American Journal of Primatology 70:510–513 (2008) BRIEF REPORT An Experimental Analysis of Ingestion Rates in an Omnivorous Species M. STAMMATI1,2, G. SABBATINI1,2, AND E. VISALBERGHI1 1 Unità di Primatologia Cognitiva, Istituto di Scienze e Tecnologie della Cognizione, Consiglio Nazionale delle Ricerche, Rome, Italy 2 Dipartimento di Biologia Animale e dell’Uomo, Università di Roma ‘‘La Sapienza’’, Rome, Italy Food intake is difficult to estimate under natural conditions. We investigated ingestion rates of 14 different food types in 26 captive capuchin monkeys (Cebus apella). The procedure consisted in weighing a piece of food and using a two alternative choice tests to present food to the subject, alone in its cage. We recorded the food chosen and the time it took the subject to consume the food entirely. Consumption time was converted into ingestion rates (g/s) for each food type. Ingestion rates of food types significantly differed, and the difference was significantly higher among foods than among subjects. In particular, ingestion rates of the fruits were higher than those of human-processed food. Interestingly, food preferences were significantly related to energy intake rate, i.e., to the amount of energy ingested per unit of time, but not with ingestion rates or energy content alone. The energy acquired by eating different types of food cannot be calculated on the basis of the time spent eating unless a correction factor for each given food (or similar ones) is applied. Future controlled studies should provide field researchers with such corrections factors, possibly using foods collected in the wild. Am. J. Primatol. 70:510–513, 2008. r 2008 Wiley-Liss, Inc. Key words: ingestion rate; Cebus apella; food preferences; energy intake rate INTRODUCTION Food intake (grams of food eaten) is difficult to estimate because the sources of measuring errors are numerous, particularly under natural conditions [Hladik, 1977]. Time spent feeding is often used as an approximation of actual food intake, but Zinner  demonstrated that feeding time has to be used with caution to estimate food intake, and Hladik  found that time spent feeding on some food items by Hanuman langurs was 10 times higher than the proportion of food intake. Therefore, an accurate estimate of food intake should take into account both time spent feeding and mass of food eaten. The ingestion rate (grams of food ingested per unit of time) depends on many factors, for example, mastication and chewing patterns. The characteristic breakage of a given food is important for the sensory feedback it can provide when chewed and, in turn, mastication patterns depend on the perceived taste and texture [van der Bilt et al., 2006]. In humans, the time until swallowing was shorter, and fewer chews were observed as palatability of the food increased [Bellisle et al., 2000]. Very recently, Schulke et al.  showed that ingestion rates have a crucial importance in determining the energy intake rates in two ecologically different populations of Hanuman langurs (Semnopithecus entellus). The authors calculated the ingestion rate of each monkey by multiplying its bite rate r 2008 Wiley-Liss, Inc. by the average bite dry weight. Bite rate was recorded from a limited number of complete direct observations of eating, and the weight of a bite was estimated from a set of food samples having similar size of that eaten by the monkey. Their complex calculations, possibly the best obtainable in nature, are still far from being precise. Although variation in both ingestion rate and chemical composition of primate food items has been investigated [e.g., Laska et al., 2000; Zinner, 1999], few studies have explored their relation with food preferences. We studied an omnivorous primate species, the capuchin monkey, in controlled laboratory conditions in which we could accurately assess ingestion rates for a variety of familiar foods and individual preferences toward these foods. For each food presented we knew its mass and its energetic Contract grant sponsor: FIRB/MIUR; Contract grant number: RBNE01SZB4; Contract grant sponsor: ANALOGY: Humans— the Analogy-Making Species; Contract grant number: 029088. Correspondence to: Elisabetta Visalberghi, Unit of Cognitive Primatology and Primate Center, Institute of Cognitive Sciences and Technologies, CNR, Via Ulisse Aldrovandi, 16/b, 00197 Roma, Italy. E-mail: email@example.com Received 1 March 2007; revised 12 December 2007; revision accepted 12 December 2007 DOI 10.1002/ajp.20523 Published online 9 January 2008 in Wiley InterScience (www. interscience.wiley.com). An Experimental Analysis of Ingestion Rates / 511 and nutritional characteristics. In this exploratory study, we aimed to reveal possible variation in ingestion rates across individuals and/or food types. If these factors turn out to play a role, it means that field studies can increase the accuracy of their estimation of food intake by taking variation in ingestion rates into account. frequency of eating behavior was scored with instantaneous sampling every 5 s. Eating was scored when the subject bit and masticated the food, licked or swallowed it, but did not include processing behaviors (e.g., peeling) occurring outside the mouth. The experiment was carried out between January 2001 and January 2002. MATERIALS AND METHODS Analysis Subjects We calculated ingestion rates (g/s) by dividing the weight of the piece of food (g) by the time (s) taken to consume it totally. As the assumptions of parametric statistics were not met, we used non-parametric tests. To determine whether ingestion rates vary among food items more than among individuals, we compared coefficient of variation (standard deviation over the mean) with Mann–Whitney U test. To assess whether there was a significant difference between the ingestion rates of the different foods, we carried out a Friedman analysis of variance test and Wilcoxon Matched Pairs test to compare pairs of foods. As many comparisons were made, we applied Bonferroni’s correction, with the corrected a level set at 0.00076. We used the Mann–Whitney U test to assess whether age and sex affected ingestion rates. Ingestion rates of the three categories of food used (fruit: tangerine, banana, pear, pineapple, grapefruit, tomato; leaf-tuber: boiled potato, lettuce; and humanprocessed food: bread, monkey chow, canned meat, boiled pasta) were compared with the Kruskall–Wallis test and post hoc analyses were conducted by means of the Mann–Whitney U test with a corrected a level set at 0.017. We obtained the item-specific energy intake rate (kJ ingested per sec) by multiplying the average ingestion rate across individuals with the energy content (kJ per 1 g) of each food item. We carried out Spearman correlations (with Bonferroni’s correction, a level set at 0.0167) to assess whether food preferences (i.e., the average number of times in which each food was chosen) were related with ingestion rates, item-specific energy intake rate and energy content of food. To exclude spurious correlations, we carried out Spearman correlations (with Bonferroni’s correction, a level set at 0.0167) between item-specific energy intake rate and ingestion rates, energy content of food and water content and between ingestion rates and water content of food. We tested 26 capuchins belonging to four groups: 10 males, 16 females; five juveniles (o4 years old), 21 adults (44 years old); the age range was 2–35; average age was 13. Twenty-one subjects were laboratory born, whereas five were of unknown origin (probably wild born). The monkeys were housed at the Primate Center of the Institute of Cognitive Sciences and Technologies (CNR) inside the Bioparco (Rome). Group 1 (N 5 12) had a total of 105 m3, group 2 (N 5 6) had a total of 135 m3, group 3 (N 5 3) had 90 m3 and group 4 (N 5 5) had 70 m3. By passing through one or more sliding doors, the subject can be separated from the rest of its group in the testing cage. Monkey chow (Altromin-A pellets, Rieper standard diet for primates), fresh fruits and vegetables were given every afternoon. Three times a week, monkeys received a mixture of curd cheese, vitamins, egg, bran, oats and sugar. The research complied with protocols approved by the Italian Health Ministry, and all procedures complied with ASAB guidelines and European law on the humane care and use of laboratory animals. Procedure The subjects were presented with all possible binary combinations of 14 foods (tangerine, banana, pear, boiled potato, bread, monkey chow, lettuce, pineapple, canned meat, boiled pasta, grapefruit, tomato, boiled string beans and savoy cabbage) using a two-alternative choice test [see Addessi et al., 2005; Visalberghi et al., 2003]. This procedure was repeated three times and led to a data set in which the tested foods were not evenly represented because during a single trial subjects could choose between two pieces of food, and most preferred foods were chosen a higher number of times than least preferred ones. Given this, string beans and cabbage could not be used for ingestion rate analyses as these items were entirely eaten only a few times. The foods were cut into pieces of similar sizes (about 1 cm3), and each piece was weighed with a digital scale (AND compact scale, CA; 0.1 g accuracy, 200 g capacity) before testing. Note that similar size leads to weight that depends on the type of food. The trial started when the subject took the food and ended once the food was totally consumed. To calculate the average time spent by each subject eating each food type, the RESULTS Coefficient of variation of ingestion rate was significantly higher across foods (Mean7SEM 5 0.8670.04) than across individuals (Mean7SEM 5 0.4570.04; U 5 29.00, Z 5 4.39, nf 5 26, ns 5 14, P 5 0.000002). The ingestion rates were significantly different across foods (Friedman analysis of variance w2 5 84.44; Po0.0001). Mean ingestion rates of foods Am. J. Primatol. 512 / Stammati et al. TABLE I. Post Hoc Comparisons Between the Foods Bread Lettuce Pear Monkey chow Potato Tangerine Grapefruit Meat Pasta Pineapple Tomato Banana Bread Lettuce Pear 0.00001 0.00028 0.00002 0.00001 0.00001 0.00001 n.s. 0.00003 0.00001 0.00016 n.s. 0.00029 0.00001 0.00004 0.00002 0.00001 0.00004 0.00029 0.00023 0.00001 0.00004 0.00039 0.00023 n.s. 0.00046 0.00064 0.00044 0.00029 n.s. 0.00081 0.00001 0.00002 n.s. n.s. 0.00003 0.00001 n.s. n.s. Monkey chow Potato Tangerine Grapefruit 0.00001 0.00001 0.00003 n.s. n.s. 0.00001 0.00003 0.00002 0.00005 0.00004 0.00002 0.00010 0.00004 n.s. 0.00003 0.00001 n.s. n.s. 0.00009 0.00004 n.s. n.s. Meat Pasta n.s. 0.00003 0.00001 0.00009 0.00004 Pineapple n.s. P-values are reported only for significant comparisons; n.s., not significant. (7SEM) from the highest to the lower value were the following: banana 0.52 (70.04), grapefruit 0.39 (70.03); tomato 0.37 (70.03), pear 0.34 (70.02), pineapple 0.32 (70.02), tangerine 0.30 (70.02), potato 0.16 (70.02), lettuce 0.14 (70.01), string bean 0.14 (70.01), savoy cabbage 0.12 (70.01), monkey chow 0.04 (70.00), pasta 0.03 (70.00), canned meat 0.03 (70.00) and bread 0.02 (70.00). Table I reports P-values for the post hoc comparisons. The ingestion rates differed across food categories (H (2, 12) 5 9.23; Po0.01), and post hoc comparisons revealed a significant difference between fruit and artificial food categories (U 5 0.00; Z 5 2.56; n1 5 6, n2 5 4; P 5 0.009), but not between fruit and leaf-tuber categories (U 5 0.00; Z 5 2.00; n1 5 6, n2 5 2; P 5 0.07), and between leaf-tuber and human-processed food categories (U 5 0.00; Z 5 1.85; n1 5 2, n2 5 4; P 5 0.13). Nevertheless, potato and lettuce had significantly lower ingestion rates than fruits, if considered one at a time (see Table I). Ingestion rates were not affected by age and sex. Preferences were the following, from the most to the least preferred food: tangerine, banana, boiled potato, pear, pineapple, canned meat, bread, grapefruit, boiled pasta, tomato, monkey chow, savoy cabbage, boiled string beans and lettuce. Neither ingestion rates (rs 5 0.39; ns) nor energy content (rs 5 0.40; ns) correlated with food preferences. In contrast, item-specific energy intake rate (kJ ingested per sec) significantly correlated with food preferences (rs 5 0.83; Po0.001). Item-specific energy intake rate did not correlate with ingestion rates (rs 5 0.53; ns), energy content of food (rs 5 0.50; ns) and water content (rs 5 0.36; ns). Moreover, water content did not correlate with ingestion rates (rs 5 0.50; ns). DISCUSSION As reported by Schulke et al.  for langurs, the ingestion rates of capuchin monkeys differed more across food items than across individuals. Am. J. Primatol. In particular, there were marked differences in ingestion rates between the most pairs of foods. Moreover, although water could facilitate mastication, ingestion rates were not significantly influenced by water content of food. Ingestion rates of humanprocessed foods were lower than those of fruits. It is interesting to note that as the human-processed foods had higher energy content than fruits and vegetables, their slow ingestion rate still resulted in a high energy intake. Our foods reflected a sufficiently broad range of mechanical properties [Lucas, 2004; see Fig. 5 in Williams et al., 2005], but unfortunately we did not measure them. Tufted capuchins exhibit craniofacial features and a broad face associated with powerful masticatory muscles that anteriorly can produce and dissipate high masticatory forces. In the wild, Cebus apella open palm nuts using the canines and the premolars [Izawa, 1979]. Moreover, Wright  found that plant tissues processed by C. apella are tougher than those processed by C. olivaceus and suggested that to adapt to a hard diet C. apella has exaggerated the ancestral condition for capuchins by increasing the robusticity of canine size. This adaptation allows the monkey to break hard fruits with the anterior dentition without precluding consumption of pulpy fruits that are easier and quicker to process and ingest. The decreased size of the more distal molars characterizing C. apella can account for the lower values of ingestion rates for leaves, tubers and human-processed foods. In fact, this feature reduces the ability to perform higher magnitude and repetitive loadings with the posterior dentition required for eating foods such as leaves, tubers and human-processed food [Spencer, 2003]. Future studies will benefit from combining data on ingestion rates with those on the food mechanical properties and primate mastication. A novel and important finding that comes from our data is that food preferences are significantly related to energy intake rate (i.e., the amount of energy ingested per unit of time), but not with ingestion rates (grams of food ingested per unit of An Experimental Analysis of Ingestion Rates / 513 time) or food energy content alone (kJ per g). This result reinforces the previous results of Nakagawa  and Barton and Whiten  that primates optimize their energy intake according to food availability and patch use. There is increasing knowledge about how primates perceive and represent food [Rolls, 1999]. Food perception begins externally with visual, auditory, olfactory and tactile cues and proceeds in the mouth with texture and taste stimuli [Dominy et al., 2001]. Finally, the primate orbitofrontal cortex is an important site for the convergence of representations of these stimuli and for their non-linear combination with visceral and satiety-related signals and with previously learned associations [Rolls, 2005]. This entire system acts on food acceptance and preferences. Future studies should investigate how primates integrate mechanical and chemical properties of food to establish preferences. We demonstrated that the same period of time spent eating corresponds to different amounts of food eaten, for example, 1 sec spent eating lettuce corresponds to 0.10 g ingested and 1 sec spent eating banana corresponds to 0.52 g ingested. As previously pointed out by Hladik , this result challenges the common view that the time spent feeding reflects the same amount of food eaten, regardless of food. This means that the energy acquired by eating different types of food cannot be calculated on the basis of the time spent eating unless a correction factor for each given food (or similar ones) is applied. Future controlled studies should provide field researchers with such correction factors possibly using foods collected in the wild, and not foods from the grocery stores as we used. ACKNOWLEDGMENTS We thank F. Natale and E. Addessi for statistical advice and the two anonymous referees for their useful comments. We are also grateful to the Bioparco Foundation for hosting the laboratory where the experiment was carried out and our keepers M. Bianchi and S. Catarinacci for their help. This research complied with ASAB guidelines and the European law on the humane care and use of laboratory animals. REFERENCES Addessi E, Stammati M, Sabbatini G., Visalberghi E. 2005. How tufted capuchin monkeys (Cebus apella) rank monkey chow in relation to other foods. Anim Welf 14:215–222. Barton RA, Whiten A. 1994. Reducing complex diets to simple rules: food selection by olive baboons. Behav Ecol Sociobiol 35:283–293. Bellisle F, Guy-Grand B, Le Magnen J. 2000. Chewing and swallowing as indices of the stimulation to eat during meals in humans: effects revealed by the edogram method and video recordings. Neurosci Biobehav Rev 24:223–228. Dominy NJ, Lucas PW, Osorio D, Yamashita N. 2001. The sensory ecology of primate food perception. Evol Anthropol 10:171–186. Hladik CM. 1977. A comparative study of the feeding strategies of two sympatric species of leaf monkeys: Presbytis senex and Presbytis entellus. In: Clutton-Brock TH, editor. Primate ecology. Studies of feeding and ranging behaviour in lemurs, monkeys and apes. London: Academic Press. p 324–353. Izawa K. 1979. Foods and feeding behavior of wild blackcapped capuchin (Cebus apella). Primates 20:57–76. Laska M, Hernandez Salazar LT, Rodriguez Luna E. 2000. Food preferences and nutrient composition in captive spider monkeys, Ateles geoffroy. Int J Primatol 21:671–683. Lucas PW. 2004. Dental functional morphology: how teeth work. Cambridge: Cambridge University Press. 355p. Nakagawa N. 1997. Determinants of the dramatic seasonal changes in the intake of energy and protein by Japanese monkeys in a cool temperate forest. Am J Primatol 41: 267–288. Rolls ET. 1999. The brain and emotion. Oxford: Oxford University Press. 367p. Rolls ET. 2005. Taste, olfactory, and food texture processing in the brain, and the control of food intake. Physiol Behav 85:45–56. Schulke O, Chalise MK, Koenig A. 2006. The importance of ingestion rates for estimating food quality and energy intake. Am J Primatol 68:951–965. Spencer MA. 2003. Tooth-root form and function in platyrrhine seed-eaters. Am J Phys Anthropol 122:325–335. van der Bilt A, Engelen L, Pereira LJ, van der Glas HW, Abbink JH. 2006. Oral physiology and mastication. Physiol Behav 89:22–27. Visalberghi E, Sabbatini G, Stammati M, Addessi E. 2003. Preferences towards novel foods in Cebus apella: the role of nutrients and social influences. Physiol Behav 80:341–349. Williams SH, Wright BW, Truong VD, Daubert CR, Vinyard CJ. 2005. Mechanical properties of foods used in experimental studies of primate masticatory function. Am J Primatol 67:329–346. Wright BW. 2005. Craniodental biomechanics and dietary toughness in the genus Cebus. J Hum Evol 48:473–492. Zinner D. 1999. Relationship between feeding time and food intake in hamadryas baboons (Papio hamadryas) and the value of feeding time as predictor of food intake. Zoo Biol 18:495–505. Am. J. Primatol.