**1. Introduction**

The Asian elephant (*Elephas maximus*) is the o fficial national animal of Thailand, classified as endangered by the International Union for Conservation of Nature (IUCN 2010), and listed in Appendix 1 of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES). Since the logging ban in 1986, thousands of elephants and their mahouts were left without work and took to the streets to beg for food. About a decade later, a new industry emerged for Thai elephants—tourism. In 2017, there were 2673 elephants working in 223 tourism venues throughout the country (National Elephant Institute, Lampang, Thailand). While a few camps o ffer observation only, most utilize elephants in a variety of scheduled activities, like riding with and without saddles, entertainment shows, and tourist feeding and bathing. Around 900 elephants reside in more than 80 venues in Chiang Mai province. Variation in elephant demographics, work activities, elephant care, and mahout managemen<sup>t</sup> is evident among the camps of northern Thailand, as recently reviewed by Bansiddhi et al. [1], all of which can a ffect how individuals cope with the tourist environment.

Assessment of animal welfare relies on measures of physiological function (e.g., health, reproduction, stress) and/or behavior, applied at individual or population levels. The most commonly used biomarkers of stress and, by extension welfare, are glucocorticoids (GC) [2]. In response to an acute stimuli, activation of the hypothalamic-pituitary-adrenal (HPA) axis causes the release of cortisol from the adrenal cortex [3], which then feeds back to inhibit further release to restore homeostasis [4]. In humans and animal models, cortisol is typically measured in blood serum or plasma; however, the potential for inducing stress due to handling and blood collection [5] is a concern for most wildlife species, especially for repeated sampling. Thus, noninvasive approaches that quantify GC metabolites in urine or feces have been developed to assess acute and chronic stress responses in many species [6,7]. Another method—saliva collection—is less invasive than blood and, with a lag time of only 20–30 min, provides almost real time results [7]. Although studies have shown the value of GC for monitoring stress and welfare, including in elephants [8,9], today it is recognized that additional indicators that incorporate multiple physiological systems o ffer more ways to assess both negative and positive welfare states [10].

Recently, salivary immunoglobulin A (sIgA) has been promoted as a potential biomarker of positive a ffects [11,12]. Immunoglobulin A is found in many secretory fluids, including saliva and breast milk, and in nasal, gastrointestinal, bronchial, and urogenital secretions [13]. In general, sIgA responds quickly to acute events (positive or negative), increasing or decreasing depending on the stressor. Positive mood inductions related to movies, music, and self-referent statements have been shown to increase sIgA [14], as well as relaxation and massages [13]. However, sIgA concentrations also alter in response to negative e ffects. For example, depletion of sIgA occurs in humans taking academic exams [15,16], while elevations have been related to mental arithmetic tasks, and reported daily hassles and work demands [13,17]. Studies in mammalian species have linked sIgA to stress as well. In dogs, noise stressors and defense training caused a decrease in sIgA [18,19], while puppies displayed increased sIgA after behavioral testing [19,20]. In pigs, sIgA increased due to restraint stress [21] and isolation [22]. However, before using sIgA as a welfare biomarker, baseline levels must be established, as well as any endogenous patterns. Two studies have measured immunoglobulin A in Asian elephants [23,24], but neither examined specific temporal patterns.

The circadian rhythm is a roughly 24 h cycle in physiological processes, which generally are endogenously generated, although they can be modulated by recurring external cues, such as sunlight, temperature, and sleep-wake and activity cycles [25]. The suprachiasmatic nucleus (SCN) in the hypothalamus serves as the master pacemaker that sets the timing of rhythms by regulating neuronal activity, body temperature, and hormonal signals [26]. Circadian patterns are important to consider when interpreting biological results, to distinguish between basal rhythms and extrinsic effects. It is generally accepted that most mammals exhibit circadian patterns in GC secretion, with concentrations being highest in the morning and lowest at around midnight [16,22,27,28]. Studies in Asian elephants [29,30] have confirmed this pattern in urine [29] and saliva [30,31] samples. IgA also has been shown to have a diurnal pattern. In humans, concentrations decline throughout the day from a morning peak at 08:00–09:00 h. By contrast, in pigs [21] and dogs [18], sIgA concentrations are lowest in the morning (09:00 h), increase during the day (11:00–15:00 h), and then decline at night (17:00 h). There are no reports of circadian rhythms in sIgA in Asian elephants, nor its relationship with the well-studied stress hormone, cortisol. Thus, the goals of this study were to assess temporal patterns of sIgA throughout the day as a potential novel biomarker aiding in the assessment of welfare in elephants, and compare the patterns to those of cortisol. We hypothesized that sIgA in Asian elephants follows a circadian rhythm that is correlated with salivary cortisol.

#### **2. Materials and Methods**

#### *2.1. Animals and Sample Collection*

All animal procedures were approved by the Institutional Animal Care and Use Committee, Faculty of Veterinary Medicine, Chiang Mai University, Chiang Mai, Thailand (license number; S2/2561).

Saliva samples were collected from five male and five female Asian elephants aged 34.5 ± 4.7 years (range, 11–54 years) and weighed 3216 ± 306.29 kg (range, 2568–3702 kg) by applying the heart girth equation [32] from an elephant facility in Chiang Mai, Thailand. Samples were collected every 4 h between 06:00 to 22:00 h for 3 consecutive days (n = 15 samples/elephant). Bull elephants were restrained with long chains (30 m) during the day and short chains (5 m) at night. Female elephants were kept unrestrained in a fenced area (1600 m2) allowing social interactions during the day (09:00–16:00 h) and on short chains (5 m) at night. Elephants at night were chained inside an open shed with other elephants in close proximity, but with no physical contact. Females participated in tourist feeding and bathing routines twice a day (from 09:00–12:00 and 13:00–16:00 h), while bulls did not interact with tourists. Saliva collection did not interfere with the normal routine of the elephants during the day. At night, lights were turned off at around 22:00 h to allow elephants to rest, so samples were not collected between 22:00 and 06:00 h. Saliva was collected using Salivette® kits (Sarstedt, AG and Co, Numbrecht, Germany) by swiping the absorbent piece inside the buccal area for 30–60 s, which took less than 5 min to complete. Samples were kept in 4 ◦C coolers for less than 8 h until centrifuged at 1500× *g* for 5 min at 15 ◦C. Two swabs were collected and the saliva pooled, resulting in an average volume of 500 μL (100–1500 μL) per sample. Saliva was stored at −30 ◦C until analysis. Samples were analyzed within 3 months as suggested by Ng et al. [33].

## *2.2. Enzyme Immunoassays*
