Effects of natural environmental conditions on faecal glucocorticoid metabolite concentrations in jaguars (Panthera onca) in Belize

We performed immunoassay and environmental validations to assess the effects of natural environmental exposure of jaguar faeces on glucocorticoid metabolites. Glucocorticoid concentrations were stable for 5 days with the corticosterone RIA. Scat aging by morphology was unreliable. Environmental validation is needed to determine field sample collection protocols for elusive species.


Introduction
Non-invasive hormone monitoring (NHM) is widely used in a variety of wildlife species, but requires rigorous testing to ensure the validity of techniques employed to assess steroid metabolites in excreta, such as faeces (Möstl et al., 2005;Palme, 2005;Touma and Palme, 2005).The applicability of this powerful tool has been demonstrated in several areas of animal research, such as ethology, reproductive physiology and animal health (Goymann et al., 2001;Wielebnowski et al., 2002;Brown, 2006;Franceschini et al., 2008;Dorsey et al., 2010).Most of the studies employing NHM, however, are performed in ex situ settings, such as zoos or research centres, where samples can be collected fresh and exposure to environmental conditions is limited.While a substantial amount of research using NHM has been conducted in free-ranging populations of primates, less has been done in other species of mammals, reptiles and amphibians (Schwarzen berger, 2007).The reasons for this disparity are attributed to limitations in observing fresh defaecations without eliciting adrenal activity by the presence of personnel performing sample collections, difficulties in assessing sample age and/or assigning collected samples to an individual, and instability of steroid hormone concentrations in faeces after defaecation (Möstl and Palme, 2002;Sheriff et al., 2011).However, recent advances in the use of remote tracking have proved valuable for following individuals to collect scat without disturbance (Ganswindt et al., 2010).Molecular techniques, such as DNA detection in animal excreta, are now employed to monitor populations non-invasively and assign individual identity or gender to animals from scat samples (Piggott et al., 2006;Barja et al., 2008).The remaining challenge is to ensure that faecal hormone metabolite concentrations are stable, within the time frame of collection, after deposited scat is exposed to natural environmental conditions.
The increased interest in employing NHM to aid wildlife conservation efforts has provided incentive to develop sampling methods that are robust in field conditions.For instance, primatologists have tested various sample storage conditions to avoid changes in hormone concentrations and facilitate transport of excreta and hormone extracts from the field to the laboratory (Khan et al., 2002;Lynch et al., 2003;Pappano et al., 2010;Shutt et al., 2012).Recent efforts in other mammals have targeted the design of 'field-friendly' techniques to increase the applicability of NHM to field endocrinology.Some of these methods include hormone extraction, sample storage and immunoassay of samples in field conditions (Freeman et al., 2010;Santymire and Armstrong, 2010).It is widely accepted that hormonal metabolites in faeces are affected by factors such as temperature, humidity and ultraviolet rays, which can influence the presence of flora that alter hormone concentrations directly and indirectly through biotransformation processes (Macdonald et al., 1983;Touma and Palme, 2005;Schwartz and Monfort, 2008).There is evidence that steroid hormones can increase or decrease in concentration after exposure to the environment (Washburn and Millspaugh, 2002;Abáigar et al., 2010), although it is not always clear whether steroid hormone metabolites are genuinely changing in concentration or if there is a conversion of metabolites to forms that have higher or lower affinity to the antibody.The lack of this information for most species hampers our ability to use NHM in in situ wildlife studies (Kelly et al., 2012).Another caveat arises in accurately ageing faecal samples exposed to environmental conditions.Most field-based studies determine faecal age through subjective judgement based on scat morphology, such as categorizing samples based on condition, colour, moisture level, odour strength, presence of mould and presence of invertebrates (e.g.Vynne et al., 2012); however, these sample characteristics can vary dramatically based on environmental conditions other than simply scat age.
Most of the research using NHM in wild populations of elusive species has occurred in temperate environments, yet the effects of the environment on hormone metabolite concentrations have not been assessed directly in the majority of these studies.For instance, a survey of faecal glucocorticoid metabolites (FGMs) of Alaskan brown bears (Ursus arctos horribilis) conducted during the summer months in Katmai National Park, Alaska, considered 2 days as prudential time for hormone concentration stability (von der Ohe et al., 2004).Other studies have been performed strictly during the snow-filled winter, thus avoiding any type of change in hormone metabolite concentrations, which are stable when frozen [e.g.elk (Cervus elaphus) and wolves (Canis lupus) in Yellowstone National Park, Creel et al., 2002;capercaillie (Tetrao urogallus), Thiel et al., 2008].While there have been some attempts to evaluate the effects of environment on hormone concentration for NHM, in general the studies have been fairly limited (Fanson, 2009;Rehnus et al., 2009;Abáigar et al., 2010;Hulsman et al., 2011).
The objectives of the present study were to validate a cortisol enzyme immunoassay (EIA), to perform biological validations of a cortisol EIA and a corticosterone radioimmunoassay (RIA) for measuring FGMs in jaguars (Panthera onca) and to conduct an environmental validation as a means of assessing the effects of a natural environment, with its climate variations (dry and wet seasons), on the stability of FGM postdefaecation in jaguar scat.

Animals and sample collection
The jaguars (Panthera onca) included in this study were kept at the Belize Zoo, Belize, Central America.All individuals were wild-born adults (five males and four females, between ~6 and 10 years of age), with the exception of one captiveborn male (2 years of age).Most jaguars were not exposed to the public, except for two wild-born adults (exposure to the public for >2 years) and the captive-born male (exposure to the public for ~1 year).Jaguars not exposed to the public were housed individually; each enclosure had two dens and a common yard, where individuals alternated yard use every 24 h, without having direct contact with each other.The diet consisted mainly of chicken carcasses and other animal products when available.Water was provided ad libitum.

Study site and environment
The natural vegetation in Belize, a Neotropical country, is predominantly moist and wet broadleaf forest.In some areas, there are pine forests and more open pine savannahs.The climate follows bimodal tropical conditions separated by a cool transitional period.The dry season, from February to Toolbox May, is characterized by higher temperatures ranging from 24 to 33°C and scarce rainfall (≤100 mm month −1 ).The wet season, from June to November, has slightly lower temperatures (18-28°C) and predominant rains, ~60% of the annual precipitation, which on average fluctuates between 1500 and 2000 mm.The cool transition period occurs between November and January.Relative humidity oscillates around 80% year-round, with some fluctuations between the dry and wet seasons (Fuller and Wilson, 2002).

Study design
Fresh faecal samples (e.g.no older than 8 h) from all jaguars included in the study were collected from each individual at the Belize Zoo in both the wet and dry seasons over a 3 day period in each season (one or two samples per individual per season).Samples were placed in plastic bags and moved to the field site at the Tropical Education Centre (TEC) within 30 min, where they were exposed to two environmental conditions: shade (17.35707N, 088.54159W; 41 m elevation); and sun (17.35735N, 088.54121W; 39 m elevation).Faecal samples assigned to the shade treatment were placed under broadleaved trees that provided shade during daylight periods, whereas samples assigned to the sun treatment were placed under direct exposure to the sun.Collected faeces were thoroughly mixed before subsampling over time.A control (first sub-sample) was immediately frozen at −20°C to arrest degradation.Thereafter, daily sub-samples were collected from each scat and frozen over a 5 day period in both seasons.Morphological scat ageing was assessed using the following parameters: colour (dark brown, light brown, white) and moisture (scale: 1 = dry to 5 = very moist).Photographs of scats were also taken daily for later reference.

Weather variables
Temperature (in degrees Celsius), relative humidity (expressed as a percentage) and dew point (in degrees Celsius) were measured every 5 min with two data loggers (Lascar Electronics, EL-USB-2) placed at ground level in each set of each environmental conditions.A hand-held unit (Omega, HHUVA1) was used to measure ultraviolet (UV) rays (in watts per square metre) every 3 h at ground level during collection periods.

Extraction of steroids from faeces
At the Smithsonian Conservation Biology Institute (SCBI), faeces were freeze-dried, homogenized and pulverized.Dry faeces were extracted (0.19 ± 0.01 g) in 5 ml of ethanol 90% (v/v) by boiling in a water bath (90-100°C) for 20 min and centrifuging at 500g for 15 min.Supernatants were recovered and pellets resuspended in 5 ml of ethanol 90% (v/v), vortexed for 30 s and centrifuged again.Supernatants were combined and air-dried overnight.Extracts were resuspended in 1 ml of methanol and placed in an ultrasonic cleaner for 10 min.Extracts were diluted (1:1 v/v) with steroid buffer (0.1 m NaPO 4 and 0.149 m NaCl, pH 7.0) and stored at −20°C until analysis.
Faecal extracts were also analysed using a double-antibody 125 I-labelled corticosterone RIA (MP Biomedicals, Orangeburg, NY, USA), previously validated for jaguar faeces (Conforti et al., 2012), according to the manufacturer's instructions, except that all reagent volumes were halved.The polyclonal antiserum was raised in rabbits against corticosterone-3-carboxymethyloxime coupled to bovine serum albumin.Faecal extracts were diluted 1:250 in steroid diluent and analysed in duplicate.The inter-assay coefficient of variation (CV) of quality control samples run in each assay was <10% (cortisol EIA and corticosterone RIA: CV for high-dose control, 6.42 and 4.08%; CV for low-dose control, 9.65 and 8.04%; n = 14 and 3, respectively).The intra-assay CV was <10% for both immunoassays.Faecal glucocorticoid metabolite concentrations are expressed as nanograms per gram dry faecal matter (i.e.ng g −1 ).

Immunoassay validation
Faecal extracts from fresh samples (i.e.controls for environmental validation) were used to conduct parallelism, exogenous corticosteroid accuracy recovery, high-performance liquid chromatography (HPLC) and a biological validation for each immunoassay.

Parallelism
A test for parallelism between the assay standards and a pool of faecal extract was performed for both the corticosterone and cortisol immunoassays.An equal amount of faecal extract from each fresh sample was pooled and diluted serially, 2-fold, in steroid buffer.Serial dilutions were analysed in duplicates for each assay.

Exogenous corticosteroid accuracy recovery
Exogenous corticosteroid accuracy recovery was tested by adding a known amount corticosteroid, according to the standard/original antigen of the immunoassay, to a pool of jaguar faecal steroid extracts.Serial dilutions of the spiked combinations were analysed in duplicate.The percentage steroid recovery was calculated based on two factors, the expected concentrations (known amount added) and the observed concentrations (observed concentration minus endogenous concentration of the pool), by incorporating them into the following formula:

High-performance liquid chromatography
Reverse-phase HPLC was performed to analyse the immunoreactivity of a pool of faecal steroid extracts with the following known radioactive glucocorticoid tracers: 3 H-cortisol, 3 H-corticosterone and 3 H-deoxycorticosterone.A pool of faecal steroid extracts, from three males and three females, was filtered, dried and spiked with radioactive hormone markers.Fifty-five microlitres of sample extract was separated on a Microsorb column (Reverse Phase Microsorb™ MV 100 C-18, 5 µm diameter particle size; Varian Inc., Woburn, MA, USA) using a gradient of 20-100% (methanol:water) over 80 min (1.0 ml min −1 flow rate, 1 ml fractions) as described by Young et al. (2004).The percentage of methanol remained constant at 20% for 5 min and increased linearly to 100% over the remaining time.A 100 µl aliquot of each fraction was used to analyse radioactivity (in counts per minute).The remaining 900 µl per fraction was air-dried overnight and reconstituted in 150 µl of distilled water for immunoreactivity analysis.

Biological validation of immunoassays
Faeces from individuals that were assumed to undergo periods of high adrenal activity, i.e. a 'stressful' situation, were compared with faecal material from those that had not been exposed to a stressor.The Belize Zoo has a jaguar rehabilitation programme, which brings free-ranging, 'problem jaguars' into captivity to avoid the killing of jaguars by ranchers or villagers in retaliation for livestock losses.Free-ranging jaguars that are brought into captivity experience extreme adrenal challenges, such as transport, confinement, appetite loss and, potentially, self-injury.As jaguars acclimate to the new captive environment, their appetite increases, aggressive behaviours towards keeper staff decrease, and the jaguars spend less time hiding (personal communication from Sharon Matola, Director of the Belize Zoo).Jaguars are thought to take <1 year to acclimate, at least behaviourally, to captive conditions (personal communication from Sharon Matola).Furthermore, jaguars that spend >1 year in captivity show little or no aggressive behaviours and even affiliative expressions, such as body rubbing and vocalizations (e.g.gurgle and prusten) towards keeper staff (personal communication from Sharon Matola).Therefore, faecal samples were collected from each jaguar in the study, and individuals were grouped based on time spent in captivity as follows: <1 year (n = 2 males); between 1 and 4 years (n = 2 males and 2 females); >5 years (n = 1 male and 2 females); and captive born (n = 1 male).Four samples were collected per individual and averaged for statistical analysis.Jaguars recently brought into captivity (e.g.≤1 year) were expected to have higher FGM concentrations than those individuals that were captive born or had been exposed to long-term captivity (e.g.>5 years).

Statistical analysis
All analyses, except for the environmental validation assessment, were implemented in the statistical software JMP Pro 10 (version 10.0.2, 2012; SAS Institute Inc., Cary, NC, USA).All data were tested for normality before applying a statistical test.Normality was assessed with the Shapiro-Wilk goodness-of-fit test (α = 0.05).Data distributed in a non-normal fashion were logarithmically transformed.Multiple comparisons were performed when appropriate using the Tukey-Kramer HSD multiple comparisons test (α = 0.05).Biological validation was assessed by grouping FGM concentrations of jaguars by time spent in captivity using a one-way ANOVA.The test for parallelism was done with a multiple linear regression of logarithmically transformed concentration and the binding percentage of both the standard and serially diluted extracts; the least-squares means of the regression were linearly contrasted by standard and serially diluted samples, as suggested by Grotjan and Keel (1996).Expected and observed concentrations of the exogenous corticoid accuracy recovery analysis were tested using a simple linear regression for each immunoassay, the cortisol EIA and the corticosterone RIA.Environmental validation was assessed using a generalized linear mixed model constructed and run in the statistical software SAS (version 9.4, 2002-2012;SAS Institute Inc.), where FGM concentrations as repeated measures were a function of sex, season, environmental condition, time and their respective interactions.Scats of individual jaguars were assigned at random to seasons and treatments.Weather variables were condensed as daily averages.Condensed data were analysed in a nested model, where every weather variable was a function of the environmental condition (sun or shade) factor, which was nested in the season factor (wet and dry).
In the HPLC for the cortisol EIA, the majority of immunoactivity was observed in fractions 4 and 7-25, indicating the presence of highly polar glucocorticoid metabolites in jaguar faecal steroid extracts (Fig. 1).Another peak was observed at fraction 35.An immunoreactive peak at fraction 46 co-eluted with 3 H-corticosterone.Similar immunoreactivity patterns were observed for the corticosterone RIA, with the exception of an additional peak at fractions 49-50.Again, most of the cross-reactivity was observed in fractions 4-25, indicating the presence of more polar glucocorticoid metabolites (Fig. 1).

Weather variables
As shown in Fig. 3, all weather variables were significantly different between seasons (wet and dry) and environmental conditions (sun or shade).

Faecal glucocorticoid metabolite measures
Faecal glucocorticoid metabolite concentrations achieved normality after logarithmic transformation for both assays (cortisol EIA, W = 0.8869, P < 0.05; and corticosterone RIA, W = 0.8971, P < 0.05).Statistical analysis indicated that FGMs, measured with the cortisol EIA, were more stable overall in the dry season than in the wet season (d.f.= 20,

Scat age assessment
We were unable to determine accurately the age of scat exposed to different environmental conditions.Moisture and, to a lesser degree, colour changed markedly after a scat was exposed to the environment.Scats gained moisture after rainfall, causing the colour to turn darker.In general, scat samples became dryer and whiter over time in the dry season, but no such trend was observed in the wet season (Fig. 5).

6
Figure 3: Climatic variables, summarized in box and whisker plots, associated with environmental validation of jaguar FGMs.The number of samples for each climatic variable in the dry and wet seasons was 40 and 64, respectively; r 2 is the value of the square root of the significant model: weather variable = environmental condition (sun or shade) nested within a season (wet or dry) + error.Different letters denote significant differences.Dotted lines represent trends in environmental conditions within a season (temperature, d.f.= 3, F = 28.16,P < 0.0001; relative humidity, d.f.= 3, F = 32.88,P < 0.0001; dew point, d.f.= 3, F = 49.10,P < 0.0001; and UV, d.f.= 3, F = 30.69,P < 0.0001).

Toolbox Discussion
A cortisol EIA (R-4866 antibody) was validated to measure FGMs in jaguar scat.The results of the corticosterone RIA (MP Biomedicals) validation were congruent with a previous immunoassay validation performed by Conforti et al. (2012).Both immunoassays showed satisfactory parallel displacement and evidence of multiple glucocorticoid metabolites based on HPLC analysis, including an important amount of immunoactivity associated with more polar metabolites, similar to that observed in other felid species (i.e.cheetah, domestic cat, clouded leopard; Young et al., 2004).Contrary to those species, jaguar extracts also contained immunoactive peaks that co-eluted with 3 H-corticosterone, which is feasible for the RIA that uses a corticosterone antibody; however, the cortisol antibody crossreacts only 0.7% with this glucocorticoid, so it is unlikely to be native steroid in that case.For this HPLC analysis, we used a pool of faecal extracts from male and female jaguars.Unfortunately, we were unable to make inferences regarding sex-related differences in HPLC profiles of FGMs.Other studies have shown that there are sex-related differences in FGMs (Touma et al., 2003).Future studies could perform HPLC or other chromatographic techniques (e.g.gas chromatography coupled with mass spectrometry) to discern any potential differences in the array of FGMs between the sexes.Biological validity of both assays was demonstrated by showing a significant difference in FGM among individuals in relationship to the time spent in captivity.As expected, individuals that had spent <1 year in captivity exhibited much higher adrenal activity than those with 1-4 or >5 years in captivity or a captive-born individual.These results are also consistent with behavioural differences among jaguars relative to time in captivity noted by the Belize Zoo director and staff (personal communication from Sharon Matola).The biological validity of the corticosterone RIA is further supported by the results of a previous study of captive jaguars that were challenged with exogenous adrenocorticotrophic hormone (Conforti et al., 2012).Although different steroid extraction techniques were used [boiling (present study) vs. shaking (Conforti et al., 2012)], the FGM concentrations assessed with the corticosterone RIA of the captive-born jaguar and the individuals that had spent <1 year in captivity at the Belize Zoo were similar to the baseline and the adrenocorticotrophic hormone-challenged jaguars, respectively, in the study by Conforti et al. (2012).Thus, based on laboratory and biological validations, both immunoassays appear to be 7 effective at measuring FGM in fresh jaguar scat samples.Moreover, our sample sizes were too small to separate by sex in the biological validation; it still remains unclear whether jaguar males or females display different FGM profiles after challenging the hypothalamic-pituitary-adrenal axis.
A few investigators have assessed the change of faecal steroid metabolite concentrations over time, although the terminology differs, e.g.washing-out experiment (Rehnus et al., 2009), effects of faecal age and seasonality (Abáigar et al., 2010), field stability experiment (Fanson, 2009) and rate of FGM degradation (Hulsman et al., 2011).Regardless of terminology, systematic experimental assessments of the effects of natural environmental conditions on hormone metabolites in excreta of wild animals are crucial to ensure methodological validity.In our environmental validation, the results differed somewhat for the two immunoassays.Faecal glucocorticoid metabolite concentrations did not differ statistically between seasons using the corticosterone RIA.In contrast, the cortisol EIA indicated that jaguar FGM immunoreactivity changes, possibly due to steroid degradation, at faster rates and with more variability during the wet season than during the dry season.Overall FGM concentrations also fluctuated more during the wet season for both immunoassays.However, a stronger effect was seen in the cortisol EIA, where FGM in older samples could mistakenly identify an individual as having higher adrenal activity.Both immunoassays showed that FGM concentrations changed over time when exposed to the environment; however, the effects were variable.Faecal glucocorticoid metabolite concentrations measured with the corticosterone RIA did not differ statistically after 5 days of exposure in either season.In contrast, FGM concentrations measured with the cortisol EIA were stable for <1 day of environmental exposure during the wet season.During the dry season, little change in FGM concentrations over time during was observed for both the cortisol EIA and the corticosterone RIA.Selection of the appropriate immunoassay, in this case, for measuring FGM in Belizean free-ranging jaguars, ultimately depends on the research question and goals.The corticosterone RIA is robust against the effect of climate seasonality, while the cortisol EIA should be used for scat field surveys only during the dry season.The contrasting outcomes from these immunoassays suggest the importance of evaluating more than one immunoassay in environmental validations.This practice could increase the chance of capturing or confirming relevant changes in steroid metabolite concentrations.

Toolbox
Even though sex differences in FGM concentrations have been shown to occur naturally in some felid species, such as Canada lynx (Lynx canadensis; Fanson et al., 2012), our environmental validation showed no evidence of sex-related effects on the change of FGM concentrations of jaguar faecal samples exposed to the environment after defaecation.
Warmer and dryer weather appeared to minimize variation in FGM concentrations.Cooler and wetter weather possibly creates a more hospitable environment for microorganisms with steroid biotransformation capacities, as previously suggested (Washburn and Millspaugh, 2002).Within season, vast differences in scat exposure to UV -rays among environmental conditions (e.g. 30 and 10 times higher UV intensity in sun vs. shade treatments in dry and wet seasons, respectively) did not appear to have an important effect on FGM concentrations.These results are perhaps explained by the low capacity of UV rays to penetrate solid structures, such as faeces (for example, Abdelrahman et al., 1979).Therefore, the effects of UV-ray exposure on FGM degradation might be limited to the surface of the scat (i.e.depth of micrometres), whereas faecal steroids are widely distributed throughout the entire faecal sample.
Scat ageing by assessment of morphology in relationship to hormone stability was not successful because of high variation in colour and moisture changes after rain events.Additionally, scats turn white very quickly in dry conditions, which is likely to lead to overestimates of sample age.Furthermore, we observed that humidity drastically altered the morphology of faecal samples, making the ageing of scats in the field by morphology inaccurate.Instead, we recommend adjusting survey regimens to incorporate the results of environmental validations by resurveying an area at intervals to ensure stability in hormone concentrations in excreta as described below.
We suggest that those researchers interested in NHM sampling for jaguars in a tropical country, such as Belize, should conduct surveys in the dry season by using either the cortisol EIA or the corticosterone RIA, while surveys in the wet season should use only the corticosterone RIA.We found that scats cannot be aged accurately by morphology, but because FGM concentrations are stable for up to 5 days, sampling can involve an initial clearing of all scats from the target area and, to be conservative, resurveying every 3-4 days to ensure accuracy in hormone concentrations.A different sampling regimen could be adopted during the wet season, and if the cortisol EIA is used, resurveying would need to take place daily; however, strict caution in the interpretation of data should be taken due to the rapid change in FGM concentrations.In addition, daily survey may be impractical in field conditions.
Assessing the degree of stress in wild jaguars ranging across areas of varying human disturbance is a timely application of these methodologies.Belize has experienced a progressive increase in human activities such as hunting, housing developments, forest eradication and land conversion for agriculture, hence increasing levels of human-jaguar conflict.Our methods will ensure physiologically relevant FGM concentrations in faeces and could advance conservation physiology for jaguars by exploration of the linkage of habitat fragmentation and human-wildlife conflict to measures of adrenal activity in free-living jaguars.

Figure 2 :
Figure 2: Biological validation of jaguar FGMs using two immunoassays, corticosterone RIA (A) and cortisol EIA (B).Mean FGM concentrations were measured in jaguars exposed to different durations in captivity after capture.Blue dots represent values for each individual.Bars with different letters are significantly different from each other; n is the number of individuals included in each category; §captive-born male not included in statistical test.

Figure 4 :
Figure 4: Faecal glucocorticoid metabolite concentrations assessed by two immunoassays, corticosterone RIA (top panel) and cortisol EIA (bottom panel) in jaguar faecal samples exposed for 5 days to two environmental conditions: sun (light line) and shade (dark line), in two seasons: wet and dry.Individual jaguars included in the dry season (shade n = 7, sun n = 9) and the wet season (shade n = 6, sun n = 6).Error bars represent 1 SEM.*Significant difference from control sample or day 0 of exposure.Horizontal red dashed lines represent the average FGM concentrations of individuals with high adrenal activity (i.e.<1 year captivity) from biological validation.

Figure 5 :
Figure 5: Morphology of jaguar scat during 5 days of exposure during the wet and dry seasons in two environmental conditions: sun and shade.An example of four scats photographed over time (row direction) is shown for every combination of environmental conditions.Morphological characteristic scores [M = moisture, from driest (1) to moistest (5); C = colour (DB, dark brown; B, brown; and W, white)] are presented in the black boxes.