Various ecological and human-induced disturbances play an important role in defining the health of an animal. To cope up with such threats or challenges to its homeostasis, an animal responds by secreting stress hormones (glucocorticoids) to mobilize the energy. In general, glucocorticoids help animals to cope up with the ‘stressors’. Prolonged exposure to any ‘stressors’ may lead to an elevation in the levels of glucocorticoids, thereby leading to reproductive inhibition, immune system suppression, neuronal death and impaired cognitive function and hence, threatening the survival and fitness of an animal. However, the long-term effects of environmental or anthropogenic disturbances on the fitness and survival are difficult to analyze in a slow reproducing, long-lived species such as elephants. Spread over 3% of India’s geographical area, India shelters around 28,000 to 30,000 Asian elephants (Elephas maximus; Bist, 2002; Santiapillai and Sukumar, 2006). Protected areas in India are estimated to form 22% of elephant habitat; while the rest of the elephant habitats extend outside the protected areas (Lenin and Sukumar, 2011) making them vulnerable to human-induced threats. Consequently, there are increased incidences of human-elephant conflicts leading to crop and property depredation and human-elephant mortality. Most of the studies, in context of such conflicts, highlight the temporal, spatial scales of conflicts majorly focusing on mitigation measures. None of the studies so far have elucidated how escalating human-elephant conflicts would influence the physiological health of Asian elephants. This study, hence, focused on addressing the proximate causation of the stress-response in free-ranging Asian elephants of the Bandipur National Park, the Nagarahole National Park and Hassan district of Karnataka, using a non-invasive technique for measuring faecal glucocorticoid metabolites (fGCM). There are several factors which could act as possible stressors to the physiological state of an elephant, but depending upon the feasibility and time limitations of conducting the study, we selected those factors which we thought are essential to address the proximate causation of stress-response in Asian elephants. Being the first detailed study to explore the stress-response in free-ranging Asian elephants, this study attempted to understand the influence of some of the fundamental ecological (seasonality, group size, and body condition) and human-induced stressors. The main body of the dissertation is divided into six chapters. The first and the last chapters discuss the general introduction and conclusion while the rest four chapters highlight the main four objectives of the study. The main objectives of the thesis were: 1. The first objective of this thesis was to validate and standardize some of the crucial parts of protocols to avoid any technique-based bias while interpreting the levels of faecal glucocorticoid metabolites. Under the first objective, the influences of some of the fundamental intrinsic and extrinsic factors on the fGCM levels were also assessed. 2. The second objective was to assess the association between body condition, seasonality and stress response. 3. Having assessed the relationship between body condition and stress-status, the third objective focused on the influence of socio-ecological correlates such as group size, lactational status and the presence of adult females in a herd on the stress status of female adult elephants. 4. With above three objectives majorly highlighting the influence of ecological factors, the fourth objective was to assess the influence of anthropogenic factors on stress-response of free-ranging elephants under which we compared the stress-response between crop-raiding in human-dominated and nonraiding elephants in forested habitats. The study was conducted during dry season (February to May) and wet season (August to December) of 2013 and 2015. The dissertation is organized in the following way: Chapter 1: General Introduction The first chapter provides a general introduction about stress response and its physiological pathway. It also discusses the importance of non-invasive technique used in our study. It highlights the significance of this study based on the available literature on both African and Asian elephants. Chapter 2: General validation and standardization of field techniques and assays; influence of intrinsic and extrinsic factors on faecal glucocorticoid metabolites This chapter is divided into two parts. The first part deals with general validation and standardization of field techniques and the second part assesses the influence of fundamental intrinsic and extrinsic factors. Being the first study on stress-response of free-ranging Asian elephants, influence of various field-based techniques and variations were studied to facilitate the reliable interpretation of fGCM levels. We first conducted the experiments to standardize the field protocol and the laboratory protocol. Under the field protocol, we found that there was no within sample variation (as observed in many other species) in fGCM levels from thoroughly mixed faecal samples and samples from the center of the same dung bolus. The time of collection of samples had a significant influence, suggesting that the concentration of fGCM remains stable until 6-9 hours. Influence of storage of lyophilized samples suggested that the concentrations of fGCM levels decline if exposed to moisture. Under the laboratory protocol, A group specific 11-oxoetiocholanolone EIA, used for analyzing fGCM levels (μg/g), was modified to enhance the repeatability, accuracy and precision. EIA was analytically validated in every hormonal analysis by calculating the inter- and intra-assay coefficients of variation, specificity and parallelism. In the second part, we assessed the influence of age, sex and season on fGCM levels and found there was no age-wise variation and sex-wise variation. Season-wise variation was significant only in female elephants. We discuss the potential physiological reasons behind such variation. These findings suggest that aspects such as time of collection, proper storage of samples and the influence of extrinsic stresses (seasonality) should be considered for reliable and unbiased interpretation of fGCM levels. Chapter 3: Assessment of season-dependent body condition scores in relation to faecal glucocorticoid metabolites in free-ranging Asian elephants. (Published in Conservation Physiology; Pokharel et al., 2017; doi:10.1093/conphys/cox039) We studied seasonal and annual changes in visual body condition scores (BCS), and assessed how these scores were related to levels of fGCM levels in free-ranging Asian elephants in the seasonally dry tropical forests of the Mysore and Nilgiri Elephant Reserves in southern India. We assessed the animals’ BCS visually on a scale of 1 to 5; where 1 represents a very thin and 5 represents a very fat elephant. To understand the influence of seasonality on BCS, we sampled the population during dry and wet seasons of 2013 and 2015 while, for annual changes in BCS, we sampled nine free-ranging adult females from different family groups that had been repeatedly sighted over seven years. To evaluate the influence of body condition on fGCM, we measured the fGCM levels from fresh fecal samples collected from the body-condition scored animals. Effect of age and season on BCS in relation to fGCM was also studied. We found that the BCS was related with season, i.e. individuals with low BCS were more frequent during the dry season as compared to the wet season and fGCM levels were negatively associated with BCS. To be precise, concentrations of fGCM were highest in individuals with the lowest BCS (BCS-1) and then significantly declined till BCS-3. fGCM levels were almost comparable for BCS 3, 4 and 5. This pattern was more conspicuous in female than in male elephants. We discussed the possible reasons behind such relationship between fGCM and BCS. The findings from this chapter suggest that season-dependent BCS, hence, reflect the stress status as measured by fGCM, especially in female Asian elephants and could be used as an important non-invasive approach to monitor the physiological health of free-ranging elephant populations. This is the first study to compare the body condition scores and stress-response in free-ranging elephants. Chapter 4: Influence of ecological and physiological correlates on stress physiology of free-ranging female Asian elephants. We investigated the influence of herd size (HS), lactational status and number of adult females present in a herd on adrenocortical function in free-ranging adult female Asian elephants by measuring their fecal glucocorticoid metabolites (fGCM) levels. We found that there was a negative association between herd size and fGCM levels when only HS and fGCM were compared. Lactating females showed higher fGCM than that observed in non-lactating females, which might reflect the nutritional stress on mother and anti-predatory challenges against the calves. Herds with only one adult female had higher fGCM levels than herds with more than one adult female which could be because of social challenges. The poor association between herd size and fGCM when all predictor variables were combined elucidate that group size may not necessarily influence the adreno-cortical function. However, the numbers of adult females present in a herd and their lactational status has a significant role in defining the stress-response in free-ranging adult female Asian elephants. This study is the first study to document the influence of lactational status with the stress-response in free-ranging elephants. Chapter 5: How physiologically costlier it is to be a crop-raider in a human-dominated landscape? Diet quality as a possible ‘pacifier’ against stress. We studied the stress-response in crop-raiding elephants in a human-dominated landscape and non-raiding elephants in the protected forested areas. While raiding agricultural crops, elephants face various associated threats such as retaliation by humans, human-induced disturbances and stress of raiding which could enhance the energetic costs, ultimately elevating their stress levels. We hypothesized that crop-raiders (in human-dominated landscape; Hassan district of Karnataka, India) will exhibit higher faecal glucocorticoid metabolite (a proxy of stress-response; fGCM) levels as compared to the nonraiders (in forested areas; Bandipur and Nagarahole National Parks). Contrary to our hypothesis, fGCM levels were found to be higher in both nonraiding female and male elephants than female and male crop-raiders. To assess the influence of benefits obtained from crop-raiding as one of the possible factors for the lower fGCM in crop-raiders, the difference in vegetation between human-dominated landscape and forested areas were further analyzed by using a remotely sensed NDVI (Normalized Differential Vegetation Index) and was ground-truthed by measuring the quality of diet (C:N ratio through fecal samples). Interestingly, the NDVI values were higher at the human-dominated and the C:N ratio was lower (higher Nitrogen content) in faecal samples of crop-raiding elephants (both in females and in males) than the nonraiders in the forested areas. Positive correlation was observed between C:N ratio and fGCM levels. These findings suggest that crop-raiding comes with the benefits of easy access to good quality of diet which may help in reducing the stress-response in elephants while being in the human-dominated landscapes, provided there is a low intensity of human disturbance. This is the first study to document and compare the stress-status of crop-raiding Asian elephants with elephants in their natural habitats using NDVI and C:N ratio. Chapter 6: Conclusion In conclusion, the work presented in this dissertation provides an overview of how ecological and anthropogenic factors could influence the physiological health of free-ranging Asian elephants. This dissertation focused on providing detailed insights about the stress-response in free-ranging Asian elephants. The objectives have been achieved by assessing the relationship between ecological stressors such as seasonality, body condition, herd size, lactational status, a presence of individuals in a herd and anthropogenic stressors such as human-induced threats or crop-raiding using well-validated and standardized laboratory and field protocols. This study provides valuable insights into the physiological health or stress-response being synergistically influenced by various ecological, social and anthropogenic factors in free-ranging Asian elephants. Findings obtained from this study could help in addressing the issues related to the management of free-ranging elephant populations.

The Asian elephant (Elephas maximus), a charismatic ‘flagship species’, is threatened by extinction in the wild, and the development of self-sustainable captive populations is a key conservation challenge. A third of the Asian elephant population is presently in captivity and information on the reproductive status, especially in females, is still lacking to a large extent. The onset of estrus in female Asian elephants is not associated with any visible physical signs, thus making the assessment of the reproductive status rather difficult. One approach to understanding reproductive cyclicity of animals is through generating profiles of reproductive hormones in blood (Wiseman et al. 1983; Brown et al. 1999, 2004). Profiles of reproductive hormones such as progesterone (P4), estrogen, luteinizing hormone and follicle stimulating hormone have already being demonstrated in Asian elephants (Brown et al. 1999, 2004; Brown 2000). In these studies, the reproductive status of females, maintained in zoos/captivity, was characterized based on circulating levels of hormones in blood samples. This is difficult to implement in the case of semi-captive or wild populations of elephants due to practical, legal and ethical considerations. In order to overcome this problem and to better understand the estrous status of female elephants, it is important to develop and validate non-invasive methods to monitor the reproductive status of female Asian elephants. An alternative approach to evaluating the reproductive status of females is to consider behavioral responses shown by males towards chemical signals produced by females to advertise their reproductive status. In order to understand the reproductive status of the individual belonging to the opposite sex, studies have shown that elephants rely on a variety of chemical signals produced in biological fluids such as urine, temporal gland secretion, inter-digital gland secretion, etc. (Krishnan 1972; Rasmussen & Schulte 1998). Chemical signaling is one of the prominent modes of communication in elephants, especially with respect to locating potential mates (Sukumar 2003). Thus, in most cases, elephants usually employ specific behavioral responses, for example sniff, check and place behaviors of trunk, for investigating the reproductive status of the conspecific individual, belonging either to the same or the opposite sex (Rasmussen et al. 1996; Schulte & Rasmussen 1999; Bagley et al. 2006). The objectives of the thesis are two-fold. First, to develop a non-invasive method of reproductive monitoring from fecal hormonal metabolites and also to understand the possible role of feces as an inter-sexual signal. The main body of thesis is divided into four chapters. 1) Development and validation of a non-invasive method to estimate progesterone metabolite in feces, to monitor the reproductive cyclicity of female elephants (Chapter 2). 2)Generation and characterization of progesterone and its metabolite, 5α-P-3-OH, profiles of semi-captive females using the developed non-invasive method to measure fecal metabolites (Chapter 3). 3) Validation of developed methodology and assay systems to a wild-population of female elephants (Chapter 4). 4) Feces as a potential source for inter-sexual chemical signaling in Asian elephants (Chapter 5). The above studies were carried out on semi-captive male and female elephants maintained in the forest camps of Mudumalai Wildlife Sanctuary (MWLS), Tamil Nadu and Bandipur National Park (BNP), Karnataka, India (Chapters 2, 3 and 5). For Chapter-4, free-ranging females of the MWLS were examined. 1. Development and validation of a non-invasive method to estimate progesterone metabolite in feces, to monitor the reproductive cyclicity of female elephants Niemuller et al. (1993) generated a profile of the progesterone metabolite, 5βpregnanetriol, to assess the estrous phase of Asian elephants based on non-invasive urine sampling. However, the collection of urine is difficult and to some extent impossible in the case of semi-captive and as well as that of wild elephants. Thus, the method of choice in this study was the development and validation of a non-invasive approach to measure fecal progesterone metabolites to assess reproductive status of females. Sampling was carried out at monthly intervals on three female elephants at the MWLS forest camp, while three other females maintained at the BNP forest camp were sampled fortnightly. An enzyme linked immuno-sorbent assay was developed to measure the concentration of the progesterone metabolite, 5α-P-3OH in the fecal samples of the semi-captive females. Using varying concentrations of the hapten (5α-P3OH), from low (0.1 mM) to high (1 mM), a standard curve was first generated, which had a linear range between 0.25 mM to 62.5 mM, with an EC50 of 1.37 mM. The linear range was then used to detect the concentrations of 5α-P-3OH in the fecal samples of females examined. The non-invasive method was further validated as there existed a positive correlation (p<0.1) between the levels of fecal 5α-P-3OH and that of concentration of circulating P4, measured in blood samples. This is the first valid documentation of a non-invasive method based on fecal progesterone metabolite pattern in order to assess the reproductive status of the female Asian elephants. 2. Generation and characterization of reproductive hormone profiles of semi-captive females using the developed non-invasive method to measure fecal metabolites With the establishment of a non-invasive method to understand occurrence of estrus in female elephants (described in Chapter 2), attempts were made to generate hormonal profiles over a longer time interval through more frequent sampling. Based on sampling at weekly intervals, the concentrations of both fecal 5α-P-3OH and that of native P4 hormone in the blood were determined, and the females were then identified as belonging to different reproductive states of hormonal cycling (follicular and luteal phases), non-cycling and pregnant categories. Of the 7 females sampled at both MWLS and BNP, three distinct categories of hormone-metabolite profiles emerged. The first category included four females that showed regular cyclicity throughout the sampling period, as monitored through the measurement of fecal 5α-P-3OH and serum P4 levels. The estrous cycle of all the four females was divided into two phases (follicular and luteal), based on the patterns of fecal 5α-P-3OH and serum P4 concentrations. The follicular phase of the estrous cycle was assigned when the values of both fecal 5α-P-3OH and serum P4 remained below 0.3 µg/gm and 0.3 ng/ml, respectively, for a considerable time period (viz. >2 wk). However, the luteal phase was characterized, when the levels of fecal 5α-P-3OH and serum P4 remained at or above 0.3 µg/gm and 0.3 ng/ml respectively, over a period of more than 2 - 3 wk. The second category had two females showing a ‘flat-lining pattern’ for the levels of fecal 5α-P-3OH and serum P4 concentrations, without any peak or dip in their concentrations. Since the reproductive pattern for both these females was flat-lined throughout their sampling period (51 wk), maintaining the levels of fecal 5α-P-3OH and serum P4 below 0.3 µg/gm and 0.3 ng/ml respectively, the females were considered to be non-cycling or anestrus. The third category had one female in which the levels of both fecal 5α-P-3OH and serum P4 measured were consistently high throughout the entire sampling period (26 wk). The levels of both fecal -P-3OH and serum P4 were above 0.3 µg/gm and 0.3 ng/ml, respectively. At the end of the sampling period, this female delivered a male calf; thus, the measured concentrations of fecal 5α-P-3OH and serum P4 signified the levels maintained during the gestational phase of this female. Sampling and hormonal analyses were also carried out for a male in the MWLS forest camps to investigate the baseline concentrations of fecal 5α-P-3OH and serum P4. The male showed consistently low concentrations of both fecal 5α-P-3OH and serum P4 throughout the sampling period (16 wk), below the margin of 0.3 µg/gm and 0.3 ng/ml, respectively. In this part of the study, it was confirmed that the reproductive status of a female elephant can be correctly assessed on the basis of measurements of fecal 5α-P-3OH alone, with repeated sampling of the female over a longer time scale. It was also shown that the strength of the positive correlation between the concentrations of the fecal 5α-P3OH and the serum P4 increased (p<0.01) for a larger sample size than that obtained for a comparatively smaller sample (described in Chapter 2). 3. Validation of developed methodology and assay systems to wild-population of female elephants So far, the non-invasive method to estimate fecal progesterone metabolite in order to predict occurrence of estrus of elephants was largely applied to semi-captive females (described in Chapters 2, 3). However, the necessity of such a method is being recognized for assessing the reproductive status of free-ranging females. Several findings have described differences in the rate and type of steroid metabolite excretion among individuals maintained under different diet regimes (Wasser et al. 1993; Smith et al. 2006). For instance, female elephants in the forest camps are provided with supplementary diet consisting of sugarcane, rice and millets. This diet is strikingly different from the feeding materials consumed by wild/free-ranging elephants (Sukumar 2003). Therefore, differences in dietary components can potentially influence the fecal steroid metabolites’ excretory patterns shown by semi-captive vis-a-vis wild females, which can affect the validity of measuring fecal 5α-P-3OH to predict females’ reproductive status. In order to examine this problem, the non-invasive method was applied to the population of wild elephants in the forests of MWLS, through random one time sampling of 30 individual female elephants. The steroid extraction efficiency (73 ± 11.0%, mean ± S.D., n =30) determined for the fecal samples collected from the wild females was not significantly different from the coefficient calculated in the case of the semi-captive females (80 ± 4.3%, mean ± S.D., n = 38). This indicated that dietary differences between wild and captive elephants did not influence levels of fecal hormonal metabolites in feces, unlike earlier observations on baboons (Wasser et al. 1993), old world primates (Wasser et al. 1988) and sheep (Smith et al. 2006). The values of the fecal allopregnanolone determined in the case of the wild females, ranged from as low as 0.06 µg/gm to as high as 23µg/gm of the sample, thus showing the heterogeneity of the samples, indicating that the females may be belonging to different reproductive phases. However, since sampling was carried out randomly, with an adult female being sampled just once, at this stage, it is not possible to identify or elaborate on the reproductive phase of the females. This is the first study reporting the values of the fecal progesterone metabolite in female Asian elephants in the wild. Further studies may be required to carry out long term monitoring of the wild females, through repeated collection of fecal samples over time from particular female. 4. Feces as a potential source for inter-sexual chemical signaling in Asian elephants In this Chapter-5, behavioral trials were conducted on male elephants to understand the role of fecal matter in conveying inter-sexual chemical signals. This was demonstrated by analyses of specific chemosensory behavioral responses shown by males towards the fecal samples of females that were strangers and belonging to different reproductive stages. Males showed four prominent behavioral responses namely ‘distant sniff’, ‘close sniff’, ‘check’ and ‘place’ towards the fecal samples of females. The sum of frequencies of these four responses (distant sniff, close sniff, check and place) was much higher for samples of the follicular (pre-ovulatory) phase females as compared to those of the luteal (post-ovulatory) phase females (p<0.005). Thus, for the first time, it was shown that male elephants are able to discriminate the different reproductive phases of females based on their specific behavioral response towards the fecal samples of the opposite sex. In conclusion, the thesis has focused on understanding and providing new insights regarding the reproductive biology of the female Asian elephants. This has been achieved through the development of the non-invasive method based on measuring the concentrations of the fecal progesterone metabolite and through the analyses of the chemosensory responses performed by the males towards the fecal samples of strange females. These methods can potentially be applied to the populations of both wild and captive/semi-captive female elephants in order to evaluate their reproductive status, through non-invasive measures. The information derived from the application of such methods will help in understanding the reproductive potential of the wild elephants under various environmental and ecological conditions. Further, the non-invasive measurement of reproductive hormones will help in monitoring the reproductive state of the individuals and thus aid in planning strategies for the welfare and management of the elephants maintained in captive or semi-captive conditions.

付記する学位プログラム名: 霊長類学・ワイルドライフサイエンス・リーディング大学院 / 京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第22298号 / 理博第4612号 / 新制||理||1661(附属図書館) / 京都大学大学院理学研究科生物科学専攻 / (主査)教授 幸島 司郎, 教授 平田 聡, 教授 伊谷 原一 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM