Arboreal and Terrestrial Foraging Preferences in New World and Old World monkeys: An Experimental Investigation Samantha Elias Natan Gervais Stephen Lu Mara Mathieu James Wang Anika Wikramanayake Research in Ecology University of Miami
Abstract We examined the foraging and sentinel behavior between a group of New World and Old World semi-free ranging monkeys in a naturalistic environment. The group of Old World monkeys were Long-tailed Macaques (Macaca fascicularis), and the group of New World monkeys was comprised of Peruvian Squirrel Monkeys (Saimiri boliviensis peruviensis) and Black-capped Capuchins (Cebus apella). We created an experimental foraging site for each of the groups of monkeys, each with an arboreal and terrestrial component, and recorded the frequencies of foraging behavior and sentinel behavior at each area for both groups of monkeys. We found that the Old World monkeys fed significantly more frequently at the terrestrial area than the New World monkeys. New World monkeys fed more at the arboreal area in the beginning of our foraging trials, but once preferred food items were depleted, they descended to the terrestrial foraging area. Old World monkeys performed more sentinel behavior than New World monkeys, and we interpreted this as a result of subordinate monkeys monitoring the whereabouts of dominant group members in a competitive feeding situation. When terrestrial feeding increased in New World monkeys, terrestrial sentinel behavior also increased, possibly related to increased vulnerability to potential predators. Our results confirmed that new and Old World monkeys preferred to feed at sites that matched their habitat preferences in nature. Introduction Many species of nonhuman primates have generalized diets and feed on fruits, leafs, and insects and are therefore described as omnivores. Foraging, searching for food and feeding, is an important and time-consuming behavior and frequently accounts for over 50 percent of nonhuman primates waking activities (Strier 2007). Our experiment studies the foraging behavior of monkeys. There are two types of monkeys, Old World monkeys and New World monkeys. Old World monkeys come from Africa and Asia. New World monkeys come from Central and South America. New World monkeys have nostrils that are farther apart and point outwards and Old
World monkeys have nostrils that are closer together and point downward. Old World monkeys have more pronounced sexual dimorphism and the males are frequently larger than the females and have bigger canines. All New World monkeys are arboreal whereas Old World monkeys can be terrestrial or arboreal. The subjects of our investigation were two species of New World monkeys (Peruvian Squirrel Monkeys [Saimiri boliviensis peruviensis] and Black-capped Capuchins [Cebus apella]) and one species of Old World monkey (Long-tailed Macaques [Macaca fascicularis]). The Long-tailed Macaques were the largest monkeys and the squirrel monkeys were the smallest (Rowe 1996). In nature, squirrel monkeys and balck-capped capuchins are almost exclusively arboreal and rarely descend onto the forest floor (Boinksi 1987, Mitchell et. al 1991, Rowe 1996) whereas the Long-tailed Macaques are found in both the canopy and the ground and are often described as semi-terrestrial and spend time both in the trees and on the ground (DeRuiter 1994). All three species are quadrupeds with good climbing abilities; Black-capped Capuchins have prehensile tails (Youlatos 1999). All three species live in reasonably large mixed male and female groups along with juveniles and infants. Depending on the habitat and the height in the canopy where the monkeys live, monkeys may either spend most of their time foraging in the trees (arboreally) or on the ground (terrestrially). Factors that may be important in determining where a species of monkey forages include their natural habitat, size of the monkey, and presence of predators. Our problem statement was what are the foraging location preferences between New World monkeys and Old World monkeys? The null hypothesis for this problem statement was there is no difference between the foraging location preferences between New World monkeys and Old World monkeys. The alternative hypothesis was there is a difference between the foraging location preferences between New World monkeys and Old World monkeys. Based on their natural history, we predicted that there would be a difference in the foraging location preferences between New World monkeys and Old World monkeys. Specifically, since all New World monkeys are arboreal and the Old World monkeys we studied, the Long-tailed Macaques, spend time in the canopy and on the ground, we predicted that New World monkeys would show less ground foraging than the Old World monkeys. The risk of predation strongly influences the behavior of primates (Gursky and Nekaris 2007). Since we were examining foraging behavior in two different groups of monkeys that live in two different habitats, we were interested in studying if one habitat seemed more risky to the monkeys. Consequently we recorded sentinel behavior at both foraging locations with our null hypothesis being that there would be no differences in the amount of sentinel behavior performed at either foraging location. Methods Location of the experiments Monkey Jungle is located in southwest Miami Dade, Florida. Of particular interest are two enclosures containing semi-free troops of monkeys in semi-natural habitats. The larger enclosure, the macaque enclosure is 2.46 hectares and the smaller rainforest enclosure is 1.45 hectares. Both enclosures contain forests known as hammocks, which consist of neotropical hardwood trees of Caribbean origin. The hammock in the rainforest enclosure has been supplemented with exotic plantings including fruit trees, palms, and shrubs transplanted from
Iquitos, Peru. The two enclosures are very naturalistic with complex supports including lianas and resemble the habitats the monkeys occupy in nature. The enclosures are both surrounded by a two-meter high electrified fence and there is a cleared area of approximately four meters between the fence and the edge of the forest. In the rainforest enclosure, there are a series of trails that allow visitors to hand feed the New World monkeys. Native wildlife (for example raccoons and snakes) lives in both enclosures, and aerial predators (most likely hawks) have been observed carrying juvenile squirrel monkeys out of the rainforest enclosure (Evans, personal communication). Subjects Approximately 75 squirrel monkeys and five Black-capped Capuchins live in the rainforest enclosure, and about 125 Long-tailed Macaques live in the macaque enclosure. All of these monkeys were born and raised in captivity and are habituated to humans. The macaques and the squirrel monkeys live in reasonably large male/female groups together with juveniles and infants. The capuchin group contains only adult females. The Long-tailed Macaques are the largest species (adult male weight 2.5-6.7 kg and adult female weight 4.7-8.3 kg), the squirrel monkeys the smallest (adult male weight 0.9-1.9 kg and adult female weight 0.7-0.9 kg) with male capuchins weighing between 3.3-4.8 kg and female capuchins weighing between 1.3-3.4 kg (Rowe 1996). The monkeys can range freely in their enclosures. They are provisioned three times a day to accompany educational presentations for park visitors and first thing in the morning in the rainforest enclosure and at special feeding sites (hoppers) in the macaque enclosure. Location of Experimental Feeding Sites There were a number of factors that we took into account when we selected the locations for the experimental foraging sites. The sites had to be easily accessible to the experimental observers; the arboreal sites had to have branches sufficiently strong to support the containers and the weight of the adult monkeys; the terrestrial sites had to have enough exposed roots and/or trunks to which to attach the foraging containers; and finally there had to be enough space and with good visibility to accommodate the team of 10 observers. The experimental foraging site in the macaque enclosure was easily accessible from the perimeter pathway. We searched for a place that the monkeys would visit frequently and chose a site near the west hopper (a feeding dispenser from which they feed every day). In the rainforest enclosure, we selected a site adjacent to a trail and located in an area park visitors regularly feed the monkeys. For an aerial view of the enclosure and experimental site locations, see Figure 1. Scheduling of the Experiments The experiments were performed twice each day; the first was performed in the early morning (between 0900h and 1100h) and the second in the late morning/early afternoon (between 1100h and 1300h). These times were selected to avoid any conflict with the visitor feeding presentations that occur at Monkey Jungle throughout the day. Experiments were conducted three times a week over a three-week period in June/July 2010. Experimental Procedure To create each experimental foraging site, we attached 10 hard plastic containers to the natural vegetation at each site to simulate a high quality food patch. Five containers were
attached to roots or the base of trees on the ground (terrestrial foraging area) and the other five were placed in the canopy on trunks or limbs of trees (arboreal foraging area). Holes were bored into the sides of the containers so that they could be secured to the trees and roots. We made sure the foraing containers were stable and secure enough so that curious monkeys would not tip them over or pull them off the trees. In the macaque enclosure, the arboreal foraging containers were attached to the trees at a mean height of 146 cm with a mean distance of 177 cm between them. In the rainforest enclosure the foraging containers were attached at a mean height of 170m and a mean distance of 174 cm between them. The containers were filled with approximately 750 cm 3 of assorted foods from the monkeys normal diet at Monkey Jungle, which included mixed fruits and vegetables and commercial monkey chow. The food for each container was carried into the enclosures in a plastic bag so that the containers could be filled quickly at the start of each experimental trial (it took approximately 40 seconds to fill all of the containers). For each experimental trial, we recorded the frequency of three foraging aspects (monkeys approaching the containers within one meter from any direction, monkeys touching the containers, and monkeys feeding from the container) and sentinel behavior onto time-ruled (30 second interval) check sheets for an observation period of 20 minutes. Because we were not able to identify individual monkeys in either enclosure, the frequency of each behavior was recorded, not the number of monkeys performing each behavior. Sentinel behavior was defined as scanning the area and emitting alarm calls. Two people were responsible for collecting frequencies of each behavior; one collected the behavior for the arboreal foraging area and the other at the terrestrial area. Running notes were made of any unusual behavior. We did not conduct inter-observer reliability tests (Martin and Bateson 1993). After each experimental trial, containers were removed to prevent monkeys from tearing them off. We tied zip ties around branches (or trunks or roots) to mark a site of attachment of the containers (for the duration of our experiments) and placed zip ties through the holes in the containers and tied these to or near the zip ties on the branches and roots. The location of the foraging containers was therefore the same throughout the trials. If monkeys were not present by the time we had completed our experimental set up, an attempt was made to lure them to the site. If after 10 minutes no monkeys appeared, the trial was abandoned. The 20-minute trials were terminated early if there was a complete absence of any monkeys at both the terrestrial and arboreal areas for three consecutive minutes or if all the food was consumed from the foraging container at either the arboreal or terrestrial foraging areas. Data analysis When data collection was completed, we tallied the total frequency for each of the four behaviors (approaching a container within one meter, touching a container, feeding from a container, and sentinel behavior) in each area (arboreal and terrestrial) for the first and second half of each experimental trial separately. We combined these total frequencies for an overall total for that particular experimental trial. We then calculated the proportion of terrestrial foraging for each of the four behaviors. Data were analyzed using the Mann-Whitney U test. All data were analyzed using SYSTAT 12.
Results We conducted 10 trials for the New World monkeys, and 12 trials for the Old World monkeys, two of which were abandoned because no monkeys were present within 10 minutes of the experimental set-up. We therefore had data from 10 trials for each experimental site. One New World monkey trial was cut short because there was no more food in the arboreal foraging containers, and one Old World monkey trial was cut short because there was a 3-minute absence of monkeys. Foraging Behavior When we compared the proportion of ground foraging behaviors between the New World and Old World monkeys we found a difference between the feeding frequencies (Mann Whitney U test, p=0.004), but no differences for approaching within one meter or touching (MWU, p=0.821 and p=0.345, respectively; Figure 2). The Old World monkeys fed more frequently on the ground than the New World monkeys did; however, New World and Old World monkeys approached within one meter of and touched the food containers on the ground with proportionally the same frequency as did the Old World monkeys. Therefore, we rejected the null hypothesis for feeding behavior, but failed to reject the null hypothesis for the other two foraging behaviors. We noted that New World monkeys were more hesitant to feed on the ground in the first half than in the second half of the trial, so to determine when the greatest difference between New World and Old World ground feeding was, we compared feeding between Old World and New World monkeys for the first half and for the second half of the trials. There was a significant difference in ground feeding between old and New World monkeys for the first half of trials, but not for the second half (MWU, p=0.000 and p=0.07, respectively; Figure 3). Although significantly less than Old World monkeys, the proportion of New World monkeys foraging on the ground was higher than we expected. As a result, we examined whether the feeding preferences changed over time and consequently, we compared the proportion of foraging behaviors between the first half and the second half of the foraging trials. We found that the New World monkeys fed proportionally more on the ground in the second half than in the first half of the trial (MWU, p=0.002; Figure 4a). There was no difference in proportion of ground feeding in the Old World monkeys between the first and second half of the trial (MWU, p=0.821; Figure 4b). There were no other differences detected for any other foraging behavior in either the new or Old World monkeys. When we reviewed our notes, it was very striking that intraspecific aggression was observed only in the Old World monkeys. We observed intraspecific aggression in 70 percent of the Old World monkey foraging trials. In the New World monkeys, there was no aggression between conspecifics, but we often noted interspecific displacement, and the squirrel monkeys would leave the foraging container if a capuchin approached (60 per cent of trials). Sentinel Behavior There was a significant difference in terrestrial sentinel behavior between New World and Old World monkeys (MWU, p=0.003; Figure 5). A more detailed analysis of the sentinel behavior revealed one other difference a comparisons between the first half and second half of the trails for New World monkeys revealed that there was a difference in sentinel behavior (U=5, p=0.001, df=1; Figure 6), with New World monkeys performing more sentinel behavior while
feeding on the ground. A similar comparison between the first half and the second half of the trails for the Old World monkeys revealed no difference. Discussion This is the first experimental investigation comparing foraging preferences between terrestrial and arboreal foraging sites between new and Old World monkeys. The semi-natural enclosures at Monkey Jungle together with the well-habituated monkey groups made this an ideal location for this type of investigation. A field experiment investigation of predator sensitive foraging in squirrel monkeys in covered and open areas in Eastern Amazonia found that squirrel monkeys visited their experimental foraging sites on only seven out of a potential 39 trials (Stone 2007) whereas monkeys visited our experimental sites on 22 out of 24 of the trials. Foraging Behavior Both Old World and New World monkeys approached, touched, and fed from the containers. Therefore, our foraging sites were easily located and exploited by the monkeys. Of all the measures of foraging behavior, we detected a significant difference in the feeding behavior only between the New World and the Old World monkeys. There was more terrestrial feeding behavior in the Old World monkeys compared to the New World monkeys. This can be explained because New World monkeys are exclusively arboreal and the Old World monkeys we studied are semi-terrestrial. However, the amount of New World monkey terrestrial feeding was more than we expected since New World monkeys are exclusively arboreal and the two species we studied are reasonably small-bodied. We considered that the New World monkeys would be less likely to descend to the ground because of their potential increased vulnerability to predators. Largebodied cats can climb trees but are too heavy to reach outermost branches and will in general have restricted mobility (Ferrari 2008), leaving these small-bodied species safer in the trees when in their natural environment. One explanation for the increased amount of terrestrial feeding is that these individuals are habituated to the rainforest enclosure, which is void of terrestrial predators that they face in the wild (i.e., large felids and snakes that prey on monkeys). The lack of difference in the proportion of terrestrial foraging between the first and second half of the experimental trials with Old World monkeys demonstrates that there was no habituation in Old World monkeys. That is, the Long-tailed Macaques were just as comfortable foraging on the ground in the second half, as they were in the first half. While there was no increase in the proportion of terrestrial foraging for approaching within a meter and touching in the New World monkeys from the first half to the second, there was a difference in actually consuming the food. One explanation for this is that the New World monkeys were wearier of predators in the first half, and so more hesitant to feed in the ground containers. After some time, they realized that there was no real danger of foraging on the ground. An alternate, more realistic explanation is that New World monkeys preferred feeding at the arboreal foraging area, but moved to the terrestrial foraging area once the preferred food items in the arboreal containers were depleted. We consider this explanation likely because we terminated one of our New World monkey foraging trials when all of the food in the arboreal containers was consumed after 12 minutes.
Sentinel Behavior The sentinel behaviors we recorded were warning calls and scanning. There was a significant difference in Old World (Long-tailed Macaques) and New World monkeys (squirrel monkeys and capuchins) for sentinel behavior. Due to the size variation between old and New World monkeys, we did not expect to find that Old World monkeys would have more sentinel behavior than New World monkeys since New World monkeys are not adapted for ground foraging, while Old World monkeys are. One explanation for the unexpected results is that Old World monkeys are competitive feeders. Competitive feeding occurs in Long-tailed Macaque society because of their marked dominance hierarchy. The dominant male and female in a troop have first choice of food. During our trials, there was no competitive food aggression observed among the New World monkeys, however this was observed in 70 percent of the foraging trials for the Old World monkeys. Certain Old World monkeys, including macaques, possess cheek pouches (to store food) as an adaptation and these are used by subordinate monkeys in competitive feeding situations. We interpret this result as suggesting that intraspecific feeding competition in the Long-tailed Macaques was responsible for the unexpectedly high frequency of sentinel behavior. The subordinate macaques (most likely juveniles) were monitoring more dominant members of their own social group and not necessarily predators. Interestingly, when we looked at the second half of the foraging trials for the New World monkeys, the increase in feeding on the ground was accompanied by an increase in sentinel behavior. This corresponding increase suggests that the New World monkeys were more wary on the ground and demonstrates support for arboreal monkeys showing more surveillance for predators when foraging on the ground. However, as we did not record the number of monkeys present at each site we cannot eliminate the possibility that there were more New World monkeys feeding on the ground in the second half and the increase in the number of monkeys was the reason for the corresponding increase in the sentinel behavior. No comparable difference was detected for the Old World monkeys. We succeeded in designing and creating experimental foraging sites that were visited regularly by our subjects and allowed us to address our problem statement. The foraging site in the rainforest enclosure was very successful, in particular, the arboreal foraging area. Monkeys visited this area on all trials, and the squirrel monkeys in particular were observed in small subgroups foraging harmoniously. One of the educational presentations at Monkey Jungle involves feeding the rainforest monkeys while visitors watch; consequently, we made a recommendation to Monkey Jungle that they consider setting up multiple feeding sites resembling our set up as the overall effect we created was very naturalistic foraging in the trees. Literature Cited Boinski S. 1987. Habitat use by squirrel monkeys (Saimiri oerstedi) in Costa Rica. Folia Primatol 49(3-4): 151-67 De Reiter, J. 1994. Behavior and genes in natural populations of Long-tailed Macaques (Macaca fascicularis). Ph.D. thesis Utrecht University, Netherlands
Ferrari, S. 2009. Predation risk and anti-predator strategies in South American primates. Edited by Garber, P A, Estrada, A, Bica Marques, J C, Heyman, E W, and Strier, K B. Springer, New York, NY, pp. 251-271. Gursky, S. L. and Nekaris K. A. R. 2007. Primate anti-predator strategies. Springer, New York, NY. Martin P and Bateson P. 1994. Measuring Behavior University of Cambridge Press, Cambridge, UK. Mitchell C. L., Boinski S., van Schaik C. P. 1991. Competitive regimes and female bonding in two species of squirrel monkeys (Saimiri oerstedi and S. sciureus). Behav Ecol Sociobiol 28(1): 55-60. Rowe N. 1996. The pictorial guide to the living primates. East Hampton (NY): Pogonias Press. Strier, K. B. 2007. Primate Behavioral Ecology. Allyn and Bacon, Boston. Stone, A L. 2007. Age and seasonal effects on predator sensitive foraging in squirrel monkeys (Saimiri sciureus). American Journal of Primatology, 69, pp 127-141. Youlatos D. 1999. Tail-use in capuchin monkeys. Neotropical Primates 7(1): 16-20. Figures
Macaque enclosure Rainforest enclosure
Figure 4. Proportion of foraging behaviors on ground in a) New World and Old World monkeys in first and second half of the trials (mean ± SEM). *p < 0.05