This article was downloaded by: [Dr Kenneth Shapiro] On: 09 June 2015, At: 08:19 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Journal of Applied Animal Welfare Science Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/haaw20 Systematic Investigation of the Stability of Food Preferences in Captive Orangutans: Implications for Positive Reinforcement Training Andrea W. Clay a b, Mollie A. Bloomsmith a b, M. Jackson Marr b & Terry L. Maple b a Yerkes National Primate Research Center, Emory University b Center for Conservation and Behavior, Georgia Institute of Technology Published online: 22 Sep 2009. To cite this article: Andrea W. Clay, Mollie A. Bloomsmith, M. Jackson Marr & Terry L. Maple (2009) Systematic Investigation of the Stability of Food Preferences in Captive Orangutans: Implications for Positive Reinforcement Training, Journal of Applied Animal Welfare Science, 12:4, 306-313, DOI: 10.1080/10888700903163492 To link to this article: http://dx.doi.org/10.1080/10888700903163492 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content ) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the
Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
JOURNAL OF APPLIED ANIMAL WELFARE SCIENCE, 12:306 313, 2009 Copyright Taylor & Francis Group, LLC ISSN: 1088-8705 print/1532-7604 online DOI: 10.1080/10888700903163492 Systematic Investigation of the Stability of Food Preferences in Captive Orangutans: Implications for Positive Reinforcement Training Andrea W. Clay, 1;2 Mollie A. Bloomsmith, 1;2 M. Jackson Marr, 2 and Terry L. Maple 2 1 Yerkes National Primate Research Center, Emory University 2 Center for Conservation and Behavior, Georgia Institute of Technology Using preference-assessment tests with humans in conjunction with behavioral modification sessions has been a regular component of almost all operant conditioning programs with mentally challenged humans. This has been very effective in improving the efficiency of behavioral training in these settings and could be similarly effective in zoological and research environments. This study investigated the preferences of 9 captive orangutans for different food items. The study used a pairwise presentation to record each nonhuman animal s preferences for 5 different foods on 6 different occasions over the course of 6 months. Results of a Friedman s 2-way ANOVA indicated that the orangutans showed a clear overall preference for apple. However, there was significant variability among different orangutans in preference ranking for the 5 foods, as shown by a Kendall s tau. In addition, there was variability in preference rankings across time for each orangutan. Because the orangutans preferences change over time and vary according to individual, regular assessments should identify items to be used as rewards in behavioral husbandry training or as part of feeding enrichment strategies. Over the past 15 years, positive reinforcement training has become very popular in zoological institutions and in nonhuman animal laboratory facilities. Animals Correspondence should be sent to Andrea W. Clay, Yerkes National Primate Research Center, Emory University, 954 Gatewood Road NE, Atlanta, GA 30329. Email: awclay@comcast.net 306
FOOD PREFERENCES IN CAPTIVE ORANGUTANS 307 can be trained using positive reinforcement to cooperate with basic husbandry procedures such as urine collection (Bassett, Buchanan-Smith, McKinley, & Smith, 2003) or to accept injections voluntarily; in both cases, positive reinforcement training has been shown to reduce the stress of procedures (Elvidge, Challis, Robinson, Roper, & Thorburn, 1976; Lambeth, Hau, Perlman, Martino, & Schapiro, 2006; Reinhardt, 1991). We can also teach the animals to use certain equipment, like computers, which enables us to do different types of cognitive research with the animals and also provides the animals with problem-solving opportunities. Positive reinforcement has also been used to increase affiliative (Schapiro, Bloomsmith, & Laule, 2003) and decrease aggressive (Bloomsmith, Laule, Alford, & Thurston, 1994) behaviors in captive primates and has been shown to improve animal-caretaker relationships (Savastano, Hanson, & Mc- Cann, 2003). There are, however, differences in how individual animals respond to training, including how quickly they progress. One reason for this variation may be differences in the response of individuals to the rewards employed as positive reinforcement. Typically, the preferences of individual primates for food rewards are assessed using nonempirical means by people conducting the training, but this assessment may be imperfect. It may be helpful to have an empirical means to determine individual preferences. Such preference research has been conducted extensively with human subjects. Therefore, we might look to the human research as a model when initiating preference research studies on nonhuman primates (Bloomsmith, Marr, & Maple, 2007). Preference research in humans has been focused largely on the determination of individual preferences for certain stimuli and the use of those preferred items in subsequent behavioral training sessions in which the preferred item functions as a reward. Often, these methods are used to enhance behavioral modification training with severely and moderately mentally challenged individuals. Methods for determining individual preferences include the Paired Choice method (PC), which involves presenting subjects with each stimulus in a set, paired with each other item in the set, and then recording the subject s choices. Based on these choices, a rank order is assigned to the stimuli. The PC method results in differentiation between stimuli; in subsequent studies, items identified as preferred stimuli (based on this procedure) function better as rewards (DeLeon et al., 2001; Fisher & Mazur, 1997), with behavioral training occurring faster and with greater success than when rewards are randomly chosen from a set of stimuli. Conducting preference-assessment tests may be helpful in zoological and research settings if, by using preference-assessment tests as a regular part of training procedures, we can determine items to use as rewards on an individual basis and accommodate any changes in these individual preferences over time. In addition, conducting preference tests with animals could be a way to improve animal-caretaker relationships. Positive caretaker-primate relationships
308 CLAY, BLOOMSMITH, MARR, MAPLE have been found to correlate with less agonistic behavior, less self-directed behavior, and increased affiliative behaviors in captive stump-tailed macaques (Macaca arctoides) and thus could be one way to improve overall captive primate welfare (Waitt, Buchanan-Smith, & Morris, 2002). Finally, by allowing the animal to make selections between different stimuli, we introduce an element of choice into their daily lives, which is believed to be important in promoting the well being of animals (Catania & Sagvolden, 1980; Rumbaugh & Washburn, 2003). To begin to explore the usefulness of preference assessments, a study was conducted to investigate preferences for different food items in captive orangutans. Basic food items from their regular diets were used because these items are often used as rewards during behavioral training. The study was designed to determine if captive orangutans had preferences for certain foods as a group, to determine if individual differences in these preferences existed, and to explore the stability of these preferences over time. Subjects METHOD Subjects for the study were 9 orangutans housed at Zoo Atlanta: 4 adult males, 1 juvenile male, and 4 adult females. There were 7 Sumatran orangutans (Pongo pygmaeus abelii), 1 adult male Bornean orangutan (Pongo pygmaeus pygmaeus), and one adult male hybrid. They were housed in an indoor/outdoor exhibit and holding area. All test sessions were conducted in the morning, before animals had been fed, in the indoor holding area. Andrea W. Clay conducted the testing on six different occasions between September and December 2006. All testing complied with protocols approved by Zoo Atlanta and by the Georgia Institute of Technology: Institute for Animal Care and Use Committee. Procedures The five items of fresh produce tested were grapefruit, orange, apple, sweet potato, and carrot; all subjects were familiar with these foods. Items were presented using a standard PC method (DeLeon et al., 2001), balancing for left and right location and randomizing the order of the pairs presented during each session. Each possible pairing in each arrangement (item (a) on left and (b) on right; item (b) on left and (a) on right) was presented once per session, resulting in 20 pairings per session. Food pieces were cut to be roughly 1 in. (2.5 cm) in diameter. The food pieces were presented on a Lexan tray with two equally sized recessed dishes at one end. This end of the tray was held
FOOD PREFERENCES IN CAPTIVE ORANGUTANS 309 up to the mesh of the orangutans enclosure, and the animals selected one of the two items by pointing. All of the animals readily pointed at one of the two items, though occasionally they pointed at both items at once. When this occurred, the tray was removed for about 5 s and then presented again, at which point the animal always pointed at only one of the two items. Six sessions for each animal were conducted over the course of 6 months. A chi-square test was used to determine if there were individual side biases by testing the total left versus total right selections for each animal across all six test sessions. If food preference was controlling the animal s selections, rather than side bias, then this score should have been approximately 50% left and 50% right. A Friedman s two-way ANOVA was used to analyze the preferences of the group of subjects; Kendall s tau tests were used to analyze the rank-order correlation between different orangutans and for individual orangutans across test sessions. RESULTS One of the female orangutans showed a significant side bias in her responding, indicated by a significant chi-square value, 2 (1, N D 120) D 17.63, p <.05, so she was dropped from further analysis. For each of the five food items, a score was obtained for each test session by using the following formula: (number of times item selected/number of times item presented) 100 D % score for each item. This was calculated for each individual, so there was a score for each food item during each test session for each individual animal. A mean score was computed for each food item as rated across all six sessions by all the orangutans. A Friedman s two-way ANOVA test on these scores indicated that the orangutans, as a whole, preferred some of these food items over others, 2 (4, N D 8) D 24.126, p <.001. Overall, the orangutans scored apple the highest, followed by orange, sweet potato, grapefruit, and carrot (Figure 1). For comparisons of individual preferences, mean percentage scores were obtained for each food item across all six sessions for each individual animal, and these scores were compared. Different animals had different preferred foods according to these scores. A Friedman s two-way ANOVA on these scores indicated that there was variation in the mean scores that the different orangutans assigned to the five foods. For apple 2 (7, N D 6) D 21.83, p <.01; see Table 1. Rank ordering of the five foods was not highly correlated between any two individuals as indicated by Kendall s tau (Table 2). To assess the stability of each individual orangutan s food preferences, individual rating scores were graphed for the foods across six sessions. Although some animals had relatively stable scores, others did not. The scores were transformed into rank orders, and a Kendall s tau test was run to determine
310 CLAY, BLOOMSMITH, MARR, MAPLE FIGURE 1 sessions. The figure shows mean rating of food items by all orangutans over all six test TABLE 1 Mean Ratings of Individual Orangutans Across All Six Test Sessions Food Item Hati Allen Sulango Madu J.T. Junior Daisy Chantek Apple 70.8 89.6 72.6 79.2 91.7 70.8 95.8 79.2 Orange 68.8 72.9 39.6 45.8 70.8 79.2 70.8 39.6 Carrot 16.7 10.4 31.3 22.9 16.7 6.3 10.4 18.8 Sweet potato 58.3 22.9 64.6 81.3 47.9 41.7 20.8 83.3 Grapefruit 37.5 54.2 33.3 22.9 33.3 52.1 52.1 37.5 TABLE 2 Rank Order Correlations Between Orangutans as Analyzed With Kendall s Tau Subject Hati Allen Sulango Madu J.T. Junior Daisy Chantek Hati Allen.800 Sulango.800.600 Madu.527.316.738 J.T. 1.000*.800.800.527 Junior.600.800.400.105.600 Daisy.800 1.000*.600.316.800.800 Chantek.600.400.800.949*.600.200.400 Note. Significant values in bold. *p <.05.
FOOD PREFERENCES IN CAPTIVE ORANGUTANS 311 TABLE 3 Rank Order Correlations Between Tests for Individual Orangutans as Analyzed by Kendall s Tau Tests Daisy Allen Sulango Chantek Junior J.T. Hati Madu 1, 2.889*.825.889*.600.738.527.126.738 1, 3.889* 1.000**.222.527.105.667.598.527 1, 4.949*.875.126.738.600.889*.252.738 1, 5.889*.671.111.527.527.527.120.894* 1, 6.889*.894*.118.600.316.527.252.738 2, 3 1.000**.825.222.527.000.738.527.667 2, 4.949*.943*.126.738.738.738.667 1.000** 2, 5 1.000**.738.000.527.667.400.316.825 2, 6 1.000**.949*.236 1.000*.444 1.000*.222.889* 3, 4.949*.875.756.889*.316.889*.738.667 3, 5 1.000**.671.444 1.000**.444.738.400.471 3, 6 1.000**.894*.943*.527.667.738.527.778 4, 5.949*.671.882*.889*.949*.527.738.825 4, 6.949*.894*.802.738.738.738.667.889* 5, 6 1.000**.800.589.527.778.400.949*.707 Note. Significant values in bold. *p <.05. **p <.01. correlation coefficients for each animal s rank order across all six sessions. This test revealed a fair amount of variability in rank order for most of the orangutans as indicated by less-than-significant correlation values (p <.05), though one animal, Daisy, had generally stable preferences (Table 3). On average, about one third of each orangutan s food rankings changed over time. DISCUSSION The results of this study indicate three things about orangutan preferences: 1. They have preferences for certain food items; 2. Individual orangutans have specific preferences that vary from other individual orangutans; and 3. Individual orangutan preferences change over time. Based on these results, we may conclude that using preference-assessment tests may have certain applications in the management of captive primate species: identification of items to use as rewards in positive reinforcement training or identification of items with which to bait puzzle feeders and other forms of food-based enrichment devices.
312 CLAY, BLOOMSMITH, MARR, MAPLE CONCLUSION These applications could be particularly important when the goal is to use items from the animal s daily diet for training and enrichment purposes. A logical progression from these findings is to use preference-assessment tests in conjunction with enrichment or training procedures to determine the effects of using preferred items as rewards, as compared with using randomly selected items. This has already been shown to increase the efficiency of training behaviors in humans (DeLeon et al., 2001; Fisher & Mazur, 1997) and would therefore be worth assessing as a method for increasing the efficiency of training behaviors in nonhuman primates. As time is such a valuable commodity in almost all captive-care environments, any method that will allow for increased efficiency for caretakers is likely to be a very useful one. In addition, due to time and budget limitations, increasing the effectiveness of enrichment items would be similarly useful. ACKNOWLEDGMENT This research was supported in part by the Yerkes National Primate Research Center base grant (P51 RR-00165) awarded by the National Center for Research Resources of the National Institute of Health. Thanks for the hard work and cooperation of Zoo Atlanta s Primate Staff, including Laura Mayo, Charles Horton, Ricca DuCharme, Beth Pailthorpe, and Becky Richardson. We also thank Ursula Anderson and Allison Martin for their statistical analysis assistance. REFERENCES Bassett, L., Buchanan-Smith, H. M., McKinley, J., & Smith, T. E. (2003). Effects of training on stress-related behavior of the common marmoset (Callithrix jacchus) in relation to coping with routine husbandry procedures. Journal of Applied Animal Welfare Science, 6, 221 233. Bloomsmith, M. A., Laule, G. E., Alford, P. L., & Thurston, R. H. (1994). Using training to moderate chimpanzee aggression during feeding. Zoo Biology, 13, 557 566. Bloomsmith, M. A., Marr, M. J., & Maple, T. L. (2007). Addressing nonhuman primate behavioral problems through the application of operant conditioning: Is the human treatment approach a useful model? Applied Animal Behaviour Science, 102, 205 222. Catania, A. C., & Sagvolden, T. (1980). Preference for free choice over forced choice in pigeons. Journal of Experimental Analysis of Behavior, 34(1), 77 86. DeLeon, I. G., Fisher, W. W., Rodriguez-Catter, V., Maglieri, K., Herman, K., & Marhefka, J. M. (2001). Examination of relative reinforcement effects of stimuli identified through pretreatment and daily brief preference assessments. Journal of Applied Behavior Analysis, 34, 463 473.
FOOD PREFERENCES IN CAPTIVE ORANGUTANS 313 Elvidge, H., Challis, J. G., Robinson, J. S., Roper, C., & Thorburn, G. D. (1976). Influence of handling and sedation on plasma cortisol in rhesus monkeys (Macaca mulatta). Journal of Endocrinology, 70, 325 326. Fisher, W. W., & Mazur, J. E. (1997). Basic and applied research on choice responding. Journal of Applied Behavior Analysis, 30, 387 410. Lambeth, S. P., Hau, J., Perlman, J. E., Martino, M., & Schapiro, S. J. (2006). Positive reinforcement training affects hematologic and serum chemistry values in captive chimpanzees (Pan troglodytes). American Journal of Primatology, 68, 245 256. Reinhardt, V. (1991). Training adult male rhesus monkeys to actively cooperate during homecage venipuncture. Animal Technology, 2, 11 17. Rumbaugh, D. M., & Washburn, D. A. (2003). Asking questions so that animals can provide the right answers. In D. M. Rumbaugh & D. A. Washburn, (Eds.), Intelligence of apes and other rational beings (pp. 147 166). New Haven, CT: Yale University Press. Savastano, G., Hanson, H., & McCann, C. (2003). The development of an operant conditioning program for new world primates at the Bronx zoo. Journal of Applied Animal Welfare Science, 6, 247 261. Schapiro, S. J., Bloomsmith, M. A., & Laule, G. E. (2003). Positive reinforcement training as a technique to alter nonhuman primate behavior: Quantitative assessments of effectiveness. Journal of Applied Animal Welfare Science, 6, 175 187. Waitt, C., Buchanan-Smith, H. M., & Morris, K. (2002). The effects of caretaker-primate relationships on primates in the laboratory. Journal of Applied Animal Welfare Science, 5, 309 319.