1. Introduction
Presently, around 10 percent of the world’s human population visits a zoo every year, and, consequently, zoo animals are exposed to large numbers of visitors [
1]. The presence of large number of visitors in zoos has led to a variety of studies, many of them related to the impact of visitor presence and sound on animal welfare—namely, the zoo visitor effect [
2,
3,
4]. Studies have shown that the attitudes (behaviour) and presence of zoo visitors are associated with changes in the behaviour and physiology of captive animals [
4,
5,
6,
7,
8,
9].
In studies that analyse the behaviour of animals to identify the visitor effect, the expression of abnormal behaviour has often been used as an indicator of stress [
4,
10]. However, changes found in normal behaviours (such as resting and auto grooming) have also been used as indicators of animal stress [
11]. Almost all studies have shown that the presence of visitors has some effect on animal behaviour, encouraging the expression of abnormal behaviour or changing the frequencies or durations of normal behaviours. However, Gorecki, et al. [
12] found that European Souslik (
Spermophilus citellus) is active and expresses normal behaviour, such as eating, when exposed to human presence and human noise. Furthermore, a study by O’Donovan, et al. [
13] found that cheetahs’ (
Acinonyx jubatus) behaviours are not affected by visitor presence.
Some studies have shown that the presence of visitors can be enriching for the animals in captivity; for example, animals which are fed by visitors express higher activity levels and disperse through their enclosure to gain an advantage in the search for food thrown by visitors (Fa, 1989, as cited in [
2]). In contrast, most studies show that the presence of the public in zoos has a negative impact on animal behaviour, such as reducing social interactions (grooming, for example [
14]). The presence of visitors caused an increase in abnormal and aggressive behaviours in
Macaca silenus [
10], an increase in agonistic behaviours in three species of primates (
Saquinus oedipus,
Cercopithecus diana and
Lemur catta), and an increase in abnormal behaviours and activity in
Mandrillus leucohpaeus [
14]; furthermore, jaguars (
Panthera onca) suffered prolonged stress and agitation from the presence of the public [
15]. Several studies with gorillas (
Gorilla gorilla) revealed that they are more agitated in a negative way in the presence of visitors and that they are more aggressive [
11,
16,
17]. Mallapur and Chellam [
18] showed that Indian leopards (
Panthera pardus) rest more in visitor presence and that they are more active in the absence of public. The same animals expressed more abnormal behaviours when they were observed by visitors.
In most of the aforementioned studies, behavioural observations were performed over a few months with a high daily sampling effort. This form of data collection has proven to be effective. However, Margulis and Westhus [
19] demonstrated that data collected at regular intervals for long periods of time have the same efficiency; and in studies that use analysis of common behaviours of animals, the possibility of statistical error is low. Based on that, this study was conducted with constant data collection over a period of one year.
Most studies in zoos have been conducted with primates or only one species as research subject [
4]. These studies are limited because they cannot undertake comparisons of factors, which can affect how different species respond to zoo visitors. Research that simultaneously analyses different species—the comparative method—are able to explain the differences between species’ responses. They can indicate which groups of species should be examined more carefully in terms of their welfare, for example [
20].
The physical and behavioural characteristics of different species may be predictive of their response to the visitor effect: herbivores may be expected to have a stronger response than carnivores due to their evolutionary history as being a prey species; larger species may feel less intimidated by humans; diurnal species may be more affected than nocturnal as disturbance would happen during their normal activity cycle; closed habitat species would be expected to have a stronger response than species from open habitats, if they cannot fully hide; and arboreal species may feel more secure than terrestrial species, if they have appropriate substrates. These predictions have not been tested before in zoo visitor studies.
Based on the previous research presented above, we intend to overcome some presented limitations. Therefore, we tested the hypothesis that species differences in diet, activity cycle, body weight, habitat, and stratum used will affect how mammal species behaviourally respond to zoo visitors. We predict that there will be differences: between carnivorous, herbivorous, and omnivorous species; between light and heavy animals; between terrestrial and arboreal animals; between animals from open and closed habitats; and between animals with diurnal and nocturnal habits, as expressed above. We did this by observing the behavioural response of 17 different species of mammals to the presence of different group sizes of zoo visitors.
4. Discussion
For the first time, it was possible to consider the different responses of mammals to the zoo visitor effect as influenced by different physical and behavioural characteristics of species. Changes in individual behaviours can be interpreted in different ways. For example, an animal that rests more in the presence of visitors can be seen as an animal that is comfortable in the environment, even with public around, or could be seen as an animal that is resting more to avoid the visitor’s presence. Based on the literature already published on different species and on the conclusions made by authors we will discuss this in our results below. However, it is important to state that this is not a definitive discussion on the subject. This study was produced with the intent of being a first attempt at discussing zoo mammals’ response to visitors based on their physical and behavioural characteristics.
From the results, it is possible to see that carnivorous animals were less affected than herbivorous or omnivorous, terrestrial animals were more influenced by visitors than arboreal ones, closed habitat animals were more affected by visitors than open habitat species, and diurnal species were more affected by visitors than nocturnal animals. This research highlights the importance of previously overlooked factors in understanding which species are most impacted by the zoo visitor effect. For example, we can now begin the prediction that a species from closed habitat, also diurnal, terrestrial, herbivore, or omnivore would be most affected by the zoo visitor effect (e.g., deer species independent of body size).
The results showed that the presence of visitors resulted in changes in mammal behaviour: for example, a significant increase in the frequency of resting and locomotion behaviours. This evidence agrees with Hosey [
35] who states that the public in zoos directly affects animals’ activity level, which could be associated with stress. We noticed that for almost all of the factors analysed, locomotor behaviour was most commonly affected; in agreement with another study, which found that large audiences increased the locomotor activity of 12 primate species by 25% [
2]. This issue should be highlighted since an increase in locomotor behaviour could be a response to the animal’s necessity developing into abnormal behaviour such as stereotypic pacing [
36]. A study by Mallapur, Sinha and Waran [
10] concludes that, in the presence of visitors, levels of pacing for
Macaca silenus increased considerably. Another behaviour that also showed significant changes in virtually all factors was resting; confirming the observation by Wells [
11] that high public density encourages an increase in grooming by gorillas (
G. gorilla).
4.1. Habitat
Species from closed habitats were more affected than species from open habitats. We can infer from this that closed habitat species are not accustomed to contact with people, since they usually live in places where they can hide such as tropical forests, and because of this, they felt more threatened by the presence of visitors. For example, Sellinger and Ha [
15] conclude that, at higher density and agitation of the public, the stress in jaguars (
Panthera onca) is prolonged. Another example is the species
Macaca silenus, which lives in tropical forests, rarely, interacts with humans in their natural habitat. For this species, the contact with visitors in zoos can be a real source of stress [
10]. However, this situation may change if adequate opportunities to use their cryptic habits were provided. Stress levels of captive leopard cats (
Felis bengalensis) reduced significantly when opportunities to camouflage themselves and/or use hiding places were provided [
37]. Open habitat species may be more habituated to people (they often live in open environments such as savannahs) and, therefore, presented fewer behavioural changes compared to closed habitat species.
4.2. Activity Cycle
Diurnal species had more of their behavioural repertoire affected when compared to nocturnal species by zoo visitors. Observations were made during the day, the hours of visitation at the zoo, coinciding with the time activity of the diurnal species. This may have caused the greatest impact for those animals who may have felt uncomfortable and invaded by the presence of visitors. Nocturnal animals usually become inactive during the day, and possibly this was the reason they showed fewer variations in behaviours by the presence of visitors that can be a cause of concern, as they were most of the time sleeping or hiding. More studies are needed of this factor such as observations of nocturnal species at night in captivity.
4.3. Diet
Carnivorous animals being predators and, therefore, naturally more aggressive possibly felt less threatened by the presence of visitors. Our results showed this trend because carnivores had fewer behaviours that could be considered a cause of concern affected by the presence of visitors compared to herbivores or omnivores. Herbivores and omnivores are more often prey species for other animals and may perceive human visitors at their enclosures as predators due to evolutionary history [
38]. The lack of other zoo visitor studies comparing the behavioural repertoire of animals with different dietary habits hinders a deeper understanding of our results.
4.4. Stratum Used
Arboreal species virtually showed no important behavioural changes in response to visitors. This could be explained by the fact that the composition of the enclosure with trees and shelter provided the animals with control over of their exposure to the public [
35]. However, only two species in this study had arboreal habit, which could have influenced the result. Terrestrial and scansorial animals, probably by being more exposed in their enclosures, lost some control over their exposition to the public. These animals showed a significantly greater impact on their behavioural repertoire. This disagrees with Hosey [
2] who indicated a tendency for arboreal animals to be more affected by the presence of people than terrestrial animals. This difference may be due to the fact that arboreal animals in this study could maintain a minimum distance from the public because of the provision of standoff barriers (around one meter). Some zoos do not have this kind of barrier that separates the public from the edge of the animals’ enclosure.
5. Conclusions
This study identified which mammal species characteristics are most affected by the visitor effect and which factors are most associated with a negative response to this effect, for example, being a species from a closed habitat or a herbivorous animal. Therefore, we propose a predictive model showing the characteristics of species that leads to most or least concern relating to visitors in a zoo environment (
Figure 1). As an example, a deer fit in the model as an animal from open area, with diurnal, herbivorous, and terrestrial habits. Thus, it is a species in the most concern category and could be focus of more attention in the zoo when thinking about the visitor effect. On the other hand, chimpanzees, that are animals from closed habitats, omnivorous, scansorial and diurnal, are of least concern to the zoos in regards to the visitor effect. The use of this tool could help zoos better design enclosures and management protocols for their species, or even create environment enrichment activities for animals and interventions of environmental education with zoo visitors.
As mentioned previously, the model we have produced to predict zoo mammals’ response to visitor presence is a first iteration and should not be treated as definitive, especially as our sample size is small. In all areas of research when producing predictive model, it is necessary to have a first iteration, which with other studies will be modified so that future iterations will be more accurate in their predictions. We hope that this model will now be subject to testing and modification by other researchers. Comparative studies such as this are important as they illuminate the different underlying factors that can affect the response being measured. However, for a complete investigation concerning the zoo visitors’ effect, many other possibilities can be approached. For example, it would be necessary that other groups of animals besides mammals be studied such as birds and reptiles. Another proposal is to investigate nocturnal animals during their time of activity to thereby make an appropriate analysis of their welfare and responses to the public. Some zoos offer nocturnal visits where this kind of study could take place or species in zoos with nocturnal houses could be studied.
During the study, it was not possible to analyse the behaviour of animals in absence of public, because even with the presence of the researcher, there was a person present. A study that could be done is to compare differences in behavioural repertoires of animals in presence of visitors and total absence of them. For this, the proposal is to use video cameras located in the enclosures of the animals to record behaviours without an audience presence. For future research, quantified sound measurements with specific equipment should be made to investigate the quantitative difference between quiet and loud visitors.
Some previous studies analysed the physiological effect of zoo visitor on animals, showing that cortisol levels can be directly related to levels of stress [
5,
39,
40,
41]. Special attention is necessary when choosing the type of sample that will be used for corticoid measurements. Some methods (corticoid measurements from blood, saliva, milk) needs intense manipulation of the animals, which can imply in a source of stress to the animal. Therefore, the most common and non-invasive method to assess the physiology of animal welfare is the use of faecal samples [
42,
43]. This type of physiological measurements could be done in futures studies to provide a complement of the behaviours analysis in relation to visitors’ presence.