Director of Animal Behavior
In the continuing quest to provide optimal care for animals in captivity, the issue of abnormal behaviors is a cause for concern. Whether we call them stereotypic, neurotic, nonadaptive, or atypical, behaviors of this kind are problematic. The co n sequences of abnormal behaviors range from unpleasant sights for the zoo visitor, as in the case of coprophagy in a resident gorilla, to serious health problems for the animals themselves, such as self-biting in a singly housed rhesus macaque. Animals e x hibit abnormal behaviors in a variety of social and environmental contexts: naturalistic and purely functional, large social groupings and singly housed, enriched and sterile. No zoo, laboratory, or breeding facility, it would seem, is exempt. Neither i s any species. One approach to dealing with abnormal behavior is through the use of behavioral management techniques. Behavioral management refers to the combined use of positive reinforcement training and environmental enrichment techniques.
A variety of strategies to address abnormal behavior have been reported, and they tend to fall into three main categories: feeding manipulations including types of food and methods of feeding (Ruempler, 1991; Bloomsmit h , 1988); environmental manipulations and provisioning of toys and apparatus (Bryant, et al., 1988; Fried, et al., 1993); and social manipulations (Reinhardt, 1987). These various strategies have produced mixed results. Unfortunately, even in the best case scenarios, abnormal behaviors are most often reduced, not eliminated. References to the use of training as an intervention for abnormal behavioral problems are still fairly rare (Morgan, et al., 1993; Kirtland, 1993).
There are good reasons why efforts to reduce or eliminate abnormal behavior have limited success. First, the causes of abnormal behavior are often subtle and complex, and tracking down how and why the problem began can be a task worthy of Magnum P.I. Second, the relationship between real or perceived stressors and abnormal behavior can result in the behavior serving as a functional coping strategy for the animal (Gould and Bres, 1986). Once utilized as coping strategies, these beh a viors are self-reinforcing and extremely tenacious in nature. Finally, problem behaviors often occur when people are not present, limiting the opportunity for direct intervention.
So, what benefits can behavioral management techniques offer personnel dealing with abnormal behavior problems? The greatest benefit they provide is the means to systematically address abnormal behavior and the underlying behavioral issues. In its p u rest form, a behavioral management approach is a practical exercise in the scientific method. The following steps illustrate the process.
Develop a hypothesis. After a careful discovery process, a list of potential causes and contributing factors should be developed. Then, it's possible to make a guess as to why the abnormal behavior is occurring. For example, we could hypothesize that disruptive behavior by a chimpanzee, like throwing feces or trying to grab caregivers, is an attention-seeking behavior, whether the resulting attention is positive or negative. Or, we could guess that social pressure by the dominant male sea lion coupled with a predictable feeding routine are the underlying causes for a low-ranking sea lion regurgitating for extended periods of time after regularly scheduled feeds. The importance of a well-developed hypothesis is that it is the starting point from which your intervention plan is developed.
Identify specific behavioral goals and initiate training and enrichment strategies. With a clear hypothesis, you can design and implement an intervention plan to address targeted behaviors from a behavioral management perspective. For example, by definition, reinforcement increases the likelihood that a behavior will recur. In the case of the primate that utilizes disruptive behavior as an attention-seeking measure, look at the situation and determine where the reinforcement is occurring. More than likely, when the animal uses the disruptive behavior it receives a great deal of attention, probably negative, but attention none the less. Indeed, "displays" put on by frustrated humans who have just been "had" can be quite entertaining. But what is perhaps more relevant is what happens when the animal is not disruptive. Chances are the human walks right by.
According to the hypothesis, human attention is the reinforcement the animal is seeking. So, when are the rewards occurring, and what are the results? Ironically, feces throwing and arm grabbing are being continuously reinforced, while non-aggressive cooperation is not being reinforced at all. The resultant behavior is consistent with the reinforcement pattern. The intervention strategy is then two fold. First, reinforce the animal when it is not disruptive. Stop for a moment, say a few words, offer a small treat or favored toy, reinforce cooperative behavior (fig. 1). Second, do not reinforce the disruptive behavior. Turn around, count to 10, or walk away, and then look for any opportunity to reinforce the desirable behavior. This is a straightforward, simplified example of the problem-solving process. However, it is amazing how often this pattern is repeated in a variety of contexts with similar results.
Fig. 1. Rewarding non-aggressive cooperation helps
eliminate disruptive attention-seeking behaviors.
In the case of the regurgitating sea lion, strategies must be devised to address both contributing factors identified in the hypothesis. First, the socialization problem can be dealt with by utilizing a training technique called "cooperative feeding." Operationally, it entails reinforcing two events simultaneously: dominant animals are reinforced for allowing subdominant animals to work and receive food or attention, while the subdominant animals are reinforced for being "brave" enough to work and accept food or attention in the presence of these more aggressive animals. In this case, training would focus on the target animal and the dominant male (Fig. 2). Second, the feeding schedule should be altered to make it less predictable. If times of feeds are on a set schedule, extra cooperative feeding sessions should be added in-between. Finally, enrichment strategies should be implemented in-between feeds to provide activity options other than regurgitation.
Fig. 2. Training can address socialization issues which are
often contributing causes to abnormal behavior.
Check the results. There are no pat answers or guarantees in dealing with behavior, problematic or otherwise. It is dynamic in nature, and so much of what we do is guesswork. However, there are two processes that increase the likelihood of making the right guesses. First: the quality and extent of information gathered prior to initiating any work--the data collection phase. That, coupled with knowledge of the individual animal, makes a carefully developed hypothesis an educated guess. Second: an on-going evaluation of information you get back from the activity. Is a particular strategy achieving the results that are anticipated? If the answer is yes, it's a good bet to stay with the current strategy. If the answer is no, it's time to reevaluate and perhaps try something else.
Adjust strategies if necessary. It is critical to maintain a realistic expectation of results, so a strategy is not abandoned too quickly, or sustained too long. If that occurs, the result can be a great deal of frustration and confusion for the animal, which may worsen the problem. When one approach has been tried for a sufficient length of time (a critical judgment call) without the desired results, try something else. Then, check results again, and continue to adjust strategies as necessary. In tough cases, it may be necessary to try several different strategies before one works. Or often, it will be a combination of strategies that finally achieves the desired results. That's why good behavioral management skills include a heavy dose of creativity, innovation, and most important, flexibility.
The following are some specific examples of how behavioral management strategies have been used to decrease or eliminate abnormal behaviors.
Over a period of 7 months, a pilot behavioral project was conducted with a group of drill baboons (Desmond, et al., 1987). The primary goal of the project was to increase positive social interactions and reproduction among the group members (fig. 3). There was also concern about a sub-adult male who had been introduced to the group 6 months previously and was shunned bythem. He often appeared stressed, and exhibited some abnormal behavior. Intervention was difficult due to the social dynamics within the group which made him and other subdominant animals inaccessible to the keepers for handling or husbandry purposes.
Fig. 3. Drill grooming behavior increased as a result of training.
A thorough information-gathering process was conducted, which included interviews with relevant personnel and assessment of behavioral observation data conducted on the group for the previous 2 years. Based on this information, a hypothesis was formulated. The hypothesis was that a long-term moderate state of sensory deprivation had existed in the exhibit. This shortage of stimulation had resulted in subtle competition among group members which, in turn, inhibited breeding, other positive social interactions, and interest in interacting with the environment. It also contributed to the presence of problematic behaviors by all group members including fence nibbling, self-biting, and examining and eating feces.
Based on this hypothesis, the following behavioral goals and strategies were developed:
Although reducing abnormal behaviors was a goal, the training strategies were developed to address the underlying causes of these behaviors - insufficient sensory stimulation, and socialization problems. Operationally, the young male was often paired with an adult female for cooperative feeding sessions. She was reinforced for ūstayingū while he was given food and personal attention. He learned to gently touch the trainer's hand and arm, approximating grooming. He slowly became less nervous and agitated when eating with the female, and increasingly more relaxed. Prior to this project, he had been observed biting his leg in a manner and frequency that indicated the potential onset of neurotic behavior. Keepers familiar with the exhibit reported a significant reduction in the observance of the behavior throughout the training project.
Documented results (table 1) showed significant increases in all forms of positive, social interactions during and following the project (Cox, 1987). Reduction in all forms of abnormal behavior was also achieved (table 2).
Note: Although aggressive behavior increased on an absolute level from before training to during training, it dropped from 34% to 25% of the total social interaction.
Another project illustrating the process of addressing abnormal behaviors involved a captive-born Bottlenose dolphin named Pepe (Laule, 1984). Living with a couple of other young animals, two separate attempts were made to integrate him into a larger social grouping of show animals. These attempts were unsuccessful, and in the process he developed several abnormal and problematic behaviors. These included an erratic appetite and attention span; biting people during unstructured play sessions; habitual regurgitation; and chronically swallowing objects that fell into the water.
In researching Pepe's situation and history, several factors were identified which may have contributed to the development of these problem behaviors. First there were health-related problems, including the presence of small ulcers which could have affected his appetite and energy level. He had also sustained an injury to his peduncle area, which caused occasional swelling and which was under constant scrutiny. This was later diagnosed to be osteomyelitis and a source of his chronically high white blood cell counts.
Another factor was his submissive behavior. Whenever he was introduced into the larger social group, despite his greater size, Pepe always aligned himself with the subdominant males and quickly became the lowest ranking animal. In that weak position, pressures were placed on him that he was apparently unable to cope with.
Differences in training regimen from one condition to the other may have also been a factor. These changes included the loss of his one primary trainer and subsequent replacement by four new individuals. Behavioral charts indicate that in the second condition he was worked inconsistently, less often, and with less challenging work. In a period of almost 2 years, he learned only one new behavior while losing several others. He also received less personal attention.
One last factor concerns the age at which Pepe was separated from his mother, and the impact that may have had on his subsequent development. Compared to the other five dolphins born at the park, Pepe's separation from his mother, at 18 months, was at a substantially earlier age. The average age of the other five animals was 29 months, with the youngest being 26 months and the oldest 33 months.
Although I found no definitive research or conclusions on optimal age of separation, Herman notes, in his book Cetacean Behavior, "Close affiliation between the newborn and mother continues for an extended period of time and dependency may persist even into adulthood" (Herman, 1980). Whether this had an impact on Pepe's later problems is uncertain. Nonetheless, it is noteworthy, in light of the extensive research done by John Bowlby on early separation of young human children from their mothers and the far- reaching effects this has on personality and behavioral development (Bowlby, 1973).
Once the potential causes and contributing factors were identified, a variety of strategies were employed to address each of the problematic conditions. Because of his delicate health, he was the first dolphin to be trained to present his tail flukes for blood sampling, and to accept a stomach tube and fecal tube insertion for sample collection. With his voluntary cooperation, it was easier and less stressful to perform these procedures and monitor his health on a regular and frequent basis.
To help stabilize his eating habits, and to curb his throwing up, different feeding schedules were employed. For a period of 2 weeks, Pepe was fed twice nightly to increase his appetite and weight. Little change was noted. Next, a schedule of eight feeds per day was begun which continued for approximately 2 months. His normal diet was fed in small amounts over the course of the day, with at least three of the feeds being training sessions. Accurate charts were kept, listing the time of day, number of pounds fed, and the amount, if any, of regurgitation seen. This provided a clear picture of Pepe's eating and regurgitating habits, and the evidence of change when it did occur. Specific intervention for the regurgitation included using a verbal "no" and short time out when Pepe would regurgitate during or after his feeds. Extra time was spent with him immediately following a feed, during which time he was rewarded with attention and play for not regurgitating.
To address both the problem biting and his habit of swallowing foreign objects, desensitization work was initiated to train him to allow us to touch his mouth, tongue, and teeth, without biting. At the same time, he was trained to retrieve safe objects like paper cups or paper towels, and then let us open his mouth and remove the objects from his mouth or throat. Third, with a trainer in the water with him, work was begun with Pepe, reinforcing him for gentle play and non-biting behavior. The reinforcement was high at first, then slowly reduced until reinforcement was no longer necessary at all.
To assist in his socialization, he was specifically worked with each of the dominant animals. Pepe was encouraged and rewarded for participating in these sessions, while the dominant animal was rewarded for allowing him to do so. Conversely, if Pepe did not work, reinforcement was withheld from the other male until he did.
Other behavioral strategies included maintaining consistency of trainers, keeping the number of daily training sessions high, and balancing sessions between individual work and work with other animals. Special care was also taken to provide Pepe with a lot of personal attention (fig. 4).
Fig. 4. Pepe received extra personal attention during interventions to address abnormal behaviors.
Finally, Pepe was moved to a different show area with four other animals. From the first day, consistency in trainers was maintained by having myself or another familiar trainer with him everyday. For the first 2 weeks, we spent all of our extra time with Pepe, just sitting with him, rubbing him down, or playing. For 2 months, we were present during shows and training sets to work exclusively with him. Concurrently, the other animals were reinforced for allowing Pepe to work, and any positive social interactions were reinforced.
The results of these strategies were quite encouraging. Pepe was successfully integrated into the show. Socially, he appeared comfortable, interacting with all of the other animals and developing a strong bond with the female pilot whale, each displaying imitative behavior learned from the other.
His biting stopped completely. He would allow us to open his mouth and remove any objects. Rubbing his mouth and tongue became his favorite tactile behavior. In fact, Pepe's overall responsiveness to people increased tremendously. He would now seek attention and interact gently and non-aggressively. His retrieval work improved, so that he would voluntarily return an item he found to the trainer, or retrieve a specific object we pointed to.
His throwing up almost completely disappeared, with only an isolated occurrence being noted. His appetite and attention span, although occasionally erratic, improved greatly overall. He maintained his repertoire of behaviors and continued to learn others.
The examples I described were ambitious attempts to address and resolve a complex set of problematic behavioral issues. These efforts required an investment of time and effort that may not be practical or possible in many situations. However, what is applicable to every situation is the process. It doesn't have to be complicated and tedious, but to some degree it has to be done. Abnormal behavior is not a simple problem, and there are rarely simple solutions. The greatest success in dealing with abnormal behavior will come from addressing the causes of the problem, not just the problem behavior itself. With that approach, there are often simple things that can be done to positively impact the situation to some degree (Bayne, et al., 1993).
Behavior is an acknowledged indicator of well-being (Petto et al., 1990). When we strive to provide optimal care for captive animals by providing for their physical and psychological well-being, reducing or eliminating abnormal behavior is an issue that cannot be overlooked or shortchanged.
For further information contact:Active Environments Inc.
Bayne, K., Dexter, S., Strange, G. (1993). "The Effects of Food Provisioning and Human Interaction on the Behavioral Well-being of Rhesus Monkeys (Macaca mulatta)." Contemporary Topics (AALAS) 32(2):6-9.
Bloomsmith, M., Alford, P., Maple, T. (1988). "Successful Feeding Enrichment for Captive Chimpanzees." American Journal of Primatology 16:155-164.
Bowlby, J. (1973). "Attachment and Loss." In Volume 2, Separation. Basic Books Inc., New York.
Bryant, C., Rupniak, N., Iversen, S. (1988). "Effects of Different Environmental Enrichment Devices On Cage Stereotypies and Autoaggression In Captive Cynomolgus Monkeys." Journal of Medical Primatology 17, 257-269.
Cox, C. (1987). "Increase In the Frequency of Social Interactions and the Likelihood of Reproduction Among Drills." Proceedings of the American Association of Zoological Parks and Aquariums Annual Conference, Portland, OR.
Desmond, T., Laule, G., McNary, J. (1987). "Training for Socialization and Reproduction with Drills." Proceedings of the American Association Zoological Parks and Aquariums Annual Conference, Portland, OR.
Fried, J., Bennett, C., Lindsey, S. (1993). "Assessment of the impact of systematic environmental manipulation on the behavior of 2.2 Western Lowland Gorillas." Paper presented at the Environmental Enrichment Conference, Portland, OR.
Gould, E., Bres, M. (1986). "Regurgitation and Reingestion in Captive Gorillas: Description and Intervention." Zoo Biology 5:241-250.
Herman, L. (1980). Cetacean Behavior: Mechanisms And Functions. John Wiley & Sons, New York.
Kirtland, J. (ed.) (Spring, 1993). "Trainers Forum." Soundings, Newsletter of the International Marine Animal Trainers Association.
Laule, G. (1984). "Behavioral Intervention in the Case of a Hybrid Tursiops sp." Proceedings of the International Marine Animal Trainers Association Annual Conference, Los Angeles.
Morgan, L., Howell, S., Fritz, J. (1993). "Regurgitation and Reingestion in a Captive Chimpanzee (Pan troglodytes)." Lab Animal, 22(8):42-45.
Petto, A., Novak, M., Fingold, S., Walsh, A. (1990). "The Search for Psychological Well-Being in Captive Nonhuman Primates: Information Sources." Science and Technology Libraries 10(2):101-127.
Reinhardt, V., Houser, W., Eisele, S., Champoux, M. (1987). "Social Enrichment of the Environment With Infants for Singly Caged Adult Rhesus Monkeys." Zoo Biology 6:365-371.
Ruempler, U. (1991). "Behavior Modification of Lowland Gorillas at the Cologne Zoo." Gorilla Gazette, 5(2).
Top of Document
About the Animal Welfare Information Center Bulletin
The Animal Welfare Information Center
U.S. Department of Agriculture
Agricultural Research Service
National Agricultural Library
10301 Baltimore Ave.
Beltsville, MD 20705-2351
Phone: (301) 504-6212
FAX: (301) 504-7125
E-mail: Contact us: http://www.nal.usda.gov/awic/contact.php
Policies and Links