International Journal of Primatology

, Volume 38, Issue 2, pp 224–242 | Cite as

Comparing the Use of Camera Traps and Farmer Reports to Study Crop Feeding Behavior of Moor Macaques (Macaca maura)



Investigating crop feeding patterns by primates is an increasingly important objective for primatologists and conservation practitioners alike. Although camera trap technology is used to study primates and other wildlife in numerous ways, i.e., activity patterns, social structure, species richness, abundance, density, diet, and demography, it is comparatively underused in the study of human–primate interactions. We compare photographic (N = 210) and video (N = 141) data of crop feeding moor macaques (Macaca maura) from remote sensor cameras, functioning for 231 trap days, with ethnographic data generated from semistructured interviews with local farmers. Our results indicate that camera traps can provide data on the following aspects of crop feeding behavior: species, crop type and phase targeted, harvesting technique used, and daily and seasonal patterns of crop feeding activity. We found camera traps less useful, however, in providing information on the individual identification and age/sex class of crop feeders, exact group size, and amount of crops consumed by the moor macaques. While farmer reports match camera trap data regarding crop feeding species and how wildlife access the gardens, they differ when addressing crop feeding event frequency and timing. Understanding the mismatches between camera trap data and farmer reports is valuable to conservation efforts that aim to mitigate the conflict between crop feeding wildlife and human livelihoods. For example, such information can influence changes in the way certain methods are used to deter crop feeding animals from damaging crops. Ultimately, we recommend using remote-sensing camera technology in conjunction with other methods to study crop feeding behavior.


Behavioral flexibility Crop raiding Ethnoprimatology Human–wildlife conflict Macaca maura Methods 


Across many areas of the world, agricultural areas continue to expand, often eliminating or encroaching on wildlife habitat and resulting in considerable human–wildlife conflict (Woodroffe et al. 2005). Some wildlife, such as primates, can flexibly adapt to such alterations in their habitat by incorporating agricultural crops into their dietary repertoire, a behavior commonly referred to as crop raiding, or more recently, crop foraging or crop feeding (Hockings et al.2015; see also Hill 2015). Multiple primate species are well-known crop feeders, e.g., capuchins (Cebus spp.: McKinney 2011), macaques (Macaca spp.: Riley 2007), baboons (Papio spp.: Strum 2010), and chimpanzees (Pan troglodytes: Hockings et al. 2009). In many cases, primates’ intelligence and agility help them gain access to agricultural areas, while large group size and diurnal activity make certain species more conspicuous crop feeders than other wildlife (Riley 2007).

Because of the detrimental impact crop feeding behavior can have on human livelihoods and the likelihood of local support for conservation efforts, this flexible behavior has become an impediment to primate conservation. Understanding the underlying causes of crop feeding and the factors that might lead to effective mitigation programs has therefore become an important concern for primatologists and conservation practitioners (Hockings and Humle 2009; Strum 2010; Woodroffe et al. 2005). To do so requires the documentation of patterns of crop feeding, thereby making the methodology behind the study of human–wildlife conflict worthy of investigation in and of itself.

To date, primate crop feeding behavior has been studied using various data collection methods, including direct observation (Priston et al. 2012); feeding traces, i.e., inspecting food remains for evidence of species-specific harvesting and consumption techniques such as ripped stems and bite marks (Riley 2007), fecal analysis (Bessa et al. 2015; McLennan 2013), and ethnographic interviews (Hill 1997; Linkie et al. 2007). Direct observation is advantageous because it can generate more accurate measures of the timing, frequency, and amount of damage caused by crop feeding, and it enables researchers to collect demographic information on participating crop feeding individuals (Hockings et al. 2009). However, it can be difficult to conduct direct observations on crop feeding primates, particularly if they are unhabituated. There are also important ethical concerns to consider when choosing direct observation as the primary strategy, namely, that habituation (which direct observation typically necessitates) may lead to increased rates of crop feeding (Fedigan 2010; Seiler and Robbins 2016), and rapport with farmers may be compromised by the fact that researchers allow crop feeding to happen during observations, thereby negatively impacting farmers’ livelihoods. For these reasons, indirect methods tend to be more frequently used. For example, Riley (2007) measured the amount and seasonality of feeding on cacao pods by unhabituated Tonkean macaques (Macaca tonkeana) and other wildlife by quantifying feeding traces. To study the foraging strategies of chimpanzees living in matrix habitat in Bulindi, Uganda, McLennan (2013) examined feeding traces and collected fecal samples, which identified four important cultivated fruits (cacao, guava, mango, and papaya) in the chimpanzee diet. The farmers themselves who suffer from crop feeding are also sources of information regarding the timing, frequency, and severity of the damage along with the species responsible for the damage (Hill 1997). However, farmer reports may conflict with data from other sources, resulting in varying information on the extent of crop damage (Naughton-Treves 1997; Riley 2007) and identification of the species responsible for the most damage (Linkie et al. 2007; Siex and Struhsaker 1999). Another potential tool for indirect measurement that is becoming increasingly popular to study human–wildlife conflict, including crop feeding behavior, are remote sensor cameras, or camera traps (Krief et al. 2014).

Camera traps are increasingly being used by primatologists, primarily because they allow researchers to conduct noninvasive observations, eliminate the need to habituate, and collect data in a variety of locations, time periods, and conditions (O’Connell et al. 2011; Pebsworth and LaFleur 2014). Accordingly, researchers have used this method to examine a suite of variables, including activity patterns, species richness, abundance, density, demography, and interactions among predators, prey, and competitors (Farris et al. 2014; Galvis et al. 2014). For example, camera traps revealed nocturnal activity patterns in the Guizhou snub-nosed monkey (Rhinopithecus brelichi), a species long assumed to be diurnal (Tan et al. 2013). They have also been useful in documenting the presence, demography, and social behavior of unhabituated primate groups (Bezerra et al.2014; Boyer-Ontl and Pruetz 2014). Camera traps may therefore be an alternative method for studying crop feeding, particularly when direct behavioral observation is difficult and/or ethically problematic because of concerns regarding increased likelihood of crop feeding and increased risk of injury or death for habituated, crop feeding primates.

Human–primate conflict due to crop feeding is particularly salient in Indonesia given that it supports the richest nonhuman primate fauna of any Asian country, i.e., ca. 44 species, 84% of which are listed as threatened with extinction (IUCN 2008). In the last 50 years, >40% of the country’s forests have been cleared, primarily for agricultural development (Barber et al. 2002), and >60% of the human population lives in rural areas relying on agriculture as a primary source of income (ILO 2012). Macaques are especially problematic because they are highly adaptable and often thrive in secondary forests, disturbed habitats, buffer zones, and along forest–agricultural edges (Brotcorne et al. 2014; Richard et al. 1989). Human–macaque interactions in Sulawesi occur predominantly at the forest–farm interface (Priston et al. 2012; Riley 2007), where crop feeding is a common behavior exhibited by all species of Sulawesi macaques (Riley 2010).

The moor macaque (Macaca maura) is one of the seven macaque species that inhabit Sulawesi (Fooden 1969). They live in the southwest region of the island in multimale–multifemale, female philopatric social groups that range in size from ca. 15–40 individuals (Okamoto et al. 2000; Watanabe and Matsumura 1996). Moor macaques in this region of South Sulawesi inhabit dipterocarp forest with limestone formations in addition to forests adjacent to agricultural areas (Supriatna et al. 1992). They are primarily frugivorous but also eat leaves, flowers, shoots, and stems from trees, shrubs, and herbaceous vegetation, with figs (Ficus spp.) comprising a large part of their diet, as well as insects (Matsumura 1991; Sagnotti 2013). Many macaque species also feed on cultivated foods. Supriatna et al. (1992) reported that moor macaques feed on the following crops: corn, banana, tomato, coconut, long bean, soybean, peanut, bushbean, eggplant, cassava, sweet potato, jackfruit, cacao, orange, cashew, guava, and mango.

We here evaluate two methods used to assess crop feeding behavior by moor macaques in South Sulawesi, Indonesia: observations via remote sensor cameras and local farmers’ reports. We collected these data as part of a broader ethnoprimatological, i.e., the study of the human–primate interface, project examining the ecological and nutritional correlates of crop feeding by moor macaques and farmers’ perceptions of and reactions to this behavior. Our objective here is twofold. First, we evaluate the effectiveness of camera traps in studying crop feeding behavior by unhabituated moor macaques. Specifically, we ask: What elements of crop feeding behavior can be measured using camera traps? What can camera traps tell us about crop feeding as a flexible behavioral strategy? Because camera traps are best suited for detecting gregarious, relatively large-bodied, ground-dwelling species (Tobler et al. 2008; Treves et al. 2010), we predicted that this method would be suitable for examining crop feeding by moor macaques, a species known for being both arboreal and terrestrial (Sagnotti 2013). Second, we examine how camera trap data compare with local farmers’ perceptions of the conflict derived from ethnographic research. Given that previous studies have found that farmers’ reports may contrast with findings from other methods (Linkie et al. 2007; Riley 2007), we predicted that farmers’ reports would deviate from camera trap results on the frequency and timing of crop feeding events, and the species most responsible.


Study Site and Study Species

This research took place at the interface of the Education Forest, a 1300-ha forest, and agricultural areas in the town of Bengo, in the southwestern province of the Indonesian island of Sulawesi (Fig. 1). Approximately 3000 people inhabit the town, which is divided into smaller villages. The majority of Bengo inhabitants practice some form of small-scale agriculture, including wet rice cultivation during the rainy season. The Education Forest is an anthropogenically modified forest comprising mostly secondary forest and pine plantations that serves as a teaching, research, and training resource for students and faculty at Hasanuddin University (UNHAS), located in the provincial capital of Makassar. Previous research has documented that at least seven groups of moor macaques live in the Education Forest (Agustinus 2011). Agricultural areas, including wet rice fields, mixed-crop gardens, and cacao plantations, occur within the Education Forest and along the eastern edge. Bengo residents maintain agricultural areas in close proximity to the forest and reported the occurrence of crop feeding by macaques at this location before the start of the project. Pilot work conducted in 2013 confirmed, via examination of feeding traces, that macaques feed on watermelon and cacao. We chose two gardens situated at the edge of the Education Forest as our study areas: a mixed-crop garden, which is ca. 1.5 ha, and a seasonally changing monocrop garden, which is ca. 4 ha (Fig. 2). Human presence in the gardens depended on the season and crop maturity. At both locations there were days during which no humans were present, except where ripe watermelon was available.
Fig. 1

Map of Indonesia indicating study site in Bengo, South Sulawesi, Indonesia.

Fig. 2

Map of study site in Bengo, South Sulawesi, Indonesia indicating eight camera trap locations in the Education Forest: three in the mixed-crop garden (a) and five in the monocrop garden (b).

Data Collection

Camera Traps

We collected camera trap data from July 2014 to March 2015. We used remote sensor cameras (Bushnell 8 MP) to collect information on the timing, frequency, and location of crop feeding behavior by macaques ≤18 m of the camera unit. The cameras, triggered by movement, were set up to record and save the following information for each photograph or video during any time of day or night: date, time, location, and temperature. We decided to use either photograph or video mode based on factors such as garden size and type, the distance of the camera unit to potential crop feeding sites, amount of surrounding vegetation, and levels of human activity in the area. We set up the cameras to capture only one image per trigger, and the time lapse between triggers was a 2-s interval. Two cameras in the mixed-crop garden were set to record a 30-s video per trigger. All cameras were housed in metal cases and secured to a substrate using a lock at a height of ca. 0.5–4 m (Fig. 3). We chose to install multiple cameras in a single garden based on the concept of spatial autocorrelation: the assumption that sampling units placed closer together will yield more similar results (Acrenaz et al.2012). We intentionally set up the cameras so that the same crop feeding event might be captured by different cameras and therefore different perspectives, offering a more comprehensive measure of crop feeding characteristics such as macaque group size and composition. The use of camera traps for studying crop feeding behavior differs from the placement, distribution, and number of cameras used for other studies of primates, such as demography (Acrenaz et al.2012).
Fig. 3

Camera trap encased in metal protective box and secured to the tree trunk with a lock in the mixed-crop garden. The camera is facing fruiting banana trees. (Photo credit: Paisal).

We defined a crop feeding event (CFE) as a situation in which one or more macaques were present in a garden. A CFE was considered independent if >1 h had passed between captured images. We chose the time designation of 1 h based on camera trap recorded CFE durations. Macaque CFEs were broken down further into two categories. A potential CFE occurred when one or more individuals were simply present in the garden. An actual CFE occurred if the photo/video indicated physical manipulation and/or consumption of human-cultivated food items. We installed three cameras in the mixed-crop garden, which is maintained by only one farmer and his family, and where cacao, coffee, bananas, ginger, chilies, short beans, and other fruits are grown. We installed two cameras along the edge of the cacao stand, facing the majority of the trees. We placed another camera within the garden, and moved it depending on the location of fruiting trees. Crops were available at this location throughout the study period. We installed five cameras along the edge of or within the monocrop garden where watermelon (available August–September), corn (available September–October), and wet rice (not available during the study period) are planted depending on the season. This garden is managed by several farmers depending on the current crop type. At the time of this project, there were zero to five farmers working at this location at any given time. We installed multiple cameras that monitored the same general area to have a better view of the areas macaques were known to access repeatedly because of the larger and more open nature of this garden. We installed and took down individual camera traps on different days, resulting in varying numbers of trap days for each unit. We moved one camera in the mixed-crop garden on two separate occasions depending on the presence/absence of macaque crop feeding activity in the area and location of fruiting trees, and we removed one camera in the monocrop garden because it was damaged. We based our decisions for camera trap location on the willingness of the farmers to participate in the project, farmer knowledge about how macaques access the gardens, and other practical considerations, e.g., wind, vegetation obstruction, desired camera height, equipment security, and so forth. We used a 32 GB Class 4 SDHC memory card in each camera and monitored the equipment monthly to ensure proper camera function and maintain memory storage.

When a camera was installed, we collected the following information: camera ID, GPS location, date, and height. We downloaded photos and videos in the field onto a laptop. We then inspected each photograph/video for presence of crop feeding wildlife or other relevant information, e.g., presence of people, dogs, and so on. Photos containing potentially crop feeding wildlife were assigned a number and placed in a digital folder based on the date downloaded. We catalogued all saved photographs/videos in a spreadsheet containing the following information: camera ID, date downloaded, photo ID, date, time, species, and number of individuals. We opportunistically collected data on human crop guarding from the camera traps in the mixed-crop garden. This was possible because of the smaller size of the garden, and therefore the increased likelihood that human presence would also be recorded by the camera traps. Guarding occurs when one or more humans are present in or near the garden and able to chase away animals that may attempt to enter.


We conducted semistructured interviews (LeCompte and Schensul 1999) between July 2014 and March 2015 to assess human perceptions of macaque crop feeding behavior and to explore potential strategies to reduce future damage. We also collected sociodemographic data (gender, age range, religion, job, ethnicity, and education level) about the participants and descriptive information about their gardens. Interview participants included six farmers (all men who ranged in age from 33 to 64 years) who agreed to have camera traps installed in their gardens and were originally selected using convenience sampling, a method by which interviewees are chosen based on their willingness and availability to participate. All participants belonged to the Bugis or Bugis–Makassar ethnic groups. Half of the farmers had stopped schooling after the equivalent of elementary school; the other half had completed high school before beginning their job as a farmer. We conducted interviews once with each farmer in the national language of Indonesia, Bahasa Indonesia, with the help of an assistant, and recorded interviews using a voice recorder. We discussed topics such as the frequency with which macaques engage in crop feeding, the resulting damage, existing deterrence methods, and potential future solutions (Table I).
Table I

Semistructured interview questions conducted with farmers in Bengo, South Sulawesi, Indonesia


Year of birth





Education level


How long have you lived here?

How big is your garden?

How long have you been farming here?

What types of food do you grow?

How far is your garden from the village?


What kinds of animals, if any, raid your garden?

How often do monkeys raid your garden?

What time of day do the monkeys raid?

Where did the monkeys come from?

How long have monkeys been raiding your garden/raiding in this area?

How many groups of monkeys live nearby?

What food items have monkeys eaten/damaged on your property?

How do you deter the monkeys from raiding your garden?

How often do you employ these deterrence methods?

How well do these deterrence methods work?

Does raiding occur more frequently now than in the past?

How does raiding affect your livelihood?

Do you think people should chase monkeys? Why or why not?

Data Analysis

We sorted and interpreted photographs and videos from the camera traps to reveal the frequency and timing of CFEs. We analyzed crop feeding data primarily using descriptive statistics. Using SPSS v. 22, we conducted chi-squared goodness of fit tests to evaluate differences in the number of crop feeding events across different time periods: 05:00–09:00, 10:00–14:00, and 15:00–18:00 h, and morning vs. afternoon. We considered results significant at P < 0.05. We transcribed and qualitatively analyzed interview responses through coding and identification of major themes (LeCompte and Schensul 1999).

Ethical Note

Farmers who agreed to have cameras installed on their property were instructed to react to CFEs as they normally would. If photographs/videos of people contained relevant information about wildlife crop feeding or guarding behavior, we made note of the event but deleted the image file to protect the privacy of the farmers. The research complied with protocols approved by the SDSU Institutional Animal Care and Use Committee, IACUC (APF #14-03-007R) and Institutional Review Board for Human Subjects (vIRB approval #1721094), and adhered to the legal requirements for foreigners conducting research in Indonesia.


CFEs Captured by Camera Traps

From July 2014 through March 2015, for a total of 231 days, eight camera traps from two gardens captured 351 photographs and videos with moor macaques in them (Table II; Figs. 4 and 5; Electronic Supplementary Material [ESM] Videos 1 and 2). The camera traps in the two locations recorded 34 actual and 33 potential macaque CFEs, totaling 67 CFEs. There were 31 actual and 26 potential CFEs, totaling 57 CFEs in the mixed-crop garden. There were 3 actual and 7 potential CFEs, totaling 10 CFEs in the monocrop garden. The recorded portions of CFEs lasted between 1 and 57 min (mean duration of 12.81 ± SD 12.63 min) in the mixed-crop garden and 31 to 74 min (mean duration of 54 ± SD 20.8 min) in the monocrop garden. The longest CFE occurred in an unguarded corn crop in the monocrop garden and was recorded by only one camera (B2), yielding a comparatively high number (N = 101) of photographs. The other cameras from the same location did not capture this CFE because they were too far away from where the corn was planted or had already been removed. Crop types fed on included banana, corn, watermelon, and cacao (both ripe and unripe pods; see ESM Video 1). Camera trap videos captured the cacao harvesting technique used by moor macaques: a combination of biting the pod stem and ripping the pod from the tree (see ESM Video 2).
Table II

Image data of moor macaques from eight camera traps located inside the mixed-crop and monocrop gardens from July 2014 through March 2015


Camera ID

Trap days

Total no. of saved photos and videos

Photos of macaques

Videos of macaques

Mixed-crop garden


















Monocrop garden


































Fig. 4

Camera trap photograph of two adult moor macaques climbing a bamboo pole to access a banana tree and an adult female with white pelage on the ground in the mixed-crop garden on December 19, 2014.

Fig. 5

Camera trap photograph of two adult moor macaques feeding on corn in the monocrop garden on October 24, 2014.

Crop Feeding Species and Crop Type Affected: Farmers’ Reports and Camera Trap Data Align

All farmers reported that macaques and wild pigs (Sus scrofa) feed on crops. Some farmers also reported that rats (N = 2) and deer (N = 1) occasionally engage in crop feeding. They reported that macaques and wild pigs cause the most severe crop damage. Camera trap data also showed that macaques and wild pigs travel into agricultural spaces most frequently (Table III). Camera traps recorded 89 photographs and 108 videos of wild pigs in the mixed-crop garden, and 10 photographs of wild pigs in the monocrop garden. They also recorded 10 videos of rats in the mixed-crop garden, and 2 photographs and 11 videos of civets (Viverra tangalunga) in the mixed-crop garden. For all nonprimate species, all CFEs were considered potential because of our inability to determine occurrence of actual crop consumption from photographic data. Farmers also reported that the following crops are damaged by wildlife in Bengo: cacao, bananas, corn, watermelon, rice, peanuts, durian, ginger leaves, cassava, young bamboo, green beans, oranges, jackfruit, squash, and long beans. They reported that chilies, tomatoes, and ginger roots are not eaten, but are sometimes damaged by wildlife, either accidentally or even intentionally. According to the farmers, monkeys will eat whatever they plant because they like to eat whatever humans like to eat. At the time of our study, the following crops were ripe and available: cacao (continuous), banana (continuous), corn (September–October), and watermelon (July–September). The camera traps recorded macaques feeding on cacao and bananas in the mixed-crop garden and watermelon and corn in the monocrop garden, in accordance with farmers’ reports.
Table III

Comparison of farmer response frequency and CFE frequency of crop feeding species from July 2014–March 2015

Wildlife speciesa (Indonesian name)

Total number of farmer responsesb (N = 6)

Frequency of CFEsc

Wild pig (babi hutan)



Macaque (monyet)



Rat (tikus)



Deer (rusa)



Civet (musang)



aFarmer-owned cattle occasionally consume cultivated food items but are not considered raiders by participating farmers

bResponse to interview question: What kinds of animals, if any, raid your garden?

cPotential and actual CFEs combined.

CFE Frequency and Timing: Farmers’ Reports and Camera Trap Data Do Not Align

Whereas farmers reported that macaque crop feeding events occur almost daily (range: 3 times/day to once a week), camera trap data show the occurrence of a CFE on only 53 of a total of 231 (23%) trap days (Figs. 6 and 7). Crop feeding rates, calculated from camera trap data, ranged between 0.08 ± SD 0.3 (monocrop garden) and 0.21 ± SD 0.51 (mixed-crop garden) CFEs per day.
Fig. 6

The number of days no moor macaque CFEs occurred vs. the number of days one or more CFEs occurred per month when crops were available in the mixed-crop garden between July 2014 and March 2015.

Fig. 7

The number of days no moor macaque CFEs occurred vs. the number of days one or more CFEs occurred per month when crops were available in the monocrop garden between July 2014 and March 2015.

Farmers reported that macaques feed on crops any time of day, whenever there is an opportunity, e.g., the garden is unguarded, because the “food is delicious and easy to access.” Farmers did not report the occurrence of crop feeding by macaques at night. Farmers also joked about when macaques crop feed: “If we are tired and want to nap, then the monkeys come!” or “If I think it has been a while since I have seen monkeys and they must be far away, that’s when they come!” While farmers reported that many CFEs occur in the morning (06:00–11:00 h), in contrast to the camera trap data, the most frequently reported time of day was dusk (17:00–18:00 h). Farmers typically leave their gardens or have already gone home for the day by dusk. Two farmers reported that monkeys will hide nearby a garden and wait for the farmers to leave before crop feeding.

According to photographic data from both locations, macaque CFEs (actual and potential combined) occurred throughout the day, but not in a uniform distribution (Chi-square goodness of fit: χ2 = 10.77, df = 2, P = 0.005). CFEs occurred significantly more often in the early and late afternoons (11:00–18:00 h) than in the morning (Chi-square goodness of fit: χ2 = 20.43, df = 1, P < 0.001; Fig. 8). In the mixed-crop garden, actual CFEs occurred primarily in the early afternoon (11:00–15:00 h). In the monocrop garden, actual CFEs occurred primarily in the morning or early afternoon. Five CFEs occurred at dusk in the mixed-crop garden, and no CFEs were observed at either location after 18:32 h.
Fig. 8

Frequency of moor macaque CFEs (potential and actual combined) by time of day by location (N = 67) between July 2014 and March 2015.

Entire Groups or Single Individuals? Farmers’ Reports and Camera Traps Align

Farmers (N = 4) report that entire groups and individuals may crop feed, depending on human presence. They report that entire groups crop feed successfully and cause more damage when crops are unguarded, but that single individuals may successfully crop feed while people are present in the garden and still cause damage. The camera traps recorded 27 individual CFEs and 30 group CFEs in the mixed-crop garden. There were seven individual CFEs and three group CFEs in the monocrop garden. There were 1–13 individuals captured in a single photograph or video.


Our results indicate that with proper attention to camera location and placement, camera traps can provide data on crop feeding species, crop type and phase targeted, daily and seasonal patterns of crop feeding activity, and whether groups and/or single individuals crop feed. Video data from camera traps can also provide critical information on the different harvesting techniques used by feeding wildlife, which could in turn inform subsequent studies that use the feeding traces method (see ESM Video 2). However, our study also revealed the limitations of this method. Although camera traps may allow for individual identification of large, solitary animals with distinct markings (see Wegge et al. 2004), we found identifying individual primates to be difficult and unreliable. We attempted to identify active crop feeders (individuals actively manipulating a food item in a garden) by age and sex class but quickly found that our identifications were likely biased toward individuals that are more conspicuous and therefore easier to identify, i.e., adult males. As has been noted in other camera trap studies, the detection probability of individuals is lower than large social groups (Treves et al. 2010), meaning that camera traps are also more likely to capture group vs. individual CFEs. For this reason, sample frequencies of group CFEs are likely to be more representative than individual CFEs. Although camera traps will likely capture evidence of a primate group’s presence in a garden, they cannot be relied on to provide consistent photographic evidence regarding actual manipulation and/or consumption of crop items. Therefore, it is likely that many of the potential CFEs in this study are actual CFEs. Finally, camera traps are not effective in helping to assess the severity of crop feeding damage. Additional methods such as examining food traces (Riley 2007) and talking directly with the farmers (Hill 1997; Linkie et al. 2007) are needed to assess post-crop feeding damage.

Previous research in Sumatra, Indonesia, and Uganda has found that farmer reports regarding crop feeding do not always coincide with results from other methods and farmers’ perceived risk of crop loss is often greater than observed crop damage (Linkie et al. 2007; Webber and Hill 2014). In our study, the fact that the placement of cameras, which was suggested by participating farmers, was successful in capturing CFEs indicates that farmers’ reports and camera trap data coincide. Farmers’ reports of specific crops targeted, the species that crop feed, and the most frequent and destructive crop feeding species, i.e., macaques and wild pigs, also coincide with camera trap data. However, we found that farmer reports differ from photographic data in identifying the timing of crop feeding events. According to farmers, macaques will raid any time of day, whenever there is an opportunity, suggesting that CFEs are more likely to occur when a garden is unguarded. Although farmers report that many CFEs occur in the morning, the most frequently reported time of day was dusk, when farmers are typically leaving their gardens. Our camera trap data show that the most actual CFEs occurred in the early afternoon, a time when most farmers reported being present in their gardens. This discrepancy might be explained by farmers’ beliefs that simply being present in the garden is an effective deterrence method. In contrast, guarding was found to be more effective for certain species when practiced by an actively patrolling, armed male (Hill and Wallace 2012), in combination with other active threats (Osborn and Hill 2005).

Our results also indicate a mismatch between farmers’ reports and camera trap data on the frequency of macaque CFEs. Although most farmers reported CFEs occurring multiple times per week, camera traps captured far fewer CFEs than would be expected based on this report. Farmers may associate more destructive crop feeding behavior with a conspicuous, large-bodied, diurnal species such as macaques simply because they encounter them more frequently (Linkie et al. 2007). During interviews, farmers shared stories about entire crops being destroyed in a single CFE. One farmer explained that he feels lucky when he harvests anything at all. In contrast, two interviewees reported that only two and three of their watermelons were damaged by macaques during the study period. Such variation in farmer reports may be due to changes in crop feeding patterns (Hockings et al. 2015) and timing of the interviews. Because primates negotiate the immediate decision to crop feed based on multiple, constantly changing factors, farmer perceptions of crop feeding are likely influenced accordingly (McLennan and Hill 2012). General differences in overall CFE frequencies at the two locations can be explained by the frequency and consistency of human guarding behavior and the varying size and amount of vegetation cover in the two gardens. The fact that these camera trap data are for a particular range of months while the participating farmers are likely integrating information from their experiences across a longer time period may help to explain some of the mismatches.

Evaluating the discrepancies between camera trap data and farmers’ reports is important to gain a broader understanding of the impact crop feeding has on human livelihoods in Sulawesi, Indonesia, and beyond. Adesina et al. (1994), for example, found that rice farmers in the Ivory Coast could not identify diseases on their plants, thus leading to a potentially serious misunderstanding about the cause of a diminished harvest. Comparative studies such as ours do not aim to prove farmers “wrong” or “right” about what is happening; rather, human perceptions of wildlife are important from a conservation perspective regardless of whether or not they align with evidence from other sources. If farmers perceive that primates are the biggest threat to their crops and their livelihood, they will act on that belief, potentially affecting the status of endangered populations through the use of lethal deterrence methods. It is therefore important to communicate the results of crop feeding research back to farmers (cf. Riley 2007). Sharing research results can help farmers have a better understanding of the effectiveness of their guarding techniques, and how they may need to alter them for the future. For example, our results suggest that farmers dealing with macaque crop feeding need to be more vigilant throughout the day and that passive guarding is an insufficient deterrent. Although it is not practical to expect one individual to actively guard for the entire day, as farmers often must attend to other tasks that may take them away from the garden, farmers may choose to coordinate active guarding responsibilities with other farmers or family members. This collaborative effort was demonstrated to be effective at the monocrop garden in this study, which experienced far fewer CFEs than the mixed-crop garden that was protected by only one farmer. Our camera traps also provided evidence that macaques consume unripe crops, e.g., cacao pods, indicating that certain crops may need to be guarded throughout their life cycle.

Comparisons of ethnographic and camera trap data are useful to researchers trying to protect endangered species and farmers trying to protect their harvest. Each method can provide new insight that may not have been previously considered. Camera traps provide information about crop feeding animal behavior, which cannot be observed directly, that may influence the use of certain deterrence methods. For example, if a farmer has never closely observed the harvesting techniques of various crop feeding species, videos from camera traps may provide detailed information about how different animals use different strategies while crop feeding. Farmers may choose to alter crop guarding techniques in light of this new information. In addition, camera traps can be used to systematically assess deterrence method success, a method that we suggest using in conjunction with discussions with farmers. In addition to the methodological and applied contributions of this study, our results contribute to our understanding of crop feeding as a flexible foraging behavior. Our findings show that macaques exhibit flexibility in the diurnal patterning of crop feeding, i.e., assessing risk on a day-to-day basis to determine what is palatable and available in the garden, whether farmers are present and guarding, and so forth, and crop phase consumed, i.e., consuming crops at various stages of maturity.

Lessons Learned and Conclusions

To help inform future studies of human–primate conflict, we offer the following recommendations. In all ethnographic interviews it is important to consider carefully the terms used in the interview schedule and how, if applicable, they will be translated into the local language. Terminology should be neutral, reflect no biases of the researcher, and should not influence farmer responses. For example, in future studies we would simply use the word eat instead of translating the already biased term crop raiding, reflecting the more recent use of foraging and feeding instead of raiding in the literature (Hill 2015). It is also extremely important to establish a trusting relationship with participants of the study and the larger community in which one works, prior to camera trap installment. If camera traps are hastily installed and/or poorly explained, conflicts between researchers and farmers may arise. For example, we had to remove a camera trap after only a couple of days when a farmer concluded that the camera we installed was attracting macaques to his watermelon garden. When a positive, collaborative relationship is established, farmers may be more likely to provide invaluable information on the past occurrence of crop feeding events within their gardens, including knowledge about where, how, and when wildlife most often enter. It is key to consult farmers about camera trap placement. Generally, it is also important to consider body size, i.e., smaller animals require cameras placed closer to the ground, and behavior of the target species, i.e., if certain areas of the habitat are frequented such as mineral licks, when deciding on camera trap height and placement (Acrenaz et al.2012).

We recommend using multiple camera traps to survey every hectare of agricultural area, installing them both along the edge of and within agricultural areas depending on surrounding vegetation and crop type targeted by crop feeding species. Cameras must be placed in areas fairly clear of surrounding vegetation to ensure clarity of the photographs and the ease and time with which data are catalogued (Acrenaz et al.2012). For example, a branch located too close to the camera unit that moves with the wind can result in thousands of photographs that must still be examined. The security boxes and locks used to secure the cameras to the substrates were effective and we had no issues with camera security, although this is a concern when using expensive equipment.

Our results demonstrate that camera traps represent a useful tool for studying many aspects of primate crop feeding behavior. However, this method is limited in what information it can provide about crop feeding when used alone. We suggest using camera traps in addition to other methods including ethnography, direct observation (if ethical and practical), fecal analyses, and feeding traces. Because both camera traps and local reports by farmers have limitations, we stress the importance of using multiple methods to triangulate results (see also Bessa et al. 2015; McLennan 2013).



We thank the Kementerian Negara Riset dan Teknologi Republik Indonesia (RISTEK) for permission to conduct research in Indonesia, and Universitas Hasanuddin (UNHAS) for permission to work in the Education Forest. We are grateful for financial support provided by a Henry Luce Award from the American Institute for Indonesian Studies (PI: E. P. Riley) and San Diego State University’s University Grant Program (PI: E. P. Riley). We are indebted to Dr. Ngakan Putu Oka and Pak Muhammad Restu for their sponsorship, and the UNHAS Forestry Faculty for providing GIS data for the Education Forest. We also thank Joanna Setchell, the guest editors, Kimberley Hockings, Matthew McLennan, Noemi Spagnoletti, and two anonymous reviewers for their helpful comments on the manuscript. Finally, we offer many thanks to our field assistants (Amir, Paisal, and Pak Pado), Cristina Sagnotti, and the farmers whose assistance and participation made this research possible.

Compliance with Ethical Standards

Conflict of Interest

The authors have no conflict of interest or competing financial interest to declare.

Supplementary material

Video 1

Shows an adult moor macaque consuming an unripe cacao pod (MPG 22138 kb)

Video 2

Demonstrates the cacao harvesting technique used by an adult male vmoor macaque. (MPG 23690 kb)


  1. Acrenaz, M., Hearn, A., Ross, J., Sollmann, R., & Wilting, A. (2012). Handbook for wildlife monitoring using camera-traps. Kota Kinabalu, Sabah, Malaysia: BBEC II Secretariat.Google Scholar
  2. Adesina, A., Johnson, D., & Heinrichs, E. (1994). Rice pests in the Ivory Coast, West Africa: farmers’ perceptions and management strategies. International Journal of Pest Management, 40(4), 293–299.CrossRefGoogle Scholar
  3. Agustinus. (2011). Daerah jelajah and potensi jenis tumbuhan pakan M. maura pada kelompok 1 di Hutan Pendidikan. Universitas Hasanuddin.Google Scholar
  4. Barber, C. H., Matthews, E., Brown, D., Brown, T. H., Currant, L., & Plume, C. (2002). State of the forest: Indonesia. Washington, DC: World Resources Institute.Google Scholar
  5. Bessa, J., Sousa, C., & Hockings, K. J. (2015). Feeding ecology of chimpanzees (Pan troglodytes verus) inhabiting a forest–mangrove–savanna–agricultural matrix at Caiquene Cadique, Cantanhez National Park, Guinea–Bissau. American Journal of Primatology, 77(6), 651–665.CrossRefPubMedGoogle Scholar
  6. Bezerra, B. M., Bastos, M., Souto, A., Keasey, M. P., Eason, P., Schiel, N., et al. (2014). Camera trap observations of nonhabituated critically endangered wild blonde capuchins, Sapajus flavius (formerly Cebus flavius). International Journal of Primatology, 35(5), 895–907.CrossRefGoogle Scholar
  7. Boyer-Ontl, K. M., & Pruetz, J. D. (2014). Giving the forest eyes: the benefits of using camera traps to study unhabituated chimpanzees (Pan troglodytes verus) in southeastern Senegal. International Journal of Primatology, 35(5), 881–894.CrossRefGoogle Scholar
  8. Brotcorne, F., Maslarov, C., Wandia, I. N., Fuentes, A., Beudels‐Jamar, R. C., & Huynen, M. C. (2014). The role of anthropic, ecological, and social factors in sleeping site choice by long‐tailed macaques (Macaca fascicularis). American Journal of Primatology, 76(12), 1140–1150.CrossRefPubMedGoogle Scholar
  9. Farris, Z. J., Karpanty, S. M., Ratelolahy, F., & Kelly, M. J. (2014). Predator–primate distribution, activity, and co-occurrence in relation to habitat and human activity across fragmented and contiguous forests in northeastern Madagascar. International Journal of Primatology, 35(5), 859–880.CrossRefGoogle Scholar
  10. Fedigan, L. M. (2010). Ethical issues faced by field primatologists: asking the relevant questions. American Journal of Primatology, 72(9), 754–771.CrossRefPubMedGoogle Scholar
  11. Fooden, J. (1969). Taxonomy and evolution of the monkeys of Celebes. Basel: S. Karger.Google Scholar
  12. Galvis, N., Link, A., & Di Fiore, A. (2014). A novel use of camera traps to study demography and life history in wild animals: a case study of spider monkeys (Ateles belzebuth). International Journal of Primatology, 35(5), 908–918.CrossRefGoogle Scholar
  13. Hill, C. M. (1997). Crop-raiding by wild vertebrates: the farmer’s perspective in an agricultural community in western Uganda. International Journal of Pest Management, 43(1), 77–84.CrossRefGoogle Scholar
  14. Hill, C. M. (2015). Perspectives of “conflict” at the wildlife–agriculture boundary: 10 years on. Human Dimensions of Wildlife, 20(4), 296–301.CrossRefGoogle Scholar
  15. Hill, C. M., & Wallace, G. E. (2012). Crop protection and conflict mitigation: reducing the costs of living alongside non-human primates. Biodiversity and Conservation, 21, 2569–2587.CrossRefGoogle Scholar
  16. Hockings, K. J., & Humle, T. (2009). Best practice guidelines for the prevention and mitigation of conflict between human and great apes. Gland, Switzerland: IUCN/SSC Primate Specialist Group (PSG).CrossRefGoogle Scholar
  17. Hockings, K. J., Anderson, J., & Matsuzawa, T. (2009). Use of wild and cultivated foods by chimpanzees at Bossou, Republic of Guinea: feeding dynamics in a human-influenced environment. American Journal of Primatology, 71, 636–646.CrossRefPubMedGoogle Scholar
  18. Hockings, K. J., McLennan, M. R., Carvalho, S., Ancrenaz, M., Bobe, R., et al. (2015). Apes in the anthropocene: flexibility and survival. Trends in Ecology & Evolution, 30(4), 215–222.CrossRefGoogle Scholar
  19. ILO (International Labor Organization) (2012). Sectoral country profile—Indonesia.
  20. IUCN (2008). The IUCN red list of threatened species. Available at:
  21. Krief, S., Cibot, M., Bortolamiol, S., Seguya, A., Krief, J.-M., & Masi, S. (2014). Wild chimpanzees on the edge: nocturnal activities in croplands. PLoS ONE, 9(10), e109925.CrossRefPubMedPubMedCentralGoogle Scholar
  22. LeCompte, M. D., & Schensul, J. J. (1999). Analyzing and interpreting ethnographic data. Walnut Creek, CA: Altamira Press.Google Scholar
  23. Linkie, M., Dinata, Y., Nofriantao, A., & Leader-Williams, N. (2007). Patterns and perceptions of wildlife crop raiding in and around Kerinci Seblat National Park, Sumatra. Animal Conservation, 10, 127–135.CrossRefGoogle Scholar
  24. Matsumura, S. (1991). A preliminary report of the ecology and social behavior of moor macaques (Macaca maurus) in Sulawesi, Indonesia. Kyoto University Overseas Research Report of Studies on Asian Nonhuman Primates, 8, 27–41.Google Scholar
  25. McKinney, T. (2011). The effects of provisioning and crop-raiding on the diet and foraging activities of human-commensal white-faced capuchins (Cebus capucinus). American Journal of Primatology, 73, 439–448.CrossRefPubMedGoogle Scholar
  26. McLennan, M. R. (2013). Diet and feeding ecology of chimpanzees (Pan troglodytes) in Bulindi, Uganda: foraging strategies at the forest-farm interface. International Journal of Primatology, 34, 585–614.CrossRefGoogle Scholar
  27. McLennan, M. R., & Hill, C. M. (2012). Troublesome neighbours: Changing attitudes towards chimpanzees (Pan troglodytes) in a human-dominated landscape in Uganda. Journal for Nature Conservation, 20(4), 219–227.CrossRefGoogle Scholar
  28. Naughton-Treves, L. (1997). Farming the forest edge: vulnerable places and people around Kibale National Park, Uganda. Geographical Review, 87(1), 27–46.CrossRefGoogle Scholar
  29. O’Connell, A. F., Nichols, J. D., & Karanth, K. U. (Eds.). (2011). Camera traps in animal ecology: methods and analyses. Tokyo: Springer.Google Scholar
  30. Okamoto, K., Matsumura, S., & Watanabe, K. (2000). Life history and demography of wild moor macaques (Macaca maurus): summary of ten years of observations. American Journal of Primatology, 52, 1–11.CrossRefPubMedGoogle Scholar
  31. Osborn, F. V., & Hill, C. M. (2005). Technique to reduce crop loss: human and technical dimensions in Africa. In R. Woodroffe, S. Thirgood, & A. Rabinowitx (Eds.), People and wildlife: conflict or coexistence? (pp. 72–85). Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  32. Pebsworth, P. A., & LaFleur, M. (2014). Advancing primate research and conservation through the use of camera traps: introduction to the special issue. International Journal of Primatology, 35(5), 825–840.CrossRefGoogle Scholar
  33. Priston, N. E., Wyper, R. M., & Lee, P. C. (2012). Buton macaques (Macaca ochreata brunnescens): crops, conflict, and behavior on farms. American Journal of Primatology, 74(1), 29–36.CrossRefPubMedGoogle Scholar
  34. Richard, A. F., Goldstein, S. J., & Dewar, R. E. (1989). Weed macaques: evolutionary implications of macaque feeding ecology. International Journal of Primatology, 10(6), 569–594.CrossRefGoogle Scholar
  35. Riley, E. P. (2007). The human-macaque interface: conservation implications of current and future overlap and conflict in Lore Lindu National park, Sulawesi, Indonesia. American Anthropologist, 109(3), 473–484.CrossRefGoogle Scholar
  36. Riley, E. P. (2010). The endemic seven: Four decades of research on the Sulawesi macaques. Evolutionary Anthropology, 19, 22–36.CrossRefGoogle Scholar
  37. Sagnotti, C. (2013). Diet preference and habitat use in relation to reproductive states in females of a wild group of Macaca maura inhabiting Karaenta forest in South Sulawesi. Master’s dissertation, Hasanuddin University, Makassar, Sulawesi.Google Scholar
  38. Seiler, N., & Robbins, M. M. (2016). Factors influencing ranging on community land and crop raiding by mountain gorillas. Animal Conservation, 19(2), 176–188.CrossRefGoogle Scholar
  39. Siex, K. S., & Struhsaker, T. T. (1999). Colobus monkeys and coconuts: a study of perceived human–wildlife conflicts. Journal of Applied Ecology, 36, 1009–1020.CrossRefGoogle Scholar
  40. Strum, S. C. (2010). The development of primate raiding: implications for management and conservation. International Journal of Primatology, 31, 133–156.CrossRefPubMedPubMedCentralGoogle Scholar
  41. Supriatna, J., Froehlich, J. W., Erwin, J. M., & Southwick, C. H. (1992). Population, habitat and conservation status of Macaca maurus, Macaca tonkeana and their putative hybrids. Tropical Biodiversity, 1(1), 31–48.Google Scholar
  42. Tan, C. L., Yang, Y., & Niu, K. (2013). Into the night: camera traps reveal nocturnal activity in a presumptive diurnal primate, Rhinopithecus brelichi. Primates, 54(1), 1–6.CrossRefPubMedGoogle Scholar
  43. Tobler, M. W., Carrillo-Percastegui, S. E., Leite Pitman, R., Mares, R., & Powell, G. (2008). An evaluation of camera traps for inventorying large- and medium-sized terrestrial rainforest mammals. Animal Conservation, 11(3), 169–178.CrossRefGoogle Scholar
  44. Treves, A., Mwima, P., Plumptre, A. J., & Isoke, S. (2010). Camera-trapping forest–woodland wildlife of western Uganda reveals how gregariousness biases estimates of relative abundance and distribution. Biological Conservation, 143(2), 521–528.CrossRefGoogle Scholar
  45. Watanabe, K., & Matsumura, S. (1996). Social organization of moor macaques, Macaca maurus, in the Karaente Nature Reserve, South Sulawesi, Indonesia. In T. Shotake & K. Wada (Eds.), Variations in the Asian macaques (pp. 147–162). Toyko: Tokai University Press.Google Scholar
  46. Webber, A. D., & Hill, C. M. (2014). Using participatory risk mapping (PRM) to identify and understand people’s perceptions of crop loss to animals in Uganda. PLoS ONE, 9(7), e102912.CrossRefPubMedPubMedCentralGoogle Scholar
  47. Wegge, P., Chiranjibi, P. P., & Jnawali, S. R. (2004). Effects of trapping effort and trap shyness on estimates of tiger abundance from camera trap studies. Animal Conservation, 7, 251–256.CrossRefGoogle Scholar
  48. Woodroffe, R., Thirgood, S., & Rabinowitz, A. (Eds.). (2005). People and wildlife: conflict or co-existence? Cambridge: Cambridge University Press.Google Scholar

Copyright information

© Springer Science+Business Media New York 2016

Authors and Affiliations

  1. 1.San Diego State UniversitySan DiegoUSA

Personalised recommendations