Annual Interdisciplinary Graduate Symposium presented by the Anthropology Graduate Student Association

Human fascination of primates is not a new construct. The first zoos were symbols of social status, filled with homages paid to kings by subordinates of neighboring countries. As the centuries ticked by, zoos evolved along with human perceptions and ideas about animals. By the 18^th century, the view of the world moved into an investigative nature from the previous reliance on faith. Collecting and cataloguing animals became a passion for many early scientists. By the turn of the 19^th century, traveling menageries were commonplace, and thus the view of animals for entertainment was born. Permanent collections began to spring up, but unfortunately it was not until the mid 20^th century that zoos began to understand their roles as educators. Today, zoos are a crucial part of education and conservation. However, they still have limitations. Animals living in captivity are provided with basic needs: food, water, and shelter. However, having these essentials leaves many hours of an animals’ day empty. In addition, captive animals are introduced to unfamiliar humans on a daily basis. These two situations can potentially contribute to physiological responses in animals related to stress. Previous research indicates that physiological stress is associated with a number of behaviors in captive primates (Sapolsky 2005). Given the negative health effects of chronic stress (Maestripieri et al. 1992) it is crucial that caretakers recognize its causes in order to circumvent them as much as possible.

Zoos are constantly working within a framework of conflicting goals (entertaining and educating visitors vs. animal welfare). On the one hand they must focus on bringing as many patrons as possible through their gates, and on the other, it has been shown that high densities and noise levels of visitors can be a source of stress for zoo-housed animals (Fernandez et al. 2009). In the last decade, numerous studies have been conducted on possible effects of visitors on captive zoo animals, and have found results that vary widely both between and within species (Hosey 2000). Although many of these investigations have provided information on visitors’ perception of the zoo environment, and on visitor interaction with captive animals (Davey 2006a), most have not included information about specific visitor characteristics such as density, noise or behaviors. This may be a critical mistake as there is some evidence that some of these variables can influence animal behavioral responses in captivity (Davey 2006b).

Now, the concept of "zoo" has evolved. Thanks to technological advances, humans can wander farther than ever before. Visiting exotic locations to enjoy nature and endemic wildlife is no longer just an adventurer’s dream, but a tourist’s reality. Ecotourism is the industry that emerged to supply these needs. They often advertise ecological, economic, and social benefits, and focus their hype around one or two wildlife species. However, ecotourism establishments may actually have negative impacts on target species. Specifically, some tourist management practices have been linked to behavioral problems and reduced fitness in primates; one example is the provisioning of food. In many locations, [e.g. Formosan macaques (Macaca cyclopis) at Shou-Shan Nature Park, Taiwan] groups are provisioned at regular intervals even though provisioning increases both the frequency and duration of agonistic interactions (Hsu et al. 2009). Range restriction is also a concern: at Mt. Huangshan, in China, Tibetan macaques (Macaca thibetana) under these management practices experience increased conspecific aggression and infant mortality (Berman et al. 2007). The detrimental impacts of these stressful events can negate positive influences of ecotourism and demonstrate the need for a better understanding of primate responses to tourism.

Organisms maintain an internal equilibrium (homeostasis) through coordinated physiological responses to stimuli that may potentially threaten their normal function. Stress is one of these responses (Sapolsky 2005). In addition to visitors, several other potential sources of stress in the environment of zoo animals have been discussed; for example, artificial lighting, loud noise levels, “uncomfortable” temperatures, or specific odors (Morgan & Tromberg 2006). Some potential stressors may be “confinement-specific”, for instance, restricted space. In addition, different stimuli (loud noises, continuous light cycles, confinement, and proximity to humans) have been found to elevate levels of both physiological (stress hormones) and behavioral arousal. Whether or not particular stimuli are sources of stress is usually determined by examining physiological responses to them.

Acute stressors typically first activate a cascade of physiological responses by the animal’s autonomic nervous system. This process is called “fight or flight” and it is seen as an adaptive response to a perceived threat to survival. Several physiological actions occur during this process: heart rate increases, digestive system slows, blood is directed into limbs, pupils dilate, and immune system activity increases. “Fight or flight” is an adaptive response because the body becomes prepared for a situation that could possibly decrease the chances of survival. Having these physiological responses to stress can increase fitness. However, to have a system in “fight or flight” mode frequently, or for long durations, can be maladaptive and disruptive to physiological processes, such as immune function and growth. This is due to its high energetic cost, as well as the hormones that are released in response to stress.

A second physiological response system involves an increase in activity in the hypothalamic-pituitary-adrenal (HPA) axis, which increases, among other things, the cortisol levels within the individual under duress. These levels can be measured in blood, saliva, and urine. Chronically high levels of this steroid can cause hypertension, diabetes, and ulcers, which can decrease fitness. Chronic stress can cause the nervous system to activate in this manner. Smith & French (1997) explored psychosocial stress and urinary cortisol levels in marmosets (Callithrix kuhli). They found that exposing marmosets to particular stimuli increased cortisol levels in a dose dependent manner: as the stressor level increased, the amount of cortisol measured also increased. Increases in urinary cortisol have been tied to the presence of visitors and captive primates. For example, spider monkeys (Ateles fusciceps) at the Chester Zoo showed an increase in cortisol levels associated with an increase in visitor density (Davis et al. 2004), although other factors can also modify the expression of cortisol, such as age, social dynamics, and reproductive status (Clarke et al. 1996). Other physiological indicators of stress can also be used; individuals under stress also show increases in heart rate, blood pressure, and cathecolamines and corticosteroids in the blood. However, collection of these biological markers can be invasive, expensive, and disruptive. For these reasons, the use of behavioral indicators may be more appropriate in zoo settings.

Stress responses to a variety of stimuli, including unavoidable and chronic exposure to unfamiliar humans, have been identified in primates and humans via specific physiological, psychological, and behavioral responses (Sapolsky 2005). Physiological stress responses are commonly measured in wild primates by assessing changes in glucocorticoid levels in feces and saliva; behavioral responses (behavioral stress indicators) emerge as an increase in self-scratching, picking, and aggression. Problems can arise when relying only on one measure of stress. For example, fecal metabolites are subject to bacterial degradation (Negrao et al. 2004), represent an average concentration over an extended period, rather than a response to a particular stressor or event (Lutz et al. 2000), and do not always correlate with known stressors (Marechal et al. 2011).  Saliva has a host of problems as well; because salivary cortisol levels are an acute measure, the timing of its collection may be problematic. It may not always be possible to obtain a sample soon after an event of interest. In addition, the techniques associated with collecting salivary samples are complex. Another issue to consider is the role of behavioral stress indicators. These are less expensive and require less processing, but have been assumed to be less accurate or sensitive than physiological indicators. However, this is not necessarily the case. Animals may respond behaviorally even when there are no detectable physiological responses (Marechal et al. 2011).

In the last 20 years, researchers have identified a number of specific behaviors that correlate with physiological stress indicators, and hence, can be used reliably as behavioral stress indicators. These links allow researchers to use these behavioral measures to infer stress levels inexpensively and noninvasively. The two most commonly studied behavioral stress indicators (also referred to as displacement activities) are self-scratching and auto-grooming, however, other measures can be utilized such as regurgitation, aggression and social activities (Maestripieri et al. 1992; Fernandez et al. 2009).

Behavioral stress indicators have been used in a wide range of primate species and contexts. For example, Palagia & Norscia (2010) explored self-scratching after stressful events in wild brown lemurs (Eulemur fulvus) on the island of Madagascar. They found that self-scratching increased after predatory attacks and intra-group aggression and decreased after reconciliation. In addition, Rowell & Hinde (1963) found that when rhesus monkeys (Macaca mulatta) were exposed to a mild (observer staring) and strong (observer wearing a grotesque mask while making quick movements) stressors they showed an increased frequency of self-scratching. Auto-grooming, defined as manipulation of one’s own fur and self, has also been used extensively as a behavioral stress indicator. For example, Mireille Bertrand (1969) observed increased levels of auto-grooming by a pair of juvenile stumptail macaques (Macaca arctoides) who were housed in unusually small cages. In addition, she noted occurrences of self-scratching and autogrooming in response to staring by observers.

Correlations have been found between increases in physiological and behavioral stress indicators. For example, Clarke et al. (1996) examined the relationship between aggression and hormonal responses associated with social change in two groups of rhesus monkeys. The control group consisted of 51 individuals and was considered to be socially stable. The study group consisted of 49 individuals that had previously resided in nine separate groups. Researchers found a consistent relationship between aggression and physiological stress indicators in the study group; specifically, both non-contact aggression and cortisol levels increased during the first 24 hours after introduction. In addition, specific social behaviors have been linked to stress and/or stress relief. For example, Rowell & Hinde (1963) found that in both mild and strong stress conditions, grooming of others ceased.

Regurgitation has also been proposed to be an indicator of stress in some captive primates, although there is much debate about the cause of it, especially in gorillas. Some researchers characterize it as coping strategy to compensate for the decrease in foraging time in captivity (Akers & Schildkraut 1985). In a study of gorillas in 14 zoos, Gould & Bres (1986) found that regurgitation occurred significantly less in mother-reared versus hand-reared individuals, and suggested that it might be a response to current stress in individuals previously exposed to the stress of early separation.
Visitor Effects on Captive Primates:

Zoos tend to focus conservation and education efforts on a number of intelligent, social and appealing species, including primates. As such, primates have been the focus of most visitor effect studies. In general, the results of these studies suggest that primates find the presence of large active groups of visitors to be stressful. Reviews of previous research on captive primates have found that the presence of visitors can lead to increased intra-group aggression, alterations in use of enclosure space, and increased abnormal behavior (Hosey 2005; Davey 2006a; Davey 2007 & Fernandez et al. 2009). Many studies on captive great apes have supported these findings. For example, Maki et al. (1987) recorded aggressive behavior in two chimpanzee (Pan troglodytes) social groups at UT MD Anderson Cancer Center in Bastrop, TX. Chi-square analysis indicated that a significant increase in the frequency of intra-group directed aggressive behaviors occurred when unfamiliar visitors were present, as opposed to when only the staff was present.

In addition to increased intra-group aggression, visitor presence can alter a great ape’s use of enclosure space. Measuring use of space is important as it can be used as an indicator of visitor avoidance (Hosey 2005). A study of two groups of gorillas (Gorilla gorilla gorilla) at Disney’s Animal Kingdom found that gorillas were more likely to be out of view in the “Large Crowd” condition than they were in the “Small Crowd” condition. Accurate crowd size was obtained from a turnstile counter, through which all patrons who viewed the gorillas had to exit. This result suggests that being in view of larger crowds is an aversive situation (Kuhar 2008). In addition, Birk (2002) examined effects of visitor density on orangutans (Pongo pygmaeus) at the Chester Zoo and found that at higher levels of visitor density, orangutans utilized paper sacks more frequently to cover their heads. This suggests that visual perception of visitors by apes plays a key role in understanding visitor effects. In order to explore the importance of visual perception, Blaney & Wells (2004) erected a visual barrier in front of a gorilla enclosure at the Belfast Zoological Gardens. They found that adding the camouflage net barrier reduced self-scratching (an indicator of physiological stress) and intra-group aggression. These results suggest that altering great apes’ perception of the presence of visitors may reduce the negative effects of crowds.

A few small-scale studies on gorillas at the Buffalo Zoo also suggest negative response to visitors. Last spring, Dominique Mendez, a student in my advisor’s Ethology Practicum course, measured levels of behavioral stress indicators, relaxation, and aggression both in the presence of and absence of visitors. Her results suggested that gorillas exhibit higher levels of behavioral stress indicators when visitors are present compared to when they are absent. However, this study included data for only 20 days and recorded visitors present at only one of four viewing windows in the exhibit. In addition, Dawn Bruffett, another student in my advisor’s Ethology Practicum course, measured levels of locomotion, behavioral stress indicators (self-grooming, self-scratching, hitting the glass, charging), grooming, and use of space as a function of changes in visitor density and noise level. Her results suggested that each gorilla acted uniquely depending on the combination of independent variables; Koga (the silverback male) was affected by both noise level and density, while Lily and Becky (2 mature females) were affected only by noise level. However, this study included data for only 10 hours.

Other primate species have also been found to display behavioral stress indicators, abnormal behaviors or changes in their use of enclosure space in the presence of visitors (Davey 2007). Gibbons (Hylobates syndactylus) at Disney’s Animal Kingdom spend significantly more time away from visitors (behind visual barriers) during periods of high attendance than low attendance. (Smith & Kuhar 2010). Another visitor effect study was performed on captive lion-tailed macaques (Macaca silenus) in two conditions (short-term & long-term) across eight zoos in India. The short-term condition measured behavioral stress indicators while the monkeys were situated in their “on-exhibit” enclosure to see the immediate effects of visitor presence. The long-term condition explored the lasting impact of visitor presence by measuring the same behaviors for a different group of monkeys during both their “on-exhibit” and “off-exhibit” days. Abnormal behaviors increased significantly when visitors were present in both conditions, suggesting that visitor presence is an aversive experience to lion-tailed macaques (Mallapur et al. 2005).

Among New World Monkeys, critically endangered cotton-topped tamarins (Saguinus oedipus Oedipus), have been found to be difficult to breed in captivity, particularly when on display to the public. A comparison of the social behavior of a display group (which was not breeding successfully) versus an off-display group (which was breeding successfully) indicated that the “on-display” breeding pair interacted with each other significantly less than the “off-display” pair (Glatston et al. 1984). These results suggest that visitor presence may be a possible source of differences in tamarin social behavior and breeding success.

Oswald and Kuyk (1977) explored visitor effects on three nocturnal Prosimian species at the Woodland Park Zoological Gardens: lesser galagos (Galago senegalensis), greater galagos (G. Crasicaudatus), and slow lorises (Nycticebus coucang). All three species showed some increase in activity during their hours on display. The lesser galagos and lorises showed an increase in autogrooming, while the greater galagos showed an increase in locomotion. It was also noted anecdotally that lesser galagos spent more time out of view of visitors while on display.

Although many visitor effect studies suggest that captive primates respond negatively to the presence of visitors, a few do not. The reasons for conflicting outcomes are unclear. Some researchers believe that they may be due to methodological design flaws or species-specific differences. Some notable design flaws include a lack of specific definitions for both visitor attributes and primate responses, as well as a lack of attention to possible interacting variables such as season, outdoor access, temperature, and group composition. For example, Deborah Wells (2005) looked at the effects of visitor density on gorilla behavior at the Belfast Zoological Gardens and found that the presence of visitors did not have a significant effect on aggression, abnormal behavior (repetitive teeth clenching, body rocking, spinning), auto-grooming, or use of space. However, she did not use an ethogram (a clearly defined set of behaviors specific to the species) of gorilla behaviors, nor did she control for possible seasonal differences. Proper delineation and classification of behavioral stress indicators may be necessary to correctly interpret the results of a visitor effect study. In another example, Carder & Semple (2008) observed two groups of western lowland gorillas and found positive correlations between stress-induced behaviors and numbers of visitors in one group, but not in another. The second group lacked a silverback (a fully grown sexually active, adult male), suggesting a possible interacting effect of group composition. Also, a study done on eastern lowland gorillas (Gorilla beringei graueri) at the Antwerp Zoo in Germany revealed that visitor presence affected only one out of the 5 adults’ use of enclosure space. However, during scans, the researchers measured only the presence or absence of visitors along with gorilla location, instead of measuring visitor density, location or other characteristics. It is possible that these other characteristics may have interacted with gorilla behavior. In fact, visitor location may have been of particular importance because the shape of the viewing area zone consisted of odd angles. For a gorilla to be considered in the zone, it must be within 6m of the zone viewing windows. Measures were taken without regard for how far away a visitor was from a gorilla. It is possible that a gorilla sitting inside the zone, but on the opposite side of the enclosure from a visitor, may have perceived itself to be less at risk than one much closer to visitors (Vrancken et al. 1990).

Tourism is an expansive industry that generates billions of dollars a year. Many tourism operations attempt to conserve eco-systems as well as to educate and benefit local populations, all under the umbrella of profit. However, these goals are not always achieved. Many "eco-tourism" locations feature primates as main attractions and may actually have negative impacts on target species. Specifically, some tourist management practices have been linked to behavioral problems and reduced fitness in primates. For example, groups of pygmy marmosets (Cebuella pygmaea) that are exposed to relatively more tourists and tourist-related activities engage in less social play and avoid the lower strata of the forest than less exposed groups (De la Torre 2000). In a four year study, it was shown that Howler monkeys in Belize, regularly exposed to tourism, have higher cortisol levels than those exposed only to researchers (Behie et al. 2010). In addition, at Mt. Huangshan, Tibetan macaques (Macaca thibetana) have their ranges restricted in spite of evidence that this practice is associated with increased aggression and infant mortality (Berman et al. 2007). Also in this group, stress related behavior is correlated with tourist density, and specific tourist behavior is related to monkey aggression (Matheson et al. 2007). Specifically, intragroup aggression sometimes follows tourists feeding monkeys, while tourist directed threats are preceded by attention getting noises and staring at the monkeys. These examples clearly show the value of pinpointing aspects of tourism that are harmful versus neutral, because it allows designers the opportunity to maximize the benefits of tourism and to minimize the detrimental effects on the target animals.

Some studies have paired behavioral stress indicators with measures of cortisol. On a study done on orangutans at the Kinabatangan Orangutan Conservation Programme, fecal cortisol levels were significantly elevated in samples collected the day after tourist visitation, among the habituated animals used for tourism. This is indicative of elevated cortisol production on the previous day during tourist visitation (Muehlbrien 2012). In addition, wild Moroccan Barbary macaques (Macaca sylvanus) show elevated rates of stress indicators (self-scratching) with increased proximity to tourists, and these rates rise when tourists speak or take pictures. Fecal glucocorticoids also rise significantly the day after aggressive encounters between macaques and tourists. Although physiological and behavioral measures of stress are generally well correlated with one another, this is not always the case (Marechal et al. 2011). When assessing the impacts of tourism in eco-tourist locations, considering more than one physiological stress measure with behavioral measures is likely to yield finer grained and more comprehensive descriptions of stress responsiveness than using any single type of measure.
Discussion:

While these examples show how tourism negatively affects the behavior of primates, some operations utilize finely tuned regulations and protocols to alleviate stress. In Volcano National Park (VPN), Rwanda, strict protocols are maintained regarding tourism group size, group/animal distance, number of tourist groups allowed per day, time limit for each visit, and dismissal of sick tourists. In addition, other fine tuned rules exist such as denying flash photography and insisting that human fecal waste be buried. These rules focus to maintain the physiological and psychological health of the local gorilla population. In fact, gorillas are checked on a daily basis, and frequent health evaluations are conducted. This extreme regulation is made possible by the government of Rwanda, which takes a primary role in the maintenance of its ecotourism industry. Not only do the gorillas remain visibly healthy, but also the social and economic impacts of Rwandan gorilla tourism are huge. Almost 1000 people are employed by VPN, and also, the generated revue has allowed several schools, water tanks, and other social improvements to be built.