Parallel and seasonal changes in gonadal and adrenal hormones in
male giant pandas (Ailuropoda melanoleuca)
DAVID C. KERSEY,* DAVID E. WILDT, JANINE L. BROWN, YAN HUANG, REBECCA J. SNYDER, AND STEVEN L. MONFORT
Smithsonian Conservation Biology Institute, Front Royal, VA 22630, USA (DCK, DEW, JLB, SLM)
China Conservation and Research Center for the Giant Panda, Wolong Nature Reserve, Wenchuan, Sichuan Province
623006, China (YH)
Zoo Atlanta, Atlanta, GA 30315, USA (RJS)
* Correspondent: dkersey@westernu.edu
The purpose of this study was to determine androgen and glucocorticoid (GC) hormonal patterns in male giant
pandas (Ailuropoda melanoleuca) by monitoring gonadal and adrenal metabolites in feces. Initial validation
experiments demonstrated comparable excretory patterns in urine versus feces for both androgen and GC
measures. Matched urinary and fecal androgen and GC were correlated strongly with each other in a single male
that was assessed over 2 years. A single pharmacological injection of adrenocorticotropic hormone caused a 15-
fold GC increase in feces above baseline within 10 h, a peak at 12 h, and a return to baseline at 20 h,
demonstrating the biological relationship between adrenal activation and GC excretion. Longitudinal androgen
and GC excretory profiles in male giant pandas housed at North American (n5 2) and Chinese (n5 3) facilities
were similar, with fecal androgens generally exceeding baseline coincident with the onset of the 5-month annual
breeding season (January?June), after which values returned to nadir. Similarly, fecal GC excretion increased
during the breeding season but was baseline thereafter. Fecal androgen and GC in a single male monitored
through transition from subadult to sexual maturity also occurred in parallel. In this individual, basal fecal
androgen and GC increased 88% and 66%, respectively, from 5 to 6 years of age. Collectively, these data
demonstrate seasonal variations in gonadal activity in the giant panda by measuring androgen metabolites in
feces, with elevations consistently occurring from January through June before a return to baseline for ,4?
6 months. Findings also reveal a similar temporal rise in adrenal GC patterns associated with breeding season
onset, perhaps a mechanism to enhance metabolism, maximize body energy stores, and provide a competitive
advantage in achieving mating opportunities. Examination of data from a single male suggests that the ability to
produce these seasonal androgen and GC elevations is age dependent and occurs coincident with puberty.
DOI: 10.1644/09-MAMM-A-404.1.
Key words: adrenocorticotropic hormone (ACTH) challenge, fecal hormones, male giant panda, puberty, reproductive
seasonality
E 2010 American Society of Mammalogists
The giant panda (Ailuropoda melanoleuca) exhibits repro-
ductive seasonality, with breeding (in situ or ex situ) occurring
from February through June, with a peak in April (Steinman et
al. 2006). Almost all assessments of variation in sexual
receptivity have focused on the female, especially its
predilection for an extremely short (24- to 72-h) annual
window of lordotic behavior (estrus) that almost always occurs
coincident with increasing day length (Snyder et al. 2004;
Steinman et al. 2006). In contrast, males are considered
capable of reproducing for a more protracted interval during
the breeding season (Howard et al. 2006; Snyder et al. 2004).
For example, under captive management conditions, male
giant pandas breed with multiple females over months of time
(Xie and Gipps 2008) while producing prodigious numbers of
high-quality spermatozoa (Howard et al. 2006). In nature,
despite females having a short estrus, males compete for a
single, estrual female, converging on her location for combat,
with the winner having the opportunity to copulate (Schaller et
al. 1985; Zhu et al. 2001). Males then presumably search for
other sexually receptive females.
The specifics of reproductive seasonality in the male giant
panda are less understood than those in the female. During the
w w w . m a m m a l o g y . o r g
Journal of Mammalogy, 91(6):1496?1507, 2010
1496
months of sexual activity adult males produce sperm-dense
ejaculates composed of mostly morphologically normal,
highly motile spermatozoa (Howard et al. 2006). More
recently, distinctive variations in sperm production have been
associated with time of year; including some (but not all)
males becoming azoospermic during periods of gonadal
quiescence (C. Aitken-Palmer, University of Florida, pers.
comm.). Urinary profiles of androgen and glucocorticoid (GC)
metabolites have revealed that androgen excretion is seasonal,
increasing coincident with the general period of anticipated
female sexual receptivity (Bonney et al. 1982; MacDonald et
al. 2006; Snyder et al. 2004). Temporal GC excretion in urine
also yields a seasonal pattern similar to that of androgens, with
GC values elevated during the breeding season and at basal
levels in late spring, summer, and fall (MacDonald et al. 2006;
Owen et al. 2005).
These collective efforts (conducted in a total of 13 male
giant pandas) have demonstrated that both androgens and GCs
appear seasonally dependent. Building on initial findings from
urinary profiles, we considered an alternative idea?that such
hormone metabolite fluctuations could be discernible in feces.
The use of feces is advantageous because feces are much more
readily available and easier to collect than urine, thereby
providing the opportunity for better data interpretation through
increased sample sizes. Additionally, once validated, assess-
ments of fecal steroids can be adapted to field investigations
where perturbed habitats are believed responsible for the
uncertain status of this endangered species in nature (Vin?a et
al. 2007). Giant pandas defecate frequently (,20 times a day),
and it is not unusual to encounter giant panda fecal boluses in
the wild (Schaller et al. 1985; Wei et al. 2000; Vin?a et al.
2007). Therefore, we suspected that feces could be used to
estimate testicular and adrenal hormone status in free-ranging
giant pandas. However, the 1st step was to validate efficacy
and correspondence with behavioral and physiological chang-
es. We were confident in this approach because fecal steroid
monitoring already had been used to evaluate androgen
metabolites in male brown (Ursus arctos?Ishikawa et al.
2002) and sun (Helarctos malayanus?Hesterman et al. 2005)
bears, and GC excretion in sloth (Melursus ursinus?Young et
al. 2004), brown (von der Ohe et al. 2004), and Asiatic black
(Ursus thibetanus?Young et al. 2004) bears.
We examined longitudinal androgen and GC metabolite
profiles in the feces of multiple giant pandas housed in ex situ
collections in both the United States and in China. Our
specific objectives were 1) to compare 2 approaches for
extracting fecal hormone metabolites and then validate
enzyme immunoassays for quantifying fecal androgen and
GC patterns; 2) to determine the excretory route and time
course of GC excretion after exogenous administration of
adrenocorticotropic hormone (ACTH); and 3) to examine the
correspondence between urinary and fecal androgen and GC
measures and characterize gonadal and adrenal hormone
seasonality across a reasonably sized cohort of giant pandas.
Although our goal was to develop a comprehensive under-
standing of testicular and adrenal function in adult animals, we
also had the unique opportunity to examine the impact of age.
This effort complemented a parallel investigation that
measured estrogen and progestagen metabolites in female
giant pandas associated with the estrous cycle and pregnancy,
and throughout the year (Kersey et al. 2010a, 2010b).
MATERIALS AND METHODS
Study animals and facilities.?The Smithsonian National
Zoological Park (39uN, 77uW) and Zoo Atlanta (33uN, 84uW)
each managed 1 male (SB458, year of birth 1997; and SB461,
year of birth 1997; respectively) throughout the course of this
study (2001?2005). Water was provided ad libitum, and the
giant pandas were fed a freshly cut bamboo diet (10?14 kg
daily) supplemented with fruit and a diet biscuit high in fiber,
vitamins, minerals, and amino acids (Edwards et al. 2006).
Each animal was housed in an indoor (50- to 100-m2)?outdoor
(100- to 300-m2) complex. Although both males were
managed within auditory, olfactory, and visual proximity of
a female conspecific, only the male at the Smithsonian
National Zoological Park had consistent physical contact with
an adult female outside of the annual estrual period.
At the time of our study (2001, 2002, 2004) the China
Conservation and Research Center for the Giant Panda at the
Wolong Nature Reserve (31uN, 103uE) managed 25 adult
male giant pandas (age range, 8?17 years), of which only 3
males (SB308, year of birth 1986; SB394, year of birth 1992;
and SB399, year of birth 1993) had sufficient fecal collection
for statistical analysis. Individuals were fed a diet composed
primarily of freshly cut bamboo supplemented with a high-
fiber biscuit with water available ad libitum. Indoor (30- to 60-
m2) and outdoor (100- to 300-m2) areas were part of each
enclosure. Physical interaction with estrual females was
permitted during the breeding season (February?June, the
interval in which females had been observed in behavioral
estrus) for mating; otherwise, males were housed individually
and limited to auditory, visual, and olfactory contact with
conspecifics of both sexes. Care and maintenance of all
animals in this study followed guidelines of the American
Society of Mammalogists for use of mammals in research
(Gannon et al. 2007).
Sample collection and processing.?Fresh, morning urine
samples were collected 3?7 days/week and stored frozen
(220uC) in capped and labeled plastic specimen tubes (12 3
75 mm) until analysis. All urine samples were aspirated from a
solid, clean substrate free from contact with feces and standing
water to avoid cross-contamination and dilution. During the
study period ,1,000 urine samples were collected from
SB458 and ,800 from SB461. Urinary hormone concentra-
tions were indexed with creatinine (Cr?Taussky 1954) to
account for variations in water excretion and expressed as
hormone mass per milligram of Cr (ng/mg Cr). In brief, urine
samples were diluted (1:10 in bovine serum albumin?free
phosphate buffer) and added (0.05 ml) to a microtiter plate
(Dynatech Laboratories, Inc., Chantilly, Virginia) along with
Cr standard (0.00625?0.1 mg/ml; Sigma-Aldrich, St. Louis,
December 2010 KERSEY ET AL.?MALE GIANT PANDA ENDOCRINOLOGY 1497
Missouri) in duplicate. To each standard and sample distilled
H2O, 0.75 N NaOH, and 0.4 N picric acid were added (0.05 ml
each), and the assay was allowed to incubate at room
temperature (25uC) for 30 min prior to assessing optical
density (reading filter 490 nm, reference 620 nm) on a
microplate spectrophotometer (Dynex MRX; Dynex Technol-
ogies). Urine samples that were too dilute (,0.1 mg Cr/ml;
,10% of samples) were not analyzed for hormone content and
were discarded.
Freshly voided feces (,1 h postexcretion), free from contact
with standing water or urine, were collected each morning (1?
7 days/week) and stored frozen (220uC) in sealable plastic bags
until processing. Fecal samples collected in China were shipped
frozen to Smithsonian Conservation Biology Institute for
processing after securing the appropriate permits for interna-
tional transfer of this biomaterial. We used ,300 fecal samples
from males at Wolong Nature Reserve and ,2,000 fecal
samples from SB458 and SB461 combined. Prior to hormone
extraction fecal samples were freeze-dried (Lyophilizer;
Labconco, Kansas City, Missouri), crushed to a powder, and
extracted using previously described methods (Kersey et al.
2010a). A subset of fecal samples collected during the ACTH
challenge trial (see below) was used to compare hormone
profiles generated from the extraction of undried (0.5 g) versus
dried (0.1 g) feces. The extraction procedure followed methods
described previously (Wasser et al. 1994), and fecal extractants
were reconstituted in phosphate buffer (1 ml) and then stored
frozen (220uC) until analysis.
Adrenocorticotropic hormone challenge.?To artificially
stimulate the adrenal cortex male SB458 was injected
intramuscularly with 2 IU/kg of a synthetic ACTH (Cortrosyn;
Wedgewood Pharmacy, Swedesboro, New Jersey) on a single
day in October (injection at ,0800 h). Because this animal
had been conditioned to enter a training cage and coopera-
tively allow injection and venipuncture, it was assumed that
this ACTH challenge was non- (or minimally) stressful. A
blood sample (,3 ml) was collected 1 h before and 1 h after
ACTH injection, centrifuged within 1 h, and the serum stored
at 220uC. A freshly voided fecal sample also was obtained 1 h
before administering ACTH, and then samples of all urine and
feces were collected over the course of the next 3 days (72 h)
and stored frozen as described above. For the first 24 h this
translated into 17 fecal and 2 urine samples (at 13.8 and
23.3 h).
Endocrine analyses.?Androgen metabolite concentrations
in diluted urine and fecal extracts were quantified with a
single-antibody testosterone enzyme immunoassay (Dloniak et
al. 2004; Munro and Lasley 1988) that cross-reacted 100%
with testosterone, 57.4% with dihydrotestosterone, 0.3% with
androstenedione, 0.04% with androsterone and dehydroepian-
drosterone, and 0.02% for all other tested analytes, including
cortisol (Dloniak et al. 2004; C. Munro, University of
California, Davis, pers. comm.). Polyclonal anti-testosterone
(R156/7?C. Munro, pers. comm.) was diluted (1:7,500),
added (0.05 ml) to microtiter plates (96 well, Nunc-Immuno,
Maxisorp; Fisher Scientific, Pittsburgh, Pennsylvania), and
allowed to equilibrate (12?18 h) at 4uC. Prior to the addition
of samples in duplicate (unprocessed urine, equivalent to
0.0002?0.0025 ml; fecal extract, equivalent to 0.0005?
0.005 ml) and standards in triplicate (0.05 ml; 47?12,000 pg/
ml; 17b-hydroxy-4-androsten-3-one; Steraloids, Newport,
Rhode Island), unadsorbed antiserum was removed with wash
solution. Enzyme-conjugated testosterone (0.05 ml; C. Munro,
pers. comm.) then was added to each well containing standard
or sample and incubated (2 h; 25uC) before unbound
components were removed. A chromagen solution was added
(0.1 ml) to each well and allowed to incubate (,30 min)
before optimal densities were determined (maximum binding
5 1.00 optical density; reading filter 405 nm; reference filter
540 nm) on the microtiter plate reader. Interassay coefficients
of variation (CVs) for 2 internal controls (n 5 98 assays) were
12.9% (mean binding, 28.9%) and 12.6% (mean binding,
67.9%), and intraassay CV was ,10%. Immunoreactivity of
serially diluted urine and fecal extracts paralleled standard
binding. A linear regression analysis of testosterone standard
added to urine (r2 5 1.00, y 5 1.2532 3.85; F1,6 5 7,388.41,
P  0.001) and fecal extracts (r2 5 0.98, y 5 0.943 2 0.01;
F1,6 5 371.19, P  0.001) to unaltered standards demonstrat-
ed significant recovery.
A single-antibody cortisol enzyme immunoassay (Munro
and Lasley 1988; Young et al. 2001) that cross-reacted 100%
with cortisol, 9.9% with prednisolone, 6.3% with prednisone,
6.2% with 11-deoxycortisol, 5% with cortisone, 0.7% with
corticosterone, 0.5% with 21-deoxycortisone, 0.3% with
desoxycorticosterone, and 0.1% with androgens, including
testosterone, dihydrotestosterone, and androstenedione (C.
Munro, pers. comm.; Young et al. 2001), was used to analyze
GC metabolite concentrations in all serum, urine, and fecal
samples. Microtiter plates (96 well, Nunc-Immuno, Maxisorp;
Fisher Scientific) were allowed to adsorb cortisol antibody
(R4866, C. Munro) for 12?18 h (4uC) before adding duplicate
samples (unprocessed urine and fecal extract, equivalent to
0.001?0.005 ml) and triplicate cortisol standards (0.05 ml;
range 0.08?1,000 ng/ml; 17-hydroxycorticosterone; Sigma-
Aldrich). Plates were incubated at room temperature (25uC;
1 h) after adding (0.05 ml) of an enzyme-linked cortisol (C.
Munro, pers. comm.). Following incubation, unbound com-
ponents were removed with a wash, and a chromagen solution
(ABTS in citrate buffer; Sigma-Aldrich) was added (0.1 ml) to
all wells. When optimal optical density (1.00) was reached, the
resulting color change was quantified on the microtiter plate
reader. Interassay CV for 2 internal controls (n 5 108 assays)
was 12.7% (mean binding, 37.1%) and 13.3% (mean binding,
71.9%), and intraassay CV was ,10%. Both urine and feces
demonstrated parallel displacement with the cortisol enzyme
immunoassay. Significant recoveries were demonstrated when
unlabeled standard was added to urine (linear regression, r2 5
0.99, y 5 1.153 2 4.2; F1,6 5 978.95, P  0.001) and fecal
extracts (r2 5 1.00, y 5 0.903 2 3.38: F1,6 5 7,320.67, P 
0.001).
High-pressure liquid chromatography.?Fecal androgen
and GC metabolites were analyzed by reverse-phase high-
1498 JOURNAL OF MAMMALOGY Vol. 91, No. 6
pressure liquid chromatography (Varian ProStar; Varian
Analytical Instruments, Lexington, Massachusetts) with po-
larity gradients as previously described (Staley et al. 2007).
For androgen metabolite identification fecal extracts from 6
samples containing elevated androgen concentrations were
pooled, evaporated to dryness, reconstituted in 0.3 ml of
phosphate-buffered saline (0.01 M NaPO4, 0.14 NaCl, 0.5%
bovine serum albumin, pH 5.0), and spiked with tritiated [3H]
testosterone (,14,000 cpm/ml) and 3H androstenedione
(,14,000 cpm/ml) to act as cochromatographic markers.
The pooled sample (0.3 ml of phosphate-buffered saline) for
GC metabolite separation was obtained from the combination
of 6 extracts of a fecal sample collected ,12 h post-ACTH
administration with ,14,000 cpm/ml of 3H cortisol, 3H
corticosterone, and 3H desoxycorticosterone added to the pool
as radiolabeled markers. Retention times of radioactive
markers were determined by combining 0.1 ml of each high-
pressure liquid chromatography fraction with 3 ml of
scintillation cocktail, and radioactivity was quantified on a
radioactive beta counter (Beckman Coulter, Inc., Brea,
California). Residual fractions were dried under air, resus-
pended in 0.3 ml of assay buffer, and assayed, and
immunoactive peaks were compared to the retention times
of radiolabeled markers.
Statistical analyses.?Baseline urinary and fecal androgen
and GC metabolites concentrations were determined through
an iterative process (Moreira et al. 2001). Briefly, for each
data set concentrations in excess of the mean plus 2 SDs were
removed until no values exceeded mean + 2 SD. The resulting
mean was considered baseline and expressed as mean 6
standard error of the mean (SE). Additionally, data sets were
evaluated by calendar month and expressed as a mean value
(6 SE). All data sets were tested for normality (Kolmogorov?
Smirnov test) before additional statistical tests were conducted
(all tests, a 5 0.05). The correspondence between hormone
standards and samples (urine and feces) spiked with known
concentrations of standard was assessed using linear regres-
sion, with all residuals examined for variation to the regression
line being ,2 standardized residuals. Associations between
data sets were evaluated using Pearson product moment
correlation (r). Comparisons between fecal androgen and GC
were conducted for both baseline mean and overall mean (all
sample hormone concentrations in the collection period)
values. Monthly trends in androgen and GC concentrations
were determined via 1-way repeated-measures analysis of
variance (ANOVA; with normality and equal variance
assumptions accepted at P. 0.05), with multiple comparisons
versus baseline values conducted by the Holm?Sidak method
(White et al. 2005). Breeding success was determined based
on a male?s ability to mate with a female and produce
offspring. The inclusion of SB458 in the multiyear portion of
the study (starting at 3 years of age as a subadult) also allowed
examining endocrine changes associated with pubertal onset.
Thus, data were examined for statistically significant changes
and the overall relationship of androgen and GC over time
with a 1-way repeated-measures ANOVA (Holm?Sidak
method). Differences between wet and dry extracts of feces,
androgen, and GC values between males (SB458 and SB461)
and between years of the same male were determined by the
Mann?Whitney U-test. The Kruskal?Wallis (H) 1-way
ANOVA test was performed to assess differences among
breeding season androgen and GC values by reproductive
success, with post hoc analysis conducted via Tukey?s test.
Baseline iterations were conducted using Microsoft Excel
2007 (Microsoft, Inc., Seattle, Washington), and all other
statistical tests were conducted using SigmaStat 3.1 (Systat
Software, Inc., Point Richmond, California).
RESULTS
High-pressure liquid chromatography.?Immunoreactive
fecal androgen peaks quantified after high-pressure liquid
chromatography separation co-eluted with 3H-testosterone and
3H-androstenedione at fractions 33 and 43, constituting 11.0%
and 2.9% of total immunoreactivity, respectively (Fig. 1A).
An additional 5 unidentified immunoreactive peaks were
detected by androgen enzyme immunoassay. The largest
unidentified peak (fractions 9?13) represented 38.0% of total
immunoreactivity, with the additional peaks occurring at
fractions 15 (10.8%), 19 (7.4%), 25 (15.6%), and 56 (2.2%).
FIG. 1.?High-pressure liquid chromatography cochromatographic
profiles of 3H A) androgens and B) glucocorticoids with correspond-
ing fecal androgen and glucocorticoid metabolites, respectively.
Arrows mark elution of tritiated markers; A) testosterone and
androstenedione, and B) cortisol, corticosterone, and desoxycortico-
sterone.
December 2010 KERSEY ET AL.?MALE GIANT PANDA ENDOCRINOLOGY 1499
Five distinct immunoreactive peaks were identified by
cortisol enzyme immunoassay after high-pressure liquid
chromatography, with the largest peak (fractions 46?52;
52.2% total immunoreactivity) co-eluted with 3H corticoste-
rone (fraction 48; Fig. 1B). The 2nd most prominent
immunoreactive metabolite (fraction 42; 12.8%) co-eluted
with 3H cortisol, whereas 3 additional unidentified peaks
occurred at fractions 18, 39, and 57, representing 6.5%, 5.8%,
and 4.1% of total immunoreactivity, respectively.
Matched urinary and fecal endocrine measures.?Androgen
patterns in urine versus feces of giant panda SB458 were
similar temporally (Pearson product moment correlation, r 5
0.61, n 5 24, P 5 0.001; Fig. 2A), as were urinary and fecal
GC profiles (r 5 0.53, n 5 24, P 5 0.008; Fig. 2B). We also
found statistical correspondence between androgen and GC
excretion in urine (r 5 0.74, n 5 24, P 5 0.00003) and feces
(r 5 0.62, n 5 24, P 5 0.001).
Adrenocorticotropic hormone challenge.?Serum, urine,
and fecal GC and androgen concentrations in SB458 before
and after ACTH administration are depicted in Fig. 3. Serum
GC concentrations increased 11-fold (from a baseline of 1.1 to
12.8 ng/ml) within 1 h post-ACTH administration, whereas
circulating androgen increased 2.3-fold (from a nadir of 1.1 to
2.6 ng/ml). Of the 2 urine samples collected within 24 h of
ACTH administration, the 1st at 13.8 h represented a GC value
(711.1 ng/mg Cr) that was 5.6-fold above the pre-ACTH nadir
of 128.7 6 11.9 ng/mg Cr (range, 91.4?176.3 ng/mg Cr). The
androgen content (8.3 ng/mg Cr) in this same sample was
within the pre-ACTH nadir range (X? , 11.0 6 0.56 ng/mg Cr;
range, 4.6?17.0 ng/mg Cr). The 2nd urine sample (at 23 h
post-ACTH) had GC (56.1 ng/mg Cr) and androgen (10.1 ng/
mg Cr) concentrations below and within the range of pre-
ACTH values, respectively. The 1st post-ACTH fecal sample
(at ,8 h) yielded a 14.9-fold increase in GC (from a baseline
of 143.6 6 15.7 to 2,133.6 ng/g) and an 8.2-fold increase in
androgen (from a baseline of 82.8 6 8.1 to 677.9 ng/g). Peak
fecal GC (3,479.6 ng/g) was detected ,12 h post-ACTH and
remained elevated for ,7 h more before declining to
preinjection concentrations at 20 h. In contrast, fecal androgen
concentrations were reduced by 8 h post-ACTH (234.1 ng/g)
and were basal (27.7 ng/g) by 9.8 h. Although fecal GC had
declined by 20 h post-ACTH, several fecal samples after this
period contained GC concentrations above baseline, with the
sample at 36 h being at nadir.
FIG. 2.?Matched A) monthly (6 SE) urinary (open triangles,
dashed line) and fecal (closed triangles, solid line) androgen
metabolites, and B) monthly (6 SE) urinary (closed circles, solid
line) and fecal (open circles, dashed line) glucocorticoid metabolites
for 24 months in a single adult male giant panda (SB458). Data are
aligned to month of the year.
FIG. 3.?Serum, urinary, and fecal A) glucocorticoid and B)
androgen concentrations before and after exogenous adrenocortico-
tropic hormone (ACTH) administration to a single male giant panda
(SB458). Data are aligned to the hour time of ACTH administration.
1500 JOURNAL OF MAMMALOGY Vol. 91, No. 6
Temporal fecal GC excretion patterns in dried versus
undried giant panda feces were highly correlated (Pearson
product moment correlation, r 5 0.92, n 5 20, P , 0.001).
However, mean metabolite concentration for the sample set (n
5 20) was 7.8-fold greater (Mann?Whitney U5 58.0, n5 20,
21, P  0.001) in dried (64.6 6 13.4 ng/g) versus undried
feces (8.2 6 1.2 ng/g).
Seasonality in male giant pandas maintained in North
America.?We observed no within-animal differences in
average endocrine values for either of the 2 giant pandas
monitored over 2 years (2003 and 2004) in North America
(SB458?androgens: U 5 12,711.0, n 5 149, 185, P 5 0.222;
GC: U 5 12,082.0, n 5 149, 166, P 5 0.724; SB461?
androgens: U 5 13,008.0, n 5 170, 173, P 5 0.065; GC: U 5
14,424.0, n 5 170, 173, P 5 0.76). Therefore, to allow a more
detailed examination of the existence of gonadal seasonality,
hormonal data for both years were combined and averaged by
month for each individual (6 SE; Fig. 4). For SB458 temporal
trends in fecal androgen and GC excretion were similar
(Pearson product moment correlation, r 5 0.74, n 5 12, P 5
0.006; Fig. 4A). Fecal androgens were elevated above
baseline (156.3 6 4.1 ng/g; F12,322 5 5.89, P , 0.001)
during February (243.5 6 29.6 ng/g; Holm?Sidak method, P
5 0.006), March (260.2 6 25.9 ng/g, P , 0.001), October
(256.5 6 21.9 ng/g, P, 0.001), November (255.66 26.4 ng/g,
P 5 0.001), and December (298.9 6 35.6 ng/g, P , 0.001).
The seasonal trend in fecal GC excretion was less
pronounced with concentrations of this metabolite only
exceeding baseline concentrations (233.3 6 5.1 ng/g;
F12,303 5 2.89, P , 0.001) in October (403.8 6 31.2 ng/g,
P , 0.001) and December (459.1 6 96.8 ng/g, P 5 0.001).
Likewise, temporal trends in fecal androgen and GC
excretion were similar (r 5 0.61, n 5 12, P 5 0.04) in
SB461 (Fig. 4B). Fecal androgen concentrations were
elevated above baseline (111.8 6 2.4 ng/g; F12,330 5
14.31, P , 0.001) in January (150.9 6 9.2 ng/g, P ,
0.001), February (184.9 6 15.2 ng/g, P , 0.001), and March
(226.1 6 14.0 ng/g, P , 0.001). Fecal GC concentrations
only exceeded baseline (159.4 6 2.8 ng/g; F12,328 5 6.35,
P , 0.001) in March (326.5 6 40.8 ng/g, P , 0.001) and
May (233.2 6 20.0 ng/g, P 5 0.002).
When overall fecal hormone concentrations were compared
between males, SB458 produced more androgen (209.1 6
7.3 ng/g; U 5 36,759, n 5 334, 343, P  0.001) and GC
(314.4 6 12.6 ng/g; U 5 27,539, n 5 315, 343, P  0.001)
than SB461 (133.6 6 3.5 ng/g and 194.4 6 5.8 ng/g,
respectively), a ,1.6-fold difference. This variation in overall
concentration over time also extended to a higher basal
excretion level for SB458 for androgens (156.3 6 4.1 ng/g; U
5 24,033, n5 268, 287, P  0.001) and GC (233.36 5.1 ng/g;
U 5 15,168, n 5 242, 284, P  0.001) compared to SB461
(androgens, 111.8 6 2.4 ng/g; GC, 159.4 6 2.8 ng/g).
Seasonality in male giant pandas maintained in China.?
Fecal androgen and GC metabolite profiles of 3 representative
males from Wolong Nature Reserve are presented in Fig. 5.
Androgen and GC measures for male SB308 (Fig. 5A) were
correlated significantly (Pearson product moment correlation,
r 5 0.76, n 5 10, P 5 0.01) throughout the collection period.
Androgen concentrations were elevated above baseline (78.8
6 2.9 ng/g; repeated-measures ANOVA, F10,61 5 3.10, P ,
0.001) only during February (294.1 6 87.9 ng/g; Holm?Sidak
method, P , 0.001); however, GC values exceeded baseline
(122.3 6 4.3 ng/g; F10,97 5 6.20, P , 0.001) throughout most
of the breeding season (February, 440.5 6 86.0 ng/g, P ,
0.001; March, 341.6 6 72.1 ng/g, P , 0.001; April, 295.2 6
68.9 ng/g, P 5 0.002). The correlation between androgen and
GC excretory patterns for male SB394 (Fig. 5B) was not
significant (r 5 0.53, n 5 10, P 5 0.10). Fecal androgen
metabolites were greater than basal values (112.8 6 7.5 ng/g;
F10,56 5 6.72, P , 0.001) for the first 2 months of the
breeding season (February, 209.4 6 17.9 ng/g, P , 0.001;
March, 256.7 6 36.8 ng/g, P , 0.001) and for 3 months
during autumn (September, 238.6 6 41.2 ng/g, P , 0.001;
October, 221.7 6 73.5 ng/g, P , 0.001; November, 207.1 6
41.3 ng/g, P , 0.001). Fecal GCs were elevated above
baseline (228.0 6 8.3 ng/g; F10,92 5 6.58, P , 0.001) for only
the month of March (638.7 6 121.8 ng/g, P , 0.001).
Although appearing similar, trends in androgen and GC
excretion for male SB399 (Fig. 5C) were not correlated (r 5
0.66, n 5 10, P 5 0.09). Androgen values were not greater
(F10,61 5 1.98, P 5 0.052) than baseline (100.5 6 5.0 ng/g)
throughout the collection period; however, monthly GC
FIG. 4.?Seasonal trends in fecal androgen (closed triangles, solid
line) and glucocorticoid (open circles, dashed line) excretion
averaged by month (6 SE) over a 2-year interval in male giant
pandas A) SB458 and B) SB461. Data are aligned by month of
the year.
December 2010 KERSEY ET AL.?MALE GIANT PANDA ENDOCRINOLOGY 1501
concentrations exceeded baseline (125.8 6 6.2 ng/g; F10,97 5
3.86, P , 0.001) during the breeding season months of
February (368.8 6 136.6 ng/g, P , 0.001) and March (348.2
6 93.7 ng/g, P , 0.001).
Relationship of androgen and GC concentrations to
breeding success.?Androgen concentrations during the
breeding season (February?June) did not differ according to
overall reproductive success in the years of evaluation or
during the reproductive life span of a given giant panda (H5 5
11.04, P 5 0.051). Conversely, GC concentrations differed
(H5 5 18.65, P 5 0.002) by reproductive success, with SB458
(a male that never mated) excreting lower (161.5 6 22.1 ng/g)
during the breeding season (Tukey?s test, P , 0.05) than
SB308 (618.7 6 66.1 ng/g) and SB394 (599.8 6 148.3 ng/g,
two males that had successfully mated).
Male pubertal onset.?On average, overall and baseline
androgen concentrations of a single male (SB458) that was
evaluated intensively for 5 years and through puberty onset
differed (overall: F4,564 5 207.83, P , 0.001; baseline: F4,420
5 741.53, P , 0.001) with age (Table 1). Most notable
increases (Holm?Sidak method, P , 0.05) in androgen
concentrations occurred between ages 3 and 4 (1.7- and 1.3-
fold increase in overall and baseline androgens, respectively),
and 5 and 6 (1.7- and 1.9-fold increase in overall and baseline
androgens, respectively).
Mean overall and baseline GC concentrations (Table 1)
varied with age (overall: F4,548 5 81.67, P , 0.001; baseline:
F4,419 5 715.56, P , 0.001). Although overall GC
concentrations increased (P , 0.05) 1.3-fold from 3 to 4 years
of age, no differences (P . 0.05) were detected in overall GC
excretion or basal values from 3 to 5 years of age. As observed
for androgen measures, overall (1.4-fold) and baseline (1.6-
fold) fecal GC excretion increased (P, 0.05) from 5 to 6 years
of age.
DISCUSSION
This work demonstrated, for the 1st time, the feasibility of
monitoring seasonal fluctuations in testicular and adrenal
hormonal metabolites in the highly fibrous feces of the giant
panda. Excreted androgens were elevated during what has
typically been identified as the 5-month-long breeding season
for the male of this species (Snyder et al. 2004; Steinman et al.
2006), which contrasted substantially to the remarkably short
(24- to 72-h) interval for the conspecific female (Czekala et al.
1998, 2003; Lindburg et al. 2001; Monfort et al. 1989;
Steinman et al. 2006). This prolonged interval of maximal
testosterone production no doubt supports protracted prodi-
gious sperm production (Howard et al. 2006) and likely
provides incentive for multiple breeding opportunities with
various females (Pan et al. 2004; Schaller et al. 1985; Zhu et
al. 2001). The synchronized elevations in GC could be related
to greater metabolism that assists wild giant pandas sustain
energy, ambulate longer distances, and express aggressive
behaviors important for winning breeding competitions. Our
results also suggested that physiological puberty appeared to
be occurring at ,4 years of age. Finally, it was evident that
male giant pandas generally expressed the same endocrine
profiles (temporally and in amplitude) regardless of whether
they lived in North American zoos or in a Chinese breeding
center, thereby confirming that ex situ conditions were
influencing endocrine patterns minimally, if at all. Male
pandas maintained in urban Washington, D.C., and Atlanta,
Georgia, and exposed to millions of visitors annually,
demonstrated endocrine profiles no different than counterparts
living in a remote, high-occupancy breeding center within a
Chinese national reserve.
The circannual fecal androgen patterns observed in our
study were similar to fluctuations measured in urine from 9
giant pandas (MacDonald et al. 2006) and other ursids,
including black (Ursus americanus?Garshelis and Hellgren
1994; McMillin et al. 1976; Palmer et al. 1988), polar (Ursus
maritimus?Palmer et al. 1988), and brown (Tsubota and
Kanagawa 1989) bears. Regardless of species or use of the
urinary or fecal approach, androgen concentrations were
highest by mating season onset and basal by end of the
breeding period. From the intensive monitoring of the pandas
in North American zoos, the seasonal androgen peak in males
(February) was preceded by a gradual rise that began as early
FIG. 5.?Representative fecal androgen (closed triangles, solid
line) and glucocorticoid (open circles, dashed line) excretion profiles
averaged by month (6 SE) in male giant pandas A) SB308, B)
SB394, and C) SB399 living ex situ in the Wolong Nature Reserve.
Shaded area represents the known interval of the species? breeding
season. Data are aligned to month of the year.
1502 JOURNAL OF MAMMALOGY Vol. 91, No. 6
as September. Although this pattern has been inconsistent in
other species, prebreeding increases in urinary androgens have
been measured in brown (Tsubota and Kanagawa 1989) and
American black (Palmer et al. 1988) bears. This persistent,
steady rise is believed important for promoting spermatogen-
esis while playing a role in renal and gastrointestinal
physiology during hibernation (Garshelis and Hellgren 1994;
Nelson et al. 1978). However, variations among species
clearly exist, because the giant panda does not experience
hibernation, perhaps because the primary food source
(bamboo) is available year-round (Garshelis 2004). Although
increased androgen production is a prerequisite for testicular
activity and spermatogenesis in the giant panda, other factors
that also likely assist in evoking a full repertoire of sexual
behaviors in this species must be examined, such as olfactory
and auditory cues or simply female presence.
Based on our experiences, assessing hormonal metabolites
in giant panda urine required less time and expense than fecal
analysis. However, urine collection can be complicated and
laborious, especially in situations requiring simultaneous
monitoring of multiple animals living on soil substrates or in
facilities where urine becomes contaminated with water (e.g.,
from cleaning). The fecal strategy is better suited to centers
managing many pandas where feces can be recovered at the
convenience of the care-givers rather than watching and
waiting for individual animals to urinate. Furthermore, giant
pandas experience rapid gut transit of only 6?7 h (Dierenfeld
et al. 1982; Edwards et al. 2006), which results in frequent
defecations that occur more often than voiding urine. Repeated
defecation is advantageous, because samples are readily
available, easy to collect, and represent discrete periods of
physiological activity (Monfort 2003; Schwartz and Monfort
2008). The disadvantage is that fecal processing and analysis
are cost- and time-intensive in the laboratory, usually
requiring 15?20 h/100 samples to generate data. Nonetheless,
for large-scale characterization and comparative investigations
(as in the present study), the fecal approach is most practical.
This assertion also was supported by our discovery that results
were unaffected by the need for preemptive drying of feces.
This process normally requires another 120?240 h for
lyophilization per batch of fecal samples. The ability to
eliminate this step reduces turnaround time and obviates the
need to purchase expensive lyophilizer devices that generally
have not been available in Chinese giant panda breeding
centers. Finally, results revealed that fecal sampling 2?3 days
per week was sufficient to identify trends in androgen and GC
excretion over time.
Although high-pressure liquid chromatography indicated
the presence of several GC metabolites, the group-specific
antibody in our cortisol enzyme immunoassay had the
capacity to quantify basal and peak GC activity. The validity
of this assay was confirmed by ACTH challenge testing.
Within hours of ACTH administration, clear evidence of
adrenal activation and a subsequent increase in GC excretion
were seen, with the lag time being ,24 h for both urine and
feces. The interval for a detectable response in feces (,10 h)
was shorter than for urine (,13 h), which no doubt reflected
the rapid movement of the bamboo diet through the gut
(Dierenfeld et al. 1982; Edwards et al. 2006). Our observations
also suggested that the ability to identify an acute perturbation
or stressful event on the basis of fluctuating GC in a giant
panda required at least once- or twice-daily fecal sampling.
However, adrenal longitudinal tendencies or patterns can be
determined by collecting only a few samples per week over a
sustained period of time.
To maximize information learned we evaluated androgen
excretion in serum, urine, and feces pre- and post-ACTH.
Although urinary androgens did not change, ACTH appeared
to stimulate androgen metabolite excretion into panda feces
within 2 h. A short-term androgen burst in response to ACTH
also has been observed in the cheetah (Acinonyx jubatus?
Wildt et al. 1984), pudu (Pudu puda?Bubenik and Reyes-
Toledo 1994), fallow deer (Dama dama?Bartos? et al. 2004),
white-tailed deer (Odocoileus virginianus?Bubenik et al.
1990), spotted hyena (Crocuta crocuta?Lindeque et al 1986),
and red squirrel (Tamiasciurus hudsonicus?Boonstra et al.
2008). For the giant panda the difference between the urine
and fecal reaction likely was due to preferential metabolism
and excretion of androgen metabolites by the gastrointestinal
tract rather than via renal clearance. Regardless, the acute rise
TABLE 1.?Comparison of overall and baseline fecal androgen and glucocorticoid concentrations (mean 6 SE, ng/g) over successive years in
giant panda male SB458.
Age
Androgen Glucocorticoid
Overall Baseline Overall Baseline
3 61.5 6 3.6a (n 5 120) 46.8 6 2.1a (n 5 99) 170.1 6 7.7a (n 5 120) 145.2 6 4.1a (n 5 105)
4 104.9 6 6.2b (n 5 145) 61.7 6 2.6a (n 5 97) 228.9 6 13.3b (n 5 150) 137.2 6 4.3a (n 5 96)
5 113.2 6 5.8c (n 5 203) 70.4 6 2.3b (n 5 143) 213.4 6 11.9ab (n 5 176) 139.0 6 4.7a (n 5 125)
6 192.5 6 9.8d (n 5 160) 132.8 6 5.0c (n 5 120) 296.0 6 14.1c (n 5 139) 231.3 6 7.1b (n 5 110)
7 182.5 6 9.9d (n 5 143) 130.0 6 4.3c (n 5 108) 297.7 6 19.4c (n 5 143) 219.0 6 7.6b (n 5 112)
F-testa 207.83 741.53 81.67 715.56
d.f. 4,564 4,420 4,548 4,419
r2b 0.88* 0.86* 0.85* 0.67
a One-way repeated-measures ANOVA of within-column means (P , 0.001 for all tests); different lowercase letters within a column indicate significant differences (Holm?Sidak
method, P , 0.05).
b Linear regression between age (explanatory variable) and mean hormone concentration (response variable).
* P , 0.05.
December 2010 KERSEY ET AL.?MALE GIANT PANDA ENDOCRINOLOGY 1503
and fall in fecal androgen content post-ACTH helped
substantiate minimal androgen cross-reactivity in the cortisol
enzyme immunoassay. Although some analyte cross-reactivity
could have occurred between the 2 assays, fecal GC values
were distinguished by a longer elevation post-ACTH than
fecal androgens (20 versus 8.9 h, respectively). If cross-
reactivity had been a factor, the GC profile would have
declined coincidentally with the androgen pattern. Also, our
findings were consistent with a wealth of information in the
human literature demonstrating differential androgen and GC
production by the adrenals after ACTH administration
(McKenna et al. 1997).
As many as 42% of adult, male giant pandas have failed to
breed in ex situ collections (Zhang et al. 2006). Our androgen
excretion data were uninformative in predicting males that
were reproductively competent versus those that were not.
Individuals that were proven breeders had androgen levels and
temporal profiles no different (and in some case substantially
less) than males failing to copulate. Such comparisons can be
challenging due to natural variations and dynamism within or
among males that are partially due to individual animal
differences in metabolic, excretory, or hormonal clearance, or
a combination of these. Yet, others have found that older
American black bears (.8 years of age) secrete less androgen
than younger counterparts, and that perhaps this condition
reduces the ability to compete for females (Garshelis and
Hellgren 1994). Although we found no indication of a
relationship between androgen concentrations and overall
breeding success, determining the linkage (if any) between
this hormone and explicit reproductive behaviors is a rich area
for future investigations. Although this topic has received
considerable attention in female giant pandas (Bonney et al.
1982; Kleiman 1984; Kleiman et al. 1979; Lindburg et al.
2001; Murata et al. 1986; Schaller et al. 1985; Snyder et al.
2004; Swaisgood et al. 2002, 2003), including being a major
factor limiting breeding programs in China (Ellis et al. 2006),
the relationship among male behaviors and reproductive
success largely is unknown. Aggression toward estrual
females is common among nonbreeding males (Ellis et al.
2006), which might reflect abnormal androgen patterns.
Greater understanding could result from longitudinal moni-
toring of androgen metabolite profiles and concentrations in
pandas that display usual versus aberrant or unusual behaviors,
especially during the period surrounding estrus.
Two earlier investigations reported a seasonal and coinci-
dental rise in urinary androgen and GC in the giant panda
(MacDonald et al. 2006; Owen et al. 2005). We observed a
similar annual rhythm via fecal monitoring, with GC peaking
during the breeding season (February?June) and returning to
baseline by summer. Seasonal GC fluctuation is believed to be
related to changes in reproductive behavior, energy mobiliza-
tion, and exposure to seasonal stressors (Romero 2002).
However, the mechanisms regulating the interplay of these
events, including seasonal adrenal activation, is not well
studied or understood. Increased GC during or just before the
reproductive season is a common feature in the gray wolf
(Canis lupus?Sands and Creel 2004), southern muriqui
(Brachyteles arachnoides?Strier et al. 1999), tufted capuchin
monkey (Cebus apella nigritus?Lynch et al. 2002), squirrel
monkey (Saimiri boliviensis boliviensis?Schiml et al. 1996),
and arctic ground squirrel (Spermophilus parryii?Boonstra et
al. 2001a, 2001b). One of the primary physiological effects of
GC is to convert stored protein and lipids into carbohydrates
(Ferin 2006), and this metabolic property may be necessary to
facilitate competition and breeding behaviors observed in
reproductively seasonal males. Ex situ studies have demon-
strated that male giant pandas generally increase activity
patterns and reduce appetite as the breeding season approach-
es, especially when nearby females enter estrus (Kleiman et al.
1979; Snyder et al. 2004). Furthermore, wild male giant
pandas engage in battles with conspecifics in pursuit of mating
opportunities (Pan et al. 2004; Schaller et al. 1985; Zhu et al.
2001). Therefore, androgens and GC might act synergistical-
ly?with anabolic effects of androgens augmented by
metabolic effects of GC?to enhance virility.
Although only 1 animal was available for exploring
pubertal transition, we measured the ontogeny of an important
hormonal change for the 1st time. It was apparent that
androgen concentrations increased steadily in male giant
pandas from 3 to 7 years of age, with adult concentrations
achieved by 6 years. The latter age also coincided with the 1st
annual increase in GC production. These endocrine data fit
well with studbook records indicating that the youngest male
ever to have mated successfully and sired a cub was 5.5 years
of age (Xie and Gipps 2008). Onset of male sexual behaviors
(e.g., depressed appetite, increased vocalizations, scent
marking, and locomotion) occurs at 5 years of age (Kleiman
et al. 1979; Snyder et al. 2004), and sperm have been collected
by electroejaculation from pandas 5.5 years of age, although it
is of lesser quality than that from older counterparts (Howard
et al. 2006). The male grizzly bear (Ursus arctos horribilis) is
considered adult at 5.5 years when spermatozoa accumulate in
the seminiferous and epididymal tubules (White et al. 1998).
By contrast, serum testosterone concentrations, mate pairings,
and body weight are greatest in American black bear males
that are 4 years of age or older (Garshelis and Hellgren 1994).
A detailed and longitudinal investigation of giant panda
puberty that integrates endocrine data with behavior and
sperm production is underway in our laboratory.
In conclusion, we have demonstrated the usefulness of fecal
monitoring for characterizing seasonal excretion patterns of
androgen and GC metabolites in the giant panda. This has
increased our fundamental understanding of male reproductive
biology, but it is only the 1st step in exploiting this
noninvasive monitoring approach for this endangered species.
Historically, important findings have been made from
conducting endocrine monitoring in free-ranging wildlife
(Monfort 2003; Schwartz and Monfort 2008). Opportunities
exist for studying wild giant pandas to determine the
variability (or similarity) among males in reproductive
function and competency, or if seasonal reproductive patterns
in nature mimic those observed in captive individuals.
1504 JOURNAL OF MAMMALOGY Vol. 91, No. 6
Particularly exciting would be determining adrenal function in
the context of reproductive fitness and if endocrine patterns
are altered by disturbances to the local environment. Such
studies now are feasible given molecular advances in
identifying individuals and paternity (David et al. 2006) and
improved abilities to locate quality fecal samples in nature,
including with dogs (Lasley and Kirkpatrick 1991; Millspaugh
and Washburn 2004; Monfort 2003; Schwartz and Monfort
2008; von der Ohe and Servheen 2002, Wasser et al. 2004).
ACKNOWLEDGMENTS
We thank N. Parker, C. Aitken-Palmer, A. Crosier, N. Presley, K.
Steinman, R. Stewart, and S. Walker for logistical support. Special
thanks are extended to C. Bazlett, J. Beckman, S. Enloe, V. Parkman,
N. Savageau, and B. von Holdt for laboratory assistance. We also
acknowledge the enormous contributions made by the keeper and
curatorial staffs at the Smithsonian National Zoological Park, Zoo
Atlanta, and the China Conservation and Research Center for the
Giant Panda in frequent and accurate sample collection. This study
was supported by The Friends of the National Zoo and the
Smithsonian National Zoological Park.
LITERATURE CITED
BARTOS?, L., G. A. BUBENIK, AND M. TOMA?NEK. 2004. The role of
adrenal androgens in antler growth of castrated fallow deer (Dama
dama): possible stimulation of adrenal secretion of testosterone by
stress. Pp. 45?53 in Advances in antler science and product
technology (J. Suttie, S. Haines, and C. Li, eds.). AgResearch,
Invermay Agricultural Centre, Mosgiel, New Zealand.
BONNEY, R. C., D. J. WOOD, AND D. G. KLEIMAN. 1982. Endocrine
correlates of behavioural oestrus in the female giant panda
(Ailuropoda melanoleuca) and associated hormonal changes in
the male. Journal of Reproduction and Fertility 64:209?215.
BOONSTRA, R., A. H. HUBBS, E. A. LACEY, AND C. J. MCCOLL. 2001a.
Seasonal changes in glucocorticoid and testosterone concentrations
in free-living arctic ground squirrels from the boreal forest of the
Yukon. Canadian Journal of Zoology 79:49?58.
BOONSTRA, R., ET AL. 2008. Plasma DHEA levels in wild, territorial
red squirrels: seasonal variation and effect of ACTH. General and
Comparative Endocrinology 158:61?67.
BOONSTRA, R., C. J. MCCOLL, AND T. J. KARELS. 2001b. Reproduction
at all costs: the adaptive stress response of male arctic ground
squirrels. Ecology 82:1930?1946.
BUBENIK, G. A., AND E. REYES-TOLEDO. 1994. Plasma levels of
cortisol, testosterone and growth hormone in pudu (Pudu puda
Molina) after ACTH administration. Comparative Biochemistry
and Physiology, A. Comparative Physiology 107:523?527.
BUBENIK, G. A., J. H. SMITH, D. K. POMERANTZ, AND D. SCHAMS. 1990.
Plasma LH, FSH, testosterone, prolactin and androstenedione in
male white-tailed deer after ACTH and dexamethasone adminis-
tration. Comparative Biochemistry and Physiology, A. Compara-
tive Physiology 95:163?169.
CZEKALA, N., D. LINDBURG, B. DURRANT, R. SWAISGOOD, T. HE, AND C.
TANG. 1998. The estrogen profile, vaginal cytology, and behavior
of a giant panda female during estrus. Pp. 111?113 in Proceedings
of the international symposium on the protection of the giant panda
(Ailuropoda melanoleuca) (A. J. Zhang and G. X. He, eds.).
Sichuan Publishing House of Science and Technology, Chengdu,
China.
CZEKALA, N., L. MCGEEHAN, K. STEINMAN, L. XUEBING, AND F. GUAL-
SIL. 2003. Endocrine monitoring and its application to the
management of the giant panda. Zoo Biology 22:389?400.
DAVID, V. A., ET AL. 2006. Parentage assessment among captive giant
pandas in China. Pp. 245?273 in Giant pandas: biology, veterinary
medicine and management (D. E. Wildt, A. Zhang, H. Zhang, D. L.
Janssen, and S. Ellis, eds.). Cambridge University Press, Cam-
bridge, United Kingdom.
DIERENFELD, E. S., H. F. HINTZ, J. B. ROBERTSON, P. J. VAN SOEST, AND
O. T. OFTEDAL. 1982. Utilization of bamboo by the giant panda.
Journal of Nutrition 112:636?641.
DLONIAK, S., J. A. FRENCH, N. J. PLACE, M. L. WELDELE, S. E.
GLICKMAN, AND K. E. HOLEKAMP. 2004. Non-invasive monitoring of
fecal androgens in spotted hyenas (Crocuta crocuta). General and
Comparative Endocrinology 135:51?61.
EDWARDS, M. S., G. ZHANG, R. WEI, AND X. LIU. 2006. Nutrition and
dietary husbandry. Pp. 101?158 in Giant pandas: biology,
veterinary medicine and management (D. E. Wildt, A. Zhang, H.
Zhang, D. L. Janssen, and S. Ellis, eds.). Cambridge University
Press, Cambridge, United Kingdom.
ELLIS, S., ET AL. 2006. Life histories and behavioural traits as
predictors of breeding success. Pp. 87?100 in Giant pandas:
biology, veterinary medicine and management (D. E. Wildt, A.
Zhang, H. Zhang, D. L. Janssen, and S. Ellis, eds.). Cambridge
University Press, Cambridge, United Kingdom.
FERIN, M. 2006. Stress and the reproductive system. Pp. 2627?2696 in
Knobil and Neill?s physiology of reproduction (J. D. Neill, ed.).
Elsevier Academic Press, St. Louis, Missouri.
GANNON, W. L., R. S. SIKES, AND THE ANIMAL CARE AND USE COMMITTEE
OF THE AMERICAN SOCIETY OF MAMMALOGISTS. 2007. Guidelines of
the American Society of Mammalogists for the use of wild
mammals in research. Journal of Mammalogy 88:809?823.
GARSHELIS, D. L. 2004. Variation in ursid life histories. Pp. 53?73 in
Giant pandas: biology and conservation (D. G. Lindburg and K.
Baragona, eds.). University of California Press, Berkeley.
GARSHELIS, D. L., AND E. C. HELLGREN. 1994. Variation in
reproductive biology of male black bears. Journal of Mammalogy
75:175?188.
HESTERMAN, H., S. K. WASSER, AND J. F. COCKREM. 2005. Longitudinal
monitoring of fecal testosterone in male Malayan sun bears (U.
malayanus). Zoo Biology 24:403?417.
HOWARD, J. G., ET AL. 2006. Male reproductive biology in giant
pandas in breeding programmes in China. Pp. 159?197 in Giant
pandas: biology, veterinary medicine and management (D. E.
Wildt, A. Zhang, H. Zhang, D. L. Janssen, and S. Ellis, eds.).
Cambridge University Press, Cambridge, United Kingdom.
ISHIKAWA, A., S. KIKUCHI, S. KATAGIRI, H. SAKAMOTO, AND Y.
TAKAHASHI. 2002. Efficiency of fecal steroid hormone measure-
ment for assessing reproductive function in the Hokkaido brown
bear (Ursus arctos yesoensis). Japanese Journal of Veterinary
Research 50:17?27.
KERSEY, D. C., D. E. WILDT, J. L. BROWN, R. J. SNYDER, Y. HUANG,
AND S. L. MONFORT. 2010a. Endocrine milieu of periestrus in the
giant panda (Ailuropoda melanoleuca) as determined by noninva-
sive hormone measures. Reproduction, Fertility and Development
22:901?912.
KERSEY, D. C., D. E. WILDT, J. L. BROWN, R. J. SNYDER, Y. HUANG,
AND S. L. MONFORT. 2010b. Unique biphasic progestagen profile in
parturient and nonparturient giant pandas (Ailuropoda melano-
leuca) as determined by faecal hormone monitoring. Reproduction
140:183?193.
December 2010 KERSEY ET AL.?MALE GIANT PANDA ENDOCRINOLOGY 1505
KLEIMAN, D. G. 1984. Social and reproductive behavior of the giant
panda (Ailuropoda melanoleuca). Pp. 45?58 in Proceedings of the
international symposium on giant panda (Ailuropoda melanoleuca)
(A. J. Zhang and G. X. He, eds.). Sichuan Publishing House of
Science and Technology, Chengdu, China.
KLEIMAN, D. G., W. B. KARESH, AND P. R. CHU. 1979. Behavioural
changes associated with oestrus in the giant panda Ailuropoda
melanoleuca with comments on female proceptive behaviour.
International Zoo Yearbook 19:217?223.
LASLEY, B. L., AND J. F. KIRKPATRICK. 1991. Monitoring ovarian
function in captive and free-ranging wildlife by means of urinary
and fecal steroids. Journal of Zoo and Wildlife Medicine 22:23?32.
LINDBURG, D. G., N. M. CZEKALA, AND R. R. SWAISGOOD. 2001.
Hormonal and behavioral relationships during estrus in the giant
panda. Zoo Biology 20:537?543.
LINDEQUE, M., J. D. SKINNER, AND R. P. MILLAR. 1986. Adrenal and
gonadal contribution to circulating androgens in spotted hyaenas
(Crocuta crocuta) as revealed by LHRH, hCG and ACTH
stimulation. Journal of Reproduction and Fertility 78:211?217.
LYNCH, J. W., T. E. ZIEGLER, AND K. B. STRIER. 2002. Individual and
seasonal variation in fecal testosterone and cortisol levels of wild
male tufted capuchin monkeys, Cebus apella nigritus. Hormones
and Behavior 41:275?287.
MACDONALD, E., N. CZEKALA, F. GUAL-SIL, AND T. NAKAO. 2006.
Urinary testosterone and cortisol metabolites in male giant pandas
Ailuropoda melanoleuca in relation to breeding, housing, and
season. Acta Zoologica Sinica 52:242?249.
MCKENNA, T. J., U. FEARON, D. CLARKE, AND S. K. CUNNINGHAM. 1997.
A critical review of the origin and control of adrenal androgens.
Baillie`re?s Clinical Obstetrics and Gynaecology 11:229?248.
MCMILLIN, J. M., U. S. SEAL, L. ROGERS, AND A. W. ERICKSON. 1976.
Annual testosterone rhythm in the black bear (Ursus americanus).
Biology of Reproduction 15:163?167.
MILLSPAUGH, J. J., AND B. E. WASHBURN. 2004. Use of fecal
glucocorticoid metabolite measures in conservation biology
research: considerations for application and interpretation. General
and Comparative Endocrinology 138:189?199.
MONFORT, S. L. 2003. Non-invasive endocrine measures of reproduc-
tion and stress in wild populations. Pp. 147?165 in Reproduction
and integrated conservation science (W. V. Holdt, A. R. Pickard, J.
C. Rodger, and D. E. Wildt, eds.). Cambridge University Press,
Cambridge, United Kingdom.
MONFORT, S. L., K. D. DAHL, N. M. CZEKALA, L. STEVENS, M. BUSH,
AND D. E. WILDT. 1989. Monitoring ovarian function and
pregnancy in the giant panda (Ailuropoda melanoleuca) by
evaluating urinary bioactive FSH and steroid metabolites. Journal
of Reproduction and Fertility 85:203?212.
MOREIRA, N., ET AL. 2001. Reproductive steroid hormones and ovarian
activity in felids of the Leopardus genus. Zoo Biology 20:103?116.
MUNRO, C. J., AND B. L. LASLEY. 1988. Non-radiometric methods for
immunoassay of steroid hormones. Pp. 289?329 in Non-radiomet-
ric assays: technology and application in polypeptide and steroid
hormone detection (B. D. Albertson and F. D. Haseltine, eds.). A.
R. Liss, Inc., New York.
MURATA, K., M. TANIOKA, AND N. MURAKAMI. 1986. The relationship
between the pattern of urinary oestrogen and behavioural changes
in the giant panda Ailuropoda melanoleuca. International Zoo
Yearbook 24/25:274?279.
NELSON, R. A., D. L. WELLIK, J. M. MCMILLIN, AND P. J. PALUMBO.
1978. Role of testosterone in hibernating black bears. Physiologist
21:84.
OWEN, M. A., N. M. CZEKALA, R. R. SWAISGOOD, K. STEINMAN, AND D.
G. LINDBURG. 2005. Seasonal and diurnal dynamics of glucocor-
ticoids and behavior in giant pandas. Ursus 16:208?221.
PALMER, S. S., R. A. NELSON, M. A. RAMSAY, I. STIRLING, AND J. M.
BAHR. 1988. Annual changes in serum sex steroid in male and
female black (Ursus americanus) and polar (Ursus maritimus)
bears. Biology of Reproduction 38:1044?1050.
PAN, W., Y. LONG, D. WANG, H. WANG, Z. LU, AND X. ZHU. 2004.
Future survival of giant pandas in the Qinling Mountains of China.
Pp. 81?87 in Giant pandas: biology and conservation (D. G.
Lindburg and K. Baragona, eds.). University of California Press,
Berkeley.
ROMERO, L. M. 2002. Seasonal changes in plasma glucocorticoid
concentrations in free-living vertebrates. General and Comparative
Endocrinology 128:1?24.
SANDS, J., AND S. CREEL. 2004. Social dominance, aggression and
faecal glucocorticoid levels in a wild population of wolves, Canis
lupus. Animal Behaviour 67:387?396.
SCHALLER, G., H. JINCHU, P. WENSHI, AND Z. JING. 1985. The giant
pandas of Wolong. University of Chicago Press, Chicago,
Illinois.
SCHIML, P. A., S. P. MENDOZA, W. SALTZMAN, D. M. LYONS, AND W. A.
MASON. 1996. Seasonality in squirrel monkeys (Saimiri scirureus):
social facilitation by females. Physiology and Behavior 60:1105?
1113.
SCHWARTZ, M. K., AND S. L. MONFORT. 2008. Genetic and endocrine
tools for carnivore surveys. Pp. 228?250 in Noninvasive survey
methods for North American carnivores (R. A. Long, P. MacKay,
J. C. Ray, and W. J. Zielinski, eds.). Island Press, Washington,
D.C.
SNYDER, R. J., ET AL. 2004. Reproduction in giant pandas. Pp. 125?132
in Giant pandas: biology and conservation (D. G. Lindburg and K.
Baragona, eds.). University of California Press, Berkeley.
STALEY, A., J. BLANCO, A. DUFTY, D. WILDT, AND S. MONFORT. 2007.
Fecal steroid monitoring for assessing gonadal and adrenal activity
in the golden eagle and peregrine falcon. Journal of Comparative
Physiology, B. Biochemical, Systematic, and Environmental
Physiology 177:609?622.
STEINMAN, K., ET AL. 2006. Endocrinology of the giant panda and
application of hormone technology to species management. Pp.
198?230 in Giant pandas: biology, veterinary medicine and
management (D. E. Wildt, A. Zhang, H. Zhang, D. L. Janssen,
and S. Ellis, eds.). Cambridge University Press, Cambridge, United
Kingdom.
STRIER, K. B., T. E. ZIEGLER, AND D. J. WITTWER. 1999. Seasonal and
social correlates of fecal testosterone and cortisol levels in wild
male muriquis (Brachyteles arachnoides). Hormones and Behavior
35:125?134.
SWAISGOOD, R. R., D. G. LINDBURG, AND H. ZHANG. 2002.
Discrimination of oestrus in giant pandas (Ailuropoda melano-
leuca) via chemical cues in urine. Journal of Zoology (London)
257:381?386.
SWAISGOOD, R. R., X. ZHOU, G. ZHANG, D. G. LINDBURG, AND H.
ZHANG. 2003. Application of behavioural knowledge to conserva-
tion in the giant panda. International Journal of Comparative
Psychology 16:65?84.
TAUSSKY, H. H. 1954. A microcolorimetric determination of creatine
in urine by the Jaffe reaction. Journal of Biological Chemistry
208:853?861.
TSUBOTA, T., AND H. KANAGAWA. 1989. Annual changes in serum
testosterone levels and spermatogenesis in the Hokkaido brown
1506 JOURNAL OF MAMMALOGY Vol. 91, No. 6
bear, Ursus arctos yesoensis. Journal of the Mammalogical Society
of Japan 14:11?17.
VIN?A, A., ET AL. 2007. Temporal changes in giant panda habitat
connectivity across boundaries of Wolong Nature Reserve, China.
Ecological Applications 17:1019?1030.
VON DER OHE, C. G., AND C. SERVHEEN. 2002. Measuring stress in
mammals using fecal glucocorticoids: opportunities and challeng-
es. Wildlife Society Bulletin 30:1215?1225.
VON DER OHE, C. G., S. K. WASSER, K. E. HUNT, AND C. SERVHEEN.
2004. Factors associated with fecal glucocorticoids in Alaskan
brown bears (Ursus arctos horribilis). Physiological and Biochem-
ical Zoology 77:313?320.
WASSER, S. K., ET AL. 2004. Scat detection dogs in wildlife research
and management: application to grizzly and black bears in the
Yellowhead Ecosystem, Alberta, Canada. Canadian Journal of
Zoology 82:475?492.
WASSER, S. K., S. L. MONFORT, J. SOUTHERS, AND D. E. WILDT. 1994.
Excretion rates and metabolites of oestradiol and progesterone in
baboon (Papio cynocephalus cynocephalus). Journal of Reproduc-
tion and Fertility 101:213?220.
WEI, F., Z. FENG, Z. WANG, AND J. HU. 2000. Habitat use and
separation between the giant panda and the red panda. Journal of
Mammalogy 81:448?455.
WHITE, D., JR., J. G. BERARDINELLI, AND K. E. AUNE. 1998.
Reproductive characteristics of the male grizzly bear in the
continental United States. Ursus 10:497?501.
WHITE, D., JR., J. G. BERARDINELLI, AND K. E. AUNE. 2005. Seasonal
differences in spermatogenesis, testicular mass and serum
testosterone concentrations in the grizzly bear. Ursus 16:198?207.
WILDT, D. E., D. MELTZER, P. K. CHAKRABORTY, AND M. BUSH. 1984.
Adrenal?testicular?pituitary relationships in the cheetah subjected
to anesthesia/electroejaculation. Biology of Reproduction 30:665?
672.
XIE, Z., AND J. GIPPS. 2008. International giant panda (Ailuropoda
melanoleuca) studbook. Chinese Association of Zoological
Gardens, Beijing, China.
YOUNG, K. M., J. L. BROWN, AND K. L. GOODROWE. 2001.
Characterization of reproductive cycles and adrenal activity in
the black-footed ferret (Mustela nigripes) by fecal hormone
analysis. Zoo Biology 20:517?536.
YOUNG, K. M., S. L. WALKER, C. LANTHIER, W. T. WADDELL, S. L.
MONFORT, AND J. L. BROWN. 2004. Noninvasive monitoring of
adrenocortical activity in carnivores by fecal glucocorticoid
analyses. General and Comparative Endocrinology 137:148?165.
ZHANG, Z., ET AL. 2006. Historical perspective of breeding giant
pandas ex situ and high priorities for the future. Pp. 455?468 in
Giant pandas: biology, veterinary medicine and management (D. E.
Wildt, A. Zhang, H. Zhang, D. L. Janssen, and S. Ellis, eds.).
Cambridge University Press, Cambridge, United Kingdom.
ZHU, X., D. G. LINDBURG, W. PAN, K. A. FRONEY, AND D. WANG. 2001.
The reproductive strategy of giant pandas (Ailuropoda melano-
leuca): infant growth and development and mother?infant
relationships. Journal of Zoology (London) 253:141?155.
Submitted 6 November 2009. Accepted 8 July 2010.
Associate Editor was Jane M. Waterman.
December 2010 KERSEY ET AL.?MALE GIANT PANDA ENDOCRINOLOGY 1507