Twenty-four hour activity budgets and patterns of behavior in captive

Applied Animal Behaviour Science 71 (2001) 67±79

Twenty-four hour activity budgets and patterns of behavior in captive ocelots (Leopardus pardalis) Sara H. Weller*, Cynthia L. Bennett Dallas Zoo, 650 R.L. Thornton Frwy., Dallas, TX 75203, USA Accepted 2 August 2000

Abstract Activity budgets of captive ocelots (Leopardus pardalis) were assessed from over 547 h of observational data obtained from six ocelots; two females at the Dallas Zoo (Dallas, TX), two females at the Caldwell Zoo (Tyler, TX) and a male and female at the Fossil Rim Wildlife Center (Glen Rose, TX). Data were examined for the percentage of active behaviors exhibited during the day and nighttime hours; temporal patterns of active, pace, exploratory and marking behavior, and for signi®cance in pacing behavior between pre- and post-feeding times. The captive cats had a bimodal pattern of active behavior similar to ®eld studies of wild ocelots, except that the timing of the active peaks were closer to the diurnal hours for the captive cats. The captive ocelots were less active than wild ocelots, and more diurnal. Also, the captive cats exhibited stereotypic pacing. When the percentage of time of active behavior was assessed for each cat, a strong variation between individuals and institution was not seen. Pacing behavior was highest prior to the feeding times for the cats. In assessing patterns of behavior, peaks in marking and exploratory behavior in the cats did not occur at the same time as the peaks in active behavior. However, we did see institutional differences in the pattern of exploratory and marking behavior, which may have been in¯uenced by differing management practices. # 2001 Elsevier Science B.V. All rights reserved. Keywords: Ocelot; Activity patterns; Zoo animals

1. Introduction Small felids have long been a part of zoo animal collections. However, as display animals these cats can be a challenge. They often remain inactive or hidden from public view. Furthermore, when active, the animals may frequently engage in apparently aberrant or stereotypical locomotor patterns (Shepherdson et al., 1993). Zoos also face inconsistent * Corresponding author. Tel.: 1-214-670-6833; fax: 1-214-670-6717. E-mail address: [email protected] (S.H. Weller).

0168-1591/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 5 9 1 ( 0 0 ) 0 0 1 6 9 - 6

68

S.H. Weller, C.L. Bennett / Applied Animal Behaviour Science 71 (2001) 67±79

results in the reproductive success of the cats. Problems such as these suggest management conditions and housing requirements of small captive cats are not adequately understood (Mellen, 1991; Carlstead et al., 1993). The activity pattern for a species is considered an adaptation to environmental in¯uences (Beltran and Delibes, 1994) and is an important basis for understanding behavior. Hence, such behavioral studies can be an important element to successful captive management of a species (Stevens and Hutchins, 1993; Kleiman, 1994). They can be even more critical for developing effective conservation strategies for many species. Studies of free ranging animals are often hampered by the dif®culty of direct observation of the subjects (Emmons, 1988; Law et al., 1997). This is especially true with felids. Thus, scienti®c study of captive cats cannot only provide insights (information, data, etc.) for improvements in zoo husbandry but can aid in the conservation of wild cats as well (Law et al., 1997). The ocelot is essentially crepuscular and nocturnal (Caso, 1994; Laack, 1991; Emmons, 1988). The current range extends from southwestern Texas to Paraguay and northern Argentina (Kitchener, 1991). It is listed as endangered by both the United States Fish and Wildlife Service (Harwell and Siminski, 1990) and International Union for the Conservation of Nature (Nowell and Jackson, 1996), and as Appendix I by the Convention for International Trade of Endangered Species (US Department of the Interior, 1992). Little information on the general behavior patterns of captive ocelots can be found. Only a few studies examined speci®c aspects of ocelot behavior (Mellen, 1991, 1993; Powell, 1997; Mansard, 1990). Fortunately, a large captive population of ocelots exists, consisting mainly of generic (unknown origin) individuals. We, therefore, began our studies by examining the activity of six ocelots held at three different institutions. 2. Methods 2.1. Subjects and site locations The subjects for this study were one male and ®ve female adult, generic ocelots. All were captive born except one wild caught female. None of the cats were pregnant during the course of this study. The three institutions used in this study were located between 95±978 longitudes, and 32±338 latitudes, and were of similar climatic conditions. Two females (DZF1 and DZF2) were housed at the Dallas Zoo (DZ), Dallas, TX; two females (CWF3 and CWF4) at the Caldwell Zoo (CW), Tyler, TX; and a male (FRM1) and female (FRF5) at Fossil Rim Wildlife Center (FR), Glen Rose, TX. Data for DZ and CW were collected between August 1994 and February 1995. For the cats housed at FR, data collection occurred from late June to December 1995. 2.2. Habitat/management The DZ cats were housed individually, in landscaped, naturalistic rock and dirt exhibits with concrete and barred holding rooms at the back. One exhibit was 4:6 m wide 6:3 m deep, and 4.1 m high, while the other was 9:1 m 6:3 m 4:1 m. There were two


69

holding rooms per cat; one was enclosed and heated, measuring 1:7 m 2:0 m 2:0 m and the other was 2:3 m 1:9 m 2:0 m. When the ambient temperature was above 48C, the cats were con®ned to their exhibits from 09.00 to 16.00 h; otherwise they had continuous access to their holding rooms. The exhibit for the CW cats included rockwork, a waterfall and stream, and one tree large enough for climbing. The exhibit was 5:5 m 7:6 m with an average height of 2.7 m. The cats alternated daily use of this exhibit. When not on exhibit, the cats were housed individually in 1:7 m 1:7 m 2:0 m indoor quarters. The cats at FR were housed together in an exhibit constructed around a heavily treed natural area, measuring approximately 18 m 15 m in length, and 5 m in height. They had continual access to two nest boxes measuring 1:2 m 0:76 m 0:4 m that were heated during the cooler months. All animals were fed once daily in the morning, between 07.00 and 08.00 h for DZ and CW, and between 09.00 and 10.00 h for FR. Location for food presentation was constant for each institution. The diet consisted of a measured amount of horse meat and whole prey (rat, mouse, quail, rabbit or chick). Water was available ad libitum from pools at DZ and CW, and from a pool and automatic drinker at FR. 2.3. Procedures The behaviors studied were derived from an ethogram (Mellen, 1989) that was later modi®ed to accommodate our own study. Data were obtained on each animal over a 24 h period using both direct and videotaped observations. Data were collected by observing focal animals for 15 min and continuously recording all occurrences (Altmann, 1974) of 15 selected behaviors (Table 1). Durations were recorded for nine of the behaviors (walk, run, climb, jump, pace, stand, sit, lie awake and lie asleep) and for when the subject was not visible. Frequencies were recorded for the concurrent behaviors (investigate, sniff, spray, scrape, sharpen claws and cheek rub). The 15 min observation sets were randomized and balanced over a 24 h period. Sampling varied from one to three observation sets during the day and two to four sets during the night. Trained research aides and staff were used to record behavioral data directly from the DZ cats during the hours of 07.00±18.00 h. All observers passed a reliability test at or above the 90% con®dence level. All other observation sets for the three institutions were taken from videotaped recordings, and were scored by the ®rst author. The cats at CW were videotaped in their exhibit, using hand held cameras during the day hours, and had a camera set up in their night quarters for nocturnal observations. The cats at FR were videotaped using a seven-camera surveillance system that was set up for both daytime and nighttime recordings. 2.4. Analysis The percentage of time the cats spent in a behavior was calculated for each 15 min observation set. The durational behaviors were combined into four categories: locomotion (walk, climb, run and jump); pace, rest (stand, sit, lie awake and lie asleep) and not visible for all analyses involving individual cats. When analyzing the cats as a group, locomotion and pace were combined to form an ``active behavior'' category.

70


Table 1 Ethogram of ocelot behaviors used in this study Behaviour Duration Locomotor Walk Run Climb Jump Pace Rest Stand Sit Lie/asleep Lie/awake Not visible Frequencies Exploratory Investigate Sniff Mark Spray Scrape Sharpen claws Cheek rub

Quadrupedal locomotion, movement of opposite limbs, two limbs on ground Rapid, forward moving locomotion, pushing off with hind limbs. All limbs may leave ground simultaneously, but no pause between next ground push off Vertical movement up or down substrate. Claws are used to grasp substrate Subject leaps from one point to another, either vertically or horizontally Slightly rapid, repetitive, unvarying ambulatory movement, unique to each cat Subject is stationary. Subject may be in a quadrupedal position, or may be in bipedal position, with front legs resting against vertical surface Subject is resting on haunches, forelegs are braced Settled on substrate, either sternally, laterally, or on back. Eyes are closed Like the above, except eyes are open, subject is alert Subject cannot be seen by observer. Location of the subject may or may not be known

Searching, pawing at, or trying to reach item Smell object Subject directs urine horizontally on an object Scraping, or rubbing hind feet alternately on ground Front claws are used to scratch on object Cheek is rubbed against an object or another cat

The frequencies of the concurrent behaviors were combined to form the categories explore (investigate and sniff) and mark (spray, scrape, sharpen claws and cheek rub). The behaviors were calculated as the rate per 15 min observation set. Data were analyzed to calculate the total time the six cats spent active, resting or not visible for the day (07.00±18.59 h) and night (19.00±06.59 h) hours. The mean for each category was calculated for each cat prior to averaging the totals for the group. The Kruskall±Wallis and Mann±Whitney U-tests were used for statistical analysis using P-value of 0.05 to accept signi®cance. Temporal behavior patterns were also determined over a 24 h period. Data were divided into 12, 2 h-time increments, starting at 01.00 h and ending at 23.00 h. For example, all samples between 01.00 and 02.59 h were grouped into the 01.00 h category. Only the active behavior category was used to determine the temporal behavior pattern for the six cats combined. For every 2 h period, the mean was calculated for each cat prior to averaging the total for the group. When determining the temporal behavior patterns for individual cats, the locomotion, pace, explore and mark categories were used. Differences in pacing behavior prior to and after feeding were determined for each cat at DZ and FR. The feeding time was kept constant to within a half-hour at each institution. All


71

observations that occurred a half-hour prior to the feeding time were compared to all that occurred the half-hour after feeding time for each cat. For example, if the feeding time occurred between 07.00 and 07.29 h, samples from 06.30 to 06.59 h were considered prefeeding, and samples from 07.30 to 07.59 h were considered post-feeding. 3. Results A total of 2,186 15 min focal observations (547 h) were analyzed for the six ocelots over the course of the study. For DZ, there were 570 focal observations of its cats, 535 of which occurred during the daytime. Because of insuf®cient numbers of nighttime observations, these cats were excluded from all analysis of nocturnal data. For CW, a total of 689 observations were taken from videotaped recordings of the cats. There were 213 observation sets for the day hours and 476 for the night. The observations for FR were successfully balanced over a 24 h period. There were 962 observations taken from videotaped recordings for these cats, 485 from the day hours, and 477 from the night. During the day, the six cats (1233 observation sets) spent 20.7% of the time in active behaviors and 69.3% resting. The remainder of the time (9.7%) the cats were not visible, and activity could not be determined. During the night (953 observation sets), the CW and FR cats spent 16.8% of the time in active behaviors, and 61.8% of the time resting, with 21.4% of the time not visible. When the cats were not visible, they were either behind an obstruction or in the small areas beyond the range of the cameras. Individual cats were compared to determine whether there were signi®cant individual differences in active behaviors. The Kruskall±Wallis test showed there were no signi®cant differences in active behavior at night, but there were for the six daytime periods (H 21:32; df 5; P 0:0007). Comparisons of individuals using the Mann±Whitney U-test for the 6 day periods showed the male FRM1 was signi®cantly less active than all the females (U 0:0; P 0:04). CWF3 was signi®cantly less active than the females DZF2, CWF4 and FRF5 (U 4:0; P 0:04). Group analysis of the temporal pattern of active behavior showed bimodal peaks for this behavior (Fig. 1). There was a peak just after dawn (07.00 h) and a second peak at 17.00 h. A drop in active behavior occurred during the midday from 11.00 to 15.00 h. Another low point was seen near midnight (23.00 and 01.00 h). Results were also obtained with one female (DZF2) removed from the data pool because of possible outlier effects (Fig. 1). The active behavior pattern for the remaining ®ve cats showed a larger decline in active behavior during the midday hours. The active behavior pattern for each individual cat was also assessed (Fig. 2). Similarities could be seen in four of the six cats (DZF1, CWF3, FRF5 and FRM1). While some variations existed between these cats, all showed bimodal peaks in active behavior; one occurring in the early morning hours, and the other in the evening. The morning peak occurrences ranged between 03.00 and 07.00 h, and the evening peak between 17.00 and 21.00 h. All showed a pronounced drop during the midday, from 11.00 to 13.00 h. For the three cats (CWF3, FRF5, and FRM1) with night data, another drop in active behavior occurred between 23.00 and 01.00 h.

72


Fig. 1. Temporal pattern of active behavior for all subjects.

Fig. 2. Individual temporal patterns of active behavior for all six subjects.


73

Fig. 3. Individual temporal patterns of locomotor, pace, mark and explore behaviors. (A) DZF1; (B) DZF2; (C) CWF3; (D) CWF4; (E) FRF5; (F) FRM1.

74


Fig. 3. (Continued ).


75

FRM1, overall, was less active than the other cats. He had a much longer range of daytime inactivity than the other cats. He spent over 90% of his time resting between the hours of 07.00 and 17.00 h. His zeniths in active behavior occurred during the hours of darkness, with peaks at 03.00 and 21.00 h. Patterns of exploratory and marking behavior were also determined for each cat (Fig. 3a±f) over a 24 h period (12 h for DZF1 and DZF2). For the most part, no relationship could be determined between these behaviors and peaks in locomotor and pacing behavior. We did see corresponding peaks in explore behavior with peaks in locomotor behavior for the male. Also, DZF1 had a peak in explore behavior at 07.00 h that corresponded to her peak in locomotor behavior. Peaks in marking behavior did not correspond to peaks in locomotor pattern for any of the cats, except for DZF1 at 17.00 h, and CWF4 at 07.00 h. In general, we found that the peaks in pacing behavior were exclusive of the behaviors of mark and explore. The pattern of pacing behavior was determined for each cat and compared to his or her respective locomotor patterns (Fig. 3) over a 24 h period (12 h for DZF1 and DZF2). Pacing behavior was found to peak at different times from the locomotor peaks for most of these cats. Only DZF1 and CWF3 had a pacing pattern that matched the locomotor pattern. Pacing behavior was also examined for the half-hour prior to, and following feeding time. Data was available only from DZ and FR. Three of the four cats showed signi®cantly higher levels of pacing during the pre-feeding hour than after feeding (DZF1, U 141:00, P 0:02; FRF5, U 96:00, P 0:008; FRM1, U 157:5, P 0:029). 4. Discussion Wild ocelots have mostly crepuscular and nocturnal activity periods (Ludlow, 1986; Emmons, 1989; Laack, 1991; Konecny, 1989; Tewes, 1986; Caso, 1994). The nocturnal peaks in activity of wild ocelots re¯ect the peak activity of their major prey species (Ludlow and Sunquist, 1987; Emmons, 1988). Further, ocelots are active for between 52 and 92% of the night hours (Navarro, 1985; Ludlow, 1986; Crawshaw, 1995). In our study the captive cats were much less active than in studies done with wild ocelots. In the wild, hunting accounts for much of the ocelots' nocturnal activity (Emmons, 1988). The captive cats studied were fed their entire daily ration of food in the morning hours. The lack of a need to hunt for their own food may account for the cats' decreased night activity. Also, decreased activity in the captive cats can be an indication of low stimulus diversity in its environment. ``. . .chronically understimulated captive animals depress their needs for stimulation by lowering their expectations of the level of stimulatory input from their surroundings'' (Carlstead, 1996, p. 326). The higher percentage of diurnal versus nocturnal behaviors seen in the captive cats may in part be a result of the dif®culty in reviewing the night videotapes. The percentage of time the cats were scored as not visible was higher at night than during the day. Active behaviors may have occurred during this time. The cats could have had a somewhat higher percentage of nocturnal activity than what has been recorded. However, even if the cats had been engaged in an active behavior the entire time they were not visible, their nocturnal activity would still be less than that seen in wild cats.

76


When comparing the cats' total amount of time in active behavior for the night hours, a signi®cant difference was found only between CWF4 and FRF5. It was interesting that we did not see a strong variation between cats or institutions. When daytime totals were compared, the male was the least active of all the cats. Since he was the only male in the study, it is not known whether this was due to gender or individual effects. Only one female (CWF3) appeared to be signi®cantly less active than the other females. Activity patterns for wild ocelots are related to sunrise and sunset (Ludlow and Sunquist, 1987; Emmons, 1988; Laack, 1991; Caso, 1994). Peaks in activity occur shortly after dusk and prior to dawn. A small resting period occurs at night, and a much larger resting period during the midday hours. The captive ocelots in this study, like their free ranging counterparts, also showed a bimodal pattern in activity. However, the morning peak occurred after dawn (between 07.00 and 08.59 h) rather than before. The second peak in activity occurred right around the dusk hours, similar to wild cats. The majority of the management of the cats' routines (feeding, cleaning and shifting the animals to or from their exhibit) occurred in the morning hours. Afternoon management routines mostly entailed whatever shifting procedures were necessary. Thus, the manipulation of the cats' routines, such as feeding, may have played a role in their morning activity peak. Their anticipation of the feeding schedule may explain this later morning peak in activity. However, no such management activity explains the afternoon peak. When not heavily managed, the captive cats still retained similarities to the activity patterns seen in wild ocelots. This was even more evident when the cat with unusually high levels of daytime activity (DZF2) was removed from the data set. Institutional differences in behavior were more evident when examining locomotor, marking and exploratory temporal patterns, and perhaps were in¯uenced by varying management styles. We also expected to ®nd that peaks in marking and exploring behavior would coincide with peaks in the locomotor behavior pattern. With few exceptions, this did not appear to be the case. Thus, exploring or marking territory did not seem to be the primary function of locomotion. At FR, the male's activity pattern most resembled that of a free ranging ocelot, with peaks of activity occurring under the full cover of darkness, and a long resting period during the day. The female in the same enclosure, however, had peaks at very different times than his. Her peaks occurred closer to the daylight hours. We did not ®nd many instances of marking behavior in these cats. Scent marking is a very overt display that usually includes a raised tail and shuf¯ing with the hind feet, which can be easily detected in videotapes. These two cats continuously shared the same enclosure. Perhaps the constant presence of another cat inured or inhibited the other from scent marking. This presence may also have resulted in their self-partitioning their most active peaks in usage of their exhibit. Removing the male and determining whether the female's activity pattern changes could test this hypothesis. At CW, both cats had peaks in marking at 07.00 h, when ®rst being given access to the exhibit. It appears each is overmarking their predecessor's scent. A peak in exploratory behavior followed the peak in marking behavior for these two cats. Scent marks in cats are continuously updated and rival's marks are usually overmarked when establishing territory (Kitchener, 1991). Also according to Kithchener, snif®ng scent marks allows the cat to gain an olfactory pro®le of other cats in it's area and be able to take the appropriate ®ght or ¯ight


77

response if it should meet up with it. The peaks in marking and exploring behaviors at CW seem to be associated with their alternating usage of the exhibit. For the DZ cats, we did not see high levels of marking behavior, nor did we see a pattern that could be correlated to their locomotor behavior patterns. These cats never shared usage of their enclosure. Perhaps the lack of a need to establish or maintain a territory explains the low levels of marking behavior in these two cats. Management practices at all three institutions may have had an effect on the temporal pattern of pacing seen. While pacing and locomotion are mutually exclusive behaviors, their temporal patterns do not have to be exclusive. If the temporal patterns had been similar for most of these cats, we may have concluded that pacing was a displacement for frustrated travel goals. However, pacing did not appear to be a substitute for locomotor behavior. Instead, pacing seemed to be related predominately to feeding times. The peak times in pacing varied from cat to cat, but all but DZF2 and CWF4 showed peaks in the early morning hours prior to their feeding times. Other studies have shown that pacing is a stereotypy found speci®cally in captive wild animals (Hediger, 1950; Rushen et al., 1993), and is many times associated with frustrated appetitive behavior (Wechsler, 1991). Pacing behaviors were especially evident when the data were broken down to the half-hour preand post-feeding for the DZ and FR cats. Three of the four cats showed signi®cant levels of pacing prior to being fed. Thus, pacing appears to be associated with frustrated appetitive behaviors for these cats. 5. Conclusion The captive cats were more sedentary than their wild counterparts and exhibited the stereotypic behavior of pacing. Despite the varying management styles and enclosure designs of the three institutions, the overall activity levels in the cats were very similar. Pacing behavior seemed to be a condition of the feeding routine used for these cats. Varying feeding times, the presentation and placement of food, and the number of feedings per day may reduce pacing behavior. The majority of the cats in this study maintained the basic activity pattern of a free ranging cat. Unfortunately, this means the ocelots are typically most active when visitors are not present. Again, altering feeding practices and routines to encourage more ``hunting'' or food procurement behaviors may increase the activity of the cats and change the times when they are most active. Differences in housing protocols of the cats may play a part in scent marking and exploratory behaviors. Alternating use of the exhibit for the CW may have led to spikes in the cats marking and exploring behavior when ®rst let into the exhibit enclosure. Increasing the use of olfactory cues from conspeci®cs and randomizing these times may lead to an increase in active behaviors for captive cats. Acknowledgements This study was supported in part by a grant from the Institute of Museum Services Conservation Project Support, IC-40044-94. We would like to thank the staff and

78


volunteers of the Caldwell Zoo, Dallas Zoo and Fossil Rim Wildlife Center, for their assistance in helping us with data collection. A special thanks goes to Dr. Jeanette Boylan for statistical and technical advise in writing the paper, and to Ken Kaemmerer for his editorial advice. We would also like to thank the reviewers of this paper.

References Altmann, J., 1974. Observational study of behavior sampling methods. Behaviour 49, 227±267. Beltran, J., Delibes, M., 1994. Environmental determinants of circadian activity of free-ranging Iberian lynxes. J. Mammal. 75, 382±393. Carlstead, K., 1996. Effects of captivity on the behavior of wild mammals. In: Kleiman, D., Allen, M., Thompson, K., Lumpkin, S. (Eds.), Wild Mammals in Captivity, University of Chicago Press, Chicago, pp. 317±333. Carlstead, K., Brown, J., Seidensticker, J., 1993. Behavioral and adrenocortical responses to environmental changes in leopard cats (Felis bengalensis). Zoo Biol. 12, 321±331. Caso, A., 1994. Home range and habitat use of three neotropical carnivores in Northeast Mexico. MS Thesis, Texas A&M University, Kingsville, TX, 83 pp. Crawshaw, P., 1995. Comparative ecology of ocelot (Felis pardalis) and jaguar (Panthera onca) in a protected subtropical forest in Brazil and Argentina. Dissertation, University of Florida, Gainesville, FL, 190 pp. Emmons, L., 1988. A field study of ocelots (Felis pardalis) in Peru. Rev. Ecol. Terre Vie. 43, 133±157. Emmons, L., 1989. Ocelot behavior in moonlight. In: Redford, K., Eisenberg, J. (Eds.), Advances in Neotropical Mammalogy, Sandhill Crane Press, Gainesville, FL, pp. 233±242. Harwell, G, Siminski, D. (Eds.), 1990. Listed Cats of Texas and Arizona Recovery Plan (With Emphasis on the Ocelot). USFWS, Albuquerque, NM, 131 pp. Hediger, H., 1950. Wild Animals in Captivity. Butterworths, London, 201 pp. Kitchener, A., 1991. The Natural History of the Wild Cats. Comstock Publishing Associates, New York, 280 pp. Kleiman, D., 1994. Animal behavior studies and zoo propagation programs. Zoo Biol. 13, 411±412. Konecny, M., 1989. Movement patterns and food habits of four sympatric carnivore species in Belize, Central America. In: Redford, K., Eisenberg, J. (Eds.), Advances in Neotropical Mammalogy, Sandhill Crane Press, Gainesville, FL, pp. 367±390. Laack, L., 1991. Ecology of the ocelot (Felis pardalis) in South Texas. MS Thesis, Texas A&I University, Kingsville, TX, 113 pp. Law, G., MacDonald, A., Reid, A., 1997. Dispelling some common misconceptions about the keeping of felids in captivity. Int. Zoo Yearb. 35, 197±207. Ludlow, M., 1986. Home range, activity patterns, and food habits of the ocelot (Felis pardalis) in Venezuela. MS Thesis, University of Florida, Gainesville, FL, 70 pp. Ludlow, M., Sunquist, M., 1987. Ecology and behavior of ocelots in Venezuela. Natl. Geogr. Res. Rep. 3, 447± 461. Mansard, Pat., 1990. Oestrous cycles and oestrous behavior in the ocelot (Felis pardalis), Ratel 17 180±183. Mellen, J., 1989. Reproductive behavior of small captive exotic cats (Felis spp.). Dissertation, University of California, Davis, CA, 161 pp. Mellen, J., 1991. Factors influencing reproductive success in small captive exotic felids. Zoo Biol. 10, 95±110. Mellen, J., 1993. A comparative analysis of scent-marking, social and reproductive behavior in 20 species of small cats. Am. Zool. 33, 151±166. Navarro, D., 1985. Status and Distribution of the Ocelot (Felis pardalis) in South Texas. MS Thesis, Texas A&I University, Kingsville, TX, 92 pp. Nowell, K., Jackson, P. (Eds.), 1996. Wild Cats: Status Survey and Conservation Action Plan. IUCN, Gland, Switzerland, 382 pp. Powell, K., 1997. Environmental enrichment programme for ocelots at North Carolina Zoological Park, Asheboro. Int. Zoo Yearb. 35, 217±224.


79

Rushen, J., Lawrence, A., Terlouw, E., 1993. The motivational basis of stereotypies. Stereotypic Animal Behaviour: Fundamentals and Applications to Welfare, CAB International, Wallingford, pp. 41±64. Shepherdson, D., Carlstead, K., Mellen, J., Seidensticker, J., 1993. The influence of food presentation on the behavior of small cats in confined environments. Zoo Biol. 12, 203±216. Stevens, E., Hutchins, M., 1993. In: Proceedings of AAZPA Omaha Annual Conference on the Applying Behavioral Research to Zoo Animal Management Workshop: An Update, 12±16 September, pp. 41±45. Tewes, M., 1986. Ecological and Behavioral Correlates of Ocelot Spatial Movements. Dissertation, University of Idaho, Moscow, ID, 128 pp. US Department of the Interior, 1992. CITES Appendices I, II and III to the Convention of International Trade in Endangered Species of Wild Fauna and Flora. Wechsler, B., 1991. Stereotypies in polar bears. Zoo Biol. 10, 177±188.