Lessons Learned

A new stress-monitoring system for the giant panda

Our noise effect studies motivated our interest in examining other potentially stress-inducing environmental challenges. We also realised that our initial methods had several limitations. For example, although the duration, pitch and amplitude of noise were known, we did not always have an idea of the source. Our data collection also was not ideal. Our behavioural observations were based on only a few hours of observation on each panda each week. Although keeper records provided useful information about what occurred outside our observation period, it was not recorded in a manner that allowed statistical analysis. Therefore, we decided to explore a better monitoring system which would capitalise on the observations of animal care staff that work with and observe the pandas intensively each day. Our goal was to provide a way for the keepers to record observations quantitatively as they went about their daily routines which, in turn, would allow statistical evaluations of environmental influences on animal well-being.

We also concluded that this new approach needed to be useful across institutions while addressing a much larger set of factors that may cause stress and aversive reactions. We first made a list of independent variables with the potential of causing stress, including:

• novel or strong odours (e.g. vehicular exhaust);

• exposure to conspecifics;

• aggressive interactions with conspecifics;

• new enrichment items;

• interactions with keepers;

• conditioning/training;

• veterinary procedures;

We also identified a host of husbandry variables, including enclosure modifications, confinement in different or smaller portions of customary enclosures, diet modifications and schedule changes (e.g. time of feeding). We recognised that the list of behavioural responses to these potential stressors was similarly diverse. For example, a panda may consume less bamboo, or it may vocalise, pace, increase general movement or engage in more abnormal, stereotypical or door-directed behaviour. Finally, the panda may attend to, retreat or hide from the disturbance or adopt some other idiosyncratic response. As an example of the latter, a 12-year-old female at San Diego Zoo (SB 371), when troubled, would typically climb a tree, 'huff' and bounce up and down while looking at the source of the disturbance.

The new monitoring system has been designed to reflect two discrete modes of the stress response, both of which require the use of a detailed ethogram. The first deals with behaviours that may change immediately upon stressor exposure. In this fight-flight response (Selye, 1946), the HPA may not be activated and, thus, adrenal corticoid production may not change substantially. Similarly, behavioural activity may quickly return to baseline if the disturbance is not protracted or if accommodation occurs rapidly. The second assessed mode involves situations when a disturbance continues unabated, with the animal's activities affected for hours or days, accompanied by a concomitant rise in corticoids. An ability to monitor both types of response simultaneously is critical to understanding what environmental dimensions cause aversion. While a record-keeping system has been created for caretakers to track environmental changes and the behavioural responses that could be provoked, urine samples continue to be collected twice daily to be analysed for corticoids as a reflection of physiological status and adaptive ability.

Experiences from daily monitoring of stress and well-being

Keepers record all changes in the panda's environment and its behaviour and health on a day-to-day basis. It is important to record this information even when no environmental changes are evident and when pandas show no signs of behavioural agitation. For example, a keeper may note that the animal appeared 'anxious' today, and that 'there were many visitors'. But by making this same anxiety observation on another day when there are few visitors, then it will eventually become possible to test statistically the impact of visitor numbers on giant panda 'anxiety'. The success of this approach, however, requires adherence to specific definitions of observable behaviours. The giant panda ethogram used by the authors and with operational definitions and guidelines for recording frequency or severity is depicted in Table 12.1.

Throughout the day, one designated keeper observes the panda's behaviour and makes notes on different behaviours from the etho-gram. At the day's end, the keeper estimates the amount of time spent monitoring each animal's activity, a reflection of actual awareness of that person for what that particular animal was doing on a given day. Once this time-frame has been established, the keeper records the approximate percentage of this time the panda spent engaged in behaviours defined in the ethogram. Directions are provided to guide the keeper's rough estimates (see Table 12.1). In conjunction with the keeper's detailed records, these data are later analysed to determine if changes in the environment have an impact on stress and well-being. For example, data may reveal that substantial changes to the enclosure increase pacing, that changes to the feeding schedule promote restless door-directed behaviour or that mucous stools are associated with elevated honking vocalisations. By contrast, it may be determined that novel odours and the number of visitors have no effect.

Monitoring acute responses to disturbances

Table 12.2 illustrates some of the variables being tested at the San Diego Zoo and Ocean Park-Hong Kong to assess a giant panda's acute response to disturbance - a sudden change in environment. An animal may

Table 12.1. Examples of behaviours recorded in the daily monitoring ethogram for the giant panda

Behaviour Definition

Appetite-bamboo Consumption of provisioned bamboo. If food is weighed, record weight of consumed food in kg. If not, estimate percentage of normal diet consumed. For example, if animal eats half of its normal amount, record '50%' Back and forth, or perimeter, travel in a repetitive, sustained, predictable pattern. Animal follows the same or a similar path repeatedly, using only a small portion of the space available for locomotion Locomotion in an anxious manner which is reminiscent of pacing but does not meet the repetitive, predictable pattern required for pace or quasi-pace Panda at a door or gate or any area where food or keeper interaction may occur. Panda's behaviour is oriented toward food, keeper or gate; waiting with mild to extreme restlessness. Behaviours may include sniffing, pawing or scratching, pushing, head butting or other manipulations A short, tonal, nasal call, falling in pitch. Honking almost always occurs in a series lasting at least several minutes Participation when shifting from one enclosure to another or during training sessions, or less formal keeper interactions. Unlike other behavioural measures in this section, rate degree of uncooperativeness on a scale from 0 to 3; 3 is most cooperative become startled and run in response to a loud noise or appear agitated and move away from a noisy, large group of people near the exhibit. Again, the keeper (or other observer) can record the exact nature of the disturbance according to defined categories, as well as the severity of response on a scale of 0 to 3 (Table 12.3).

Such an approach adds a more detailed and temporal dimension to data collection. Furthermore, it has the potential of identifying the precise source of the disturbance while pairing it with the immediate behavioural response, thus providing better information on cause and effect. It also allows us to 'catch' ephemeral responses. For example, a brief honking bout after a visitor waves his arms may not influence honking rate for that day, yet the fact that it occurred suggests that it


Anxious walk

Door-directed behaviour



Table 12.2. Examples of environmental disturbances or husbandry changes recorded for monitoring responses to acute disturbances



Strong novel odours Enclosure restrictions Activity of visitors Monitoring noise with sound meter Subjective noise monitoring

• Approximation of pitch

• Approximation of amplitude

Mucous stool

Exposure to conspecific

Describe odour or known source. Sources may include fresh paint or cleaning products Panda restricted to a smaller portion of the enclosure than normal

Visitor(s) wave arms, jump up and down, lunge forward, move rapidly or throw objects Methods vary according to equipment used

Evaluate noise level based on his/her experience during course of the day. Describe noise events and note approximate duration If noise source is outside the panda facility (e.g.

construction), the proximity of the disturbance is noted Low Examples: large truck or deep human voice Medium Example: group of adults talking in normal speaking voices High Examples: shrill, piercing, metal grinding against metal, a small dog barking or a small child's scream

Quiet Not louder than normal speaking voice Moderate: Not louder than voices in a heated argument (not quite yelling) Loud Ambient noise such that it would be difficult to hear someone talking at a normal speaking voice Panda passes faeces containing large quantities of mucus, often accompanied by signs of physical discomfort Any substantial change in sensory access to other panda individuals. Examples include exposure to the odours of another panda, visual access through cage bars, visual access with limited physical access through cage bars or direct physical access in the same enclosure. Indicate frequency with which panda is subjected to this exposure may have been an important event. Similarly, an animal retreating from a noise disturbance would not be recorded under the daily monitoring system. The acute response system does not, however, allow us to determine whether a panda displays a chronic stress response, has a

Table 12.3. Examples of a behavioural ethogram for the response to acute disturbance

Behaviour Definition/scoring

Attends Panda orients towards the disturbance

0 = Not observed

1 = Animal calmly focuses attention in the direction of the disturbance (e.g. sniffs air or looks toward disturbance). May be brief or last several seconds

2 = Animal intensely focuses attention in the direction of the disturbance (e.g. eyes riveted). May stand or otherwise reorient body position. Posture may appear rigid, and response will be held for at least several seconds

3 = Animal is focused on, and moves towards, the disturbance in a deliberate manner (at least two body lengths or approaching the limit of the enclosure)

Hides Behaviour that places visual barrier between panda and source of disturbance

0 = Not observed

1 = Animal covers eyes/head with paws

2 = Animal withdraws to a corner or crouches behind available cover, such as a log pile that offers a partial visual barrier

3 = Animal withdraws to another space, including behind a full visual barrier (e.g. wall or enters another room)

Retreats Behaviour that increases distance between panda and source of disturbance

0 = Not observed

1 = Animal moves less than three body lengths away in a calm manner

2 = Animal moves greater than three body lengths away in a calm manner or less than three body lengths in rapid fashion

3 = Animal moves greater than three body lengths away from disturbance rapidly or climbs a tree or other vertical structure delayed response to a stressor or if it responds to less abrupt environmental changes. The two methods together offer complimentary information and a more complete picture of the potential effects of stress on well-being in captivity as well as simply more biological information about this fascinating species.

Our data collection for the 'acute response system' is not yet complete, but we foresee its potential to provide concrete answers to questions about which aspects of the typical zoo environment pandas may find aversive. Examples of potential findings that these data collection protocols could produce include:

1. A loud noise (especially at a high pitch) commonly startles a giant panda causing it to retreat and hide, often followed by pacing and/or honking.

2. A giant panda often honks after an aggressive interaction with a conspecific.

3. A giant panda often attends to arm-waving visitors but shows no other overt response.

4. Most episodes of stereotypic behaviour are in response to groups of large visitors (i.e. more than 100 people).

Preliminary findings from Ocean Park-Hong Kong are already providing useful feedback for making management changes. The giant panda enclosures at this facility are entirely indoors, primarily for temperature regulation. While this allows the animals to enjoy the benefits of appropriate temperature and decreases external noise, indoor crowd noise may be amplified. Although we have found no dramatic physiological evidence indicating severe distress, factors such as husbandry changes, noise, crowd size and illness appear to be associated with slightly elevated levels of stress-related behaviours throughout the day. More intensive monitoring has helped identify the source of noise disturbances. For example, on four occasions, this particular panda pair responded to high-pitched, moderately loud drilling from inside the facility with increased pacing, restless behaviour at the door leading to the night quarters and, in one case, immediate rapid retreat from the disturbance. Similar responses were also observed following low-pitched, medium-amplitude noises from outside the facility.

0 0

Post a comment