Tool use in urban populations of capuchin monkeys Sapajus spp . ( Primates : Cebidae )

Capuchin monkeys, Sapajus Kerr, 1792, are known for their flexible behavior, including tool use, and their ability to survive in urban forests. We observed capuchin juveniles using wood as hammer and anvil and different materials as sponges (four tool-use events) in two geographically distinct urban populations in Brazil, in 2012: two in Goiânia, Central Brazil and two in Foz do Iguaçu, Southern Brazil. In Goiânia, a male used a detached tree branch as a hammer and a buttress root as an anvil to pound a seed of Terminalia Linnaeus. Another male used a small branch with leaves as a dipping tool to access water inside a tree trunk hole. In Foz do Iguaçu, the capuchins used a small branch and a piece of bread to obtain water by dipping them into tree trunk holes. This latter event might be interpreted as a case of self-control, with a familiar food item used as a tool to reach a resource that is difficult to access otherwise. Our observations contribute to the knowledge on the tool-kit of capuchins and we propose that these urban populations should be conserved for scientific evaluations of behavioral flexibility in non-human primates.

Capuchin monkeys, Sapajus Kerr, 1792, are extractive foragers, and their behavioral flexibility allows them to inhabit a variety of habitats, including forest fragments within many Brazilian cities (FRAGASZY et al. 2004).Although anthropic environments can provide some advantages to animals (e.g., partial food provisioning, low predation risk and increased free time), they also impose new challenges (LOWRY et al. 2013, SOL et al. 2013).For example, in anthropic forest fragments, capuchins must adjust their foraging patterns by exploring new feeding resources (artificial or exotic foods) (VISALBERGHI & ADDESSI 2003).At the same time, they have to cope with stressful situations, such as crowding, roadkill, electrocution, intoxication, and conflicts with humans.Behavioral flexibility is widely believed to be one of the mechanisms by which animals deal with these challenges (SOL et al. 2013).Tool use, an example of flexible behavior found among capuchin monkeys (FRAGASZY et al. 2004), has been documented in some urban populations (ROCHA et al. 1998, OTTONI & MANNU 2001).
Habitual tool use by wild capuchins has been increasingly documented in the last decade (OTTONI & IZAR 2008).Most accounts concern Sapajus libidinosus Spix, 1823, a species that inhabits savannah-like environments in Brazil, and describe the use of tools for food extraction (sensu BENTLEY-CONDIT & SMITH 2010), for instance stones for digging tubers, and pounding tools for cracking open encapsulated fruits (MOURA & LEE 2004, MANNU & OTTONI 2009, DE MORAES et al. 2014).In addition, the tool-kit used by S. libidinosus comprises tools of other categories, such as utensils to transport resources, including probes (MANNU & OTTONI 2009), and for communication (FALÓTICO & OTTONI 2013).The use of probes was also observed in a group of blond capuchin monkeys (Sapajus flavius Schreber, 1774) inhabiting a forest fragment in the Atlantic Forest Domain (SOUTO et al. 2011).
The differences observed among tool-kits of capuchins from different populations illustrate their flexibility in tool use.Although genetic and ecological factors may contribute to the development of tool use behavior (MEULMAN et al. 2013), the social environment appears to provide a basis for naive capuchins to learn technical skills (FRAGASZY et al. 2013).
Here we report the use of tools by capuchins from two urban populations in geographically distant locations in Brazil, with two different purposes: first, to pound seeds with a wooden hammer and anvil and second, to obtain water inside tree trunk holes using different kinds of sponges.The use of ability to survive in urban forests.We observed capuchin juveniles using wood as hammer and anvil and different materials as sponges (four tool-use events) in two geographically distinct urban populations in Brazil, in 2012: two in Goiânia, Central Brazil and two in Foz do Iguaçu, Southern Brazil.In Goiânia, a male used a detached tree branch as a hammer and a buttress root as an anvil to pound a seed of Terminalia Linnaeus.Another male used a small branch with leaves as a dipping tool to access water inside a tree trunk hole.In Foz do Iguaçu, the capuchins used a small branch and a piece of bread to obtain water by dipping them into tree trunk holes.This latter event might be interpreted as a case of self-control, with a familiar food item used as a tool to reach a resource that is difficult to access otherwise.Our observations contribute to the knowledge on the tool-kit of capuchins and we propose that these urban populations should be conserved for scientific evaluations of behavioral flexibility in non-human primates.Our observations took place in two urban parks in Brazil: Areião Municipal Park (16°42'18.4"S,49°15'16.1"W),14 ha large, in the city of Goiânia (Central Brazil, Cerrado Domain) and the Jardim Ipê Forest (25°28'56.1"S,54°33'55.9"W),3.5 ha large, in the city of Foz do Iguaçu (Southern Brazil, Atlantic Forest Domain).In both areas, capuchins are semi-free ranging and are provisioned weekly, although they can also forage for resources in the forest and consume human leftovers.Both areas are open the public and humans frequently feed the capuchins.Contrasting with the Foz do Iguaçu Park, where water is more limited, water sources are continuously available in the Goiânia Park (a stream and a lake).Stones that are large and hard enough to be used as hammers and anvils are unavailable in both locations.
A native population of S. libidinosus has been living in the Goiânia Park at least since 1938, when the reserve was created.Today, this population consists of three groups.There have been local accounts of these monkeys using probes, which we aimed to ascertain.Capuchins from different groups were observed from August to November 2012, three days a week, at different hours (8:00-11:00 or 13:00-16:00 h) of the day, adding up to a total of 126 h in 52 days.Tool use was observed in two distinct groups: one of 24 individuals [13 adults (five males and eight females), two subadults, four juveniles and one infant], and another of 34 individuals [13 adults (three males and 10 females), four subadults, eight juveniles and one infant)].
In Foz do Iguaçu, tool-use events were registered during a social behavior study of a group of 17 individuals (J.P. Back, unpubl.data) that inhabit the forest [six adults (one male and five females), three subadults, three juveniles and five infants].Although Sapajus nigritus (Goldfuss, 1809) occurs naturally in this region, the coats of individuals in this group are blond and they rather resemble Sapajus cay (Illiger, 1815).Therefore, they are here referred to as Sapajus sp.(LYNCH-ALFARO et al. 2014).According to local people, these capuchins have lived there for 30 years, and are most likely introduced.The group was observed for three days each month between 7:00 and 17:00 h from January 2012 to June 2013 for a total of 530 hours.
We followed the definition of tool use by SHUMAKER et al. (2011).We considered as a tool-use event the usage of an object to obtain a resource until the abandonment of that object.These events were counted using the all-occurrences sampling method (ALTMANN 1974), and were recorded and described ad libitum in a field notebook.
We observed four tool-use events: two events in Goiânia (1.6 event/100 h) and two in Foz do Iguaçu (0.4 event/100 h).
All tool-use events were observed in 2012 and were performed by juveniles.In all cases, no other individual was within five meters of the subject.
The first event was recorded in Goiânia (September 7 th , at 9:47 h): a male juvenile on the ground used a detached tree branch as a hammer and a buttress root as an anvil to pound an encapsulated seed of Terminalia L. (Fig. 1).One of us (E.C.Carneiro) became aware of this behavior when it was already being performed, and for this reason we did not record its duration.We were also unable to verify whether the use of this tool was successful because the animal abandoned the site carrying the seed.The second event (September 12 th , 14:58-15:00 h), also in Goiânia, took place in the canopy, ten meters above ground: another male detached a small branch with leaves from a tree of an undetermined species and dipped it into a tree trunk hole (Fig. 2).The capuchin left the branch inside the hole for three seconds and then kneaded the leaves with two hands and sucked them.Subsequently, it moved one meter away and discarded the branch.
In Foz do Iguaçu, we observed the use of two different tools for dipping into tree trunk holes of the forest understory (approximately two meters above ground): 1) a small piece of bread provided by humans (January 7 th , 14:45-14:51 h) and 2) a small branch with leaves picked up from the ground (September 6 th , 16:21-16:22 h).During both events, the animal sucked the object repeatedly after multiple insertions into the cavities.During the first event, the capuchin ate the bread after dipping it and sucking from it five times.During the second event, the capuchin left the branch after drinking from it three times.It was not possible to identify the juveniles in Foz do Iguaçu, and therefore they may have been performed by the same individual or not.
Tool use was a rare behavior at our study sites compared with other sites (e.g., PHILLIPS 1998, MANNU & OTTONI 2009), but we observed differences in tool-use behavior in both areas.In Goiânia, tool use can be split into two categories: food extraction and resource transportation (here: water).In Foz do Iguaçu, we only observed one of those categories, resource transport, but two different techniques were employed: the use of a natural item from the forest (branch) and of a man-made food item (bread).This may indicate a temporary transition of a familiar food item into a dipping tool, as discussed below.
Capuchins can maintain the use of tools through social learning, and adult individuals tend to be more skilled users (OTTONI et al. 2005).However, we only observed tool use by juveniles at our observation sites, and adults were not involved.The observed categories of tool use are common among capuchin juveniles (PHILLIPS 1998, MANNU & OTTONI 2009) and have been associated with their great learning capacity, propensity for innovation and flexible behavior (KUMMER & GOODALL 1985).LOWRY et al. (2013) noted that juveniles can express a greater foraging plasticity in urban environments, and tool use can be considered as a foraging strategy (BENTLEY-CONDIT & SMITH 2010).
In Goiânia, a capuchin used a hammer and anvil in a feeding context.According to FRAGASZY et al. (2013) these are complex tools for animals and require practice.Young capuchins can combine non-stone items for pounding (MANNU & OTTONI 2009), as observed by us where stones are not available.Thus, the use of wooden hammer and anvil suggests a pounding technology analogous to the more common lithic tools observed at other sites, since the basic principles of tool use can be applied to both organic and inorganic materials (WESTERGAARD & SUOMI 1994).
The urban capuchins used dipping tools as sponges but in a simpler fashion than other capuchin populations, which perform object modifications (PHILLIPS 1998, MANNU & OTTONI 2009).The use of bread as a tool by a capuchin was noteworthy and the bread was only consumed after repeatedly using it as a sponge for drinking water, suggesting that the user is capable of self-control (EVANS & WESTERGAARD 2006).Apparently, the capuchin was able to postpone the immediate consumption of a food item to access water, which is a scarce and valuable resource in this urban forest.The use of food as a tool, including bread, has been recorded for captive capuchins (MCGREW & MARCHANT 1997), and our observations indicate that capuchins in a non-captive context share this ability.Alternatively (but not to the exclusion of the first explanation), it is possible that the capuchin meant to soften the bread instead of using it as a tool.However, the action pattern and the repetition of combined insertion and drinking indicate that the bread was used as a tool.
Our observations were made on semi-free ranging and partially provisioned monkeys, a situation that can interfere with the foraging behavior of the species (LOWRY et al. 2013).However, it is difficult to evaluate these influences on the urban capuchins from Goiânia because wild populations of S. libidinosus frequently use tools.Yet, the continuous availability of food and water suggests that these animals live with a surplus of time and energy, and that the observed tool use be-havior could result from the opportunistic habits of the species (HASLAM 2013).Anthropic food are abundantly available to the capuchins in Foz do Iguaçu, while water sources are scarce, which could explain the self-control and opportunistic habit of using a common food item as a tool.
Finally, tool use behavior is absent in the majority of other species of Sapajus, especially in forested environments (OTTONI & IZAR 2008).There are only four sites in the Atlantic Forest Domain where capuchins were observed using tools (this study, ROCHA et al. 1998, OTTONI & MANNU 2001, SOUTO et al. 2011), three of which are urban forests.Monkeys in urban forests are often more accessible for primatological field studies and urban populations of tool using capuchins can be viewed as scientific subjects for accessible studies of primate behavior.In addition to surveying tool-use techniques by capuchin populations, we encourage environmental administrators, managers and scientists to use the so-called Ecologistic-Scientific Human Value of Nature (KELLERT 1993) as a motivation to better protect these animals in cities.
Our observations contribute to the knowledge on the tool-kit of capuchins and we propose that these urban populations should be subjected to scientific evaluations.Standardized field studies will help to catalog, compare and contextualize the social and ecological significance of the tool-kit of these urban monkeys, while providing a better understanding of animal behavioral flexibility and interactions with humans.

Figures
Figures 1-2.Capuchin monkeys, S. libidinosus, from Goiânia using tools: (1) a juvenile using a wooden hammer and anvil to open a Terminalia sp.seed; (2) a juvenile inserting a small branch with leaves into a tree trunk hole to obtain water.