**1. Introduction**

Animals kept in zoos are constantly influenced by a number of factors that a ffect their welfare. In order to ensure the welfare of captive animals, it is essential to avoid excessive and, in particular, prolonged stress [1]. Even general husbandry tasks, such as handling, transport and veterinary treatment can be potential stressors [2]. These may cause a stress response, especially in more sensitive individuals [2].

Several methods are commonly used to evaluate stress response. They are largely based on the determination of glucocorticoid hormone levels of cortisol and corticosterone or their metabolites [3,4]. A commonly used method for determining the glucocorticoid concentration is to test the blood of the animal. However, in some zoos and in wild animals, the possibility of blood collection is limited and sometimes cannot be performed at all [5,6]. Blood collection brings about many risks, such as injuries caused during catching or sedation of the animal [6]. A more suitable technique is a non-invasive method of assessing stress levels [6], such as the determination of glucocorticoid hormone levels from urine, saliva, fur, milk or animal faeces. This non-invasive assessment of animal welfare is a practical method used for an increasing number of species not only of wild animals, but also of laboratory or farm animals and pets [7]. Collection of faeces seems to be the most appropriate in cases of stress assessment in exotic animals that are not used to handling by humans [8].

Reactions to stressors often vary among di fferent animal species [7,9,10]. Detailed knowledge of the individual species helps to understand their natural behaviour and responses to stressful situations [11]. According to Mason [12], di fferent animal species perceive the conditions of captivity di fferently. For example, small primates were more stressed by the presence of zoo visitors than larger primates. Di fferences between ground and tree species were also observed [12]. Weingrill et al. [13] compared the glucocorticoid hormone levels as stress markers in Bornean (*Pongo pygmaeus*) and Sumatran orangutans (*Pongo abelii*). Higher levels of glucocorticoids in the Bornean orangutans were predicted, due to the much more solitary lifestyle and higher frequency of male aggression directed towards females in this species in the wild, which was subsequently confirmed [13]. Furthermore, Cocks [14] found that the hybrids of Bornean and Sumatran orangutans were less resistant to stress than both subspecies, suggesting that the lower survival rate of hybrid orangutans is due to reduced fitness associated with hybridization. Given the variability in species' stress responses and the variety of animals held under human care, understanding species-specific trends in stress responses to daily stressors in captivity can be useful for guiding managemen<sup>t</sup> practices.

To achieve reliable results via non-invasive monitoring in individual species, knowledge of its metabolism and the mechanism of glucocorticoid hormone excretion is essential [1]. According to Möstl and Palme [1], the concentration of cortisol metabolites in faecal samples reflects the stress-induced glucocorticoid levels with a delay. The species-specific time delay depends on the metabolism and gu<sup>t</sup> passage time of the animal [9]. Steroids are metabolised in the liver and excreted via the bile into the gut. Therefore, measured concentrations of faecal glucocorticoid metabolites (FGMs) reflect an event occurring a certain time ago. This lag time may range from less than 30 min to more than one day, depending on the species and its activity rhythms [5,7,9]. Cambell et al. [15] reported, for instance, that the digestive tract structure varied among di fferent lemurs. They generally have a simple stomach with a slightly prolonged small intestine, sac-like appendix and large intestine of varying length. Black-and-white-ru ffed lemurs (*Varecia variegata*) are reported to have a fast passage of the digestive tract (2 to 4 h). Given the very similar structure of the digestive tract in ring-tailed lemurs (*Lemur catta*) [15], the time of intestinal passage will probably be similar. It can be assumed that for similar species belonging to the same family as in this case, after the method has been validated for one species, the same method can be used for a related species, as demonstrated by the following studies evaluating the stress response in lemurs. Balestri et al. [16] compared the faecal glucocorticoid metabolite (FGM) levels among a group of collared brown lemurs (*Eulemur collaris*) living in fragments of forests with a group living in better preserved areas. For the determination of glucocorticoid hormones in the faeces, a successfully validated analysis for this species targeted against the metabolite 11-oxo-etiocholanolone was chosen. Similarly, glucocorticoid levels in faeces during the reproduction season were analysed in red-fronted lemurs (*Eulemur fulvus rufus*) [17] using the analysis targeting the metabolite 11-oxo-etiocholanolone. The suitability of the analysis of the metabolite 11-oxo-etiocholanolone has also been confirmed by studies of the stress level in ring-tailed lemurs and black-and-white-ru ffed lemurs by non-invasive methods [18,19].

*Lemuridae* are generally classified as endangered species according to the International Union for Conservation of Nature (IUCN) Red List and their populations in the wild are decreasing although they are frequently kept in captivity. In order to ensure their stable population, health and welfare in captivity, it is important to reduce the negative effects of stress. A commonly measured endocrine response to stress is the secretion of glucocorticoids [20]. Consequently, there is a need for validation of the methods for assessing the changes in glucocorticoid concentrations in the individual animals non-invasively [1,10].

The aim of this study was to compare the changes in glucocorticoid concentrations in response to various types of potential stressors occurring during a 30-day period in a temporary housing facility in the three selected species of *Lemuridae*, namely ring-tailed lemurs (*Lemur catta*), white-headed lemurs (*Eulemur albifrons*) and collared brown lemurs (*Eulemur collaris*). Given their small body size and fast metabolism, changes in FGMs were expected to be measurable within one to two days after exposure to a stressor [1,10] as confirmed by several studies assessing changes in FGM concentrations resulting from stress, for example, in zoo-living orangutans (*Pongo* spp.) [13], common marmoset (*Callithrix jacchus*), long-tailed macaque (*Macaca fascicularis*), Barbary macaque (*Macaca sylvanus*), chimpanzee (*Pan troglodytes*) and gorilla (*Gorilla gorilla*) [21], Western lowland gorilla (*Gorilla gorilla gorilla*) [22], pileated gibbons (*Hylobates pileatus*) [23], wild gray mouse lemurs [24], ring-tailed lemurs (*Lemur catta*) [18] and black-and-white ruffed lemurs (*Varecia variegata*) [19].

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

## *2.1. Ethics Statement*

This study was carried out in strict accordance with the Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes and Czech national legislation (i.e., Act no. 246/1992 Coll.) on the protection of animals against cruelty, as amended. All samples were collected non-invasively and animals did not undergo any experimental procedures. Sampling was carried out during temporary housing of three species of *Lemuridae* in an approved holding facility. Since only non-experimental clinical veterinary practices were performed and no handling of animals related to research was carried out, a formal ethics approval from the Animal Welfare Body of the University of Veterinary and Pharmaceutical Sciences Brno with regard to the EU Directive 2010/63/EU was not required.

#### *2.2. Study Subjects and Housing*

For the analysis, faecal samples from 9 ring-tailed lemurs, 6 collared brown lemurs and 2 white-headed lemurs were collected for a period of 30 days. The lemurs originated from European zoos and were temporarily housed in a holding facility in Poštovice in the Czech Republic operated by a certified international transportation company, The Nature Resource Network, that was in charge of their transport to other zoos. Details on individual animals monitored in the study are given in Table 1.

All lemurs were fed primate dry formula (pellets), vegetables (carrot, celery, cucumber, lettuce, red beat, onion, boiled rise, tomatoes) and fruits (banana, apple, peach, pineapple, grapes, kiwi) two times in a day, with green leaves, and with cheese and boiled egg once a week. Drinking water was available ad libitum. The lemurs were housed in indoor enclosures (natural light provided from windows) without access to an outdoor area. Changes in the groups during the monitored period are described below.

In ring-tailed lemurs, 9 adult individuals were monitored (8 females and 1 male). There were 4 offspring in the ring-tailed lemurs but they were not included in the analysis. Four adult females (R1–R4) were housed in cage 1 in room 1. Two of them had offspring (R3 had two offspring, R4 had one offspring). The two lactating females (R3 and R4) together with their offspring were separated on day 16 of monitoring to another cage in the same room due to their agonistic behaviour. Female R3 with her offspring was housed in cage 2 and female R4 with her offspring was housed in cage 3. The remaining two adult females (R1 and R2) were left in cage 1. Subsequently, female R1 was isolated in a separate cage 4 in the same room due to injury on the 23rd day of observation. Another four adult ring-tailed lemurs (R5–R8) were housed in room 2. Each adult was individually housed in separate cages, except for female R6, who was co-housed with her o ffspring. The first day of monitoring one more adult female ring-tailed lemur (R9) arrived at the facility and was placed in a cage in room 6. Rooms 1 and 6 were interconnected. In white-headed lemurs, 2 individuals (both males, W1 and W2) were housed together throughout the whole monitoring period in room 1 in cage 5. In collared brown lemurs, 6 individuals were assessed: 2 males (C1 and C2) and 4 females (C3–C6). Originally, they were housed together in cage 1 in room 5. During the monitored period, however, they were divided into smaller groups due to mutual attacks and relocated to di fferent cages within the same room (room 5). Namely, on day 16 two males (C1 and C2) were relocated to cage 2, on day 18 one female (C3) was relocated to cage 3, on day 23 one female (C4) was relocated to cage 4 and on day 29 one female (C6) was relocated to cage 5. While the ring-tailed lemurs and white-headed lemurs were housed in the same room, the collared brown lemurs were kept on the other side of the facility. A scheme of the floor plan of the facility and relocation of individuals is provided in Figure 1. All lemurs had visual, olfactory and auditory contact with other animals of the same species during the entire monitored period. All animals were clinically healthy without any signs of health problems or behavioural disorders.

#### *2.3. Faecal Sampling and Analysis*

The monitoring was carried out for 30 days when three lemur species were kept in the facility. Lemurs were exposed to the same husbandry activities (cage cleaning, feeding and watering, visual control of the animals by caretakers). All activities related to the operation of the facility were recorded (e.g., construction work, increased noise, newly incoming animals, transfers within the facility, lactating animals, etc.) by the caretakers during the monitored period in a systematic way with the use of pre-prepared sheets. The faecal samples for the determination of FGMs were collected from the floor once a day between 8:00–12:00 a.m. from each animal, ensuring that they were not contaminated with urine, water or other material. A total of 510 faecal samples were taken. There was no variability in the number of samples taken from di fferent animals. The collected samples were put in plastic sealable bags within 1 h of defecation, labelled and immediately frozen at −20 ◦C. The subsequent sample processing including extraction was performed at the University of Veterinary Medicine in Vienna, Austria. The samples were thawed, individually homogenised and then successively put into test tubes at weights between 0.480 and 0.520 g. Afterwards 1 mL of distilled water and 4 mL of 100% methanol were added to the weighed samples. Each sample with water and methanol was stirred on a shaker for 30 min and then centrifuged at 4 ◦C (3750× *g*, 10 min). The detailed extraction process is described in other methodological publications [1,9,25]. Concentrations of immunoreactive glucocorticoid hormone metabolites were measured in the obtained extracts by using an enzyme immunoassay. The polyclonal antibodies used in this assay were directed against the metabolite 11-oxo-etiocholanolone I (laboratory code: 72a, first described by Palme and Möstl [26]), based on previously conducted studies.

## *2.4. Statistical Analysis*

The results were statistically evaluated using the statistical program Unistat 5.6 (Unistat Ltd., London, UK). Data normality was tested using the Shapiro–Wilk test [27]. Since the data were not distributed over the Gaussian curve, the statistical significance of the FGM level di fferences between the monitored lemur species during the 30 days of observation was determined on each day using a non-parametric Kruskal-Wallis ANOVA test [27] and subsequently nonparametric multiple comparisons with t-distribution using rank sums [27]. For each lemur species, the mean value of the FGMs measured on the first day of the observation was also compared to each subsequent day to determine statistical significance of the di fferences in the FGM levels by means of the Friedman ANOVA test [27] and subsequently the Dunnett test using rank sums [27]. A value of *p* < 0.05 was considered significant. The first day of sample collection was the day when FGM levels were expected not to be a ffected by stress, not even in those individuals who arrived on the first day of observation due to a time delay in excretion of FGMs [1]. Therefore, the data from the first day were used as the reference point. FGM levels measured during the monitored period were compared with the timeline of events recorded in the facility in order to determine possible factors eliciting the increase in FGM levels.
