chacma baboons (Papio ursinus) avoided a large mole snake. (Pseudaspis cana) while Isemonger (1962) observed pythons attacking baboons and a puff adder biting a young baboon. In contrast to Mitchell (1912), Kortlandt and Kooij (1963) reported that capuchins eat snakes. Rainer Lorenz (1971) reported that the following New World monkeys have shown no evidence of fear of snakes: marmosets (Callithrix jacchus, C. penicillata, C. humeralifer, Saquinus midas, S>. oedipus), cebus monkeys (Cebus ape11a, C. spirifer, C. fatuellas, C% capuchinus), and spider monkeys (Ateles sp., Brachyteles arachnoides). In addition, he observed Goeldi's marmoset (Callimjco goeldii) killing and eating a small eastern ribbon snake (Thamnophis s. sauritus). Experimental studies on primates1 fear of snakes have used apes and Old World monkeys as subjects. Adult chimpanzees avoided snakes and snake-like objects (Hebb, 1946; Yerkes and Yerkes, 1936; Haslerud, 1938) while infant chimpanzees tended to avoid the most animate objects, regardless of their appearance. Schiller (1952) adapted adult chimpanzees to a snake by presenting the snake in a glass box on top of which a banana was placed. He also noted that in contrast to infants, adult chimpanzees exhibited a marked fear of snakes.

Tinklepaugh and Hartman.(1932) reported a similar age difference for rhesus monkeys when they noted that rhesus monkey mothers had to restrain their infants from touching a live 3 garter snake. Green (1965) found that juvenile and adult feral rhesus monkeys took significantly longer to grab food in the presence of a snake than in the presence of other stimuli. Butler (1964) discovered that a rhesus monkey, would learn to pull a chain to keep his box illuminated when garter snakes were placed in the box with him. In contrast, Bernstein and Mason (1962) tested rhesus monkeys ranging in age from one month to 27 months and concluded that the snake model elicited less emotionality from the subjects than the models of a beetle, a grasshopper, or a dog. Wolin, Ordy, and Diliman (1966) found no significant differences between feral, zoo, and laboratory rhesus monkeys' avoidance responses to a live garter snake. Singh (1966) discovered that urban and forest dwelling rhesus monkeys avoided a stuffed cobra, a probable predator when alive. Finally, Jos1in, Fletcher, and Emlen (1964) reported that feral rhesus monkeys took longer to grab food in the presence of a snake model than did the laboratory reared rhesus monkeys, but the latencies of the two groups did not differ significantly for the straight tube, wood block, and food only conditions. Rumbaugh (1968) is the only researcher to date who has reported that squirrel monkeys (Saimiri sciureus) fear snakes. He found that squirrel monkeys, golden marmosets (Leontideus

4 rosalia) and cotton-topped tamarins (S. oedipus) were afraid of prehistoric animal models and a live snake in a jar, but exhibited no fear of such neutral stimuli as unpainted blocks of wood of irregular shape and size. In summary, feral chimpanzees, gibbons, orangutans, baboons, macaques, guenons, mandrills, squirrel monkeys, spider monkeys, and howler monkeys exhibited fear behavior in the presence of snakes» The evidence was mixed for golden marmosets, tamarins, and capuchins. Goeldi's marmoset and capuchins were observed eating snakes. Finally, the only pro- simian observed in this fashion, the lemur, did not avoid the snake. Except for Rumbaugh (1968), investigators in the laboratory have studied only apes' and Old World monkeys' fear of snakes while neglecting New World monkeys. The present study had several purposes. The first was to extend the literature on laboratory studies of primates' fear Of snakes to include New World monkeys, in particular, squirrel monkeys. Squirrel monkeys make good subjects for the study of fear because they are preyed upon in the wild by rainbow boas (Epicrates cenchris). Eimerl and DeVore (1965) presented the evidence in a color photograph of a rainbow boa wrapping itself around a squirrel monkey. Only Singh (1966) has used a probable natural predator for a stimulus and he used an inanimate stuffed version. In the present study a common boa (Boa constrictor constrictor) was the stimulus because of the unavailability of a rainbow boa. However, Kauffeld (1969)

reported that the common boa preys on monkeys in the wild and Parker (1963) has stated that both the common and rainbow boas have almost the same geographical range from Mexico to Central Argentina. This area includes the geographical range of the squirrel monkey who can be found from the west coast of Central America, the foothills of the Andes in Bolivia, Peru, Ecuador, and Columbia, to south and east of the Orinoco River in the Guianas through the Amazon basin (Napier and Napier, 1967). In addition, Curran and Kauffeld (1937) have stated that the common boa can grow to a length of twelve feet while the rainbow boa does not grow beyond four to six feet in length. Therefore, it seems logical to assume that the common boa is likewise a natural predator of the squirrel monkey since the common boa is a larger snake than the rainbow boa which is a known squirrel monkey predator, the common boa does eat monkeys, and all three animals inhabit the same geographical range. Finally, Rumbaugh (1968) has cited evidence that the squirrel monkey has poor discriminability of red light so that the markings of the reddish rainbow boa and the brown and gray common boa probably appear more similar to the squirrel monkey than they do to humans. Second, this study employed, in the laboratory reared group, squirrel monkeys who were raised with their mothers and peers in an enriched social environment. Previous studies (Bernstein and Mason, 1962; Joslin, Fletcher, and Emlen, 1964; Green, 1965; WOlin, Ordy, and Dillman, 1966) using laboratory

animals, used animals who had been reared either in social or sensory isolation or with their mother alone in a restricted laboratory environment. Third, this experiment sought to determine the point at which squirrel monkeys exhibit fear of stimuli that are ordered according to their similarity to a natural predator. Finally, by repeatedly presenting the stimuli in the presence of food, one can determine if squirrel monkeys can become desensitized or less fearful (as measured by a decrease in food-taking latency) of a natural predator.

METHOD Subjects. Six feral and eight laboratory raised squirrel monkeys served as subjects The feral group and the laboratory infant group each originally consisted of two females and four males. Prior to the testing of the laboratory infant group, one male infant died, reducing the number of subjects in that group to five. All the feral animals were adults while the laboratory animals raised in an enriched social environment were nine months old at the time of testing. The infants were removed from their mothers at seven months and housed in a separate room with other squirrel monkeys in the colony. Because such investigators as Jones and Jones (1928), Haslerud (1938), Schiller (1952), and Green (1965), have noted an increase in fear of snakes in older primates (which could be due to social learning in the case of children or differences in early experience with monkeys and apes), a third group of older non-feral squirrel monkeys was added as a control for age. This group (n=3) consisted of a five year old adult male who had been hand-reared, a two year old female who had been raised by her mother but who lacked early peer group experience, and a two year old male raised by his mother in a zoo. ' 7

Apparatus The following objects were presented to each subjects 1, a live, moving two foot long boa constrictor (Boa constrictor constrictor) 2. a two foot long.coiled rubber snake painted with a close approximation to a boa constrictor's colors and markings 3o a two foot long coiled rubber snake painted gray 4. a two foot long coiled rubber tube painted with a close approximation to a boa constrictor's colors and markings 5. a two foot long coiled rubber tube painted gray 6. a 6% inch long straight rubber tube painted gray 7» a wooden block Jh inches long, 1% inches wide, and 3/4 inches high painted gray 8, a clear plastic box with no cover and 7 inches long, 5 inches wide, and 2h inches deep. (This box was also used as the container for the other seven stimuli.) The stimuli were arranged on an ordinal scale from those most like tile natural predator, the live boa constrictor, to those least like it, the plastic box (also used during the adaptation trials). Presenting each stimulus in the open, clear plastic container allowed the snake to move about, but prevented it from crawling off the tray. The experimenter made no attempt to control the snake's movements. The wood

9 block served as a control for novelty, since none of the subjects had seen an object of that size or color before, The testing of all subjects took place in a modified Wisconsin General Test Apparatus (WGTA) described by Harlow and Bromer (1938) The subjects were tested individually and remained in a stainless steel transport cage (13 inches wide, 15 inches long, and 14 inches high) that was placed behind a manually operated screen. The plastic box containing nothing or one of the stimuli was placed on an object tray (14 inches wide by 5-3/4 inches deep) behind a one inch diameter foodwell that contained either a meal worm or a raisin, depending upon the subject's preference. A one-way vision screen prevented the subject from seeing the experimenter. An electric timer operated by a foot pedal measured the subject's food-taking latency. A frequency duration counter recorded the frequency and duration of the following fear behaviors : visual scanning, standing erect, withdrawal, locomotion, vocalization, and turning away from the stimulus. Design This study utilized a three-way mixed design in which subjects were nested in three groups and crossed with stimuli and trials. All eight stimuli were presented to each subject so that each subject served as its own control, and each subject had 42 trials with each stimulus. All subjects experienced 14 consecutive days of testing.

10 The order of the stimuli was arranged so that all possible pairs of one stimulus following another could be accounted for. The purpose was to evenly distribute residual effects from any of the stimuli. With eight stimuli, there are 56 possible pairs since order is important and the same stimulus did not appear twice in succession. Although the stimuli were presented singly, the pair concept was used to designate the stimulus presented and the one immediately following it. All eight stimuli were presented before any stimulus was repeated. Consequently, the order of presentation can be conceptualized as eight stimuli in a row and seven rows. An experimental session consisted of 24 trials so that each stimulus was presented three times on a given day. On the first day of testing, the experimenter presented the sequence in the first three rows; the second day, rows four through six; the third day of testing rows seven, one, and two; the fourth day of testing rows three through five; the fifth day of testing rows six, seven, and one; the sixth day rows two through four; and the seventh day rows five through seven. The first week's order was replicated the second week. In this way, the sequence in each row served once as the first eight stimuli, once as the middle eight stimuli, and once as the last eight stimuli for each week of testing.

11 Procedure All adults were deprived of food for 23 hours before each session while the infants were deprived of food for 18 hours (they required more frequent feedings than the adults) All animals were fed immediately following each test session. Prior to the start of the experiment, all subjects were adapted to the apparatus. Adaptation consisted of presenting on the tray, a raisin or a meal worm (depending on individual preference) in the foodwell in front of the empty plastic box. The adaptation criterion was a food-taking latency of 3.00 or fewer seconds for three consecutive trials. The adaptation procedure preceded the start of the experimental session on each of the 14 test days. Adaptation arid testing took place in a darkened room. The house light of the WGTA provided illumination for the stimulus. A second light placed behind the transport cage provided enough illumination for the experimenter to observe the subject's behaviors. The frequency-duration counter which operated throughout the adaptation and testing trials emitted a sound loud enough to mask any external noises. Actual testing consisted of placing the food in the foodwell and the stimulus in the plastic container behind the foodwell on the tray, pushing the tray up to the opaque screen, raising the screen and starting the timer when the screen was fully raised, recording any fear behaviors with the frequencyduration counter, releasing the foot pedal to stop the timer

at the end of the trial, lowering the screen, bringing the tray back to the start position, and removing the stimulus from the tray. The trial ended when the subject grabbed the food or when 6 seconds had elapsed. If a subject never touched the food in a given trial, his latency to grab the food was recorded as 6 seconds for that trial. The food- taking latency determined the maximum duration of any fear behavior for a given trial (the duration could be less than the maximum or zero if a particular fear behavior did not occur). The intertrial interval was 60 seconds.

RESULTS The mean.latency (and mean duration for each fear behavior) of the three daily stimulus presentations was computed for each subject for each stimulus. An unweighted means analysis of variance (Winer, 1962) was applied to the latency data. In addition, an unweighted means analysis of variance was applied to.the data for day 1 and day 14 separately to assess the effects of habituation. No analysis of variance was performed on the fear behaviors because their duration was positively related to the latency measure (as in visual scanning) or they occurred only in the feral group (withdrawal) and not in the other two groups, thus violating the homogeneity of variance and covariance assumptions in the analysis of variance. Group mean latencies for each stimulus were analyzed according to Scheffe's (1953) multiple comparison method. Figure 1 compares the 14-day mean food-taking latency for each stimulus for the three groups. The graph clearly indicates that the three groups did not differ greatly on their latency to take the food in the presence of the empty box, the block, the straight tube, the plain coiled tube, and the painted coiled tube. The lab infant and the control (older captivity reared) groups did not differ significantly 13:

30-. ADULT FERAL CONTROL 25 LAB INFANT 20 - Z LU I- < '5- < LlI 5 LIVE SNAKE PAINTED PLAIN PAINTED PLAIN RUBBER RUBBER COILED COILED SNAKE SNAKE TUBE TUBE STRAIGHT TUBE BLOCK EMPTY BOX Figure 1. Fourteen-day Mean Group Latency to Grab Food per Stimulus Condition.

15 t \ ' ( >.20) from each other in the presence of the live snake, the painted rubber snake, or the plain rubber, snake. The feral animals exhibited a significantly ( >.01) longer mean latency in the presence of the live snake than the other two groups. This same pattern held ( <.05) for the two rubber snake stimuli although the difference was not as great. When the feral group alone are considered, they took longer ( <.01) to grab the food in the presence of the live snake than in the presence of any of the seven other stimuli, and they took longer ( <.10) to grab the food in the presence of either of the two rubber snakes than in the presence of any of the other stimuli. Table 1 shows the analysis of variance summary table for the entire 14 day analysis. The main effects of groups, stimuli, and days were all significant. All interactions (groups x stimulus, groups x days, stimulus x days, and groups x stimulus x days) were significant ( <. 001). Table 2 contains the 14-day group means for each stimulus which Figure 1 also shows graphically. The type of stimulus used failed to produce significant differences in the latency to grab the food for either the control of the lab infant groups ( >.20). However, the feral group took sig- nificantly longer ( <.01) to grab the food in the presence of the live snake than for any of the other stimulus conditions and took significantly longer to grab the food in the painted rubber snake condition than in the painted coiled

16 TABLE 1 ANALYSIS OF VARIANCE SUMMARY TABLE FOR LATENCY TO GRAB FOOD Source of Variance Degrees of Freedom Mean Square F Ratio & Between Groups Groups 2 8,287.42 5.32 <.025 Subjects within Groups 11 1,557.41 Within Groups Stimulus 7 2,552.26 5.63 <.001 Groups x Stimulus 14 1,905.21 4.20 <.001 Subjects within Groups x Stimulus 77 453.66 Days 13 448.18 7.07 <.001 Groups x Days 26 291.40 4.59 <.001 Subjects within Groups x Days 143 63.42 Stimulus x Days 91 61.77 2.90 <.001 Groups x Stimulus x Days 182 45.94 2.15 <.001 Subjects within Groups x Stimulus x Days 1001 21.33 Total 1567

17 TABLE 2 FOURTEEN-DAY GROUP MEANS PER STIMULUS CONDITION Adult Feral Control Infant Live Snake 30.67 3.93 1.07 Painted Rubber Snake 12.85 1.66 0.64 Plain Rubber Snake 12.50 1.68 0.49 Painted Coiled Tube 3.50 1.49 0.57 Plain Coiled Tube 2.30 1.74 0.64 Straight Tube 1.61 1.08 0.50 Block 1.55 1.13 0.52 Empty Box 1.44 1.10 0.54

tube to the empty box conditions (jd<.05). Using an =.10 level as recommended by Scheffe (1953), it can be concluded that the feral group's food-taking latency was significantly (jd<.10) longer for the plain rubber snake condition than for the painted coiled tube to empty box conditions. The feral group showed no significant (p>.20) differences in their foodtaking latencies for the empty box, block, straight tube, plain coiled tube, or painted coiled tube conditions, or between the two rubber snake conditions. In summary, the feral adult group exhibited the following response pattern from the longest food-taking latency (greatest fear of the stimulus) to the shortest latency (least fear): (1) live snake (2) painted and plain rubber snakes (3) all other stimuli. Different stimuli failed to produce significant changes in food-taking latencies for either the control or lab infant groups. The Scheffe test also showed that the latencies for the control and lab infant groups did not differ (p-^.20) for any stimulus and that the adult feral latencies did not differ significantly (pp-,20) from the control or lab infant groups for the empty box, block, straight tube, plain coiled tube, and painted coiled tube stimuli. The feral group took significantly longer to grab the food than the control or lab infant groups in the presence of the live snake (p<c.01) and the painted and plain rubber snakes (p<j.05).

19 Figure 2 demonstrates the decrease in food-taking latencies as a function of days for the following stimulus conditions: live snake, painted rubber snake, plain rubber snake, and painted coiled tube. Graphs for the other four stimuli were not included because the curves were all essentially flat. Habituation after repeated exposure to a fear stimulus in the presence of food was most apparent in the live and rubber snakes conditions for the feral group. The decrease in food-taking latencies in the presence of these stimuli as a function of time largely accounted for the significance of the days main effect, the stimulus x. days interaction, and the groups x days interaction. The groups x stimulus x days interaction was significant due to the feral. monkeys habituating to the live and rubber snakes. The feral group took longer to adapt to the live snake than to the rubber snakes. This, in part, accounted for the greater mean latency in the presence of the live snake as compared to the rubber snakes when one considers the 14-day means. By day 8, the feral group's latencies in the two rubber snake conditions were not significantly different from the control or lab infant group's latencies. Table 3 shows an analysis of variance summary ta6le for day 1 food-taking latencies. The main effects of groups and stimuli and the groups x stimulus interaction were all significant (p<.001).

20 6 0 6 0 5 5 - \ 5 5 - \ 5 0 - \ 5 0-4 5 - \ 4 5 - \ x/v, 4 0-4 0-3 5-3 5-3 0-3 0-2 5 - A 2 5-2 0 - \ X. 2 0 - MEAN LATENCY (SEC) 15- \ \ 1 5-10- 10- / X / X A 5-5 - 0 - - 0-45 -i 4 0-3 5-30 2 4 6 8 10 12 14 DAYS LIVE SNAKE 4 5-4 0-35 3 0 - i [ ( [ ( - \,,, 2 4 6 8 10 12 14 DAYS PAINTED RUBBER SNAKE ADULT FERAL CONTROL LAB INFANT A 2 5-20- 15-10- 5-0- X/ \ ' I ' I'' I ' I 2 4 6 8 DAYS PLAIN RUBBER SNAKE \ r-1-! 10 12 14 2 5-20- 15-10- 5-0- V. 1 I 1 I 1 I 1 I 2 4 6 8 DAYS PAINTED I ' I ' I 10 12 14 COILED TUBE Figure 2. Changes in Group Mean Food-taking Latencies as a Function of Days for Live Snake, Painted and Plain Rubber Snakes, and Painted Coiled Tube.

21. TABLE 3 ANALYSIS OF VARIANCE SUMMARY TABLE FOR. DAY 1 LATENCY TO GRAB FOOD Source of Variance Degrees of Freedom Mean Square F Ratio E Between Groups Groups 2 4,650.78 18.89 <.001 Subjects within Groups 11 246.21 Within Groups Stimulus 7 838.05 15.91 <.001 Groups x Stimulus 14 668.04 12.69 <.001 Subjects within Groups x Stimulus 77 52.66 Total 111 I

Table 4 contains the group latency means per stimulus 22 for day 1 only. The Scheffe test showed that the stimulus conditions failed to differ significantly ( p*1.20) for either the control or lab infant groups. The feral group latency in the presence of the live snake was significantly longer than their latency in the presence of the painted rubber snake f e e.10) and their latencies in the presence of the other six stimuli (pc.ol). For the feral group, the latencies for both rubber snake conditions did not differ significantly (p>.20) from each Other nor did the latencies from the painted coiled tube to the empty box condition differ significantly (ppr.20). The feral group's latencies in both rubber snake conditions were significantly ( p <.01) longer than their latencies in the painted coiled tube, plain coiled tube, straight tube, block, and empty box conditions. There were no group differences in the empty box, block, straight tube, plain coiled tube, and painted coiled tube conditions, and no significant differences between the control and lab infant groups for the plain rubber snake, painted rubber snake, or live snake stimuli (p-y.20). The feral group had a significantly (p<.01) longer food-taking latency than either the control or lab infant groups in the live snake, painted rubber snake, and plain rubber snake conditions. Habituation was further assessed by analyzing the data for day 14, the final day of testing, as shown in Table 5.

23 TABLE 4 MEAN GROUP LATENCIES TO GRAB FOOD PER STIMULUS ON DAY 1 Adult Feral Control Infant Live Snake 56.81 4.74 1.30 Painted Rubber Snake 47.63 6.09 1.82 Plain Rubber Snake 41.84 4.62 1.17 Painted Coiled Tube 9.94 3.89 1.77 Plain Coiled Tube 8.39 3.13 0.83 Straight Tube 6.97 2.31 0.69 Block 3.87 1.13 0.61 Empty Box 2.58 1.03 0.34

24 TABLE 5 ANALYSIS OF VARIANCE SUMMARY TABLE FOR LATENCY TO GRAB FOOD ON DAY 14 Source of Variance Degrees of Freedom Mean Square F Ratio E Between Groups Groups 2 38.57 1.47 n.s. Subjects within Groups 11 26.17 Within Groups Stimulus 7 28.76 1.45 ' n.s. Groups x Stimulus 14 27.89 1.41 n.s. Subjects within Groups x Stimulus 77 19.84 Total 111

25 By day 14, there were no significant differences between groups or between stimulus conditions. The groups x stimulus interaction was not significant. Figure 3 is a graph of one of the fear behaviors, visual scanning. This figure demonstrates that the visual scanning duration was positively related to the food-taking latency (compare with Figures 1 and 2 and Table 4). Visual scanning occurred most often in the feral group in the presence of the live snake and rubber snakes conditions for the first few days. By day 7, the feral animals were adapted to the rubber snake stimuli, but not until the final day of testing did the feral group's duration of visual scanning decrease enough in the presence of the live snake to be no longer significantly different from the durations of the control and lab infant groups. The Appendix contains the group means per stimulus for the first two days, middle two days, and final two days of testing for the five other fear behaviors recorded: standing erect, locomotion, withdrawal, vocalization, and turning away from the stimulus. The habituation pattern for standing erect, locomotion, and vocalization paralleled that for visual scanning, but the feral duration in the live and rubber snakes conditions was not as long as that for visual scanning. The non-feral groups failed to exhibit withdrawal behavior or turn away from the stimulus. The feral group did not turn away from the stimulus for a long enough duration to warrant further analysis of this behavior.

4 0 -, A - L IV E SNAKE B - PAINTED RUBBER SNAKE C -P L A IN RUBBER SNAKE D-PAINTED COILED TUBE E -P L A IN COILED TUBE F-STRAIGHT TUBE G- BLOCK H- EMPTY BOX 0 ADULT FE R A L CONTROL E3 LAB INFANT 3 5-3 0-2 5 - MEAN DURATION (SEC) 5-5 - 3 0 -i 2 5 - A B DAY I DAY 2 DAY 7.(0) B g io ) Z ti( O ) = ( 0 ) g, 0 ) a f i t o i 20-20- 10-15- 10-5 - (ocl KWo) OL-to) w J3(0) _d») loi Mq. DAY 8 DAY 13 DAY 14 Figure 3. Visual Scanning: Group Mean Duration per Stimulus for Selected Days. to

DISCUSSION 1. The present study indicated that feral squirrel monkeys fear a live boa constrictor and rubber snake models, while laboratory raised squirrel monkeys do not. This finding supports Yerkes and Yerkes (1936) thesis that monkeys fear of snakes is learned by means of prior experience with an aversive stimulus. It is also possible to conclude that adult squirrel monkeys do not exhibit greater caution or fear than younger squirrel monkeys in the presence of a novel stimulus unless the stimulus resembles a natural predator. Proof was provided by the lack of significant differences in the food-taking latencies in the empty box to the painted coiled tube conditions between subject groups, the lack of significant differences between the older non-feral group and the infant non-feral group in the rubber and live snake conditions,and the disappearance of the significant differences between the feral and non-feral groups in the live and rubber snakes conditions as a result of habituation. Yerkes and Yerkes (1936)» Haslerud (1938), and Menzel (1964) noted that adult chimpanzees feared snake-like objects while infant chimpanzees feared those objects that moved the 27

28 most. Green (1965) and Bernstein and Mason (1962) presented their stimuli on a moving belt. The present study attempted to control for movement. All of the stimuli were inanimate except for the live boa constrictor, and the opaque screen was not raised until the tray had been pushed up to it. None of the infants exhibited fear of the live snake. The boa constrictor's movement was not sufficient to arouse fear in the "infant squirrel monkeys. However, the boa constrictor's movement delayed the feral adults' habituation to this stimulus. Inspection of the day 1 means in Table 4 leads to the conclusion that the live snake's movement was responsible for the significantly longer latency in the live snake condition and that the feral group feared the live snake more than they feared the inanimate snake models. The feral group trial 1 data for each stimulus revealed that all six feral animals had a maximum score of 6 seconds in the live snake condition and five of the six feral animals had a maximum score of 6 seconds in both rubber snake conditions. Only one feral animal failed to show fear of the inanimate snake models on trial 1. His latencies were 6.54 seconds in the plain rubber snake conditions and 3.33 seconds in the painted rubber snake condition. The faster habituation to the inanimate rubber snake models (see Figure 2) was the reason for the significant difference in food-taking latencies between the live and rubber snakes conditions.

29 Since there were no significant differences between the infants who were reared in an enriched social environment in the lab and the older non-feral group, who were either hand-reared or reared by their mothers in a comparatively restricted environment (lab and zoo), one can conclude that an early enriched social environment is not essential for testing lab reared squirrel monkeys on their fear of a natural predator with which they have had no prior experience, according to the methodology employed in this study. This conclusion should not be generalized to other primates without empirical evidence for those species. Joslin, Fletcher, and Emlen's (1964) lab reared rhesus monkey subjects exhibited longer food-taking latencies in the presence of the live snake than in the presence of the other stimuli, while in the present study, the stimuli did not produce significantly different latencies in the non-feral groups. In addition, all of the older non-feral squirrel monkeys had received both social and sensory experience even though some of the early social interactions were only with humans. The feral group's habituation to their predator, the boa constrictor, by the end of testing demonstrated that squirrel monkeys can adapt to a feared stimulus. The boa con-, stricter used in the experiment was too small to kill a squir-. rel monkey. any subject. Although the snake moved around, it never attacked Each subject was reinforced for grabbing the preferred food with the opportunity to consume it.

30 Consequently, the food's positive incentive value and the absence of any predatory behavior in the live snake (and of course in the snake models) caused the feral subjects to become less afraid of the snake stimuli with repeated exposures. Schiller (1952) and Green (1965) described feral chimpanzees' and macaques' habituation to snake-like stimuli following repeated presentations in the presence of food.' Bernstein and Mason (1962) and Wolin, Ordy, and Diliman (1966) did not scale their stimuli according to the degree of approximation to a live snake. his stimuli, but did not use a live snake. Green (1965) scaled Only Joslin, Fletcher, and Emlen (1964) ordered the stimuli according to the degree of similarity to a live snake to determine the point of similarity at which the stimulus produced avoidance behavior. However, they used a harmless garter snake as the live snake stimulus for their rhesus monkey subjects instead of a snake that was potentially harmful to rhesus monkeys in the wild. For a feral squirrel monkey, a snake-like stimulus must resemble a snake (color and markings are not important) to produce avoidance behavior. The large significant difference between the painted coiled tube and the plain rubber snake (especially when day 1 data alone were considered) showed that the colors or markings of the snake were not the crucial variables. The stimulus had to resemble a natural

' - '.31 predator.6 Only the rubber snakes and the live snake had this property. The stimulus means can be broken into two parts. The non-feral monkeys did not fear any of the stimuli used in this study and the feral group did not fear stimuli ranging from the empty box to the painted coiled tube throughout the test period. Overall, the feral group exhibited fear of the live snake and the rubber snakes, but they eventually adapted to these stimuli. They took longer to adapt to the live snake than to the rubber snakes. This difference in adaptation time accounted for the significantly greater latency in the live snake condition in comparison to the rubber snakes, for the entire period of testing.

APPENDIX FEAR BEHAVIORS t GROUP MEAN DURATION PER STIMULUS FOR SELECTED DAYS 32

33 TABLE 6 STANDING ERECT; GROUP MEAN DURATION PER STIMULUS Day Group Live Snake Paint Rub. Snake Plain.Rub. Snake Paint Coil Tube Plain Coil Tube Str. Tube Block Empty Box 1 Feral Cont 14.44 0.13 7.03 0.28 0.10 6.81 0.89 0.10 3.28 2.89 1.83 0.81 0.56 2 Feral 11.78 0.11 0.30 6.64 0.22 0.03 9.81 0,00 2.56 0,00 1.28 0.50 0.17 1.36 0.28 0.07 0.86 O.oo 7 Feral 7.39 0.17 2.72 3.03 0.56 0.42 0.11 0.03 0.31 : o.o o ; 0.31 8 Feral 5.50 0.44 0.33 3.92 0.67 2.97 0.64 0.08 ; 0.39 0.14 1 0.17 0.14 13 Feral 3.78 0.44 0.20 1.08 0.13 0.13 0.08 0.20 0.07 0.11 0.13 14 Feral 3.25 0.07 0.17 0.13 0.06 0.20._

34. TABLE 7 LOCOMOTION: GROUP MEAN DURATION PER STIMULUS Day Group Live Snake Paint Rub. Snake Plain Rub. Snake Paint Coil Tube Plain Coil Tube Str. Tube Block Empty Box 1 Feral 23.18 0.11 21.11 0.44 16.28 2.11 0.42 0.03 0.06 2 Feral 16.31 3.83 0.37 8.72 6.06 0.14 0,00 7 Feral 6.22 0.50 3.56 4.39 0.06 o.po 0.06 0,00 0,00 0.11 8 Feral 7.44 3.81 6.36 0.11 13 Feral 5.39 0.44 0.13 1.58 0.33 0,00 14 Feral 4.31 0.44 0.07 0.11 0.11 0.10 0.06

35 TABLE 8 VOCALIZATION: GROUP MEAN DURATION PER STIMULUS Day Group Live Snake Paint Rub. Snake Plain Rub. Snake Paint Coil Tube i iplain Coil Tube Str. Tube Block Empty Box 1 Feral 11.81 0.06 7.33 0.07 8.97 0.17 0.06 0.03 2 Feral 12.42 0.78 4.17 2.39 o.oo ; 7 Feral 6.31 1.03 0.67 0.03 1 : 8 Feral 4.78 0.13 1.53 0.67 0.17 0.11 0.11 o.oo : 13 Feral 4.64 0.67 0.20 0.75 0.13 0.17 0.06 0.27 0.07 14 Feral 9.19 0.07 0.06 0.07 0.06 0.06 0,00 0.03 0.13

36 TABLE 9 WITHDRAWAL; GROUP MEAN DURATION PER STIMULUS Day Group Live Snake Paint Rub. Snake Plain Rub. Snake Paint Coil Tube Plain Coil Tube Str. Tube Block Empty Box 1 Feral 14.94 15.28 11.67 0.17 0.64 3.25 0.08 2 Feral 12.89 7.17 3.53 1.11 7 Feral.7.28 1.33 0.97 o.oo 8 Feral 15.36 3.89 0.72 13 Feral 4.19 14 Feral 2.97

37 TABLE 10 TURNING AWAY FROM THE STIMULUS: GROUP MEAN DURATION PER STIMULUS Day Group Live Snake Paint Rub. Snake Plain Rub. Snake Paint Coil Tube Plain Coil Tube Str. Tube Block Empty Box 1 Feral 2.03 2.61 1.18 0.17 0.03 2 Feral 1.44 0.39 1.67 7 Feral 0.83 0.19 0.17 8 Feral 0.78 0.19 0.33 13 Feral 14 Feral o.oo

REFERENCES Antonins, 0, Uber die Schlangenfurcht der offen. Tierpsycholoqie, 1939,. _2, 293-296. Bernstein, Stephen and Mason, William. The effects of age and stimulus conditions on the emotional responses of rhesus monkeys: responses to complex stimuli. Journal of Genetic Psychology, 1962, 101, 279-298. Brown, A. E. Fear of snakes in primates. Nature, 1878a, 18, 227. Brown, A. E. The serpent and the ape. The American Naturalist, 1878b, 12, 225-228. Butler, Robert A. The reactions of rhesus monkeys to fearprovoking stimuli. Journal of Genetic Psychology, 1964, 104, 321-330. Curran, Charles H. and Kauffeld, Carl. Snakes and Their Ways. New York: Harper and Brothers, Publishers, 1937. Einter 1, Sarel and DeVore, Irven. The Primates. New York: Time-Life Books, 1965. Green, Phillip C. Influence of early experience and age on expression of affect in monkeys. Journal of Genetic Psychology, 1965, 106, 157-171. Hall, K. R. L. Variations in the ecology of the chacma baboons, Papio ursinus. Symposium of the Zoological Society of London, 1963, If), 1-28. Harlow, H. F. and Bromer, J. A. A test-apparatus for monkeys Psychological Record, 1938, _19, 434. Haslerud, G. The effect of movement of stimulus upon avoidance reactions in chimpanzees. Journal of Comparative Psychology, 1938, 25, 507-528. Hebb, D. O. On the nature of fear. Psychological Review, 1946, 53, 259-276. 38

39 Isemonger, R. M. Snakes of Africa, Southern, Central, and East. Johannesburg s Nelson, 1962. Jones, Harold Ellis and Jones, Mary Cover. Fear. Childhood Education, 1928, 5, 136-143. Joslin, J., Fletcher, H., and Emlen, J. A comparison of the responses to snakes of lab and wild reared rhesus monkeys. Animal Behaviour, 1964, 12, 348-352. Kauffeld, Carl. Snakes: The Keeper and The Kept. Garden City: Doubleday & Company, Inc., 1969. Kortlandt, A. and Kooij, M. Protohominid behaviour in primates. Symposia of the Zoological Society of London, 1963, 10, 61-88. Lorenz, Rainer. Goeldi's monkey Callimico goeldii Thomas 1904 preying on snakes. Folia Primatologica, 1971, 15, 133-142. Menzel, Emil W. Responsiveness to object-movement in young chimpanzees. Behaviour, 1964, 24, 147-160. Mitchell, P. C. The Childhood of Animals. New York: Frederick A. Stokes Company, 1912. Napier, J. R. and Napier, P. H. A Handbook of Living Primates. New York: Academic Press, 1967. Parker, H. W. Snakes. New York: W. W. Norton and Company, Inc., 1963. Rumbaugh, Duane M. The learning and sensory capacities of the squirrel monkey in phylogenetic perspective. Chapter 10 in Rosenblum, Leonard A. and Cooper, Robert W., eds. The Squirrel Monkey. New York: The Academic Press, 1968. Scheffe, H. A. A method for judging all possible contrasts in the analysis of variance. Biometrika, 1953, 40, 87-104. Schiller, P. Innate constituents of complex responses in primates. Psychological Review, 1952, 5_9, 177-191. Singh, S. D. The effects of human environment upon the reactions to novel situations in the rhesus. Behaviour, 1966, 26, 243-250.

Tinklepaugh, O. L. and Hartman, Carl G. Behavior and maternal care of the newborn monkey (Macaca mulatta "M. Rhesus"). Journal of Genetic Psychology, 1932, 40, 257-286. Warwick, J. E. The habits and manners of the female Borneo orang-utan (Simla Satyrus), and the male chimpanzee (Simla troglodytes) as observed during their exhibition at the Egyptian Hall in 1831. Magazine of Natural History, 1832, j>, 305-309. Winer, B. J. Statistical Principles in Experimental Design. New York: McGraw-Hill Book Company, Inc., 1962. Wolin, R. L., Ordy, J. M., and Diliman, A. Monkeys' fear of snakes: a study of its basis and generality. Journal of Genetic Psychology, 1966, 103, 207-226. Yerkes, R. and Yerkes, A. Nature and conditions of avoidance (fear) responses in chimpanzees. Journal of Comparative Psychology, 1936, 2JL, 53-66. 40

6 5.19