Hormones are the keys to reproductive success as well as the regulators of animal well-being. In this chapter, we described the unique endocrinology of the giant panda, from the peri-oestrus interval, pregnancy, pseudopregnancy and seasonality to adrenal responsiveness to a physiological 'stressor' (anaesthesia). Data revealed the extraordinary potential that non-invasive endocrine methods provide for biologists to understand the fundamental mechanisms associated with endocrine control of general health and reproductive success.
During a giant panda ovarian cycle, it is now known that urinary oestrogens increase gradually over one to two weeks, rapidly accelerate to pre-ovulatory concentrations, decline precipitously coincident with ovulation and return to baseline two to five days later (Bonney et al., 1982 ; Hodges et al., 1984; Chaudhuri et al., 1988; Monfort et al., 1989; Mainka et al., 1990; Lindburg et al., 2001; McGeehan et al., 2002; Czekala et al., 2003). Thus, peak oestrous behaviours, including sexual receptivity (e.g. tail-up and lordosis) and mating, usually occur when circulating and excreted oestrogens are declining (Bonney et al., 1982; Kleiman, 1985; Monfort et al, 1989). Although mating has been observed when oestrogens are increasing (or are at peak concentrations), those incidences generally have been associated with the first of successive copulatory bouts (Bonney et al., 1982; Monfort et al., 1989). In 1999, the San Diego Zoo female (SB 371) became pregnant following AI conducted on three successive days; AI was initiated only after oestrogen concentrations were declining. Likewise, the cub produced at the Smithsonian's National Zoo in 2005 was born to SB 473 who was artificially inseminated one time after urinary oestrogen levels were clearly already declining. Thus, optimal fertility (to mating or AI) occurs in the late peri-ovulatory interval in an endocrine atmosphere characterised by declining oestrogen influence (Lindburg et al., 2001), a phenomenon also observed in the gray wolf (Seal et al., 1979), domestic dog (Wildt et al., 1979) and mink (Pilbeam et al., 1979).
Urinary oestrogen patterns, in combination with reproductive behaviours and vaginal cytology (see Chapter 9), provide the best means of accurately evaluating reproductive status of the female giant panda. For example, when only behaviours were monitored, 80% of captive giant pandas were classified as experiencing a 'weak' or 'silent' oestrus (Zheng et al., 1997). However, subsequent evaluations of oestrogen excretion indicated that most of these females actually ovulated (McGeehan et al., 2002). Nonetheless, more work is warranted to compare 'weak' versus 'strong' oestrual females, and to determine if variations in behavioural intensity at oestrus are associated with fertility and fecundity.
Intermittent oestrous acyclicity or anoestrus has been correlated to lactational suppression of ovulation in the giant panda (Schaller et al, 1985). However, for the first time, we report that non-lactating females with a history of regular ovarian cyclicity can experience idiopathic acyclicity (SB 332; see Fig. 8.2c). Skipping an annual ovarian cycle does not appear to be rare in giant pandas, and we have not detected any substantive changes in management procedures that would suggest that psychosocial or physiological stress contribute to this phenomenon. Periodic absence of ovarian activity indicates that the giant panda may not be evolutionarily programmed to sustain pregnancy and lactation on an annual cycle due to the long-term maternal investment required to nurture young until weaning (18-24 months; Schaller et al, 1985).
Exogenous gonadotrophins have been used to stimulate follicular development and/or induce ovulation in many wildlife species in captivity (Howard, 1999), including the giant panda (Chaudhuri et al, 1988). Our findings revealed that the sequential injection of FSH and/or hCG could cause serious hyper- and protracted oestrogen secretion. Even hCG alone caused this effect, probably by stimulating too many ovarian follicles and/or interfering with normal steroidogenic pathways in lu-teinised follicles or the corpus luteum (CL). More work is needed to develop appropriate ovulation induction regimens for the giant panda. Promising new research to clone and express giant panda FSH and LH (L. Mingjuan, pers. comm.) is in progress. Regardless, non-invasive hormone monitoring will continue to be valuable for tracking the effectiveness and safety of new endocrine therapies.
Our data also revealed that faecal steroid analysis was effective for assessing ovarian status via both oestrogens and progestins. The strong correlation between urinary and faecal steroids, both of which coincided with appropriate behaviours associated with oestrus and ovulation, provided strong incentive for further applying this technique. However, because faecal analysis requires a time-consuming extraction procedure, urinary analysis provides a quicker evaluation of oestrual activity for captive management. Faecal steroid measures are better suited for longitudinal assessments, perhaps eventually including studying free-living giant pandas.
For the giant panda, an initial rise in excreted progestins occurs immediately after the peri-ovulatory oestrogen peak (Hodges et al., 1984; Monfort et al., 1989; Mainka et al., 1990; McGeehan et al., 2002; Czekala et al., 2003). This initial post-ovulatory rise in urinary progestins, although not as robust as the later secondary rise, signals the occurrence of ovulation (Mainka et al., 1990; McGeehan et al., 2002). Due to temporal variation in speed and amplitude of this initial rise, however, the urinary progestin profile is most effective as an ovulation marker when used in concert with the dramatic pre-ovulatory oestrogen surge.
A secondary progestin excretion rise (74-122 days post-ovulation) occurs 40 to 50 days before parturition, suggesting that the giant panda experiences delayed implantation (Hodges et al, 1984; Chaudhuri et al., 1988; Monfort et al., 1989; Mainka et al., 1990; McGeehan et al., 2002). At the onset of the secondary progestin rise, it is believed that the blastocyst becomes implanted in the uterine wall with foetal development further supported by continued progestin production. Delayed implantation occurs in all bears (Sandell, 1990), the mink (Cochrane & Shackelford, 1962), badger (Bonnin et al., 1978) and western spotted skunk (Foresman & Mead, 1974), presumably to ensure that birth occurs at a time most conducive to offspring survival. This 'delay' is believed to be due to an insufficiency of the CL which, when exposed to the pituitary hormone prolactin, becomes reactivated to produce significant amounts of CL-derived progesterone (Mead, 1993; Sato et al, 2001). Under progesterone influence, the conceptus implants to resume development.
Delayed CL reactivation in bears is obligate once ovulation occurs, whether or not a female conceives. This is why some nonpregnant females are classified as pseudopregnant (Mainka et al., 1990; Sato et al., 2001). Females have no way of assessing their own pregnancy status, so pseudopregnancy may function as a hormonal 'insurance policy' protecting pregnancy should a female be carrying a diapausing embryo awaiting implantation. Qualitative and temporal progestin (and oestrogen) excretion patterns are similar between the pregnant and non-pregnant giant panda (Hodges et al, 1984; Chaudhuri et al, 1988). There also appears to be no difference in hormonal patterns among a pubertal pseudopregnancy, pregnancy and adult pseudo-pregnancy. It is likely that some females classified as pseudopregnant actually have been pregnant but suffered undetected pre-implantation embryonic or post-implantation foetal loss. Therefore, we suspect that the incidence of pseudopregnancy may be overestimated. We recommend that, until improved pregnancy detection methods are developed, this term should be limited to describing unmated/ovulatory females that exhibit concurrent endocrine and/or behavioural signs of pseudopregnancy. Comparisons between known pregnant and non-mated, ovulatory females - a rare group given the emphasis on breeding all adult giant pandas - could help to establish endocrine differences between these two reproductive states.
More research is also needed regarding specific factors for identifying pregnancy in the giant panda. For example, embryo-derived platelet activity factor (EDPAF) has recently been shown to be useful for predicting pregnancy in this species (R. Hou, pers. comm.). However, this approach requires a minimum of two blood samples collected during the peri-ovulatory interval (immediately before and after presumed conception); thus, it may be considered too invasive, especially if anaesthesia or physical restraint is required for venipuncture. Despite the lack of a definitive pregnancy diagnostic test, a decline in the secondary urinary progestin surge (occurring at the end of pregnancy or pseudopregnancy) is a valuable sentinel measure. The approaching return to baseline either signals an impending birth (allowing managers to mobilise resources to monitor the female) or, in the case of pseudopregnancy, indicates the end of a neccessary pregnancy watch (Chaudhuri et al., 1988; Monfort et al., 1989; Czekala et al., 2003).
Physiological or psychosocial stress increases corticosteroid secretion, long a marker for 'stress' in diverse mammals (Morton et al, 1995). The blood collection procedure itself (restraint and/or anaesthesia plus venipuncture) can be a stressor. Therefore, serum corticoid values may not accurately represent undisturbed, baseline adrenal activity (Mon-fort, 2003). For the first time, we have demonstrated the ability to assess adrenal activity in the giant panda non-invasively by assessing urine. A cause and effect between administering anaesthesia (a physiological 'stressor') and subsequent increased corticosteroid excretion was observed. The brief elevation (~10 hours) in this stress hormone suggested that more frequent sampling (e.g. multiple urines per day) would better document the impact of short-term, acute stressors. Infrequent sampling could incorrectly lead to the conclusion that a physiological stressor had no impact on corticosteroids. For example, if samples in our study had only been collected after ten hours, anaesthesia would not have appeared to increase corticosteroid excretion. As the giant panda defaecates more frequently than it urinates, faeces may be more useful for these type of studies; the potential of faecal corticosteroid measures is now being examined.
In a preliminary study (data not shown), an injection of adreno-corticotrophic hormone (ACTH) to a single giant panda male markedly increased both faecal and urinary corticoids in <24 hours (Kersey, unpublished data). Interestingly, the post-ACTH corticosteroid increase occurred earlier in faeces (nine hours) than urine (19 hours). Now that faecal monitoring has been shown to be a physiological reflection of true adrenal activity, there is a need to determine if the giant panda exhibits seasonal fluctuations in corticosteroid activity, as reported for other bears (Palumbo et al., 1983; Harlow et al., 1990). In preliminary studies (data not shown), there were no differences (p > 0.05) in urinary corticosteroids (1) within females over the four-month breeding season (February to May), (2) in spontaneously ovulating versus gonadotrophin-treated females, or (3) among mated, non-mated, anovulatory, pregnant or pseudopregnant females. The latter findings are important because they suggest that anovulation or 'weak oestrus' in the giant panda is probably unrelated to stress.
In the only previous such study in a male giant panda, urinary andro-gens increased during the reproductive season when the mate was sexually receptive, but not during the next year when the female failed to exhibit sexual behaviours (Bonney et al., 1982). Anticipation of mating or mating itself can induce androgen production in the rat (Taleisnik et al., 1966), rabbit (Saginor & Horton, 1968) and bull (Katon-gole et al, 1971). Our studies of multiple male giant pandas revealed that urinary androgens were elevated consistently during the normal mating season (February to April) regardless of whether the male was sexually active. The extent to which the male giant panda is responding to photoperiodic cues is unknown, but occasional out-of-season matings (SB 390 mated with a female in autumn 2001) and sperm production (see Chapter 7) suggest that testicular function in this species is not strictly seasonal. More research is needed to understand the role of androgens in modulating spermatogenesis and the expression of reproductive behaviours in the giant panda.
Hormonal monitoring for applied management now and in the future
It is well accepted that endocrine monitoring is an essential tool in the routine management of a breeding pair (or population) of giant pandas. The technology is critical for establishing the receptive period of the female to optimise timed mating or AI. This is important because paired animals can be seriously antagonistic outside peak sexual receptivity. Knowing the precise status of ovarian function (combined with behavioural and vaginal cytology data) can be a comfort to managers who are responsible for the tricky introduction of such individuals at the best time to achieve successful mating. Additionally, the technology is now sufficiently advanced to provide informative endocrine data within three to four hours of recovering a urine sample. The same advantage holds for using urinary progestins at the end of gestation to estimate day of parturition accurately or, in the case of pseudopreg-nancy, end an intensive pregnancy watch so that staff can return to normal duties. With the advent of EIA systems, endocrine laboratories can (and have been) established on-site at giant panda facilities with minimal cost and instrumentation.
Presently, studies are in place or planned to use non-invasive endocrine monitoring in captive giant pandas to evaluate:
1. more reproductive cycles for more individuals of diverse ages (e.g. pubertal onset or reproductive senescence) or in the breeding versus non-breeding season (e.g. what triggers seasonal ovarian activity);
2. behavioural cues predictive of reproduction;
3. optimal timing for AI using fresh versus thawed sperm;
4. pregnancy and expected parturition;
5. the influence of social and husbandry conditions (e.g. single vs.
multiple animals) or medical procedures (e.g. restraint or anaesthesia) on animal well-being and reproductive fitness.
Particularly exciting is the possibility of rapidly monitoring physiological indices of stress, information that will assist managers of captive giant pandas in optimising husbandry protocols to improve both quality of life and reproductive success. For example, enclosure space could be modified (or enriched) according to corticosteroid patterns until baseline 'stress' levels are reached. Stress could actually be measured between introduced male and female giant pandas in an attempt to understand the causes of behavioural sexual incompatibility and perhaps to develop remediation approaches.
This technology is also highly relevant for future field-related studies, including examining the influence of the potentially controversial procedure of radiocollaring wild pandas or of human disturbance or environmental disrupters (i.e. forestry practices, agriculture, pollutants and toxicants) on animal well-being. Faecal endocrine monitoring, in particular, holds great promise for integrating endocrinology with giant panda field ecology studies. The possibility of measuring adrenal activity to assess levels of human disturbance and encroachment in natural habitats is intriguing. For example, non-invasive corti-costeroid monitoring has been used to assess stress in free-living spotted owls (Wasser et al., 1997), and wolves and elk (Creel et al., 2001) subjected to logging and recreational snowmobile activity, respectively. This type of information can inform policies to minimise disruption of wild breeding areas to increase chances for species survival. Such technology may well be applicable to the giant panda, especially in the future when reintroduction might be used for repopu-lating reserves. Knowing how individual giant pandas react to specific reintroduction techniques or habitats could substantially enhance the emerging field of reintroduction science.
Was this article helpful?