Environmental complexity and feeding enrichment can mitigate effects of space constraints in captive callitrichids

Abstract

Non-human primates housed in zoos and laboratories often exhibit reduced activity and this poses welfare concerns. We examined the effects of enclosure types of differing size and environmental complexity on the activities of two species of callitrichids. We found that cotton-top tamarins housed in an enclosure of larger size and more environmental complexity showed higher activity levels, which was mainly contributed by more feeding/foraging activity. By contrast, Goeldi’s monkeys housed in an enclosure of larger size and more environmental complexity showed lower activity levels, which was mainly contributed by less locomotory activity. In both species, off-exhibit groups housed in smaller enclosures did not show significantly less locomotory activity which would have been expected, as larger availability spaces should allow more opportunities for locomotion. Furthermore, the feeding enrichment had significant effects on increased feeding/foraging activity for both cotton-top tamarins and Goeldi’s monkeys, irrespective of enclosure type. These results suggested that environmental complexity (or application of feeding enrichment) that provided more opportunities for natural foraging could have a larger effect on overall activity levels compared with larger enclosure sizes that should provide more locomotion opportunities. More importantly, it showed that even when enclosure space and complexity were limited, increased opportunities for foraging through the application of enrichment could increase species-typical behaviours. Such inexpensive, easy to implement enrichment methods should be applied to provide more complex environments for captive non-human primates, particularly in situations where there are logistical and/or cost constraints to the modification of physical exhibits.

Keywords

cotton-top tamarin Goeldi’s monkey enclosure size environmental complexity feeding enrichment

Non-human primates in captivity are housed with much more limited space and environmental variability compared with their wild counterparts. Captive management also restricts their natural foraging opportunities. As a result, activity levels and naturalistic behaviours of captive non-human primates are often reduced.¹ The effects of captive environments have been shown, for example, in reduced locomotion of long-tailed and pig-tailed macaques housed in smaller cages² and in lower activity levels in common marmosets housed in smaller and less complex enclosures.³

To mitigate these problems, captive animal managers implement various enrichment techniques to provide more challenging environments for captive animals and to promote natural behaviours.⁴ Enrichment is particularly important for off-exhibit animals that are often housed in less optimal conditions compared with on-exhibit animals. Physical enrichment, for example, involves increasing the usable space and complexity of the animal’s enclosure or adding accessories to the enclosure such as objects, substrate, or permanent structures.⁴ Physical enrichment has been shown to have positive effects such as increasing general activity, locomotory behaviour, and exploratory behaviour.^5–7 Feeding enrichment has also been widely and effectively used as a means of promoting naturalistic explorative behaviours in animals.^8–10

In this study, we examined the effects of enclosure type and feeding enrichment on the activity levels of two groups each of captive cotton-top tamarins (Saguinus oedipus) and Goeldi’s monkeys (Callimico goeldii) housed in on-exhibit and off-exhibit conditions that differed in physical space and environmental complexity. We applied a simple enrichment device with reduced visibility and accessibility of presented food that required increased foraging efforts to access. First, we hypothesized that activity levels would be higher for animals housed in larger and more complex enclosures in on-exhibit conditions, and the differences would be more marked where within-species enclosure conditions had wider variations. Secondly, we hypothesized that simple feeding enrichment could effectively compensate for lower activity levels in off-exhibit conditions by increasing active behaviours, and the effect would be more marked for off-exhibit animals in smaller and less complex enclosures. Lastly, we clarified how enclosure size, environmental complexity and feeding enrichment affected species-typical activities.

Materials and methods

Animals

This study was approved by the management of Wildlife Reserves Singapore, according to institutional ethics and welfare guidelines. Two groups each of cotton-top tamarins and Goeldi’s monkeys at the Singapore Zoo were selected for this study (Table 1). The different housing enclosures for the two groups are shown in Figure 1. Off-exhibit cotton-top tamarins were housed in holding dens (2.5 × 1.5 × 2.5 m) furnished with branches and cemented flooring. A nest box was placed at the top-right of the den (Figure 1a). On-exhibit tamarins were provided with outdoor free-ranging environment (9 × 5 m) with natural tree plantings of varying heights (maximum 6 m) and a nest box was placed at the left corner of the exhibit (Figure 1b). Housing for off-exhibit Goeldi’s monkeys consisted of an indoor holding den (3 × 1.2 × 2 m) sheltered from natural weather conditions. It was furnished with a nest box at the back of the den, branches and cemented ground (Figure 1c). On-exhibit Goeldi’s monkeys were held in a semi-indoor enclosure (5.2 × 1.5 × 1.5 m) partially sheltered from weather conditions. The exhibit adopted a naturalistic design with dirt substrate, leaves plantings, small trunks, branches, rock structures and a nest box (Figure 1d).

Table 1.

Summary of study groups.

Cotton-top tamarins		Goeldi’s monkeys
Exhibit*	Off-exhibit	Exhibit*	Off-exhibit
1 female 3 males (All captive bred)	1 female 1 male (All captive bred)	1 female (Wild caught) 2 males (Captive bred)	1 female (Wild caught) 1 male (Captive bred)

Family group consisting of father, mother and juvenile(s).

Figure 1.

Housing for (a) cotton-top tamarins off exhibit; (b) cotton-top tamarins on exhibit; (c) Goeldi’s monkey off exhibit; (d) Goeldi’s monkey on exhibit.

Feeding and diet

Off-exhibit cotton-top tamarins and both off-exhibit and on-exhibit Goeldi’s monkeys were fed thrice daily, at 08:00 h, 11:00 h and 15:00 h. On-exhibit cotton-top tamarins were fed at 08:30 h, 11:45 h and 17:00 h, with additional token feeding at 10:30 h and 14:30 h in between their main meals. Token feedings consisted of small amounts of food to showcase feeding behaviour to visitors. Water was available at all times from a water dispenser near their nest box. The subjects were fed with staple diets which consisted of long beans, carrots and apples. In addition, special fruits such as grapes, bananas and oranges were included but only one type was given each day. Boiled meat or hard-boiled eggs (egg white only) were provided four times a week, as well as supplements of mealworms, ZuPreem® Marmoset Diet Canned (Allpets Asia Pte Ltd, Singapore) and Mazuri® High Fiber Callitrichid Diet (Migoto Aquaworld Pte Ltd, Singapore).

Enrichment

Hessian cloths were cut and stitched together to form pockets with flaps folded down (6 cm × 10 cm) (Figure 2). The pockets were secured onto branches using cable ties. The number of devices was determined by group size, such that one was available for each individual animal. This was to reduce the effect of intra-group competition that could influence individual subject use of the enrichment device. A half cup of cut leaves and one-quarter cup of mealworm mixture were used to fill the pockets, divided into equal portions according to the number of pockets placed. The positions of the enrichment devices were random, but an unhindered view of the devices was always ensured for the observer. Positional placement of the device was changed each time an enrichment treatment was replicated. This enrichment device was tested on the subjects prior to the study. This was to ensure that the device was not novel (as individual and species responses to novel objects could differ) and to confirm that the device could be suitably manipulated by the subjects. A preliminary study of one observation set for each of the four different exhibits was conducted.

Figure 2.

(a) hessian cloths used for enrichment were secured using cable ties; (b) enrichment device being used by a Goeldi’s monkey.

Data collection

Observations were conducted over two months from February to March 2012. Five sets of observations were used for each treatment. A Latin square design¹¹ was used to randomize the days for baseline and enrichment treatments. Both baseline and enrichment treatment data collections were performed during non-feeding hours, from 11:00 h to 12:00 h and 14:00 h to 15:00 h. These constituted one observation set. We used a balanced design by replicating five sets of observations for each group for each treatment. All enrichment devices were removed from the exhibit space after the treatment had been completed. On days when enrichment treatments were performed, enrichment foods were proportionately reduced. Scan sampling¹² at one-minute intervals was used to record behaviours according to the ethogram shown in Table 2 which was adapted from a previous study on cotton-top tamarins.¹³

Table 2.

Ethogram of behaviours recorded for the study subjects.

Behavioural category	Behavioural description
Feeding/foraging	Eating or chewing or using hands or mouth to manipulate a food item
Locomotion	Any movement (walking, running, climbing, jumping, etc.) of more than one body length
Resting	Any inactivity in a stationary and relaxed posture including sleeping with eyes closed

Statistical analysis

We used two-way analysis of variance (ANOVA) to test for effects of enclosure type and enrichment on the following activities observed in the study groups: (1) feeding/foraging, (2) locomotion and (3) resting. Due to differences in the size and composition of the study groups, the data was standardized by dividing the total number of scans for each activity category for all individuals over the total number of scans for all individuals in each observation set. These data thus represented average group activity. Percentages for each activity category for each group were calculated for each set and used for analysis. Enclosure type (off-exhibit/on-exhibit) and enrichment (baseline/enrichment) were assigned as independent factors and the activity categories were assigned as dependent factors separately in each analysis. All data were checked for normality using the Kolmogorov–Smirnov test, and were normally distributed. All analyses were conducted using the Statistical Package for the Social Sciences version 20.0 (SPSS Inc, Chicago, IL, USA). Statistical significance was set at P < 0.05.

Results

Effects of exhibit type and enrichment on the activity levels of cotton-top tamarins

The activities of the cotton-top tamarins in on-exhibit and off-exhibit conditions are shown in Figure 3. Overall activity level (feeding/foraging and locomotion) was higher and resting was lower for on-exhibit animals compared with off-exhibit animals (on-exhibit: 47.7%; off-exhibit: 25.6%; difference: 22.1%). The application of enrichment increased overall activity levels (feeding/foraging and locomotion) and decreased resting, with the percentage change higher for off-exhibit animals compared with on-exhibit animals (on-exhibit: 7.2%; off-exhibit: 19.0%).

Figure 3.

Activity levels (percentage ± SEM) of on-exhibit and off-exhibit cotton-top tamarins during baseline and enrichment treatments. Asterisks denote significant differences between treatments.

Feeding/foraging activity was higher for the cotton-top tamarins on exhibit than those off exhibit (on-exhibit: 14.3%; off-exhibit: 0%). Feeding/foraging activity was higher during enrichment treatment compared with the baseline for on-exhibit animals (enrichment: 38.4%; baseline: 14.3%) and off-exhibit animals (enrichment: 22.9%; baseline: 0%). There was a significant main effect of exhibit type on feeding/foraging activity (F_1,18 = 30.396, P < 0.001); and a significant main effect of enrichment on feeding/foraging activity (F_1,18 = 74.389, P < 0.001). There was a non-significant interaction between exhibit type and enrichment on feeding/foraging activity (F_1,18 = 0.058, P = 0.813).

Locomotory activity was higher for the cotton-top tamarins on exhibit than those off exhibit (on-exhibit: 33.4%; off-exhibit: 25.6%). Locomotory activity was lower during enrichment treatment compared with baseline for on-exhibit animals (enrichment: 16.5%; baseline: 33.4%) and off-exhibit (enrichment: 21.7%; baseline: 25.6%). There was a non-significant main effect of exhibit type on locomotory activity of cotton-top tamarins (F_1,18 = 1.098, P = 0.313); and a non-significant main effect of enrichment on locomotory activity (F_1,18 = 2.270, P = 0.154). There was a non-significant interaction between exhibit type and enrichment on locomotory activity (F_1,18 = 0.182, P = 0.677).

Resting activity was lower for the cotton-top tamarins on exhibit than those off exhibit (on-exhibit: 62.3%; off-exhibit: 74.4%). Resting activity was lower during enrichment treatment compared with baseline for on-exhibit animals (enrichment: 45.1%; baseline: 62.3%) and off-exhibit animals (enrichment: 55.5%; baseline: 74.4%). There was a significant main effect of exhibit type on resting activity (F_1,18 = 7.012, P = 0.019); and a significant main effect of enrichment on resting activity (F_1,18 = 18.098, P = 0.001). There was a non-significant interaction between exhibit type and enrichment on resting activity (F_1,18 = 0.043, P = 0.839).

Effects of exhibit type and enrichment on the activity levels of Goeldi’s monkeys

The activities of the Goeldi’s monkeys in on-exhibit and off-exhibit conditions are shown in Figure 4. Overall activity level (feeding/foraging and locomotion) was lower and resting higher for on-exhibit animals compared with off-exhibit animals (on-exhibit: 13.6%; off-exhibit: 23.3%; difference: 9.7%). The application of enrichment increased overall activity levels (feeding/foraging and locomotion) and decreased resting, with the percentage change higher for off-exhibit animals compared with on-exhibit animals (on-exhibit: 15.4%; off-exhibit: 23.3%).

Figure 4.

Activity levels (percentage ± SEM) of on-exhibit and off-exhibit Goeldi’s monkeys during baseline and enrichment treatments. Asterisks denote significant differences between treatments.

Feeding/foraging activity was higher for the Goeldi’s monkeys on exhibit than those off exhibit (on-exhibit: 0.98%; off-exhibit: 1.39%). Foraging activity was higher during enrichment treatment compared with baseline for on-exhibit animals (enrichment: 20.0%; baseline: 0.98%) and off-exhibit animals (enrichment: 20.6%; baseline: 1.39%). There was a non-significant main effect of exhibit type on feeding/foraging activity (F_1,18 = 1.007, P = 0.686); and a significant main effect of enrichment on feeding/foraging activity (F_1,18 = 274.830, P < 0.001). There was a non-significant interaction between exhibit type and enrichment on feeding/foraging activity (F_1,18 = 0.003, P = 0.957).

Locomotory activity was lower for the Goeldi’s monkeys on exhibit than those off exhibit (on-exhibit: 12.6%; off-exhibit: 21.9%). Locomotory activity was lower during enrichment treatment compared with baseline for on-exhibit animals (enrichment: 9.0%; baseline: 12.6%) but higher for off-exhibit animals (enrichment: 23.3%; baseline: 21.9%). There was a significant main effect of exhibit type on locomotory activity of Goeldi’s monkeys (F_1,18 = 31.975, P < 0.001); and a non-significant main effect of enrichment on locomotory activity (F_1,18 = 0.194, P = 0.667). There was a non-significant interaction between exhibit type and enrichment on locomotory activity (F_1,18 = 1.210, P = 0.290).

Resting activity was higher for the Goeldi’s monkeys on exhibit than those off exhibit (on exhibit: 86.8%; off-exhibit: 76.7%). Resting activity was lower during enrichment treatment compared with baseline for on-exhibit animals (enrichment: 71.0%; baseline: 86.8%) and off-exhibit (enrichment: 56.1%; baseline: 76.7%). There was a significant main effect of exhibit type on resting activity (F_1,18 = 34.144, P < 0.001); and a significant main effect of enrichment on resting activity (F_1,18 = 72.647, P < 0.001). However, there was a non-significant interaction between exhibit type and enrichment on resting activity (F_1,18 = 1.262, P = 0.280).

Discussion

The higher activity levels (feeding/foraging and locomotion) and lower resting of cotton-top tamarins housed in a larger sized and more complex exhibit compared with subjects housed in smaller and less complex environments supported our hypothesis. The on-exhibit group showed much higher rates of feeding/foraging activity compared with the off-exhibit group, with the presence of a significant main effect of exhibit type on feeding/foraging activity. This was likely due to increased opportunities for natural foraging in a more complex and naturalistic environment with ample foliage for the animals on exhibit. Locomotory activity was also higher for on-exhibit animals than off-exhibit animals but there was no significant effect of exhibit type on locomotory activity. In addition, considering that higher locomotory activity of on-exhibit groups could also be influenced by social interactions in a larger group of on-exhibit animals, these results showed that opportunities for locomotion afforded by additional space for on-exhibit animals was likely not an important factor influencing differences in activity levels. Increased foraging opportunities in a more complex exhibit could be more important than absolute exhibit space differences used for locomotion, which supported similar results reported in another study.¹⁴ However, the effects of exhibit space limitations on overall lower activity levels in non-human primates have been otherwise widely shown.^15–17

For the Goeldi’s monkeys, subjects housed in a smaller sized and less complex off-exhibit enclosure showed higher activity levels (feeding/foraging and locomotion) and lower resting compared with subjects housed in a larger and more complex enclosure, which was contrary to our hypothesis. Higher activity for this group was contributed mainly by higher locomotory activity, since there was a significant effect of exhibit type on locomotory activity but a non-significant effect on feeding/foraging activity. This result was contrary to what was found in the cotton-top tamarin, but this could be explained by the differences in complexity and opportunities for natural foraging: the cotton-top tamarins on exhibit had much more complex environments in terms of floristic composition and space compared with the Goeldi’s monkeys on exhibit; and the differences were much larger between the cotton-top tamarins in on-exhibit and off-exhibit conditions, compared with the Goeldi’s monkeys in on-exhibit and off-exhibit conditions. However, more marked differences in activity levels between the cotton-top tamarin groups compared with between the Goeldi’s monkey groups supported our hypothesis that wider variations in within-species enclosure conditions had an effect on differences in activity levels. Height differences between the Goeldi’s monkey exhibits, whereby the off-exhibit enclosure was higher than the on-exhibit enclosure, could have contributed to the unexpected results for Goeldi’s monkeys, as vertical space has been suggested as an important factor for promoting naturalistic locomotory behaviours in marmosets and tamarin species.¹⁸ Although the height difference between the Goeldi’s monkey exhibits was small (0.5 m) compared with between the cotton-top tamarin exhibits (3.5 m), the height difference for the cotton-top tamarin exhibits was influenced mainly by a few tall trees in the on-exhibit enclosure which had limited contribution to usable vertical space compared with the meshed cages under the off-exhibit condition which provided more vertical clinging space. In addition, the effects of additional vertical space may have a more marked effect only in small sized enclosures where differences in total space are less. Remarkably, the effect of exhibit type on locomotory activity was found only for Goeldi’s monkeys and not for cotton-top tamarins. The contribution of absolute versus vertical space in influencing locomotory activity should be further examined under more appropriately controlled conditions.

Overall, we could not conclude whether differences in exhibit space at baseline conditions had an effect on activity levels due to the conflicting results we found for the different species. This could be due to inherent species or individual group differences whereby baseline activity levels of the animals could differ irrespective of the enclosure type they were housed in, and could only be resolved by within-group comparisons under different exhibit conditions that were however not logistically possible in this study. The sample size in each condition was also small and due to limited availability of study subjects in sufficiently controlled conditions of enclosure size and complexity, we could not make direct comparisons of subjects under more similar housing conditions, which resulted in the Goeldi’s monkey exhibits being much more similar compared with the larger differences between the cotton-top tamarin exhibits.

Despite these limitations, our findings suggested that the effects of enclosure size could have a smaller effect on activity levels compared with environmental complexity. This was because for both species, the off-exhibit groups did not show significantly lower locomotory activity which would have been expected as larger enclosure spaces should allow more opportunities for locomotory activity (the off-exhibit Goeldi’s monkey group showed significantly higher locomotory activity instead). This finding could be supported by the consistent results we obtained for the effects of feeding enrichment, which had significant effects on increased activity for both species irrespective of enclosure type. Feeding enrichment functionally increases opportunities for feeding/foraging, which is similar to the effects of a more complex natural environment. There was no significant main effect of enrichment on locomotory activity, and differences in rates of locomotory activity was lower for on-exhibit animals in both species (particularly for the cotton-top tamarins) during enrichment treatment compared with baseline but rates were similar for off-exhibit animals under both conditions. This indicated that on-exhibit animals traded off locomotory activity for foraging activity during enrichment, and increased foraging during enrichment was the main contributing factor for higher overall activity levels in both species.

Overall, our results suggested that even when exhibit space and complexity are limited, increased opportunities for foraging through the application of enrichment could increase species-typical activities. This is particularly beneficial for off-exhibit animals where the percentage increase in overall activity levels during enrichment is higher than for on-exhibit animals. In fact, for the off-exhibit cotton-top tamarins, feeding/foraging activity during enrichment exceeded the baseline rates for on-exhibit animals, demonstrating that feeding enrichment could effectively compensate for lack of environmental complexity. The application of simple, inexpensive feeding enrichment methods has been advocated as a useful tool for captive non-human primate management,^1,19 and should be more widely implemented, particularly in situations where there are logistical and/or cost constraints to the modification of physical exhibits, in order to provide for larger and more complex enclosures.

Footnotes

Acknowledgements

We would like to thank the animal keepers from Fragile Forest and Gibbon Island, especially Ms Sabrina Jabbar and Mr Rajan Thanapal for accommodating and facilitating data collection for this project amidst their busy schedules; and the management of Wildlife Reserves Singapore for their support.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Honess

Marin

. Enrichment and aggression in primates. Neurosci Biobehav Rev 2006; 30: 413–436.

Crockett

Bowers

Shimoji

. Behavioral responses of longtailed macaques to different cage sizes and common laboratory experiences. J Comp Psychol 1995; 109: 368–383.

Kitchen

Martin

. The effects of cage size and complexity on the behaviour of captive common marmosets Callithrix jacchus jacchus. Lab Anim-UK 1996; 30: 317–326.

Young

. Environmental enrichment for captive animals, Oxford: Blackwell Science, 2003.

Hienz

Pyle

Frey

. Enrichment use by baboons during long-term vs intermittent availability. Lab Primate Newsl 2000; 39: 1–3.

O’Neill

Novak

Suorni

. Normalizing laboratory-reared rhesus macaque (Macaca mulatta) behavior with exposure to complex outdoor enclosures. Zoo Biol 1991; 10: 237–245.

Westergaard

Fragaszy

. Effects of manipulable objects on the activity of captive capuchin monkeys (Cebus apella). Zoo Biol 1985; 4: 317–327.

Glick-Bauer

. Behavioral enrichment for captive cotton-top tamarins (Saguinus oedipus) through novel presentation of diet. Lab Primate Newsl 1997; 36: 1–4.

Phillippi-Falkenstein

. Responses of singly-housed white-crowned mangabeys (Cercocebus torquatus lunulatus) to different enrichment devices. Lab Primate Newsl 1993; 32: 5–8.

10.

Voelkl

Huber

Dungl

. Behavioral enrichment for marmosets by a novel food dispenser. Lab Primate Newsl 2001; 40: 1–4.

11.

Martin

Bateson

. Measuring behaviour. An introductory guide, Cambridge: Cambridge University Press, 1993.

12.

Altmann

. Observational study of behavior: sampling methods. Behavior 1974; 49: 227–267.

13.

Edwards

Sorkin

Richard

. Observational methods to measure behaviors of captive cotton-top tamarins (Saguinus oedipus). Zoo Biol 2010; 29: 416–431.

14.

Wilson

. Environmental influences on the activity of captive apes. Zoo Biol 1982; 3: 201–209.

15.

Chamove

Scott

. Forage box as enrichment in single- and group-housed callitrichid monkeys. Lab Primate Newsl 2005; 44: 13–17.

16.

Chamove

Rohrhuber

. Moving callitrichid monkeys from cages to outside areas. Lab Primate Newsl 1989; 8: 151–163.

17.

Clarke

Juno

Maple

. Behavioral effects of a change in the physical environment: a pilot study of captive chimpanzees. Zoo Biol 1982; 1: 371–380.

18.

French

Fite

Enrichment for non-human primates. In: Bayne

(ed). Marmosets and tamarins, Bethesda, MD: Office of Laboratory Animal Welfare, National Institutes of Health Publication, 2005, pp. 1–13.

19.

Dishman

Thomson

Karnovsky

. Does simple feeding enrichment raise activity levels of captive ring-tailed lemurs (Lemur catta)? Appl Anim Behav Sci 2009; 116: 88–95.