Dr. Moore is an animal behaviorist and wildlife biologist (CWB, The Wildlife Society) who has helped to renovate and manage small zoos for over 25 years. He is very interested in "behavioral enrichment" and serves the AZA as vice-chair of its Animal Welfare Committee, chair of the mammal standards committee, chair of the Bear Advisory Group, and member of the Behavior Advisory Group, and several taxon advisory groups including Bear TAG. He currently serves the World Conservations Union on the Deer Specialist Group, Reintroduction Specialist Group, and Small Carnivore Specialist Group.

_Case study. Gus, a male polar bear at the Central Park Zoo, was infamous for his stereotypic behavior. Moore chronicled the efforts made to decrease Gus

In 1994, the media observed Gus pacing, and the public grew more vocal and concerned about his welfare. In 1998 the Zoo became more proactive in its attempts to solve the problem, and put more soft substrates and frozen food toys into the exhibit. Gus continued to pace.

The Zoo then tried different types of enrichment, such as sprayers, hay, logs, and male fox scent. None of the enrichment trials produced the desired result—in fact, Gus' pacing increased 33% with the log and 121% with the male fox scent.

In 2002 the Zoo created a new polar bear exhibit, The Arctic Stream. Still, Gus displayed no significant decrease in stereotypic behavior.

In 2003, Gus was given almost 24/7 access to the back den. This change seems to have resulted in a no-pacing trend. In this particular case, access to the den seems to have made the most significant difference; other enrichment did not.

Recommendations:

• Offer diverse, science-based enrichment.
• Use consistent techniques for evaluation.
• Consistently evaluate program results.
• Publish results.
• Readjust enrichment as needed.

Stereotypic behavior in polar bears has several functions. It increases stimulation in a boring environment, decreases arousal in a stressful environment, provides a response to frustration (for example, when a specific behavioral need such as mating is stymied), and satisfies the motivation to be active.

As defined for this study, stereotypic behavior is "a pattern of locomotor behavior that is repeated in an identical fashion in the same place and observed on five or more occasions." Although stereotypic behavior is known to occur in some captive polar bears, is repetitive behavior an indication of compromised welfare? What factors predispose a polar bear to engage in stereotypic behavior?

The study's objectives were to describe stereotypic behavior, investigate the relationship between physiologic measures and stereotypic behavior, and understand the effect of environmental and husbandry methods on stereotypic behavior. The sample size was 55 polar bears (24 males and 31 females) in 20 partner zoos; this represented close to 50% of the North American captive polar bear population.

To assess animal well-being, the sample population was monitored for a period of three years. Stress response was gauged by measuring the secretion and excretion of glucocorticoids (blood cortisol). The study also included behavior and temperament assessments via standardized behavioral tests and keeper questionnaires.

Summary of results:

• Of the bears in Shepherdson's study population, some exhibited high rates of stereotypical behavior while others had almost none. There is some evidence linking stereotypical behavior to stress, but results are not conclusive.
• Of the bears that did exhibit stereotypical behavior, there seemed to be a seasonal effect that was sex-dependent: Males engaged in stereotypical behavior more in the first two quarters of the year, while females engaged in stereotypical behavior more in the first and fourth quarters of the year.
• The study found no correlation between stereotypical behaviors and environmental factors such as den access, exhibit complexity, or wild-caught versus captive—except that larger pool surface area (not volume) has a significant positive effect.
• The more bears in a social group, the less stereotypic behavior. The more females in a group, the less average time spent stereotyping.
• When bears were rated on behavior and interaction, the more-dominant males and more-dominated females were found to have higher corticoid levels.
• The addition of numerous daily enrichment activities decreased stereotypical behavior in females. A formal training program based on positive reinforcement also reduced stereotypical behavior in females. These same trends were seen in males, but they were not statistically significant.
• Increased enrichment seems to decrease inactivity and increase diversity of behaviors such as biting, licking, carrying, and manipulation.
• Most-effective types of enrichment (in order of greatest effectiveness): food-related enrichment, discs, bone, bucket, browse, and ice block. Balls came out fairly low, although they are one of the most popular enrichment items in use.
• Husbandry is the key to decreasing stereotypical behaviors.

Future recommendations:

• Perform further experimentation on husbandry factors.
• Focus on improving the effectiveness of husbandry.
• Give more attention to bears' social grouping and interaction.

The objectives of this study were to: 1) document baseline pacing and stereotypic behavior, 2) speculate about the cause or onset of such behavior, 3) document the types of enrichment that might eliminate the behavior, and 4) document the effectiveness of enrichment in decreasing stereotypic behavior. Bucciarelli's hypothesis was that enrichment should decrease stereotypic behavior and pacing.

Bears were observed for 15 minutes per hour, in one to seven sessions per day. At the beginning of the study, the male bear paced 29%, and engaged in other stereotypic behavior 38.7% of the time observed. The female paced 6% and engaged in stereotypic behavior 7.7% of the time observed. The male bear was off public display 16% of the time, while the female bear was off display 21.8% of the time.

During the four distinct phases of the study, a number of factors were examined. Phase 1, the baseline phase, allowed observers to document the baseline behaviors of each bear. In Phase 2, the bears were allowed to move from exhibit area to back dens (off display) any time they chose, through an always-open access door. In Phase 3, fish popsicles were given as enrichment items. In Phase 4, bears were taken off display.

Upon the completion of all four phases, Bucciarelli found that none of the environment and enrichment modifications had had a desirable effect on the male bear. She observed some short-term desirable effects (that is, reduced stereotypical behavior) on the female bear, when fish popsicles were offered as enrichment and when the female was allowed to go off display.

At the conclusion of the study, the polar bear exhibit was physical restructured, with new foraging pits, underwater boulders added to the pool, and a newly installed water-filtration system. After those exhibit changes, the male bear reduced his pacing significantly, and reduced other stereotypic behaviors by 50%.

_Case study. Cronin presented the results of her research into the stereotypic behavior of 14 polar bears at 7 zoos in the 1990s, and documented the effect of a new exhibit design on one zoo

At the time of this study, all 14 polar bears exhibited stereotypic behavior—independent of age, sex, season, zoo, or captive-born versus wild-caught. Polar bear enclosures at the subject zoos consisted of concrete, rock, and water. Some zoos had "starve" days two to three days per week, in an effort to simulate conditions in the wild. Bears still had olfactory food-related cues around them, due to the presence of concession stands, and exhibited increased stereotypic behavior on starve days.

The zoos tried various techniques to decrease stereotypic behavior. A mechanical feeder that allowed bears to eat when they wanted increased aggression by 29%. Giving bears a large variety of different feed items was somewhat better. When they began putting food inside various items to increase cognitive stimulation, the bears participated in object manipulation and play—independent of age or sex. Males had more play sessions of shorter duration than females; females had longer, more-complex play sessions in which they combined objects. Result: the more active the bear, the less stereotypic behavior.

Exhibit Modification & Design
Cronin recommends that we abandon preconceived ideas about what the Arctic is like. She referenced the summer climate and flora of the Hudson Bay region, and suggested that Hudson Bay tundra and beach habitats might better suit the needs of captive bears. Hudson Bay females head 30 miles inland to get away from males and dig soil dens. On the coastline, males forage for grass and kelp, and build large daybeds in sand.

One of the zoos she studied created a new enclosure based on the tundra/coastal model exhibited in the Hudson Bay region. In the new enclosure, levels of feeding and foraging increased. Visual barriers allowed males and females to separate from each other, and natural substrates were incorporated. (Cronin recommends that zoos consider natural substrate for inside dens as well.) With the new enclosure, all stereotypic behavior decreased—and the male stopped his repetitive behaviors entirely.

Stereotypic behaviour: In 78% of the responding zoos at least one animal stereotyped. As most elements of stereotypic behaviors are also part of normal appetetive behaviors, it can be assumed that the animals are "searching for something". In 50% of the locations the behavior correlates with the time prior to feeding. Enrichment "with feeding value" reduced the stereotypic behavior of the bears at Cologne zoo, but could not prevent its development in a male cub. After removal of optical and olfactory stimuli advertising the food the problem could be avoided in the next cub. In European zoos more females than males exhibit the behavior. In both sexes the proportion of affected animals is higher in harems compared to pairs, the difference is most pronounced in males, indicating a that social stress might be involved. Social stress was the major factor triggering stereotypic behaviour in the lowest ranking female at Cologne zoo. The behaviour dropped immediately from 53% to 11% after removal of the other two females into the dens.

Skin problems: The extent of skin problems is much lower, but if they occur they are more prevalent in the presence of stereotypic behavior. Low rank and large group size seem to favor the development of skin problems. 67% of the zoos with harems report the problem compared to 28% keeping just a pair. Here social stress seems to be a component too.

Reproduction: Of 67 total adult females, 52% had successfully mated, while 28% had successfully reared cubs. Looking at social grouping and reproduction, only 48% of the females in harems gave birth, whereas 71% of those kept in pairs produced cubs. Successful rearing did not differ between females in harems and in pairs.

Conclusions: there is evidence that stress - either related to feeding and/or to grouping patterns - contributes to the development of stereotypic behaviors and skin disorders and the low reproductive output in captive polar bears. It can be assumed that polar bears are adapted to the strong seasonality of food availability in the wild by seasonally varying feeding motivation and set points for food intake, which persist in captivity and are not necessarily met by constant feeding regimes. Outside of the mating season the sexes segregate and the adult females particularly those accompanied by cubs actively avoid males. In captivity they are forced to social life and might miss opportunities for retreat which result in search for places distant from conspecifics.

Solutions:

• Aim at much larger enclosures
• Offer a lot of retreat and separation opportunities
• Consider distances when planning the distribution of resources
• Consider seasonality in feeding
• Consider individual distances when scattering or hiding food
• Separate breeding from maintenance facilities

With the opening of its state-of-the-art polar bear exhibit in 1996, San Diego Zoo bears exhibited pacing, illness, and an onslaught of national media attention. When the Zoo received two very young, orphaned cubs and those cubs, too, began to exhibit early stereotypic behavior, Zoo professionals decided to test new behavioral management strategies on them.

When the cubs began showing early signs of stereotypic behavior, Zoo staff came up with a theory they wanted to test. In the wild, cubs faced with novel stimuli show a startle response and initial stress reaction, followed by bonding or reassurance from their mother. Were zookeepers coddling them too much, blocking the learning of coping skills, and unintentionally reinforcing the cubs' stress-related reactions? Were they entertaining the bears too much, and not helping them entertain themselves?

The cubs were crate trained. Trainers encouraged their natural curiosity and inquisitiveness, but taught cubs to make the connection that their behavior influenced whether they got what they wanted. Trainers provided no food enrichment.

Trainers created situations that would startle the cubs, in the hopes they would learn to develop their own coping skills. For example, trainers introduced the cubs to large vehicles because those are often driven around the exhibit area. The cubs reacted to the novel stimuli with a lip jut, and then investigated the situation, with each cub showing individual differences in how they chose to explore. Keepers reinforced that investigative behavior and then reassured the cubs.

Results:

• Increased time in play behavior, from less than 10% to more than 20%
• Rate of stereotypic behavior decreased from 45% in 1997 to less than 0.08% in 2003
• Percentage of time engaged in stereotypic behavior decreased from 25% to 0%

Factors for success:

• Keeper-polar bear interactions are the most important factor.
• Everything is consistent on the bears' correct behavior.
• Environment is kept very calm.
• No more coddling or protecting from environmental noises.
• More training sessions were instituted.

Can we apply some of these same principles to older bears that exhibit stereotypical behavior?

Stereotypical behaviors arise when the animal is motivated to perform a behavior or reach a goal that is somehow restricted. They are repeated because the animal is trying to do something but is unable to (for example, explore, forage, seek a mate, roam, or escape other bears). Other factors responsible for stereotypical behavior are stress, lack of stimulation, and lack of control.

Stereotypic behavior may vary with the seasons, and also varies with the individual. The more frequent the behavior, the less variation in how the bear performs it. More severe stereotypical behavior decreases less with enrichment.

In treating stereotypies, it's important to identify the source behavior or goal—that is, what the bear is trying to do, but cannot—taking clues from the stereotypy itself (for example, its form, location, or timing). Tips for treating stereotypies:

Treat as early as possible.
Treat the CAUSE of the behavior—don't expect random enrichment to decrease it.
Reduce the sterotypy's underlying motivation, by removing the motivating stimulus, enabling the bear's desired behavior or goal, enabling a substitute behavior or goal, or substituting other distractions.
Fully assess the effects of changes made.
Apply one enrichment at a time. Give changes time to work, or to stop working.
Publicize your findings, even negative ones.

In June 1996 the North Carolina Zoo did a formal observational study of its three polar bears (2 males, 1 female) and their stereotypical behaviors. After hearing that some zoos were giving their bears open access to the holding area as an attempt to reduce stereotypies, the Zoo tried giving its bears open access to the holding area.

Result: 24/7 access decreased stereotypical behaviors significantly. The males' stereotypies decreased by 62 and 66%, while the female's decreased by 68%. In addition, the bears' object manipulation increased, and social interaction increased slightly.

When the Zoo later made minor pool modifications—smoothing rough surfaces and a cobbled beach area, and turning off a loud waterfall—stereotypes decreased even more. In addition, the bears exhibited increased social interaction.