Areas Of Investigation

Captive environments and abnormal behaviours

Suboptimal captive environments for wildlife are associated with abnormal behaviours, stress and poor reproduction (Mason, 1991; Carlstead & Shepherdson, 1994; Shepherdson et al., 1998). Besides often being too small, more importantly zoo enclosures lack complexity and opportunities to perform natural behaviours. Unlike wild counterparts, the captive animal has little control over its environment. Attaining 'natural' goals of food and shelter are not required in zoos (e.g. the animal has no mandate to work for food), so there is no need to perform normal behaviours. Additionally, while avoidance behaviours are critical to surviving in nature, a captive animal may have few options to avoid unpleasant situations, such as crowds or noise. As a result, well-being may be compromised.

A prevailing theory, which helps to explain the motivational basis of suboptimal animal well-being in captivity, is the ethological needs model of motivation (Hughes & Duncan, 1988). This model proposes that animals are motivated not only to obtain important resources such as food and shelter, but also to perform the behaviours that have evolved within the species. Thus, an animal given a bowl of processed food fails to use its evolved behavioural strategies to search for, extract and handle food for ingestion. The lack of opportunity to perform these behaviours may lead to poor psychological well-being and overt abnormal behaviours. Feeding-related behaviours appear to be among the most strongly motivated, and a disproportionate amount of behavioural anomalies are observed while an animal spends time waiting for food delivery (Falk, 1977). A major class of abnormal behaviour is called stereotypies - highly repetitive behaviours that do not vary in form and have no apparent function or goal (Mason, 1991).

According to our empirical evidence (Swaisgood et al, 2001), the ethological needs model of motivation fits the giant panda. It is widely acknowledged that this species has been kept in suboptimal enclosures at many breeding facilities. Interestingly, the giant panda's rather simplistic ecology and behaviours in nature (e.g. solitary, few predators, dietary specialist) should make abnormal behaviours unlikely (Swaisgood et al., 2001). On the contrary, pandas in captivity tend to display a high incidence and impressive array of stereotypies (Swaisgood et al, 2001; Lindburg et al., in press). Of 47 individuals observed at six facilities (in China, Japan, Mexico and the USA), 62% displayed stereotypic behaviours, especially pacing, head tossing, pirouetting and cage climbing. Figure 11.1 demonstrates examples of pacing routines. In extreme cases, these activities persist for hours (pacing) or occur at rates of nearly 80 bouts per hour (head tossing). More males than females are afflicted, and the difference between sexes is statistically significant in the adult, but not the subadult population. Well over half of the wild-caught individuals in the Lindburg et al. (in press) sample (10 of 17) displayed stereotypies (although several of these individuals had been captured from nature at an early age). As might be expected, the incidence was higher among captive-born adults (78% of males and 29% of females). These findings reflect the pervasiveness of these anomalous behaviours and the need for remedial measures.

Environmental enrichment

Environmental enrichment is the behaviourally relevant modification of a captive milieu to optimise physical and psychological well-being. The guiding step in developing an effective enrichment programme is to look at nature (Swaisgood et al., 2003a). How do individuals of a species 'make a living' in nature? In the case of the giant panda, what are the critical resources and how are they distributed, located and obtained? Thus, what is the size of the home range and how far does the animal travel daily? What is the form and frequency of interactions with conspecifics, and how do individuals locate and choose mates? Much of this basic natural history is known, but many gaps in knowledge continue to leave us guessing. We do know that the giant panda spends 98% of its life eating and sleeping, leaving little time for locomotion, social interaction or other behaviours (Schaller et al., 1985). Diet consists almost exclusively of bamboo, which generally is prevalent throughout the home range. Therefore, there is little incentive to search for food proactively, with only occasional movement to an alternative bamboo patch.

Giant Panda Life Cycle Diagram

Figure 11.1. Diagrams depicting pacing routines and embedded stereotypies. (a) The routine for an adult female that paced in a circle, using only a small fraction of existing space. With each lap she paused, sat, rolled backward to the ground and then rocked into the seated posture in a movement resembling a 'sit-up' (A). Sit-ups invariably occurred at the cage bars, in one case at the location where food was delivered. The pattern of pacing and sit-ups was often repeated more than 30 times consecutively and was especially prevalent during peri-oestrus.

Figure 11.1. Diagrams depicting pacing routines and embedded stereotypies. (a) The routine for an adult female that paced in a circle, using only a small fraction of existing space. With each lap she paused, sat, rolled backward to the ground and then rocked into the seated posture in a movement resembling a 'sit-up' (A). Sit-ups invariably occurred at the cage bars, in one case at the location where food was delivered. The pattern of pacing and sit-ups was often repeated more than 30 times consecutively and was especially prevalent during peri-oestrus.

Our recent studies have involved evaluating and modifying enclosure design to affect the ex situ management of giant pandas. Indeed, design of the captive habitat is an important factor, which is not surprising as this issue is important for many wildlife species held in zoos (Forthman et al, 1995). More than 20 enclosures at the Wolong Centre are relatively small, approximately 9 x 11m (indoor plus outdoor portions). Size is important, but clearly it is not practical or financially feasible to build enclosures nearing home range size in nature (i.e. >400 ha; Schaller et al., 1985). Nonetheless, it is possible to test the value of larger captive pens and, through enrichment, to introduce some of the complexity and variability normally encountered in the wild. Toward this end, the Wolong Centre has constructed seven large (>0.25 ha each), semi-natural (containing native vegetation) enclosures on a slope along the Pitiao River. More such 'enclosed habitats' are planned at this location, and 16 similar enclosures are under construction in another mountain range south of the reserve. Each has the capacity of maintaining up to four or five subadults or two adults. Individual giant pandas from the centre are rotated through these enclosures, but the emphasis is on providing access to young animals, hoping that such special exposure will prevent the behavioural problems seen in pandas housed in smaller pens. An additional development is construction of multiple, even larger (1-30 ha) 'reintroduction pens' on the mountain above the centre. The plan is to release pregnant

(b) The routine of an adult male that paced back and forth by the cage bars adjoining a neighbouring female. At the turnaround point in the corner, he often stood bipedally and rotated the body about 180° (B: a 'pirouette') before pacing back to the opposite corner. The male often head tossed (C), an abrupt exaggerated upward swing of the head, before or instead of the pirouette. This behaviour increased, as did the speed of locomotion, when a neighbouring female was in oestrus. (c) The stereotypical routine of an adult male housed in a small enclosure. This male spent nearly one-third of his time following a specific travel path. He invariably stopped at three distinctive points along the path, sat down, scratched himself with his hind paws and looked outside the enclosure (D) for 30 to 60 seconds. One habitual stopping point was at the site of food delivery, and another overlooked the preparation area where he could observe keepers arriving with food. These behaviours accelerated in frequency as the normal feeding time approached.—, solid wall;----, cage bars, , animal's path of travel; A, sit-up; B, pirouette; C, head toss; D, sit and scratch.

females at these sites to allow resulting cubs to be reared in the absence of most human contact. Such offspring are likely to be better candidates in the event of future reintroduction studies.

Since 1996, we have collected quantitative behavioural data on giant pandas living in small versus large enclosures. Preliminary evidence has revealed a marked reduction in stereotypies among individuals maintained in the larger pens. There have been other behavioural changes indicative of enhanced psychological well-being too, including reduced pacing and increased diversity of natural behaviours (Swaisgood, unpublished data). We are learning that even very subtle differences in enclosure design can influence behavioural measures of well-being. For example, the female giant panda maintained at the San Diego Zoo (SB 371) displays more signs of behavioural distress in one pen than another (p < 0.01). Both are open to the public, but the larger of the two has less vegetation, which may mean that its design contributes to aversive behaviours. Other giant pandas at this zoo (SB 415 and 487) have shown no behavioural preferences between enclosures.

Presently, less than one-third of the giant pandas at the Wolong Centre are still housed in small enclosures. Initially, these pens were rather barren and contained no structural complexity, manipulable objects or visual barriers. Biologically meaningful aspects of complexity include vertical dimension, variety of textures, mobile and non-mobile furnishings and temperature and light gradients. These, in turn, provide behavioural opportunities for climbing, resting at elevated sites, locomotor play, exploration, resting and digging on different substrates, hiding from conspecifics, humans or disturbances and behavioural thermoregulation. Such enrichments give the animal behavioural options, allowing it some control over its interaction with the environment. That is, an animal's behaviour influences whether or not it gains access to important resources, even if it is as simple as moving from a sunny to a shaded area. To increase behavioural control options for giant pandas at the Wolong Centre, we have begun to equip enclosures with permanent furnishings. Table 11.1 summarises changes for some or all of the pens. These alterations have been successful in that pandas spend a great deal of time interacting with or using the furnishings while displaying more diverse and naturalistic behaviours. Importantly, stereotypic behaviour has been reduced significantly for animals living in these revitalised enclosures, but these measures appear insufficient to eradicate stereotypies altogether. These altered

The science of behavioural management Table 11.1 Use and function of structural alterations to small enclosures

Pen enrichment

Predominant behaviours observed and/or purpose

Trees, logs, stumps, wooden climbing structures (~3 x 3 x 2 m) Stones of various sizes

Digging pits with sand

Woven metal baskets in bedroom Railing outside enclosures Bushes, tall bamboo outside enclosures

Climb; locomotor play; rest; body rub; scent mark; tactile and oral explore; sniff; hide

Object play and manipulation; body rub;

scent mark; tactile and oral explore; sniff; hide Dirt bath; locomotor play; rest; dig; scent mark; tactile explore; sniff Rest/play area secluded from public;

locomotor play; rub; scent mark; sniff Minimise public interaction with pandas Minimise public interaction with pandas; temperature and light gradient (shade); rest; hide environments have also offered a more complex topography for presentation of supplemental enrichment (e.g. hidden or scattered food and the placement of novel objects; see below).

The essence of the giant panda's existence is eating bamboo. This fundamental fact of panda life pervades and consumes a remarkable 14 hours of each individual's day (Schaller et al., 1985). But this is also a valuable piece of information for enrichment, and we have exploited this inherent need to spend a lot of time feeding into the design of our studies. For example, the recent panda population boom at the centre has mandated the need to ensure adequate bamboo provisioning for all animals. This has been no trivial matter as more bamboo harvesters had to be hired and local farmers paid to grow bamboo as a crop. Winter shortages, which plagued the centre's early years, have been reduced and now, regardless of season, each giant panda receives multiple feedings of good-quality bamboo daily. This altered husbandry practice has increased opportunities for pandas to engage in behaviours related to bamboo consumption (removing the leaves to make a wad for consumption, stripping the outer inedible shell of the culm and, of course, lots and lots of chewing). Still, because one cannot be certain that all nutrients are available in the provisioned bamboo, the diet must be supplemented. The problem is that dietary supplements are quick, easy-to-process energy sources (see below) resulting in overall less time spent feeding than in wild counterparts. Thus, a captive giant panda under such conditions can have 'empty time' that needs to be filled with other activities, preferably related to feeding and foraging. If this does not occur, the development of stereotypies is likely.

In our studies, this available time has been used to test various enrichment strategies, usually settling on functional analogues. Giant panda foraging strategy entails little searching, but requires a great deal of handling to strip leaves and peel culms before consumption (Schaller et al., 1985) and processing to chew this fibrous food source. We have argued that the ethological needs model does not imply that animals are solely motivated to perform specific behaviours (Swaisgood et al., 2001). Rather, functional analogues may suffice - opportunities that allow animals to engage in goal-directed behaviours to secure resources. In this respect, we have developed multiple feeding-enrichment strategies for adult giant pandas. One of the most important has been replacing milk and low-fibre bread with a high-fibre biscuit (see Chapter 6). This has nutritional benefits, but the hard biscuits also require protracted time to chew and ingest. Mastication also can be promoted by freezing food in a large block of ice. Scatter feeding (hiding food around the enclosure) encourages searching behaviour. Likewise, several feeding devices have been used, all of which involved highly prized food items (apples, carrots and biscuits) and encourage handling and processing before consumption (Swaisgood et al., 2001; Tepper et al, 2001). The strategy can be made even more complex by hiding the food in a puzzle feeder. In one of these, a small hole is cut in a large section of bamboo (about 10 cm in diameter; Fig. 11.2) and apples inserted. The panda must roll the feeder with its paws or head, or pick it up and manipulate to extract the food. Similar devices have been created using a PVC pipe (Fig. 11.3) and hard plastic balls, also designed to increase handling time. Food contained in a cardboard box forces the panda to tear with teeth or claws, whereas food wedged into holes (drilled into logs or hidden in a pile of rocks) requires clawing. Enrichment devices hung with chain or rope from climbing structures will promote a different class of handling behaviours. Even nonfeeding enrichment can promote feeding-like behaviour. For example, fresh evergreen branches are handled, stripped and partially chewed by most giant pandas in a fashion remarkably similar to bamboo consumption (Swaisgood et al., 2001).

Figure 11.2. A panda at the China Conservation and Research Centre for the Giant Panda (Wolong) uses its muzzle to manipulate a bamboo puzzle feeder, rolling it back and forth until apple pieces fall out. Made from a section of large bamboo, the feeder contains a single, small hole from which apple pieces must be extracted. Encouraging pandas to work for food in such ways can lead to improved signs of well-being.

Figure 11.2. A panda at the China Conservation and Research Centre for the Giant Panda (Wolong) uses its muzzle to manipulate a bamboo puzzle feeder, rolling it back and forth until apple pieces fall out. Made from a section of large bamboo, the feeder contains a single, small hole from which apple pieces must be extracted. Encouraging pandas to work for food in such ways can lead to improved signs of well-being.

Figure 11.3. A PVC puzzle feeder with several large holes to make access to food treats easier. This design works well with a 'lazy' individual who refuses to work for food using the more challenging bamboo feeder.

Because animals often develop stereotypies due to predictable food delivery schedules, we also have altered the temporal feeding pattern and frequency at the centre. Before 1997, individuals were fed twice per day at 1000 and 1600 hours. At these times, a common sight at the feeding locations was pacing, weaving back and forth and head tossing (Swaisgood, unpublished data). Now, there are as many as seven feedings per day with food given on a more varied schedule. The result has been an apparent reduction in food anticipation and abnormal behaviours.

Our supplemental enrichment programme for giant pandas involves the occasional providing of novel and manipulable objects. Items are chosen on the basis of varying physical properties that facilitate different behavioural opportunities (e.g. sniff, pick up or swat with paws, push with head, bite, shake, throw, roll with or on, jump on, carry, chase) or needs that may well vary with sex, age or individual temperament. In general, the enrichment item is left in the panda's enclosure for 24 hours and then removed. The list of items utilised is continually growing and changing and has included: burlap sacks stuffed with straw, sand or mulch; plastic manipulable objects; piles of evergreen branches, bark, rocks and mulch; boxes; tyres; straw mats; and logs. Such an array of items results in fascinating results. For example, we know that style of presentation can be important. The male (SB 381) at the San Diego Zoo preferred substrate presented in a tall, narrow pile over the same amount of substrate presented in a flat, wide pile (Tepper et al, 2001). We also have discovered that giant pandas in general have a significant preference for a plastic bobbin over a 26-cm-diameter ball (Swaisgood et al., 2000). The bobbin, which is shaped like a thread spool, stands 35 cm high and is 30 cm in diameter at the two flat ends, tapering to 18 cm in the middle. We have consistently observed that a panda can readily pick up the bobbin, which is easy to hold, manipulate and explore it with all four paws while lying on its back. A ball, apparently more conducive to swat and chase, is of less interest. This should not be surprising in that the giant panda is a consummate 'manipulator', using its famous pseudothumb to assist in grasping and handling bamboo.

Enrichment is only as valuable as the ability to measure an animal's responsiveness quantitatively, thereby permitting efficacy to be accurately evaluated. A variety of methods has been used to measure enrichment in the giant panda, some relying on labour-intensive, formal behavioural observations. Others have capitalised on casual observations by keepers going about their daily activities. Regardless ofapproach, it is evident that enrichment has a major impact on several indices of psychological well-being. Most importantly, these giant pandas are more active and engaged in diverse play and non-play behaviours than when not enriched. They also have an overall reduced incidence of stereotypy while spending less time expressing abnormal behaviours (especially associated with predicted feeding schedules)

(Swaisgood et al., 2001). Particularly exciting is that enriched giant pandas express signs of improved well-being, even in the absence of direct interaction with the enrichment item. This suggests that the effect is not just the result of less time available to engage in stereo-typies. Rather, the animal's motivation to perform abnormal behaviour is diminished. This discovery is also consistent with the ethological needs model of motivation. That is, the enrichment objects only provide an opportunity to perform a behaviour without offering the panda an important biological resource. Another important implication is that the animal's time spent interacting with enrichment may not reflect how enriching the experience truly is. Studies at the San Diego Zoo demonstrate that enrichment occupying as little as 2 to 3% of a panda's time significantly improves well-being later in the day (Tepper et al., 2001). Thus, although the enrichment opportunity may be brief, it may be of long-term importance.

Biologically relevant behavioural management for reproduction

In developing breeding management protocols at the Wolong Centre, we capitalised on the scientific studies of wild giant pandas emphasising the solitary nature of the species punctuated by occasional contacts outside the mating season (Schaller et al, 1985; Schaller, 1993; Lu et al., 2000). Although more intensive social housing arrangements may sometimes work (Kleiman, 1984; Hoyo Bastien et al, 1985), we ascribe to the philosophy that the best approach is to mimic nature as much as possible and then modify if methods fail. For example, field researchers have noted that males have large home ranges that encompass several females' home ranges. Thus the male in nature has access to these females' scents but is unlikely to spend much time in any given female's range. Direct encounters between male and female giant pandas are relatively rare (Schaller et al., 1985). Lu and colleagues (2000) suspected that these social traits of wild pandas and the lack of similar opportunities in captive counterparts likely contribute to reproductive failure in the latter group. Thus, management activities at the Wolong Centre have been altered to ensure that adult pandas are housed individually, as fights can occur outside the mating season even between males and females. These pandas have access to neighbours through cage bars but can avoid proximity and generally most visual, olfactory and acoustic contact if desired. Animals also are moved to a different enclosure several times each year, where they encounter new conspecific odours and neighbours. These conditions probably afford opportunities for social exchange not too disparate from patterns in the wild.

Olfactory management studies

One of the hallmarks of the centre's programme is the use of species-appropriate olfactory management to create a more natural context for intersex communication. We designed a series of studies to elucidate the functions and practical application of the chemical signalling system in the giant panda (Swaisgood et al, 2004). Because these animals live a solitary existence, they must rely heavily on chemical signals to communicate without requiring face-to-face encounters. Chemical signals, including scent marks from the anogenital gland and urine, are unique and persist in the environment after the signaller has left. Anogenital marks last about three months while urine marks are sustained for two weeks (Swaisgood et al, 2004). The species appears to have developed a system of traditional communal scent-mark stations, or 'community bulletin boards', which provide individuals with reliable locations to visit, and where they can deposit signals and investigate signs left by other giant pandas.

By presenting conspecific odours, we have begun experimentally to tease apart the meaning of these chemical signals. First, it is clear that pandas can discriminate between the odours of individuals: each panda's scent mark leaves a 'chemical signature' (Swaisgood et al, 1999). Secondly, pandas 'date' these chemical messages, that is, they are able to distinguish between scents deposited at different times (Swaisgood et al., 2004). Thirdly, recipients extract a great deal of information from these chemical messages, including the sex (Swaisgood et al, 2000; White, 2001), age (White et al, 2003), reproductive status (Swaisgood et al., 2000, 2002) and competitive status of the signaller (White et al, 2002). This information is clearly important. For example, we now know that opposite-sex odours appear instrumental for priming sexual motivation before face-to-face encounters (Swaisgood et al., 2000, 2002). Giant pandas in captivity are notorious for aggressiveness and/or indifference when placed in mating situations (Zhang et al., 2004). We have found that individuals show increased sexual arousal solely in response to odours of the opposite sex. At the Wolong Centre, pandas are exposed to odours of potential mating partners for days to weeks before the mating introduction. The result is less aggression and more sexual interest.

Reproductive behaviour studies

Reproductive behaviour studies have been instrumental not only at the Wolong Centre (Swaisgood et al., 2003b; Zhang et al., 2004), but elsewhere (Kleiman, 1983; Snyder et al, 2004) to improve management. A common approach is to merge behavioural and morphological signs of oestrus through parallel endocrine monitoring to identify the onset of the fertile period (see Chapter 14). In studies at the San Diego Zoo, we have found that behaviour, hormonal profiles and vaginal cytology patterns are all similarly predictive of the timing of impending ovulation (Lindburg et al., 2001; see also Chapter 9).

At the Wolong Centre, two complimentary methods are used to record various behaviours indicative of oestrus beginning in February at the onset of the breeding season. One is a systematic, quantitative protocol for keepers to record casual observations, whereas the other involves formal observations made by trained researchers during five specific 45-minute intervals each day during morning and afternoon periods of high activity. Table 11.2 lists some of the most important behavioural and morphological signs used to track oestrus (adapted from Kleiman, 1983). By monitoring behaviours closely, we can recognise early signs of impending oestrus, and move the female to a pen adjacent to an appropriate male so that communication and familiarisation occur well in advance of the mating introduction. These early signs can often be identified 15 to 20 days before actual introduction for copulation. Visual, acoustic, olfactory and limited tactile contact through cage bars is promoted in this 'protected contact' arrangement. This method is supplemented with frequent 'pen swaps' to expose the pair more directly to each other's odours (Swaisgood et al, 2000). This approach of promoting an optimal level of familiarity between the potential pair before introduction is also advocated at the Chengdu Research Base of Giant Panda Breeding (see Chapter 14). We are now also using supplemental recording protocols to collect more intensive data for the female as well as the neighbouring male (Swaisgood et al., 2003b), eventually allowing enough information to seriously consider making the mating introduction.

Introduction of animals for mating

A useful tool for deciding to attempt or terminate a mating introduction is a 'decision tree' that covers various contingencies. Figure 11.4

Table 11.2 Behavioural and physical signs used to monitor oestrus

Behavioural and physical signs Description

Active Scent mark Urinate




Water play Backwards walk

Tail up Lordosis

Proximity to male Appetite Vulvar colour

Vulvar swelling and opening

Percentage time spent in non-rest behaviour Rubbing anogenital gland on a surface Elimination of urine; often accompanies scent marking High-pitched twittering vocalisation of one to three seconds duration, sounding much like a goat's call; 'friendly' and promotes contact with male Short (<1 second) high-pitched, harmonic vocalisation descending in pitch at the end of each call; 'friendly', promotes contact with male

Lying on back and rolling from side to side Splashing and rolling in water Walking backwards for more than four steps, sometimes for >20 minutes, often backing into a male or inanimate object with tail up

Tail is raised uncovering vaginal area Sexual posture with tail up, hind quarters raised, back arched and kneeling on forelegs

Female approaches male and/or spends time located adjacent to male enclosure Consumption of bamboo and dietary supplements declines Coloration of external genitalia changes from grey to red Enlargement of external genitalia and size of the vaginal opening

Adapted with permission from Kleiman, 1983.

offers a simplified version of such a decision tree, which can be expanded or tailored to the dispositions of individual pairs. The following rules of thumb are for giant pandas housed separately in relatively small enclosures. The ultimate decision to place two individuals together must be based on observations of behaviour indicating

Figure 11.4. Mating introduction decision tree. This simplified version shows how behavioural signs are used by management staff to make decisions to maximise mating opportunities while taking care to avoid injurious aggression. Accompanying such a tree are detailed definitions based, in part, on behaviours included in Table 11.2.

Figure 11.4. Mating introduction decision tree. This simplified version shows how behavioural signs are used by management staff to make decisions to maximise mating opportunities while taking care to avoid injurious aggression. Accompanying such a tree are detailed definitions based, in part, on behaviours included in Table 11.2.

pair compatibility. Other signs (i.e. vulvar colour, vaginal cytology and urinary oestrogen patterns) are important, but more so for identifying the physiological period of highest fertility. Behavioural traits, however, determine the level of sexual interest and the potential capacity for aggressiveness, which can be injurious to one or both individuals. At the Wolong Centre, the staff waits for a strong bout of sexual interaction in protected contact before introducing the animals together. Ideally, this is expressed as a male being excited and bleating very frequently, whereas the female is bleating, chirping and demonstrating a lordotic stance. It has been useful to gently touch the female's genitalia with a piece ofbamboo, stimulating lordosis as the male enters the female's pen. If the gate separating the pair is opened at this time (while the female is lordotic), it is more likely that the male will mount and copulate quickly. This seems to minimise fighting.

Figure 11.5. Aggression is not uncommon in mating introductions. Here the male (above) was rough with the female, and she turned in defence. Approximately two hours later, after a brief separation, the pair successfully mated.

Even when not apparent during protected contact, aggression can occur during the mating introduction (Fig. 11.5). Mild hostility is common, is no cause for alarm and should be tolerated for a few minutes before attempting to separate the pair. It is not unusual for the male to bite the female's back and perhaps paws when trying to mount, resulting in some fur removal. A portion of this activity is related to positioning the female for copulation (Fig. 11.6) and is not motivated by aggression. If the biting becomes dangerous and/or the female responds with excessive fear or belligerence, the pair should be separated immediately. Some aggressive vocalisations (bark, moan and squeal) are common for the female during courtship and mating, and are not signals for pair separation. However, excessive and hostile vocalisation from the male is unusual and reflects a potential attack on the female. Under normal conditions, a male should continue bleating frequently while with the female. An absence of this activity may indicate insufficient sexual interest. Similarly, the female should bleat and/or chirp frequently during pairing, although barking, squealing and moaning are also normal. If the pair is separated due to aggression, a re-attempt at introduction should be made within a few

Figure 11.6. Mating in species-typical positions.

hours if adequate sexual interest is apparent and the earlier bout did not result in injuries. Alternatively, a mating introduction with a different male may be attempted.

Aggression is most likely to occur in one of three circumstances and can be dangerous, especially in small enclosures where escape options are limited:

1. Despite signs of sexual arousal before pairing, upon introduction the male may attack the female immediately, especially if she is overly defensive or uncooperative. This situation requires immediate pair separation.

2. After making a series of unsuccessful mounts over more than 15 minutes, the male becomes 'frustrated' and attacks the female, often escalating to a dangerous encounter. In general, both the male and female giant panda appear to tire after 20 or 30 minutes of attempted mounts and failed copulations, a situation that can evolve into aggression. Prior to significant hostility, the pair should be separated. (This stamina issue underscores the need to keep animals in good physical condition, including by providing enrichment opportunities as discussed above.)

3. Immediately after copulation, it is common for the female to be aggressive, which is expressed as breaking away from the male, barking, growling and lunging. The post-copulatory male generally shows little interest in the female, but some individuals may retaliate; serious attacks are unusual. In any case, a pair that has just mated should be separated unless there are indications of additional sexual arousal. For example, very short copulations

(<1 minute) may be followed by another successful copulation within 10 to 30 minutes.

Separating giant pandas that are together during the nonbreed-ing season is generally easy through conditioned training and the use of favoured treats (e.g. sugar cane or apple). This is not the case during the breeding season when individuals are often uncooperative. Usually it is easier to provoke the male to move from an enclosure because a female in peak oestrus is often unresponsive, assumes the lordotic stance and will not budge even in response to a favourite treat. Alternative techniques for separating or moving individuals include spraying water with a hose or swatting the ground in front of or behind the panda with a large stick of bamboo.

Neonatal management

Our behavioural research has also been targeted to birth and maternal caregiving, which are both critical to successful captive management. It is at this time that many critical decisions are required, often on short notice and with little time. For example, if the mother is deemed inadequate, it may be necessary to remove the infant for hand-rearing (see Chapter 13). Also, weak or sick neonates must be identified as early as possible for human intervention and intensive care. Judicious decision-making at this critical juncture necessitates a clear understanding of the normal range of successful maternal as well as neonatal behaviours.

Although hand-rearing has increased the total number of offspring surviving at the Wolong Centre (and other facilities), it may well be that mother-reared individuals could be more valuable. Although more studies are needed, we can assume that dam-reared infants may have fewer behavioural abnormalities in adulthood and be better candidates for natural mating or reintroduction. Therefore, new methods are being used to encourage reluctant dams to rear their own offspring. In one case, a mother initially abandoned her neonate, which was taken to the nursery for hand-rearing. During the following weeks the mother was exposed to the odours and vocalisations of her cub and was given a surrogate toy panda. These actions evoked normal maternal care behaviours, reduced her fear and gradually allowed the cub to be reintro-duced to the dam, who eventually assumed all caregiving activities (Zhang et al., 2000). The importance of investigating the mother-infant relationship does not end with the initial maternal response. In studies similar to those described in Chapter 14 by Snyder and colleagues, we have followed the lives of 39 cubs from birth to weaning, generating a large database on successful and less successful maternal caregiving styles.

Young at the Wolong Centre are routinely weaned from the dam at five to six months of age, which is more than a year earlier than that observed in the wild (Zhu et al, 2001). It has been speculated that precocious weaning may have serious consequences for adult reproductive performance (Lu et al, 2000; Snyder et al., 2003; see also Chapter 14), although our experiences have not revealed such an interaction. Sixty-four cubs have been produced at the centre since 1991. Of these, seven were adults of more than six years of age in 2003 (thus having at least one to two reproductive seasons of experience). An additional female rescued from the wild at less than six months of age and reared in Wolong was also included in our study group. Of these eight adults, six have copulated (more than 70 times). Of the four females that have mated, all have been successful mothers and reared their own young. One explanation is that a mother in nature interacts with her cubs relatively infrequently after the first few months of life, often leaving them behind in a tree while the dam forages at a distance for hours or even days (Lu et al., 1994; Zhu et al., 2001). Captive-born cubs probably grow up to be successful breeders because they actually have more interaction with their dam and also, generally, are reared with same-age peers that provide ample social stimulation (see Chapter 14). Therefore, in this case, the need for imitating nature is probably not essential. Nonetheless, it remains worthwhile to investigate further even potentially subtle impacts of peer exposure on behavioural development and subsequent reproductive success (see Chapter 14). Moreover, mother-rearing may be essential for producing a giant panda capable of surviving reintroduction.

Another challenge to the management of newborns is the production of twins and the rejection of one neonate. Following the practice of the Chengdu Research Base of Giant Panda Breeding, the Wolong Centre has also found that a dam can be encouraged to raise both cubs through the technique of twin swapping (see Chapter 13). In brief, one cub is removed at a time and maintained in the nursery for two to four days, then traded for its twin, which is being mother-reared (a rotation system that results in a combination of dam- and hand-rearing with both offspring receiving maternal milk and colostrum as well as formula). Using this technique at the Wolong Centre, survival to six months has been 100% for 24 offspring (one triplet was lost). Prior to initiating the twin-swapping strategy, only one nursery-reared cub had ever survived for more than a few days (in this case, to five months). We are also exploring the value of using the dam's scent to reduce behavioural stress in the early days post weaning (Swaisgood et al., 2004). We suspect that intense stress at this time could provoke development of stereotypies. More recently, we have experimented with an artificial mother for freshly weaned cubs, using a soft pillow covered with artificial fur and equipped with a nipple. The idea is that the latter will satisfy the cub's need to suckle (much like a human pacifier) and stave off development of oral stereotypies (e.g. paw sucking).

Was this article helpful?

0 0

Post a comment