The visitor effect in zoo-housed apes: the variable effect on behaviour of visitor number and noise

Abstract

Human visitors have the potential to impact heavily upon the welfare of zoo-housed animals, and the study of the effect has become an established research area in the modern zoo. This effect can be caused not just through the presence of visitors, but also through their behaviour. This study sought to test the hypothesis that visitor number and the associated noise level significantly affected the behaviour of three zoo-housed primate species. This was studied through behavioural observations and measurements of visitor numbers and noise levels around enclosures, as primate species are particularly sensitive to large, noisy crowds of zoo visitors. Changes in behaviour relating to visitor number and noise levels were investigated on a species and individual level. Noise levels had a significant positive relationship with visitor number, and both factors had significant positive and negative effects on stereotypic, locomotory, inactive and feeding behaviours on an individual and species level. However, levels of individuals sitting with their back to the window was unaffected by visitor number or noise. Individual and species differences were seen in reactions to the visiting public, emphasising the complex nature of the study of the visitor effect. The increase in stereotyping and clinging behaviours, and decrease in inactivity suggest a potential negative influence on the welfare of these primates. The mixed results reinforce the notion that the visitor effect is moderated and influenced by many factors, such as husbandry and personality. The current study highlights the need for off show areas for captive primates, and the importance of considering individual differences when attempting mitigation of unwanted behaviours.

Full text

OPEN ACCESS JZAR Research arcle
Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
268
OPEN ACCESS
Research arcle
The visitor eect in zoo-housed apes: the variable eect on behaviour
of visitor number and noise
Anita Hashmi1 and Mahew Sullivan1
1Manchester Metropolitan University, All Saints Building, All Saints, Manchester M15 6BH
Correspondence: Anita Hashmi, email; anita.hashmi@stu.mmu.ac.uk
Keywords: apes, behaviour, visitor
eects, welfare
Arcle history:
Received: 11 Feb 2020
Accepted: 06 Oct 2020
Published online: 31 Oct 2020
Abstract
Human visitors have the potenal to impact heavily upon the welfare of zoo-housed animals, and the
study of the eect has become an established research area in the modern zoo. This eect can be
caused not just through the presence of visitors, but also through their behaviour. This study sought
to test the hypothesis that visitor number and the associated noise level signicantly aected the
behaviour of three zoo-housed primate species. This was studied through behavioural observaons
and measurements of visitor numbers and noise levels around enclosures, as primate species are
parcularly sensive to large, noisy crowds of zoo visitors. Changes in behaviour relang to visitor
number and noise levels were invesgated on a species and individual level. Noise levels had a
signicant posive relaonship with visitor number, and both factors had signicant posive and
negave eects on stereotypic, locomotory, inacve and feeding behaviours on an individual and
species level. However, levels of individuals sing with their back to the window was unaected by
visitor number or noise. Individual and species dierences were seen in reacons to the vising public,
emphasising the complex nature of the study of the visitor eect. The increase in stereotyping and
clinging behaviours, and decrease in inacvity suggest a potenal negave inuence on the welfare
of these primates. The mixed results reinforce the noon that the visitor eect is moderated and
inuenced by many factors, such as husbandry and personality. The current study highlights the need
for o show areas for capve primates, and the importance of considering individual dierences when
aempng migaon of unwanted behaviours.
Introducon
One prevalent factor of life for zoo-housed animals is visitors
to the instuon which houses them. Since the 1980s, the
study of the visitor eect has become an established research
area (Fernandez et al. 2009; Collins and Marples 2016). Visitors
can present as a welfare issue that is not easily remedied.
Two main hypotheses exist regarding zoo visitors: ‘visitor
aracon’, whereby visitors are more aracted to more acve
animals (Mitchell et al. 1992a), and ‘visitor eect’, whereby
the presence of visitors changes animal behaviour. The ‘visitor
eect’ exerts diering inuences dependent on various factors.
Visitors can be enriching for some species (Markowitz et al.
1981; Moodie and Chamove 1990; Hosey 2000; Hosey 2005)
or have no eect (Fa 1989; Mather 1999; Collins et al. 2017).
Interesngly, the COVID-19 pandemic has prompted reports of
animals ‘missing’ visitors. However, visitors may be dened as
a ‘stressful inuence’ (Hosey 2000) and detrimental to welfare.
This eect can cause negave behavioural responses in zoo-
housed animals, for example, decreased acvity (Chamove
et al. 1988; Mitchell et al. 1992a; Wells 2005), increased
aggression (Chamove et al. 1988; Mitchell et al. 1991; Blaney
and Wells 2004; Wells 2005; Kuhar 2008; Collins and Marples
2016).
The visitor eect is mulfaceted. Hosey (2000) argues that
primates are parcularly sensive to the visitor eect, and the
majority of previous literature suggests a stressful inuence; a
mix of posive, neutral and negave results have been noted
in non-primates (Fernandez et al. 2009). Varied responses to
human presence and behaviour have been observed across

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
269
Hashmi and Sullivan
primates (Chamove et al. 1988; Clark et al. 2011). Suggesons for
this discrepancy include social structure (Hosey 2005) and body size
(Chamove et al. 1988); however, there is no clear evidence for the
cause of these dierences. Furthermore, quesons persist about
the inuence of age on the eect of visitor presence and noise
levels, exemplied by studies including infant-specic behaviours
(e.g., Birke 2002; Cheyne 2006; Collins and Marples 2016). Sex,
environment and personality may all inuence the visitor eect
and its expression. Environment is a complex variable, as it diers
signicantly across instuons. However, a key requirement for
reducing negave reacons to visitors is seemingly free access
to private areas. Blaney and Wells (2004) observed reduced
aggression and abnormal behaviours in gorillas aer the provision
of a camouage net, which reduced direct visual contact with
visitors. Similarly, the use of privacy screens reduced negave
vigilance behaviour in a second group of gorillas (Clark et al. 2011)
and aggression in capuchins (Sherwen et al. 2015). Furthermore,
Bornean orangutans displayed increased avoidance behaviour at
mes of high visitor density (Birke 2002). It is unclear whether the
inuenal factor is privacy or the choice to use it. When given the
choice to use private areas, polar bears performed less stereotypic
pacing and increased social play (Ross 2006), and pandas showed
lower levels of behavioural agitaon (Owen 2004). Freedom of
choice reducing visitor stress behaviours has also been seen in
orangutans: the provision of choice led to infrequent observaons
of stereotypic, abnormal and aggressive behaviours (Bloomeld
et al. 2015). Addionally, visitor behaviour also appears to exert
diering eects: passive audiences do not elicit the same response
from capve primates as acve audiences (Hosey and Druck 1987;
Mitchell et al. 1992a; Birke 2002).
Previous studies have established certain behaviours as
stress indicators. Stereotypic behaviours, dened as “repeve
behaviours that are abnormal when compared to the animal’s
natural behaviour paerns” (Wielebnowski 1998), are an indicator
of stress or an inability to cope with a stressor. In primates, common
stereotypies include self-harming, for example, scratching and
mulaon (Cooke and Schillaci 2007; Hosey and Skyner 2007;
Carder and Semple 2008), and atypical ingeson behaviours such
as coprophagy (Bloomsmith et al. 2007). Regular display of these
can allude to underlying welfare issues. Further situaon-specic
behaviours include visitor avoidance, animals sing facing away
from visitors (Collins and Marples 2016), and increased aggression
(Chamove et al. 1988; Mitchell et al. 1991; Kuhar 2008; Bortolini
and Bicca-Marques 2011), and locomoon (Chamove et al. 1988;
Mitchell et al. 1992a; Wells 2005). Many of these behaviours have
been treated as indicaons of fear of humans in domesc animals
(Hemsworth et al. 2018) and of a negave welfare state (Botreau et
al. 2007; Mellor et al. 2009; Hosey 2013). Furthermore, it has been
suggested that relatedness to humans may inuence behavioural
reacons to visitors. Direct eye contact is a threatening gesture in
some species (de Waal 2003; Fuentes and Gamerl 2005).
Increased aggression as a result of increased visitor presence
has been seen across primate species: mandrills and mangabeys
showed increased aggression following increased visitor
numbers (Chamove et al. 1988; Mitchell et al. 1991); siamangs
and orangutans displayed more aggressive behaviour when
threatening gestures (e.g. yawning) were performed by visitors
(Nimon and Dalziel 1992; Birke 2002). Conversely, aggression in
capuchins was reduced when direct visual contact with visitors
was hindered (Sherwen et al. 2015).
This study examines the eect of visitor number and noise levels
upon the behaviour of three ape species: western lowland gorillas
Gorilla gorilla gorilla, Bornean orangutans Pongo pygmaeus, and
pileated gibbons Hylobates pileatus. The inclusion of three species
allows for comparisons of reacons across the family Hominoidea.
The aim of the study was to establish whether visitors aected
the ape groups, and how this potenal eect manifested in
behavioural change.
Methods and materials
Animals and enclosures
The study subjects were six western lowland gorillas with an
average age of 12.92 years ±10.60, ve Bornean orangutans with
an average age of 19.20 years ±9.18 and four pileated gibbons
with an average age of 12.08 years ±13.19 (Table 1). All animals
were housed at Blackpool Zoo, UK. The gorilla enclosure consisted
of an indoor and outdoor area, between which constant access
was provided except during cleaning. Access was given to ‘Gorilla
Mountain’, an addional outdoor enclosure, on an ad-hoc basis.
Orangutans and gibbons were housed in similar indoor-outdoor
enclosures, with the orangutans housed in the same building as the
gorillas and the gibbons in the ‘Small Primate House’. All normal
husbandry and feeding rounes were observed for the duraon
of the study, with parcipants maintained on a typical diet. One
gorilla and orangutan feed was provided during educaonal talks.
Gorilla and gibbon groups were well-established at the me of
the study; the zoo had received one orangutan (Jingga) in October
2017. Enclosures had remained unchanged for several years, with
the most recent enclosure upgrade completed in 2014.
Data collecon procedure
Data were collected twice per week, one species per session.
Observaons took place between 1000 and 1500, April-August
2018. Data for each species were collected on a rotang schedule,
with three sessions of 10 min per individual daily. Prior to each
session a 10-min habituaon period was observed to allow
parcipants to acclimase to the researcher’s presence (Mitchell
et al. 1992b). Instantaneous sampling was used every 2 min to
record the focal animal’s behaviour (Table 2), the number of
visitors present, the noise level (using a Precision Gold N05CC
decibel meter), and any addional informaon, for example,
parcipants in a social interacon. A sampling interval of 2 min was
Table 1. All individuals included in behavioural observaons.
Species Names Sex Age at beginning of study
Gorilla gorilla gorilla Bukavu M 20
Gorilla gorilla gorilla Miliki F 23
Gorilla gorilla gorilla Njema F 24
Gorilla gorilla gorilla Meisie F 7
Gorilla gorilla gorilla Moanda M 3
Gorilla gorilla gorilla Makari M 6 months
Pongo pygmaeus Ramon M 19
Pongo pygmaeus Vicky F 33
Pongo pygmaeus Cherie F 21
Pongo pygmaeus Summer F 15
Pongo pygmaeus Jingga F 8
Hylobates pileatus Chamoa M 16
Hylobates pileatus Ivy F 29
Hylobates pileatus Dobby M 3
Hylobates pileatus Baby M 4 months

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
270
The visitor eect in zoo-housed apes
selected to allow for a count of visitor numbers, recording of the
noise level, and following of the focal animal between enclosures
where necessary. Sampling order of individuals was determined
by assigning each animal a number and using a random number
generator before each sampling session. This resulted in a total of
180 observaons per individual and a total of 1080 observaons
for gorillas, 900 for orangutans and 720 for gibbons.
Talks took place once daily for orangutans throughout the
duraon of the study, and once daily for gorillas from the start
of the peak season (June). Addional talks took place throughout
the day at neighbouring enclosures and aected the noise levels
around the ape enclosures, therefore such talks were noted when
they occurred. Data were collected during talks and analysed
separately. Noise specic to visitors – inclusive of talks – was
recorded, while other environmental sounds, for example, vans
driving past enclosures, were excluded from analysis.
Stascal analysis
One orangutan (Summer) was removed from the dataset prior to
analysis due to a veterinary procedure and her subsequent removal
from the orangutan group interrupng data collecon, resulng
in 720 observaons for orangutans, which were split into ‘before’
and ‘aer’. All tests performed on orangutan data were performed
on the ‘before’, ‘aer’ and full datasets. Data were analysed using
RStudio version 1.1. Tests were performed on complete datasets
and with outliers removed: outliers were considered important as
they consisted of visitor groups relevant to the zoo seng, such
as school groups; removal of outliers allowed for comparison of
results of data with and without these outlying social groups. A
Shapiro-Wilk test was applied to visitor number and noise data
to test for normality, aer which the correlaon coecient was
calculated for visitor number and noise and a linear regression
model built to test whether visitor number was a signicant
predictor of noise levels. A Kruskal-Wallis test with post-hoc
Wilcoxon was applied to test for dierences in visitor number and
noise levels between species. Wilcoxon tests were then applied
to invesgate dierences in visitor number and noise during talks
and feeds. Gibbons were excluded from this analysis as there were
no scheduled talks or feeds for this species. Kruskal-Wallis tests
were performed on complete gorilla and orangutan datasets and
when split into ‘during talk’ and ‘no talk’, to establish whether talks
had a signicant eect on behaviour. Analysis of back to window
behaviour was performed to test for a relaonship with visitor
number and noise using a generalised linear model (GLM). A GLM
was also applied to the orangutan-specic behaviour of covering
the head with a sack or bedding, to examine relaonships with
visitor number and noise. Visitor aenon behaviour was tested
to invesgate the potenal link between human-directed vigilance
behaviours and increased visitor number and noise. On a species
level, logisc regression was used to examine the eect of visitor
number and noise on select behaviours (inacvity, locomoon
and feeding). For all logisc regression tests, visitor number and
noise were treated as connuous variables. A Kruskal-Wallis test
was performed to invesgate species dierences in inacvity and
locomoon, and a Pearson’s Chi-squared test applied to examine
whether feeding behaviour showed signicant associaon
with scheduled feeds. Locomotory behaviour was invesgated
alongside inacve behaviour as decreased inacvity may not
Table 2. An ethogram of all behaviours observed across the three species. Species-specic behaviours are denoted by 1(gorilla), 2(orangutan) and 3(gibbon).
Adapted from Braendle & Geissman (1997), Cheyne (2006), Kuhar (2008), Collins & Marples (2016).
Behaviour Descripon
Aggression (conspecic) Bing, hing, chasing (non-play) threatening to bite3, charging1,2, chest-beang1
Feed Looking for/handling food, eang, drinking
Grooming Scratching, picking, licking
Inacve Sing, lying down, sleeping
Aliave Non-aggressive conspecic interacons; play, allogrooming, touching
Baby interacon Playing with baby1,3, feeding baby1,3
Locomoon Walking, non-chasing running, climbing, brachiang3
Play Playing with objects, rolling2
Visitor aenon Staring
Stereotypy Abnormal behaviours; hair-plucking1,2, hands over ears1, coprophagy1,2, urophagia1,2, regurgitate & re-ingest2, repeve
swinging3, self-harm3
Other Engaging in any behaviour other than those listed above
Back to window Sing with back to window or viewing area
Out of sight Unable to see

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
271
Hashmi and Sullivan
necessarily lead to increased locomoon but, for example,
aggression or vigilance. Stereotyping was observed in four gorillas
but only one orangutan, therefore logisc regression was applied
to the gorilla dataset and descripve analysis performed for the
orangutan data. Similarly, only two infants were observed in the
study, so clinging behaviour was analysed descripvely. Analysis
of Makari’s clinging behaviour used the locaon ‘inside’ only, as
when outside or on Gorilla Mountain his mother prevented him
from walking; this was also applied during analysis of overall
behaviour paerns. A Pearson’s Chi-squared test was applied to
the complete dataset by species, to examine dierences in the
behaviour paerns across visitor number and noise levels. For
this analysis, visitor number and noise were grouped into three
categories: ‘low’ (visitor number: ≤20, noise: ≤55.40 dB), ‘medium’
(visitor number: 21≤40, noise: 55.41≤79.90 dB) and ‘high’ (visitor
number: ≥41, noise: ≥79.91 dB).
Results
Visitor number and noise
Visitor number and noise levels showed signicant posive
correlaon (Figure 1), and a linear regression model showed
that visitor number was a signicant linear predictor of noise
levels (Table 3). The species received dierent visitor numbers:
gorillas received a mean of 8.6 and a maximum of 50, orangutans
a mean of 9.5 and a maximum of 60, and gibbons a mean of 3.6
and a maximum of 30. There were signicant dierences in visitor
numbers between gibbons and both gorillas and orangutans, but
no signicant dierence between gorillas and orangutans. Noise
levels between all three enclosures were signicantly dierent
(Table 3, Figure 2).
A signicant relaonship was seen between talks and
visitor number (Figure 3); conversely there was no signicant
relaonship between talks and noise (Figure 4, Table 4). There
was no signicant relaonship between scheduled feeds and
visitor number or noise (Table 4). Kruskal-Wallis tests found that
talks had a signicant relaonship with behaviour in gorillas
(X2=24.524, df=11, P=0.01069) but not in orangutans (X2=9.1594,
df=9, P=0.4227).
Visitor avoidance and aenon
Time spent with back to the window (BW) was not signicantly
inuenced by visitor number or noise level (Table 5). When
orangutan data were categorised as ‘before’ and ‘aer’, analysis
of data for BW showed no signicant relaonship with visitor
number or noise levels in the ‘before’ dataset; aer Summer’s
removal, a signicant relaonship was seen between BW and
noise levels (Table 5). Applicaon of the GLM showed a signicant
negave relaonship between visitor number and orangutans
covering their heads in the ‘before dataset’, inclusive of outliers,
but this was not observed in the ‘aer’ dataset or with noise levels
(Table 5).
Figure 1. The relaonship between visitor number and noise levels: A. Dataset containing outliers (r=0.53, t=31.48, df=2517, P<0.001); B. Dataset with
outliers removed (r=0.54, t=32.088, df=2329, P<0.001).

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
272
The visitor eect in zoo-housed apes
Feeding behaviour
In gorillas there was a signicant posive relaonship between
feeding behaviour and visitor number, and a signicant negave
relaonship between feeding and noise levels with outliers
removed (Table 7). In orangutans, there was a signicant posive
relaonship between feeding behaviour in the ‘before’ dataset
with outliers, and the ‘aer’ dataset (Table 7). Pearson’s Chi-
squared test showed no signicant associaon between schedules
feeds and feeding behaviour (P>0.05).
Acvity
Inacvity showed a signicant negave relaonship with
visitor number in gorillas when outliers were included; analysis
of orangutan and gibbon inacvity showed no signicant
relaonships with visitor number or noise (Table 6). Analysis of
locomoon showed no signicant relaonship with either factor
in any species (P<0.005), and no signicant dierences in inacvity
or locomoon were observed between species.
Tes t With outliers Without outliers
Correlaon (visitor number and noise) r=0.5307664, t=31.479, df=2517, P<0.001*** r=0.5336781, t=32.088, df=2329, P<0.001***
Linear regression (visitor number and
noise)
r2=0.2814, f=987.2, df=2517, P<0.001*** r2=0.3063, f=1030, df=2329, P<0.001***
Kruskal-Wallis (visitor numbers) X2=283.46, df=2, gorilla & orangutan P=0.09,
gibbon & others P<0.001***
X2 =216.93, df=2, gorilla & orangutan P=0.68, gibbon &
others P<0.001***
Kruskal-Wallis (noise levels) X2=168.69, df=2, gorilla & orangutan P=0.02*,
gibbon & others P<0.001***
X2 =182.27 df=2, gorilla & orangutan P=0.013*, gibbon
& others P<0.001***
Table 3. Test results for visitor number and noise. Signicance levels are denoted by *(P<0.05) and ***(P<0.001).
Figure 2. Dierences in visitor numbers and noise levels between enclosures across the duraon of the study (n=2520): A. Visitor number dataset containing
outliers (X2=283.46, df=2, gibbon and gorilla P<0.001, gibbon and orangutan P<0.001, gorilla and orangutan P=0.09); B. Visitor number dataset with outliers
removed (X2=215.93, df=2, gibbon and gorilla P<0.001, gibbon and orangutan P<0.001, gorilla and orangutan P=0.68); C. Noise dataset containing outliers
(X2=168.69, df=2, gibbon and gorilla P<0.001, gibbon and orangutan P<0.001, gorilla and orangutan P=0.02); D. Noise dataset with outliers removed
(X2=182.27, df=2, gibbon and gorilla P<0.001, gibbon and orangutan P<0.001, gorilla and orangutan P=0.013).

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
273
Hashmi and Sullivan
Figure 3. Dierences in visitor number by enclosure when scheduled talks and feeds were taking place: A. Visitor number dataset for gorillas containing
outliers; B. Visitor number dataset for gorillas with outliers removed; C. Visitor number dataset for orangutans containing outliers; D. Visitor number
dataset for orangutans with outliers removed.
Figure 4. Dierences in noise level by enclosure when scheduled talks and feeds were taking place: A. Noise dataset for gorillas containing outliers; B. Noise
dataset for gorillas with outliers removed; C. Noise dataset for orangutans containing outliers; D. Noise dataset for orangutans with outliers removed.

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
274
The visitor eect in zoo-housed apes
Table 4. Test results for visitor number and noise in associaon with talks and feeds. Signicance levels are denoted by *(P<0.05), **(P<0.01) and
***(P<0.001).
Tes t With outliers Without outliers
Wilcoxon (visitor number and talks) W=210690, P=0.0002903*** W=202980, P=0.0002856***
Wilcoxon (visitor number and talks: gorillas) W=90448, P=0.009673** W=70831, P=0.02757*
Wilcoxon (visitor number and talks: orangutans) W=22052, P=0.02011* W=12514, P=0.001828**
Wilcoxon (noise and talks) W=181700, P=0.769 W=180300, P=0.7875
Wilcoxon (noise and talks: gorillas) W=80710, P=0.9852 W=80174, P=0.9587
Wilcoxon (noise and talks: orangutans) W=18009, P=0.6967 W=17169, P=0.5045
Wilcoxon (visitor number and feeds) W=12904, P=0.289 W=12320, P=0.4118
Wilcoxon (visitor number and feeds: gorillas) W=7029.5, P=0.4222 W=8945.5, P=0.3515
Wilcoxon (visitor number and feeds: orangutans) W=351, P=0.2102 W=370.5, P=0.9209
Wilcoxon (noise and feeds) W=13969, P=0.5782 W=13850, P=0.5602
Wilcoxon (noise and feeds: gorillas) W=8050.5, P=0.9584 W=8050.5, P=0.9434
Wilcoxon (noise and feeds: orangutans) W=928, P=0.4757 W=223, P=0.5288
Table 5. Test results for visitor avoidance and aenon behaviours (BW, hidden under sacks or bedding/IH and visitor aenon/V). ‘Before’ and ‘aer’
refer to the datasets before Summer’s removal from the orangutan group and aer her removal. Signicance levels are denoted by *(P<0.05), **(P<0.01)
and ***(P<0.001).
Tes t With outliers Without outliers
GLM (visitor number: gorillas BW) z=0.843, P=0.3994 z=1.128, P=0.259
GLM (noise: gorillas BW) z=-0.792, P=0.4282 z=-1.316, P=0.188
GLM (visitor number: orangutans BW) z=0.003, P=0.998 z=0.329, P=0.742
GLM (noise: orangutans BW) z=-1.503, P=0.133 z=-1.569, P=0.117
GLM (visitor number: orangutans ‘before’ BW) z=-0.737, P=0.4612 z=0.02139, P=0.268
GLM (noise: orangutans ‘before’ BW) z=-0.030, P=0.9763 z=-0.431, P=0.667
GLM (visitor number: orangutans ‘aer’ BW) z=0.564, P=0.5730 z=-0.007, P=0.9941
GLM (noise: orangutans ‘aer’ BW) z=-2.632, P=0.0085* z=-2.391, P=0.0168*
GLM (visitor number: gibbons) z=-1.347, P=0.1780 z=-2.451, P=0.0142
GLM (noise: gibbons) z=-0.422, P=0.6732 z=0.629, P=0.5297
GLM (visitor number: orangutans IH) z=-1.586, P=0.113 z=-2.054, P=0.040*
GLM (noise: orangutans IH) z=-1.324, P=0.186 z=-0.523, P=0.601
GLM (visitor number: orangutans ‘before’ IH) z=-2.131, P=0.0331* z=-1.326, P=0.185
GLM (noise: orangutans ‘before’ IH) z=-0.657, P=0.5110 z=-0.347, P=0.729
GLM (visitor number: orangutans ‘aer’ IH z=1.101, P=0.271 z=0.422, P=0.673
GLM (noise: orangutans ‘aer’ IH) z=-1.436, P=0.151 z=-1.117, P=0.264
GLM (visitor number: gorillas V) z=1.253, P=0.210 z=1.438, P=0.151
GLM (noise: gorillas V) z=-1.415, P=0.157 z=-1.520, P=0.128
GLM (visitor number: orangutans V) z=-0.006, P=0.9950 z=0.636, P=0.5247
GLM (noise: orangutans V) z=2.347, P=0.0189* z=2.046, P=0.0408*
GLM (visitor number: orangutans ‘before’ V) z=0.465, P=0.6418 z=1.051, P=0.2930
GLM (noise: orangutans ‘before’ V) z=2.495, P=0.0126* z=1.778, P=0.0755
GLM (visitor number: orangutans ‘aer’ V) z=0.076, P=0.939 z=-0.144, P=0.886
GLM (noise: orangutans ‘aer’ V) z=-0.202, P=0.840 z=-0.138, P=0.890
GLM (visitor number: gibbons V) z=3.063, P=0.00219** z=4.293, P<0.001***
GLM (noise: gibbons V) z=0.678, P=0.49777 z=1.328, P=0.184

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
275
Hashmi and Sullivan
Stereotypic behaviour
Four gorillas and one orangutan showed stereotypic behaviour.
In gorillas, visitor number and noise showed no signicant
relaonship with stereotyping (P>0.05). Stereotyping in the
orangutan did not appear to be consistent with higher visitor
numbers or noise levels.
Clinging behaviour
Incidence of clinging behaviour was examined with regards to
visitor number and noise. There was a roughly even distribuon of
clinging behaviour across visitor number and noise levels in both
infants.
Changes in behaviour during talks
Datasets were split into ‘during talk’ and ‘no talk’ and tested by
species: behaviours tested were BW, hiding under bedding or
sacks, visitor aenon, inacvity, locomoon and feeding (Table
8).
There was no signicant associaon between acve or inacve
behaviour and visitor number or noise in gorillas, and the complete
and ‘aer’ orangutan datasets; the ‘before’ dataset showed a
signicant relaonship between acve and inacve behaviours
and noise (Table 9). Gibbon data showed a signicant relaonship
between acve behaviour and noise (Table 10).
Discussion
The visitor eect on zoo-housed primates ranged from no eect
to detrimental as in previous literature (no eect: Mather 1999;
Collins et al. 2017; increased stress: Mitchell et al. 1992b; Wormell
et al. 1996; Birke 2002; Davis et al. 2005; Collins and Marples
2016). In this study, visitor number and noise had signicant but
contrasng relaonships with several behaviours; the extent
diered on a species level, as seen previously (Quadros et al.
2014). Furthermore, there were marked dierences in the number
of visitors at each enclosure in this study. Locaon may explain this
Table 6. Test results for inacvity. Signicance levels are denoted by **(P<0.01).
Tes t With outliers Without outliers
GLM (visitor number: gorillas) z=-2.749, P=0.00597** z=-1.572, P=0.116
GLM (noise: gorillas) z=0.688, P=0.49167 z=0.299, 0.765
GLM (visitor number: orangutans) z=-1.797, P=0.0724 z=-1.142, P=0.254
GLM (noise: orangutans) z=-0.027, P=0.9783 z=-0.166, P=0.868
GLM (visitor number: orangutans ‘before’) z=-1.564, P=0.118 z=-0.951, P=0.341
GLM (noise: orangutans ‘before’) z=-0.530, P=0.596 z=-0.675, P=0.499
GLM (visitor number: orangutans ‘aer’) z=-0.984, P=0.325 z=-0.804, P=0.421
GLM (noise: orangutans ‘aer’) z=1.050, P=0.294 z=1.361, P=0.174
GLM (visitor number: gibbons) z=-0.628, P=0.530 z=-1.211, P=0.226
GLM (noise: gibbons) z=-1.447, P=0.148 z=-0.111, P=0.912
Tes t With outliers Without outliers
GLM (visitor number: gorillas) z=3.135, P=0.00172** z=3.389, P=0.000701***
GLM (noise: gorillas) z=-2.406, P=0.01615 z=-2.574, P=0.010057*
GLM (visitor number: orangutans) z=2.892, P=0.00383** z=2.999, P=0.00271**
GLM (noise: orangutans) z=-0.061, P=0.95148 z=-0.931, P=0.35211
GLM (visitor number: orangutans ‘before’) z=2.156, P=0.03110* z=1.750, P=0.0802
GLM (noise: orangutans ‘before’) z=0.144, P=0.88539 z=0.015, P=0.9882
GLM (visitor number: orangutans ‘aer’) z=2.150, P=0.0316* z=2.840, P=0.00451**
GLM (noise: orangutans ‘aer’) z=-0.440, P=0.6603 z=-1.202, P=0.22949
Table 7. Test results for feeding behaviour. Signicance levels are denoted by *(P<0.05), **(P<0.01) and ***(P<0.001).

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
276
The visitor eect in zoo-housed apes
dierence: the gorilla and orangutan enclosures neighbour each
other in a busy area of the zoo. The gibbon enclosure is in a lesser-
visited locaon, so the species draws fewer visitors. Talks are held
daily for the orangutans and, during the summer, the gorillas. This
is not the case for gibbons; therefore, less aenon is acvely
drawn to the species.
Animals sing with their back to the window is a visitor
avoidance behaviour and has been denoted a stress indicator
(Collins and Marples 2016). In this study, all individuals sat with
their back to the window. However, there was no signicant
relaonship between physical orientaon and visitor number and
noise in gorillas and gibbons. A signicant relaonship between
BW behaviour and noise levels was seen in orangutans aer
Summer’s removal. Contrary to Birke (2002), the orangutans in
this study decreased their use of sacks or bedding as cover when
visitor number and noise levels increased. The hypothesis that an
increased visitor number or noise level would lead to increased
visitor aenon behaviour (from here referred to as vigilance)
as opposed to avoidance behaviours was tested alongside BW
behaviour and use of sacks and bedding in orangutans. Vigilance
showed no signicant relaonship with increased visitor number;
however, as noise increased, vigilance behaviour increased. This
suggests that, alongside the decrease in BW behaviour, vigilance
is promoted above visitor avoidance behaviours in this orangutan
group.
As visitor numbers increased, inacvity decreased. Again,
species dierences were evident: gorilla inacvity levels
declined signicantly with increasing visitor number but showed
no signicant relaonship with noise. Orangutan and gibbon
inacvity levels were unaected by visitor number and noise. This
suggests visitor number alone aected inacvity. Furthermore,
there was no signicant eect of visitor number or noise level on
locomotory behaviour. Decreased inacvity with increased crowd
size has been repeatedly observed in primates (Hosey and Druck
1987; Chamove et al. 1988; Mitchell et al. 1992a; Wells 2005) and
a change in inacvity may be used as a baseline for invesgang
other behaviours that may be performed instead. An increase
in locomoon does not necessarily follow declined inacvity
levels, but instead increased aggression or vigilance behaviour,
for example. Species dierences have been suggested to migate
potenal eects of visitors; for example, gibbons are more acve
than gorillas (Collins and Marples 2016), and this will impact on
inacvity – and reacons to visitor presence and behaviour – in
both species. However, in this study, there was no signicant
dierence in overall levels of inacvity and locomoon between
species.
Excitaon was not tested directly in this study, rather
through the tesng of visitor number and noise with inacvity
and locomoon; however, whether this impacts on welfare is
dependent on baseline acvity levels. These may be dicult to
ascertain: establishing accurately at what noise level background
noise may begin to cause health, welfare or behavioural problems
is problemac, especially within a zoological instuon which
would only allow for data collecon of background noise before
and aer closing. Furthermore, the strength of causaon on each
side is unknown: if visitor number, noise levels and animal acvity
are bidireconal (Margulis et al. 2003), is increased acvity due
to increased visitor numbers and/or noise, or are visitor numbers
and/or noise increasing due to increased acvity? Previous
research has shown that zoo visitors are more aracted to more
acve animals (Bitgood et al. 1988; Altman 1999; Margulis et al.
2003; Moss and Esson 2010). Wild animal acvity budgets may
Tes t With outliers Without outliers
Noise: orangutans ‘aer’; BW; no talk z=-2.536, P=0.0112* z=-2.165, P=0.0304*
Noise: orangutans; V; no talk z=2.129, P=0.0333* z=1.807, P=0.0708
Noise: orangutans ‘before’; V; no talk z=2.323, P=0.0202* z=1.385, P=0.165950
Visitor number: gorillas; I; no talk z=-3.318, P=0.000905*** z=-2.105, P=0.0354*
Visitor number: gorillas; L; no talk z=2.461, P=0.0.0139* z=0.650, P=0.5255
Noise: gorillas; L; no talk z=-2.300, P=0.0215* z=-2.111, P=0.0348*
Visitor number: gorillas; F; no talk z=3.277, P=0.00105** z=3.559, P=0.000372***
Noise: gorillas; F; no talk z=-2.671, P=0.00757** z=-2.890, P=0.003848**
Visitor number: orangutans; F; during z=2.331, P=0.0197* z=-0.003, P=0.998
Visitor number: orangutans; F; no talk z=0.2.269, P=0.02325* z=2.721, P=0.0065**
Visitor number: orangutans ‘aer’; F; no talk z=1.417, P=0.157 z=2.445, P=0.0145*
Table 8. Back to window (BW), visitor aenon (V), inacvity (I), locomoon (L) and feeding (F) behaviours were tested aer the dataset was split by when
talks were taking place (‘during’) and when talks were not ongoing (‘no talk’). Signicance levels are denoted by *(P<0.05), **(P<0.01) and ***(P<0.001).
This table includes signicant results only; for full results, see Supplementary Materials.
Acve Inacve
Aggression (conspecic) ‘AC’ Inacve ‘I’
Aliave ‘AF’ Inacve hidden ‘IH’**
Aliave with mother ‘AFM’* Visitor aenon ‘V’
Aenon to baby ‘B’*
Feeding ‘F’
Grooming ‘G’
Locomoon ‘L’
Other ‘O’
Play ‘P’
Stereotyping ‘S’
Table 9. Acve and inacve behavioural categories were created prior to
analysis. Species-specic behaviours are denoted by *(gorilla and gibbon)
and **(orangutan).

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
277
Hashmi and Sullivan
be used for comparison where available, but this comes with
its own dicules and may lead to inaccurate comparisons
(Veasey et al. 1996; Howell and Cheyne 2019). It is dicult to
say whether decreased inacvity alongside rising visitor numbers
and/or noise levels indicates decreased welfare, especially if
decreased inacvity is accompanied by a rise in benign acvies
(e.g. aliave behaviours) rather than the increased aggression
observed in some studies (Chamove et al. 1988; Mitchell et al.
1991; Fa 1992; Mitchell et al. 1992a; Wells 2005; Kuhar 2008;
Stoinski et al. 2012; Collins and Marples 2016). As stated by
Birke (2002) it is dicult to judge the eect of increased acvity
levels, even in species where we hold baseline data. However,
if increased locomoon is expressed in stereotypic pacing, this
cannot be considered a desirable outcome. Previous studies have
indicated the importance of o-show areas for primates (e.g.
Kuhar 2008) to avoid decreased inacvity and related increases
in negave behaviours, suggesng that increased visitor numbers
may be a welfare concern in instuons whose enclosures do not
contain freely accessible o-show areas.
There was a signicant posive relaonship between visitor
number and feeding behaviour in gorillas and the complete
and ‘aer’ orangutan datasets. However, a signicant, negave,
relaonship between noise and feeding behaviour was seen only
in gorillas. Previous studies have shown that feeding decreased
with increased visitor number (Chamove et al. 1988; Mitchell et al.
1991; Fa 1992; Wells 2005; Kuhar 2008; Collins and Marples 2016);
however, it has also been suggested that feeding enrichment may
help to reduce the visitor eect (Birke 2002; Carder and Semple
2008; Clark et al. 2011). Feeding enrichment includes scaer
feeding, which was used at Blackpool Zoo. At least one feed daily
was conducted during the gorilla and orangutan talks. There was
a signicant eect of talks on visitor number at both enclosures,
suggesng that visitors are drawn to enclosures by talks (Mitchell
et al. 1992a). Talks did not have a signicant eect on noise levels
and scheduled feeds had no signicant eect on visitor numbers
or noise levels. This is suggested to be because, although one
feed was adversed in conjuncon with the talk, the other daily
feeds were not adversed and visitors may not be aware that
they were taking place. Talks showed a signicant relaonship
with behaviour in gorillas but there was no signicant associaon
between scheduled feeds and feeding behaviour. In this study,
there is no evidence that feeds reduced the incidence of unwanted
behaviours through increased feeding behaviour. However, the
increase in feeding alongside increased visitor number may be
explained by the visitor aenon hypothesis. The reducon of
feeding in gorillas with increased noise, however, highlights a
potenal detrimental eect of acve visitors in this species.
A more indicave measure of behavioural change due to visitors
is infants clinging to their mother. Clinging is a fear response
and may be a more reliable indicator of the visitor eect; other
indicators, such as aggression, may be caused by circumstances
other than visitor presence or noise (e.g. food- or resource-related
aggression). Increases in clinging may indicate that infants nd
visitor presence and/or noise stressful, perhaps due to perceived
threat (Birke 2002; Kuhar 2008; Collins and Marples 2016). This
was not seen in this study: clinging was not aected by visitor
number or noise. Further research into this eect is required; the
current study only examined two infants of dierent species, who
were not monitored from birth. A linking hypothesis suggests that
the birth of an infant may be enriching for other group members,
reducing the visitor eect (Smith and Kuhar 2010; Collins and
Marples 2016). These two topics may be studied concurrently to
provide more data on the visitor eect on infants and adults post-
birth. In this study, interacons were observed between infants
Table 10. The results from the Pearson’s Chi-squared tests for all datasets. Acve and inacve datasets are as those dened in Table 9. Not all behaviours
in each dataset were present for each species: aside from the species-specic behaviours denoted in Table 9, conspecic-directed aggression was not
recorded for orangutans. Signicance levels are denoted by *(P<0.05) and **(P<0.01).
Tes t Results
Gorilla acve behaviour and visitor number X2=19.831, df=18, P=0.3424
Gorilla acve behaviour and noise X2=20.546, df=18, P=0.3029
Gorilla inacve behaviour and visitor number X2=0.32906, df=2, P=0.8483
Gorilla inacve behaviour and noise X2=0.36069, df=2, P=0.835
Orangutans acve behaviour and visitor number X2=12.568, df=12, P=0.4012
Orangutans acve behaviour and noise X2=17.483, df=12, P=0.1323
Orangutans inacve behaviour and visitor number X2=1.7163, df=4, P=0.7877
Orangutans inacve behaviour and noise X2=6.1232, df=4, P=0.1901
Orangutans ‘before’ acve behaviour and visitor number X2=19.002, df=12, P=0.08848
Orangutans ‘before’ acve behaviour and noise X2=23.102, df=12, P=0.02687*
Orangutans ‘before’ inacve behaviour and visitor number X2=1.105, df=2, P=0.05755
Orangutans ‘before’ inacve behaviour and noise X2=10.506, df=4, P=0.03272*
Orangutans ‘aer’ acve behaviour and visitor number X2=9.3801, df=12, P=0.6702
Orangutans ‘aer’ acve behaviour and noise X2=12.113, df=12, P=0.4367
Orangutans ‘aer’ inacve behaviour and visitor number X2=0.62488, df=4, P=0.9603
Orangutans ‘aer’ inacve behaviour and noise X2=1.3477, df=4, P=0.8532
Gibbons acve behaviour and visitor number X2=2.363, df=9, P=0.9843
Gibbons acve behaviour and noise X2=36.16, df=18, P=0.00673**
Gibbons inacve behaviour and visitor number X2=0.32906, df=1, P=0.8943
Gibbons inacve behaviour and noise X2=1.0341, df=2, P=0.5963

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
278
The visitor eect in zoo-housed apes
Figure 5. The graphs show the proporon of me each species displayed acve behaviours: conspecic-directed aggression (AC), aliave (AF), clinging
(AFM), aenon to the baby (B), feeding (F), grooming (G), locomoon (L), other (O), play (P) and stereotyping (S). ‘Low’, ‘medium’ and ‘high’ visitor number
and noise condions are as dened in Methods: A. Proporon of me gorillas spent performing acve behaviours across dierent visitor condions; B.
Proporon of me gorillas spent performing acve behaviours across dierent noise condions; C. Proporon of me orangutans spent performing acve
behaviours across dierent visitor condions; D. Proporon of me orangutans spent performing acve behaviours across dierent noise condions; E.
Proporon of me gibbons spent performing acve behaviours across dierent visitor condions; F. Proporon of me gibbons spent performing acve
behaviours across dierent noise condions.

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
279
Hashmi and Sullivan
Figure 6. The graphs show the proporon of me each species spent displaying inacve behaviours: inacve (I), hidden under sacks or bedding (IH)
and visitor aenon (V). ‘Low’, ‘medium’ and ‘high’ visitor number and noise condions are as dened in Methods: A. Proporon of me gorillas spent
performing inacve behaviours across dierent visitor number condions; B. Proporon of me gorillas spent performing inacve behaviours across
dierent noise condions; C. Proporon of me orangutans spent performing inacve behaviours across dierent visitor number condions; D. Proporon
of me orangutans spent performing inacve behaviours across dierent noise condions; E. Proporon of me gibbons spent performing inacve
behaviours across dierent visitor number condions; F. Proporon of me gibbons spent performing inacve behaviours across dierent noise condions.

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
280
The visitor eect in zoo-housed apes
and all group members. Whether this replicated the eect seen
in Collins and Marples (2016) cannot be said due to the lack of
data prior to the infants’ birth. However, this may be useful
to invesgate further as the eect of birth on social groups is
currently understudied (Collins and Marples 2016).
Stereotypic behaviour has been idened as a stress response
in apes (Blaney and Wells 2004; Wells 2005; Carder and Semple
2008; Stonski et al. 2012; Collins and Marples 2016). There was
diering prevalence of stereotyping across the species, with four
gorillas and one orangutan showing stereotypies, whereas gibbons
showed no stereotypy. Furthermore, stereotypies varied between
species: gorillas most commonly performed coprophagy where
the orangutan showed mostly regurgitaon and reingeson.
This illustrates dierent reacons to visitors and noise across
individuals and species, which could come from a range of factors,
for example, life history, personality and housing (Hosey 2000;
Hosey 2005; Choo et al. 2011; Collins and Marples 2016; Sherwen
and Hemsworth 2019), as well as diering levels of visitors
between species. The results of this study align with those of Smith
and Kuhar (2010), in which zoo-housed white-cheeked gibbons
and siamangs showed no abnormal or unwanted behaviours.
This may be because these animals had freely accessible o-show
areas to ‘escape’ visitors (Smith and Kuhar 2010); that creang
freely-accessible areas of privacy within enclosures reduces stress
and the incidence of abnormal or unwanted behaviours has been
seen in other primates and taxa, even if these areas are not used
(Blaney and Wells 2004; Fernández et al. 2009; Bloomeld et al.
2015). In this study, freely accessible o-show areas were not
provided; a potenal eect of this is the increase in stereotyping
in the orangutan individual and decreased inacvity in gorillas and
orangutans with increased visitor number. However, that not all
animals in the current study displayed stereotypies illustrates the
complexity of factors controlling responses to visitor number and
noise. Furthermore, the gorillas that did not show stereotypies
were mother and ospring, suggesng the eect of a new infant
may have been present.
However, the eects of factors such as life history, previous
husbandry or environment, and personality should not be
understated. Reacons to visitor number and noise varied greatly
between species and individuals, with gibbons appearing most
able to cope with capvity. Whether this is due to life history,
being more habituated to human presence, or simply receiving
fewer visitors cannot be discerned; however, this aligns with
the ndings of Smith and Kuhar (2010), who found that other
Hylobates species showed few behavioural dierences in response
to visitors. Conversely, great apes have been repeatedly judged as
negavely aected by visitor number and noise (e.g. Birke 2002;
Blaney and Wells 2004; Carder and Semple 2008; Collins and
Marples 2016). One explanaon is the evoluonary proximity
of humans to non-human primates, creang the propensity for
acons displayed by human visitors (e.g. staring, yawning) to be
interpreted as threatening by great apes (Birke 2002) and other
primates such as siamangs and capuchins (Nimon and Dalziel
1992; Sherwen et al. 2015). However, the potenal eects of
species dierences on reacons to visitors and noise in zoo
sengs have been understudied, as have those of personality
(Sherwen and Hemsworth 2019). Both of these areas deserve
aenon, as a deeper understanding of species’ and individuals’
reacons to visitors may allow more targeted approaches to
migate the eects of visitors and noise, for example, designing
enclosures where visitors are ‘below’ animals for arboreal species
(e.g. Chamove et al. 1988; Choo et al. 2011).
Regarding drivers of stress in capve primates, with the capve
environment comes a lack of control, and adding o-show areas or
equivalent, for example, privacy screens, to be used at will returns
some control to animals, potenally reducing the incidence of
unwanted behaviours. Visitor presence and noise are factors that
animals cannot control, adding to or perhaps causing the stress
associated with visitors. Lack of control is linked to anxiety and
stress (Morgan and Tromborg 2007), with some suggesng that
feeling in control is essenal to animal well-being (O’Neill 1989;
Friend 1991) and that lack of control may impact physiological
measures of welfare, for example, faecal corsol (Mineka and
Kelly 1989). Providing animals with the ability to control their
environment has been suggested as a method to ameliorate
the eects of stress caused by visitors and/or noise (Hanson et
al. 1976; Wemelsfelder 1993; Wiepkema and Koolhaas 1993;
Sambrook and Buchanan-Smith 1997; Hosey 2005; Smith and
Kuhar 2010; Collins and Marples 2016).
To add control to the environment, it is suggested animals be
given free access to o-show areas; however, not all instuons
currently have enclosures with open access to appropriate o-
show facilies, and the cost of renovang enclosures to provide
o-show areas is prohibive for many collecons. Alternave
modicaons to enclosure design may create the percepon of
reduced body size of visitors, such as raising viewing windows so
that only a visitor’s head is visible (Chamove et al. 1988); although,
unless enclosures are due for or undergoing renovaon, the cost
of these modicaons may again prove too expensive for many
collecons. Alternave low-cost soluons may prove eecve in
reducing stress: previous studies have trialled soluons such as
the use of cargo nets over windows to reduce direct visual contact
between animals and visitors (Blaney and Wells 2004), and privacy
screens (Kuhar 2008; Smith and Kuhar 2010; Bloomeld et al.
2015) or foliage (Kuhar 2008) as visual barriers. In this study, the
only area with foliage as a barrier was Gorilla Mountain; however,
foliage did not obstruct visual contact around the enre perimeter
of the enclosure and the gorilla group rarely had open access
to this area. Furthermore, none of these soluons, bar creang
o-show areas, have the ability to reduce noise levels around
enclosures. This is important as, in this study, some behaviours
were signicantly inuenced by noise only. For this reason, zoos
must monitor the behaviour of their visitors as far as praccable.
This may be achieved through the staoning of sta or volunteers
in the vicinity of enclosures, as their presence alone may help to
reduce incidents of disrupve behaviour. This is seen at many
walkthrough exhibits, although it is prohibive in terms of cost
and sta me for many zoos.
Eye-level signage, aimed at modifying visitor behaviour in a
posive, rather than negave, manner may prove eecve in
reducing noise levels and random noise events, for example,
banging on the glass (Kratochvil and Schwammer 1997), which may
in turn reduce stress in capve primates. Furthermore, the use of
neng over viewing windows posively inuenced the behaviour
of animals and visitors, who spoke less and more quietly when the
net was in place, with fewer recorded incidents of visitors banging
on the glass (Blaney and Wells 2004). The results of the current
study suggest the introducon of a freely accessible o-show
area may benet the apes, whether this is achieved through the
creaon of a dedicated o-show area of the employment of low-
cost visual barriers.
Acknowledgements
The authors would like to thank the primate keeping sta and
Research Department at Blackpool Zoo, as well as Dr. Hannah
Mossman, Dr. Danny Norrey and Prof. Richard Preziosi at
Manchester Metropolitan University. Further thanks go to the
anonymous reviewer whose comments helped to improve this
manuscript.
The authors declare no conicts of interest.

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
281
Hashmi and Sullivan
References
Altman J.D. (1998) Animal acvity and visitor learning at the zoo.
Anthrozoös 11(1): 12-20. doi: 10.1080/08927936.1998.11425083.
Birke L. (2002) Eects of browse, human visitors and noise on the behaviour
of capve orang utans. Animal Welfare 11(2): 189-201.
Bitgood S., Paerson D., Beneeld A. (1988) Exhibit design and visitor
behaviour: empirical relaonships. Environment and Behaviour 20:
474-490. doi: 10.1177/0013916588204006.
Blaney E.C., Wells D.L. (2004) The inuence of a camouage net barrier on
the behaviour, welfare and public percepon of zoo-housed gorillas.
Animal Welfare 13(2): 111-120.
Bloomeld R.C., Gillespie G.R., Kerswell K.J., Butler K.L., Hemsworth P.H.
(2015) Eect of paral covering of the visitor viewing area window on
posioning and orientaon of zoo orangutans: a preference test. Zoo
Biology 34(3): 223-229. doi: 10.1002/zoo.21207.
Bloomsmith M.A., Marr M.J., Maple T.L. (2007) Addressing nonhuman
primate behavioral problems through the applicaon of operant
condioning: is the human treatment approach a useful model?
Applied Animal Behaviour Science 102(3-4): 205-222. doi: 10.1016/j.
applanim.05.028.
Bortolini T.S., Bicca-Marques J.C. (2011) The eect of environmental
enrichment and visitors on the behaviour and welfare of two capve
hamadryas baboons (Papio hamadryas). Animal Welfare 20: 573-579.
Botreau R., Veissier I., Buerworth A., Bracke M.B.M.; Keeling L.J. (2007)
Denion of criteria for overall assessment of animal welfare. Animal
Welfare 16: 225-228.
Braendle C., Geissman T. (1997) Behavioural development of a pileated
gibbon (Hylobates pileatus). Internaonal Zoo News 44(1): 4-16.
Carder G., Semple S. (2008) Visitor eects on anxiety in two capve groups
of Western lowland gorillas. Applied Animal Behaviour Science 115(3-
4): 211-220. doi: 10.1016/j.applanim.2008.06.001.
Chamove A.S., Hosey G.R., Schaetzel P. (1988) Visitors excite primates in
zoos. Zoo Biology 7(4): 359-369. doi: 10.1002/zoo.1430070407.
Cheyne S. (2006) Unusual behaviour of capve-raised gibbons: implicaons
for welfare. Primates 47(4): 322-326. doi: 10.1007/s10329-006-
0190-z.
Choo Y., Todd P.A., Li D. (2011) Visitor eects on zoo orangutans in two
novel, naturalisc enclosures. Applied Animal Behaviour Science
133(1): 78-85. doi: 10.1016/j.applanim.2011.05.007.
Clark F.E., Fitzpatrick M., Hartley A., King A.J., Lee T., Routh A., Walker S.L.,
George K. (2011) Relaonship between behavior, adrenal acvity, and
environment in zoo-housed Western lowland gorillas (Gorilla gorilla
gorilla). Zoo Biology 31(3): 306-320. doi: 10.1002/zoo.20396.
Collins C.K., Marples N.M. (2016) The eects of zoo visitors on a group
of Western lowland gorillas Gorilla gorilla gorilla before and aer the
birth of an infant at Dublin Zoo. Internaonal Zoo Yearbook 50(1): 183-
193. doi: 10.1111/izy.12111.
Collins C., Corkery I., Haigh A., McKeown S., Quirke T., O’Riordan R. (2017)
The eects of environmental and visitor variables on the behavior of
free-ranging ring-tailed lemurs (Lemur caa) in capvity. Zoo Biology
36(4): 250-260. doi: 10.1002/zoo.21370.
Cooke C.M., Schillaci M.A. (2007) Behavioral responses to the zoo
environment by white handed gibbons. Applied Animal Behaviour
Science 106: 125-133. doi: 10.1016/j.applanim.2006.06.016.
Davis N., Schaner C.M., Smith T.E. (2005) Evidence that zoo visitors
inuence HPA acvity in spider monkeys (Ateles georoyii ruventris).
Applied Animal Behaviour Science 90(2): 131-141. doi: 10.1016/j.
applanim.2004.08.020.
Fa J. (1989) ‘Inuence of people on the behaviour of display primates’. In
Segal, E.F. (ed.) Housing, care and psychological well-being of capve
and laboratory primates. Park Ridge, NJ: Noyes Publicaons, pp. 270-
290.
Fa J.E. (1992) Visitor-directed aggression among the Gibraltar macaques.
Zoo Biology 11(1): 43-52. doi: 10.1002/zoo.1430110106.
Fernández E.J., Tamborski M.A., Pickens S.R., Timberlake W. (2009) Animal-
visitor interacons in the modern zoo: Conicts and intervenons.
Applied Animal Behaviour Science 120(1-2): 1-8. doi: 10.1016/j.
applanim.2009.06.002.
Friend T.H. (1991) Behavioral aspects of stress. Journal of Dairy Science
74(1): 292-303. doi: 10.3168/jds.S0022-0302(91)78173-3.
Hanson J.P., Larson M.E., Snowdon C.T. (1976) The eects of control over
high intensity noise on plasma corsol in rhesus monkeys. Behavioral
Biology 16(3): 333-340. doi: 10.1016/S0091-6773(76)91460-7.
Hemsworth P.H., Sherwen S.L., Coleman G.J. (2018) ‘Human contact.’ In
Appleby, M.C., Olsson, I.A.S. & Gilando, F. (eds.) Animal Welfare. 3rd
ed., Wallingford, U.K.: CAB Internaonal, pp. 294-314.
Hosey G.R. (2000) Zoo animals and their human audiences: what is the
visitor eect? Animal Welfare 9: 343-357.
Hosey G.R. (2005) How does the zoo environment aect the behaviour of
capve primates? Applied Animal Behaviour Science 90(2): 107-129.
doi: 10.1016/j.applanim.2004.08.015.
Hosey G. (2013) Hediger revisited: how do zoo animals see us?
Journal of Applied Animal Welfare Science 16: 338-359. doi:
10.1080/10888705.2013.827916.
Hosey G.R., Druck P.L. (1987) The inuence of zoo visitors on the behaviour
of capve primates. Applied Animal Behaviour Science 18(1): 19-29.
doi: 10.1016/0168-1591(87)90251-6.
Hosey G.R., Skyner L.J. (2007) Self-injurious behavior in zoo primates.
Internaonal Journal of Primatology 28: 1431-1437. doi: 10.1007/
s10764-007-9203-z.
Howell C.P., Cheyne S.M. (2019) Complexies of using wild versus
capve acvity budget comparisons for assessing capve primate
welfare. Journal of Applied Animal Welfare Science 22(1): 78-96. doi:
10.1080/10888705.2018.1500286.
Kratochvil H., Schwammer H. (1997) Reducing acousc disturbances
by aquarium visitors. Zoo Biology 16(4): 349-353. doi: 10.1002/
(SICI)1098-2361(1997)16:4<349::AID-ZOO7>3.0CO;2-A.
Kuhar C.W. (2008) Group dierences in capve gorillas’ reacon to large
crowds. Applied Animal Behaviour Science 110(3-4): 377-385. doi:
10.1016/j.applanim.2007.04.011.
Margulis S.W., Hoyos C., Anderson M. (2003) Eect of felid acvity on zoo
visitor interest. Zoo Biology 22(6): 587-599. doi: 10.1002/zoo.10115.
Markovitz H., Stevens V., Mellon J., Barrow B. (1981) Performance of a
mandrill (Mandrillus sphinx) in compeon with zoo visitors and
computer on a reacon-me game. Acta Zoologica et Pathologica
Antverpiensia 76: 169-180.
Mather L. (1999) Response of capve orang utans to a human audience.
MPhil, University of Manchester (Bolton Instute).
Mellor D.J., Paerson-Kane E., Staord K.J. (2009) The Sciences of Animal
Welfare. Oxford, U.K.: Wiley-Blackwell (UFAW Animal Welfare Series).
Mineka S., Kelly K.A. (1989) ‘ The relaonship between anxiety, lack of
control and loss of control’. In Steptoe, A. & Appels, E. (eds.) Stress,
personal control, and health, Oxford, U.K.: Oxford University Press, pp.
163-191.
Mitchell G., Herring F., Obradovich S., Tromborg C., Dowd B., Neville
L.E., Field L. (1991) Eects of visitors and cage changes on the
behaviors of mangabeys. Zoo Biology 10(5): 417-423. doi: 10.1002/
zoo.1430100505.
Mitchell G., Tromborg C.T., Kaufmann J., Bargabus S., Simoni R., Geissler
V. (1992a) More on the ‘inuence’ of zoo visitors on the behaviour of
capve primates. Applied Animal Behaviour Science 35(2): 189-198.
doi: 10.1016/0168-1591(92)90009-Z.
Mitchell G., Herring F., Tromborg C.T., Dowd B., Steiner S., Obradovich
S. (1992b) Targets of aggressive facial displays by golden-bellied
mangabeys (Cercocebus galeritus chrysogaster) at the Sacramento
Zoo. Applied Animal Behaviour Science 33(2-3): 249-259. doi: 10.1016/
s0168-1591(05)80012-7.
Moodie E., Chamove A. (1990) Brief threatening events benecial for
capve tamarins? Zoo Biology 9: 275-286.
Morgan K.N., Tromborg C.T. (2007) Sources of stress in capvity. Applied
Animal Behaviour Science 102(3-4): 262-302. doi: 10.1016/j.
applanim.2006.05.032.
Moss A., Esson M. (2010) Visitor interest in zoo animals and the
implicaons for collecon planning and zoo educaon programmes.
Zoo Biology 29(6): 715-730. doi: 10.1002/zoo.20316.
Nimon A., Dalziel F. (1992) Cross-species interacon and communicaon:
a study method applied to capve siamang (Hylobates syndactylus)
and long-billed corella (Cacatua tenuirostris) contacts with humans.
Applied Animal Behaviour Science 33: 261-272. doi: 10.1016/S0168-
1591(05)80013-9.
O’Neill P. (1989) ‘A room with a view for capve primates: issues, goals,
related research and strategies’. In Segal, E.F. (ed.) Housing, care,
and psychological wellbeing of capve and laboratory primates. Park
Ridge, N.J., USA: Noyes Publicaons. pp. 135-160.
Owen M.A., Swaisgood R.R., Czekala N.M., Steinman K., Lindburg
D.G. (2004) Monitoring stress in capve giant pandas (Ailuropoda
melanoleuca): behavioral and hormonal responses to ambient noise.
Zoo Biology 23: 147-164. doi: 10.1002/zoo.10124.
Ross S.R. (2006) Issues of choice and control in the behaviour of a pair
of capve polar bears (Ursus marimus). Behavioural Processes 73(1);
117-120. doi: 10.1016/j.beproc.2006.04.003.
Sambrook T.D., Buchanan-Smith H.M. (1997) Stereotypic animal behaviour.
Animal Welfare 6: 207-216.

Journal of Zoo and Aquarium Research 8(4) 2020
hps://doi.org/10.19227/jzar.v8i4.523
282
The visitor eect in zoo-housed apes
Sherwen S.L., Harvey T.J., Magrath M.J., Butler K.L., Fanson K.V., Hemsworth
P.H. (2015) Eects of visual contact with zoo visitors on black-capped
capuchin welfare. Applied Animal Behaviour Science 167: 65-73. doi:
10.1016/j.applanim.2015.03.004.
Sherwen S.L., Hemsworth P.H. (2019) The visitor eect on zoo animals:
implicaons and opportunies for zoo animal welfare. Animals 9(6):
366. doi: 10.3390/ani9060366.
Smith K.N., Kuhar C.W. (2010) Siamangs (Hylobates syndactylus) and
white-cheeked gibbons (Hylobates leucogenys) show few behavioral
dierences related to zoo aendance. Journal of Applied Animal
Welfare Science 13(2): 153-163. doi: 10.1080/10888700903579895.
Stoinski T.S., Jaicks H.F., Drayton L.A. (2012) Visitor eects on the behavior
of capve western lowland gorillas: the importance of individual
dierences in examining welfare. Zoo Biology 31(5): 586-598. doi:
10.1002/zoo.20425.
Veasey J.S., Waran N.K., Young R.J. (1996) On comparing the behaviour
of zoo housed animals with wild conspecics as a welfare indicator,
using the girae (Giraa camelopardalis) as a model. Animal Welfare
5(2): 139-153.
Wells D.L. (2005) A note on the inuence of visitors on the behaviour and
welfare of zoo-housed gorillas. Applied Animal Behaviour Science
93(1-2): 13-18. doi: 10.1016/j.applanim.2005.06.019.
Wemelsfelder F. (1993) ‘ The concept of animal boredom and its
relaonship to stereotyped behaviour’. In Rushenm A.B.L.a.J (ed.)
Stereotypic Animal Behaviour. Wallingford, U.K.: Cab Intl, p. 31.
Wielebnowski N. (1998) ‘Contribuons of behavioral studies to capve
management and breeding of rare and endangered mammals.’ In
Caro, T. (ed.) Behavioral Ecology and Conservaon Biology. Cary, USA:
Oxford University Press, pp. 130-150.
Wiepkema P.R., Koolhaas J.M. (1993) Stress and animal welfare. Animal
Welfare 2(3): 195-218.
Wormell D., Brayshaw M., Price E., Herron S. (1996) Pied tamarins Saguinus
bicolor bicolor at the Jersey Wildlife Preservaon Trust: management,
behaviour and reproducon. Dodo Journal of the Wildlife Preservaon
Trusts 32: 76-96.