t i l Variations of Fecal Cortisol and IgA Levels with Reproductive Status in Wild Taihangshan Macaque ( Macaca mulatta tcheliensis )

Fecal cortisol has been suggested an indicator of preceding physiological stress, while the secretory IgA acts as the first defense against pathogens, viruses and bacteria in animals. Fresh fecal samples of adult Taihangshan macaques ( Macaca mulatta tcheliensis ) were collected during mating and non- mating periods, respectively, to understand the influences of mating status on physiological condition of wild macaques, through variation in fecal cortisol and IgA levels. The results showed that. (1) mean concentration of fecal cortisol in mating period (AM: 1751.12 ± 199.16 ng/g; AF: 1366.38 ± 117.85 ng/g) were significant higher than that in non-mating period (AM: 705.97 ± 82.86 ng/g; AF: 699.51 ± 56.06 ng/g). (2) there was significant difference in fecal IgA levels between mating and non-mating periods in AF, but not in AM. (3) there were obvious negative correlations between fecal cortisol and IgA levels both in AM ( P = 0.04) and AF ( P = 0.01) in mating period, but not in non-mating period (AM: P = 0.19; AF: P = 0.21). We concluded that mating-related stress influences not only stress hormone level but also immunological response of adult Taihangshan macaques, and that the stress-linked immunosuppression hypothesis was supported in mating period.


INTRODUCTION
S tabilized internal environment, vital for survival and reproduction of wildlife, is especially important to endangered animals living in complex environment. As a responding hormone to environmental and physiological stresses, cortisol usually enhances individual's susceptibility to infections through immunosuppression (Pihl and Hau, 2003;Lantz et al., 2018), while cortisol is an available indicator in body fluids for stress (Beehner and Bergman, 2017;Birnie-Gauvin et al., 2017). To examine cortisol concentration, blood is believed the most reliable tested-sample, but capture-recapture method for blood collection is usually stressful and leads to increased level of plasma cortisol (Rehbinder and Hau, 2006). Feces have been demonstrated non-invasive and efficient for measuring cortisol as well as immunoglobulin A (IgA) (Schatz and Palme, 2001;Lantz et al., 2018;Gesquiere et al., 2020), and fecal cortisol concentrations were successfully used in assessing adrenocortical function of cat and dog (Schatz and Palme, 2001), quantitative analysis of internal stress levels of rats (Pihl and Hau, 2003;, and evaluating reproductive states and seasonal factors on cortisol levels of non-human primates (Stavisky et al., 2003;Weingrill et al., 2004;Paramastri et al., 2007;Hoffman et al., 2010;Gesquiere et al., 2020).
Acting as a first defense against pathogens, viruses and bacteria, and being presented in saliva, tears, milk, respiratory and intestinal secretions, IgA is the primary antibody responsible for mucosal defense in mammals and has been used as an indicator of individual immunocompetence (Huang et al., 2014;Lantz et al., 2018;Gesquiere et al., 2020). Being exposed to diseasecausing agents and encounter diverse antigens from natural environment, wild animals use mucosal tissues as the primary entry point for antigens (Hart, 2011). Therefore, mucosal tissues have evolved a chronic but noninflammatory immune response without killing microorganisms, and prevent them from crossing epithelial barriers (Russell and Kilian, 2005). A key component of the mucosal immune defense is the production of secretory immunoglobulin A (SIgA).
Being available signals of stress response in mammals,

O n l i n e F i r s t A r t i c l e
cortisol and IgA levels maintain balance under healthy physiological conditions in animals (Marketon and Glaser, 2008). However, activating immunological response to invaders and maintaining a hormonal balance are energy costed for animals, thus tradeoffs between stress hormone and immune system would be expected (Hasselquist and Nilsson, 2012). Although physiological stress of animals caused by disturbance to their environment is adaptive in the short term, chronic stress inevitably impact animals health, immune system and reproductive functions (Lochmiller and Deerenberg, 2000;Rauw, 2012). The immunocompetence handicap hypothesis suggests that androgenic hormone (i.e., testosterone) could stimulate the development of male secondary sexual characters but suppress their immunocompetence (Folstad and Karter, 1992). The plausible explanation behind the hypothesis has relied on the costs and benefit of allocating limited energy/nutrient to various physiological traits. Given that measuring stress hormones are very common in the study of monitoring the physiological condition of wildlife (Schwarzenberger, 2007;Wang et al., 2009), using stress hormones instead of testosterone could broaden the window of the immunocompetence handicap hypothesis (Behringer and Deschner, 2017). Published works indicated that higher glucocorticoid concentrations can inhibit IgA production (Griffin and Thomson, 1998;Pihl and Hau, 2003;Paramastri et al., 2007), and several studies found a negative correlation between stress level and IgA level in mammals, such as dogs (Skandakumar et al., 1995;Takahashi et al., 2009), rats (Royo et al., 2004) and humans (Deinzer, 1998;Ng et al., 1999), which all supported the immunocompetence handicap hypothesis. However, data from Sichuan snubnosed monkey (Rhinopithecus roxellana) showed a positive correlation between fecal cortisol concentrations and fecal IgA concentrations in both sexes (Huang et al., 2014), while study on lachrymal IgA in chickens failed to demonstrate a correlation between the level of IgA and stress hormone (Florence et al., 1995).
Taihangshan macaque (Macaca mulatta tcheliensis), an endemic subspecies within rhesus macaque to China, distributed in south part of Mt. Taihangshan area, northern China (Lu et al., 2007. Acting as strict seasonal breeders, adult Taihangshan macaques usually mate from September to December, and the peak in October (Lu et al., 2009;Tian et al., 2013;Guo et al., 2020). Published works focused on behavioral laterality (Guo et al., 2020), geographical distribution (Qu et al., 1993;Lu et al., 2009), food habits (Guo et al., 2010), social structure and female reproductive parameters (Tian et al., 2011(Tian et al., , 2013, habitat selection (Xie et al., 2012) and macronutrients regulation (Cui et al., 2018), but less attention was paid on physical metabolism of cortisol and IgA. By quantitatively analyzing on variation of fecal IgA, fecal cortisol and their interaction in adult macaques, we aimed to understand the tradeoffs between fecal cortisol and IgA in different reproductive status. We predicted that: (1) concentrations of fecal cortisol and IgA in adult Taihangshan macaques vary with mating status; and (2) a negative relationship between fecal cortisol and IgA levels is expected in mating period.
This site exhibits a temperate continental climate with an average annual temperature of 14.3℃, average temperature 0.1℃ in January and 27.3℃ in July (Guo et al., 2020). The annual precipitation is 695 mm with the main characteristic of uneven spatial and temporal distribution, which made wet hot summer and dry cold winter in this area (Hou and Liang, 2001;Guo et al., 2010Guo et al., , 2020. The main vegetation type is shrub with only a small area of natural secondary forest in Wulongkou area, which made food resources relatively scarce in winter and the macaques experience low food availability after mating period (Chai, 2014). During the period of data collection, our study group was provisioned with a small amount of maize and peanuts (Tian et al., 2016).

Fecal samples collection
From September 2016 to March 2017, fresh fecal samples of adult male (AM) and adult female (AF) were collected and each were preserved in 50 mL EP tube. The harvested samples were immediately stored in -20℃ freezer in study site. Fecal samples were then transported to Zhengzhou University with dry ice in 5 h for further tests. Freezing samples has been reported to have minimal effects on fecal cortisol and IgA levels (Lantz et al., 2018; O n l i n e Gesquiere et al., 2020). Totally 89 fecal samples of adult Taihangshan macaques had been collected, in which 48 and 41 samples were tested in mating and non-mating periods, respectively. When multiple samples were collected by the same individual in the same mating/ non-mating period, the average concentration was taken as the level of this individual.

Sample processing
The weighing bottle (50×30 mm) was weighted and marked, and then amount of each fecal sample (ca. 6 g) was placed into the weighing bottle and weighted. Then the samples were placed in the drawers of a SJIA-10N vacuum freeze-dryer (Ningbo SJialab Equipment Co., Ltd, Ningbo, Zhejiang, China) pumped by a 2XZ-2 direct-drive rotary vane vacuum pump (Linhai Tanshi Vacuum Equipment Co., Ltd, Linhai, Zhejiang, China). The samples were dried under -60℃ for 24 h, weighted again and grinded within clean mortar for ca. 5 min. The powdered samples were weighted for further measurements.

Fecal cortisol detection
Amount of 0.2000 (± 0.0050) g powdered feces was taken from each sample and was put into a 10 mL EP tube. Fecal cortisol was extracted following the 'Ethanol-Methanol' method with little adjustment (Straling and Wasser, 1988;Ziegler et al., 2000;Girard-Buttoz et al., 2009). For each sample, 1 mL absolute ethanol and 1 mL double distilled water was added and was homogenized with a vortex (1000 rpm) for 10 min. The suspension was placed under 4℃ for 2 h to maximize the dissolution of cortisol. It was then homogenized again with a vortex (1000 rpm) for 3 min, and centrifuged at 4000 rpm under 4℃ for 10 min. 1 mL of the supernatant was pipetted to 5 mL EP tube. Then 2 mL absolute ethanol was added to the remaining sediment and was homogenized at 1000 rpm for 10 min. The suspension was centrifuged again at 4000 rpm under 4℃ for 10 min. 2 mL of the supernatant was added to the previous tube with 1 mL supernatant. The 3 mL supernatant was homogenized and then was centrifuged at 12000 rpm under 4℃ for 10 min, before 2 mL supernatant was harvested. This supernatant was dried with 60℃ water bath. The dried sample was stored under -20℃ freezer for later measurement of fecal cortisol.
Fecal cortisol concentration was measured with general cortisol competitive-ELISA kit (Elabscience Biotechnology Co., Ltd, Wuhan, Hubei, China). The dried fecal extract was dissolved with 500 μL 0.04 M PBS (pH= 7.2) under 4℃ for 24 h. Then the measurement was conducted following the manual provided by the manufacturer. Duplicate standard and fecal samples were measured. Based on preliminary experiment, 100 μL of fecal sample solution instead of 50 μL fecal sample solution recommended in the manual, was added to the plate cell. When stop solution was added to the plates, Bio-Tek Power Wave XS microplate spectrophotometer (Bio-Tek Instruments, Inc, Winooski, USA) was used to read the OD values at 450 nm and 570 nm. The coefficient of variation for all kits assay were less than 20%.

Fecal IgA detection
Another 0.5000 (± 0.0050) g powdered feces from the same sample was placed into a 10 mL EP tube for extracting fecal IgA following the methods of Hau et al. (2001) and Lantz et al. (2018). Briefly, 5 mL 0.04 M PBS (pH = 7.2) and 25 μL Tween-20 (Beijing Solarbio Science and Technology Co., Ltd, Beijing, China) was added to the EP tube. This mixture was homogenized with a vortex (1000 rpm) for 10 min. The suspension was placed under ambient temperature for 30 min to maximize the dissolution of IgA. It was then centrifuged at 4000 rpm under 4℃ for 10 min. 2 mL of the supernatant was pipetted to 5 mL EP tube. The supernatant was centrifuged again at 10000 rpm under 4℃ for 10 min. Then 1.5 mL of the supernatant was harvested in 2 mL EP tube and stored at -20°C freezer for further tests.
Fecal IgA level was measured with monkey IgA double antibody sandwich ELISA kit. The stored fecal IgA extracts and the ELISA kits were placed under ambient temperature for ca. 30 min. The measurement was obtained following the manual provided by the manufacturer. Duplicate standard and fecal samples were measured. Based on preliminary experiment, 150 μL of fecal IgA solution instead of 100 μL fecal IgA solution recommended in the manual, was added to the plate cell. When stop solution was added to the plates, Bio-Tek Power Wave XS microplate spectrophotometer (Bio-Tek Instruments, Inc, Winooski, USA) was used to read the OD values at 450 nm and 570 nm, respectively. The coefficient of variation for all kits assay were less than 20%.

Data analysis
The OD value for each sample, either standard or treatment, was calculated as the mean OD value of duplicated tests which subtracted the values of 570 nm from those of 450 nm. Then the standard curve for each ELISA kit was modeled using Curve Expert Professional (version 2.6.3) (Hyams Development), and the optimized model was selected for further analyses. For fecal cortisol (ng/g), the DR-Gamma model (f (x) = γ +(1-γ) * (α, β * x)) was adopted according to the score. Rational model (f (x) = γ +(a+b * x) / (1+c * x + d * x 2 )) was used for calculating fecal IgA concentration (ng/g).

O n l i n e F i r s t A r t i c l e
The original data was tabulated in Microsoft Excel 2016, and then the statistics and figure plotting were conducted with IBM SPSS Statistics (version 24.0) and GraphPad Prism (version 7.0). Independent samples t-test was used to determine the differences in the concentration of fecal cortisol and IgA between mating and non-mating periods. Pearson Product-Moment correlation was used to analysis the relationship between fecal cortisol and IgA in dried feces. The values were presented as mean ± SEM. All P values are considered significant under the level of 0.05.

Correlation between fecal cortisol and IgA levels
We analyzed the correlation of fecal cortisol and IgA

O n l i n e F i r s t A r t i c l e
Fecal Cortisol and IgA of Taihangshan Macaques 5 levels in mating and non-mating periods, respectively. The results showed a trend that the fecal cortisol level was significant negatively correlated to IgA level in mating period (AM: r = -0.52, df = 16, P = 0.04; AF: r = -0.45, df = 32, P = 0.01) ( Fig. 2A), whereas no significant correlation was found in non-mating period (AM: Pearson r = 0.31, df= 19, P = 0.19; AF: Pearson r = -0.28, df = 22, P = 0.21) (Fig. 2B). Fig. 2. Correlation between fecal cortisol (ng/g) and IgA (ng/g) levels in adult Taihangshan macaques in mating period (A) and in non-mating period (B). 95% CI are also included in the graph.

DISCUSSION
The results of this study demonstrated that, in AM Taihangshan macaques, the fecal cortisol level in mating period was significantly higher than that of non-mating period, whereas no obvious difference was found in fecal IgA levels between these two periods. However, in AF Taihangshan macaques, the fecal cortisol and IgA levels in mating period were significantly higher than those of nonmating period. Moreover, there was a significant negative correlation in adult macaques between fecal cortisol and IgA levels in mating period, but not in non-mating period. Consequently, our prediction that concentrations of fecal cortisol and IgA of adult macaques varied with reproductive status was supported, while the prediction of a negative relationship between fecal cortisol and IgA levels was supported in mating period.
Cortisol, being an important stress hormone, affects many physiological functions of animals (Capitanio et al., 2004;Weingrill et al., 2004;Girard-Buttoz et al., 2009). When subject to stress, the hypothalamic-pituitary-adrenal (HPA) axis is activated, releasing excess glucocorticoids (e.g., cortisol and corticosterone) into the bloodstream which mobilize energy reserves that are important for "fight or flight" responses (Creel, 2001). Therefore, as a responding hormone to environmental and psychological stresses in non-human primates, levels of cortisol varied with reproductive status, aggressive competition, and mating activities (M. mulatta, Gordon et al., 1976;Cebus apella, Lynch et al., 2002;M. fuscata, Barrett et al., 2002;Propithecus verreauxi, Fichtel et al., 2007). For adult macaques, mating season is a particularly challenging period with several conspecific spouse competition in a polygamous society. In order to cope with the increased physical and energetic demands of conspecific competition in mating period, males often increased food intake, sexual dimorphism, and a marked rise in testosterone and glucocorticoid levels (Romero, 2002;Girard-Buttoz et al., 2009), and females showed clear cyclic solicitation and copulation peaks (Kim et al., 2013). As strict seasonal breeder, adult Taihangshan macaques usually mate from September to December annually (Lu et al., 2009;Tian et al., 2013;Guo et al., 2020), with increased mating rates and aggressive behaviors in both sexes (Sun, 2019), which reasonably explained the higher cortisol levels during this period. The enhancement of cortisol level in adults in mating period can be attributed to physiological preparations in anticipation of aggressive behavioral and energetic demands of reproduction (Sapolsky et al., 2000;Bercovitch and Ziegler, 2002). Our findings were consistent with the results of studies on sifaka (Propithecus verreauxi) (Fichtel et al., 2007), long-tailed macaques (M. fascicularis) (Girard-Buttoz et al., 2009), chacma baboons (Papio hamadryas ursinus) (Weingrill et al., 2004), and Sichuan snub-nosed monkey (Huang et al., 2014). From January to May, most AF Taihangshan macaques are pregnant or postpartum, while the mating activity and conflicts among adult AMs and AFs significantly reduce during this period (Sun, 2019). As a result, cortisol level became lower in adult macaques during non-mating period than that of mating period.
Antibodies are the primary component of adaptive immune system in animals, whereas IgA is usually presented in excreted body fluids (e.g., saliva, urine, feces, milk, sweat, and tears). Studies have found that immunoglobulin levels profoundly affect wildlife health and a significant decrease in IgA can lead to increased susceptibility to infectious diseases (Decaro et al., 2004;Huang et al., 2014;Lantz et al., 2018). IgA enzyme immunoassays (ELISAs) have allowed for a direct indication of immunocompetence in mammals, such as long-tailed macaques (Paramastri et al., 2007), Sichuan snub-nosed monkeys (Huang et al., 2014), baboons (Gesquiere et al., 2020), and reindeers (Rangifer tarandus) (Rehbinder and Hau, 2006). Published works have shown that mean IgA concentration varied among individuals but not by sex or reproductive status in adult chimpanzees (Lantz et al., 2018), and there were seasonal differences in mean IgA concentration among Sichuan snub-nosed monkeys but not by sex (Huang et al., 2014). However, in this study, we found that mean IgA concentration differed significantly between the sexes in non-mating period.
At the population level, we found that there was no O n l i n e

F i r s t A r t i c l e
Y. Guo et al. significant difference of mean fecal IgA between mating and non-mating periods in adult males, whereas the mean fecal IgA level in mating period was significantly higher than that in non-mating period of adult females. After mating season, though some adult females are in the state of pregnancy or postpartum, they still have to face severe environment with cold and food shortage, which made the immune level of AF decreased than that in mating period. At different reproductive stages, the lower fecal IgA concentrations in AF may suggest a stage-dependent tradeoff between reproduction effort and immune system. Therefore, further work on AF macaques is needed to explore the relationship between reproductive status (in pregnancy, lactation, and after lactation) and immunity levels. Under natural conditions, predation and competition may adversely affect animal physical status through neuroendocrine disruption and stress-induced immunosuppression (Folstad and Karter, 1992;Sapolsky et al., 2000). Several studies paid attention on elevated stress hormone levels regarding social aggression, reproductive status, dominance and environmental challenges of nonhuman primates (Girard-Buttoz et al., 2009;Hoffman et al., 2010;Fürtbauer et al., 2012;Beehner and Bergman, 2017), but a few involved immune system (Huang et al., 2014;Lantz et al., 2018;Gesquiere et al., 2020), and the tradeoffs between stress hormone and immune level in nonhuman primates. Our finding that the negative correlation between fecal cortisol and IgA levels of adult Taihangshan macaques in mating period supported the stress-induced immunosuppression hypothesis.
Life history theory predicts tradeoffs among costly physiological traits due to competitive allocation of limited resources (Patterson et al., 2014;Kavanagh and Kahl, 2016). Previous literature showed that an acute stress response can induce an increase in immune responses (Hau et al., 2001;Pihl and Hau, 2003), while chronic stress response tends to suppress immune levels (Marketon and Glaser, 2008;Martin, 2009). During mating period, the most direct and chronic stressor for adult Taihangshan macaques were competition/aggression caused by mating activities (Zhang, 2015). Under this context, mating-related stress continued months, which lead to increased level of fecal cortisol and suppressed immune activity due to the tradeoffs between stress hormone and immune response. These responses can be used to monitor stress and immune system function as important indicators of health status in wild populations.
In conclusion, the concentrations of fecal cortisol and IgA of adult Taihangshan macaques differed between mating and non-mating periods. Mating-induced stress influences not only hormone level but also immunological response. The immunosuppression hypothesis was supported in mating period.