Nutritional Characteristics of Wild and Cultivated Foods for Chimpanzees (Pan troglodytes) in Agricultural Landscapes
- First Online:
Primate habitats are being transformed by human activities such as agriculture. Many wild primates include cultivated foods (crops) in their diets, calling for an improved understanding of the costs and benefits of crop feeding. We measured the macronutrient and antifeedant content of 44 wild and 21 crop foods eaten by chimpanzees (Pan troglodytes schweinfurthii) in a mosaic habitat at Bulindi, Uganda, to evaluate the common assertion that crops offer high nutritional returns compared to wild forage for primates. In addition, we analyzed 13 crops not eaten at Bulindi but that are consumed by chimpanzees elsewhere to assess whether nutritional aspects explain why chimpanzees in Bulindi ignored them. Our analysis of their wild plant diet (fruit, leaves, and pith) corresponds with previous chemical analyses of primate plant foods. Compared to wild food equivalents, crops eaten by the chimpanzees contained higher levels of digestible carbohydrates (mainly sugars) coupled with lower amounts of insoluble fiber and antifeedants. Cultivated fruits were relatively nutritious throughout the ripening process. Our data support the assumption that eating cultivated foods confers energetic advantages for primates, although crops in our sample were low in protein and lipids compared to some wild foods. We found little evidence that crops ignored by the chimpanzees were less nutritious than those that they did eat. Nonnutritional factors, e.g., similarity to wild foods, probably also influence crop selection. Whether cultivated habitats can support threatened but flexible primates such as chimpanzees in the long term hinges on local people’s willingness to share their landscape and resources with them.
KeywordsAgroecosystems Crop foraging Cultivars Dietary flexibility Human-dominated landscapes Nutritional ecology
Conversion of forests for subsistence and commercial agriculture is continuing apace throughout the world’s most biodiverse regions (Gibbs et al. 2010; Laurance et al. 2014; Tilman et al. 2001). While agricultural expansion erodes wild foods, ecologically and behaviorally flexible species may exploit these new environments and their novel foods (McLennan and Hockings 2014). Crop feeding by wildlife (commonly termed crop raiding) receives considerable attention because it can cause conservation conflicts through impacts on local livelihoods (Conover 2002; Hill 1997; MacKenzie and Ahabyona 2012; Redpath et al. 2013). Understanding the attractiveness of crops, i.e., cultivated foods, to wildlife thus has strong relevance for conservation management (Dostaler et al. 2011; Osborn 2004; Rode et al. 2006).
Nonhuman primates (hereafter primates) feature prominently in the literature on crop damage by wild tropical vertebrates (Paterson and Wallis 2005). The propensity of generalist primate foragers to exploit areas of human settlement and cultivation is well documented, e.g., members of Macaca, Papio, and Chlorocebus in Asia and Africa (Brennan et al. 1985; Hill 2000; Priston and McLennan 2013; Strum 2010), and Alouatta, Cebus, and Sapajus in the Neotropics (Bicca-Marques and Calegaro-Marques 1994; McKinney 2011; Spagnoletti et al. 2017). However, with the expansion of agroecosystems in primate habitats a broad range of other taxa have been found to eat crops (Estrada et al. 2012). These include species not usually regarded as generalist, omnivorous feeders, e.g., Trachypithecus vetulus (Nijman 2012), Procolobus kirkii (Nowak and Lee 2013), and Gorilla beringei beringei (Seiler and Robbins 2016), suggesting that more “specialist” primates can also respond flexibly to agricultural encroachment, albeit if only in the short term (Nowak and Lee 2013).
Humans have selected agricultural foods to be easily digestible, energy rich, and low in plant secondary compounds that impede digestion or include harmful toxins (Milton 1999). Including crops in the diet has far-reaching consequences for primates. Frequent crop consumption is associated with major changes in activity budgets, with primates typically spending more time resting and in social behavior and less time traveling and foraging, apparently due to energetic benefits of crops that allow metabolic demands to be met sooner, e.g., Papio cynocephalus (Altmann and Muruthi 1988), P. anubis (Eley et al. 1989; Strum 2010; Warren et al. 2011), and Chlorocebus aethiops (Saj et al. 1999). Crop feeding has further been linked to reduced physiological stress (P. anubis: Lodge et al. 2013) and possibly enhanced immune responses (Colobus guereza: Chapman et al. 2006; P. anubis: Eley et al. 1989; Weyher et al. 2006). Despite significant costs, i.e., injury or mortality from pest management, frequent crop consumption may confer life history and reproductive advantages to primates, e.g., improved body condition and increased adult weight, reduced infant mortality, shorter interbirth intervals, and earlier reproductive onset (Macaca fuscata: Sugiyama and Ohsawa 1982; P. anubis: Lodge et al. 2013; Strum 2010; Warren et al. 2011). Even so, elevated serum insulin and cholesterol levels in refuse foraging P. anubis and P. cynocephalus have been reported (Kemnitz et al. 2002).
High nutritional returns of crops compared to wild forage are usually assumed. Few studies have quantified nutritional characteristics of both wild and cultivated foods in diets of crop foraging primates. In one study, cultivated cacao (cocoa) eaten by Macaca tonkeana was higher in digestible carbohydrates and lower in insoluble fiber compared to wild fruits in their diet (Riley et al. 2013). Similarly, maize and potato eaten by Papio anubis had markedly lower insoluble fiber and thus greater digestibility compared to many of their wild plant foods (Forthman-Quick and Demment 1988).
Chimpanzees (Pan troglodytes) offer a useful model for examining nutritional attributes of “natural” vs. cultivated foods in diets of wild primates. Although varying by habitat and season, their natural diets are consistently dominated by ripe fruits that they seek out even when scarce, leading some authors to label them ripe fruit specialists (Ghiglieri 1984; Watts et al. 2012; Wrangham et al. 1998). In general, chimpanzee food selection reflects a preference for higher levels of macronutrients, particularly easily digestible sugars, and lower amounts of insoluble fiber and digestion-inhibiting antifeedants, i.e., polyphenols and condensed tannins, which characterize ripe fruit (Hohmann et al. 2010; Matsumoto-Oda and Hayashi 1999; Remis 2002; Reynolds et al. 1998; Sommer et al. 2011; Wrangham et al. 1998). Unripe fruits may be eaten but are usually lower in sugar and higher in fiber and antifeedants than ripe ones (Houle et al. 2014; Wrangham and Waterman 1983), although chimpanzees seem to tolerate moderate levels of tannins (Remis 2002; Reynolds et al. 1998; Sommer et al. 2011). Fibrous piths and stems provide an additional source of carbohydrate energy, particularly during fruit shortages (Matsumoto-Oda and Hayashi 1999; Wrangham et al. 1991, 1998). Young leaves are probably selected for high protein content (Carlson et al. 2013; Takemoto 2003), which is generally low in fruits. High concentrations of tannins in leaves are avoided (Takemoto 2003). Overall, chimpanzees are considered to have high-quality diets (Conklin-Brittain et al. 1998).
Chimpanzees are found in habitats transformed by agriculture across their geographic range in equatorial Africa (Hockings and McLennan 2012, 2016). Crop feeding by these great apes reflects their species-typical preference for ripe sugary fruits, though a variety of nonfruit crops are also exploited (Hockings and McLennan 2012). At the borders of large uncultivated habitats, chimpanzees target particular crops in adjacent farmland, e.g., mango and sugarcane around Budongo Forest Reserve, Uganda (Tweheyo et al. 2005) and maize and banana around Kibale National Park, Uganda (Krief et al. 2014; Naughton-Treves et al. 1998). In some areas, chimpanzees survive in mosaic habitats within agroecosystems (Bessa et al. 2015; McLennan 2008) where crops can become integral to their feeding ecology (Bossou, Guinea: Hockings et al. 2009; Bulindi, Uganda: McLennan 2013).
Assimilation of cultivated foods into chimpanzee diets is a dynamic process (Takahata et al. 1986), and intriguing differences exist among populations regarding which crops are eaten and which are ignored, even where local crop assemblages are similar (McLennan and Hockings 2014). The extent to which nutritional factors drive chimpanzee foraging decisions in cultivated habitats, including which crops they exploit, remains unknown.
We here examined nutritional composition in a broad selection of wild and cultivated foods consumed by a population of wild East African chimpanzees (Pan troglodytes schweinfurthii) inhabiting a farm–forest mosaic habitat in Bulindi, Uganda. Our primary objective was to identify potential nutritional benefits of eating crops over wild foods for these chimpanzees. We first examined macronutrients and antifeedants in major categories of wild foods (fruits, piths, and leaves) to characterize nutritional properties of their natural diet. We then compared wild and cultivated foods eaten by these chimpanzees. A secondary aim was to determine if nutritional factors explain why they ignore certain crops exploited by one or more chimpanzee populations elsewhere. Thus, we compared nutrient and antifeedant concentrations in crops eaten and not eaten. We predicted that crops eaten would offer nutritional advantages over wild food equivalents, i.e., by being higher in digestible carbohydrates such as sugars and lower in insoluble fiber and antifeedants. We also predicted that crops fed on by the chimpanzees would likewise offer nutritional advantages over those crops that they ignored.
Bulindi (1°28′N, 31°28′E) is situated in Hoima District, western Uganda, midway between the Budongo and Bugoma Forest Reserves: two main forest blocks with >500 chimpanzees each (Plumptre et al. 2010). These reserves are separated by ca. 50 km. The intervening landscape is densely populated by people (>150 persons per km2; Uganda Bureau of Statistics 2014) and dominated by subsistence and commercial agriculture (McLennan and Hill 2015). A genetic survey revealed that 260–320 chimpanzees from 9 or more resident “communities” inhabit small fragments of unprotected forest across this cultivated landscape (McCarthy et al. 2015). Chimpanzees in Bulindi represent one of these communities. Local farmers practice a combination of subsistence farming and cash-cropping. Staple food crops include cassava, potato, maize, and groundnuts, while major cash crops are tobacco, rice, and sugarcane (McLennan and Hill 2015). Domestic fruits including mango, jackfruit, banana, and papaya are grown around homes. Since the 1990s, forest clearance for timber and farming has been extensive throughout the landscape separating Budongo and Bugoma (McLennan and Hill 2015; Mwavu and Witkowski 2008; Twongyirwe et al. 2015). Primates including chimpanzees are not traditionally hunted for food in western Uganda, which enables them to persist in modified habitats near people. Consumption of agricultural crops by chimpanzees occurs throughout this region (McLennan 2008).
Although the chimpanzees’ diet is dominated numerically by wild plants, they forage frequently on cultivated foods in gardens and by homes, as well as from abandoned or naturalized sources (McLennan 2013; McLennan and Hockings 2014). Local tolerance of chimpanzees varies from person-to-person but crop loss to the apes is considered a worsening problem by many villagers (McLennan and Hill 2012). The chimpanzees have never been actively provisioned.
Plant Food Collection
We collected plant foods during January–April 2014, September–November 2014, March–June 2015, and October–December 2015. The chimpanzee diet at Bulindi has been well studied using a combination of indirect methods (fecal analysis and feeding trace evidence) and direct observation. At least 139 different plant food items from 103 identified species have been recorded eaten to date (McLennan 2013 and unpubl. data). During daily tracking we observed feeding behavior opportunistically and did not record feeding rates. We avoided observing chimpanzees feeding on crops, unless these were from abandoned or naturalized sources, though we sometimes encountered them foraging in gardens. During observations, we paid careful attention to food items selected and how these were processed. Similarly, we examined chimpanzee feeding traces carefully to determine the part consumed. We confirmed that the chimpanzees ate certain fruits by fecal analysis. Methods used to analyze chimpanzee feeding traces and fecal samples are detailed in McLennan (2013).
We collected 78 plant foods for this study including 44 wild and 34 cultivated items (Appendix Tables II–IV). Wild foods are predominantly native plants that are not usually planted or domesticated by humans; exceptions in the sample include native figs (Ficus natalensis and F. thonningii) that are sometimes planted around homes and paper mulberry (Broussonetia papyrifera), an exotic shrub introduced previously into nearby Budongo Forest. Its occurrence in Bulindi is presumably the result of dispersal by birds; thus we treated it as wild. Cultivated foods (synonymous with crops, cultivars, or cultigens) are domesticated plants selectively bred by people; several in our sample also occur as naturalized specimens in Bulindi, e.g., guava and tamarillo (see Spencer and Cross 2007 for a discussion of cultivated vs. wild plant definitions).
We collected three major categories of plant food: fruits (ripe and unripe), leaves (young and emerging), and piths (terrestrial herbaceous stems and leaf petioles or stems). Although chimpanzees usually ate fruits ripe, they consumed some fruits throughout the ripening process, including fully unripe. For eight such fruits, we collected ripe and unripe samples. Though the precise stage of maturity varied (Houle et al. 2014), unripe fruits were small compared to mature fruits, firm, and/or with green or pale skin and pulp. We considered leaf petioles and stems piths when the manner of processing by chimpanzees corresponded to that of terrestrial stem feeding rather than leaf feeding, i.e., leaves discarded and only the inner part of the petiole/stem eaten. Other minor food categories, e.g., seeds, tubers, flowers, and cambium, were represented by one or two foods only. Life forms of plants sampled included trees, shrubs, climbers and vines, herbs, and grasses.
Plants collected included both commonly and occasionally eaten items (as indicated by fecal analysis, direct observation, and feeding trace records; McLennan 2013). Thirteen items were crops grown at Bulindi that are reportedly eaten by one or more populations of wild chimpanzees elsewhere (Hockings and McLennan 2012), including several eaten by nearby chimpanzee communities in Hoima District (M. McCarthy, pers. comm.), but for which no evidence suggests Bulindi chimpanzees eat them (Appendix Table IV). An exception is tamarillo fruit, for which feeding traces were twice attributed to chimpanzees in 2007 (McLennan 2013). However, no further evidence has suggested the chimpanzees eat tamarillo, e.g., absence of seeds in feces and absence of feeding traces at numerous naturalized tamarillo shrubs in the forest. Thus, we consider it very unlikely that chimpanzees ate tamarillo in the present study. For all other crops not eaten (including fruits such as pineapple and staple food crops such as cassava and maize cob), there has been no evidence of consumption by the chimpanzees since research was initiated. Moreover, local farmers maintain chimpanzees do not eat these crops in Bulindi (McLennan and Hill 2012).
Wherever possible, we collected samples from actual plants that chimpanzees ate from, including intact items from feeding patches after chimpanzees fed or that fell to the ground incidentally while they fed, e.g., a fruiting or leafing branch, and partially eaten items such as large cultivated fruits, e.g., jackfruit, which are often not consumed in their entirety. We collected all partially eaten items in the same morning that chimpanzees ate them. Otherwise, we collected samples from conspecific plants showing a similar phenophase. We collected intact cultivated foods from local gardens with permission. For several crops, we failed to obtain a sample in the desired stage of maturity from local gardens, so we bought them at a market in Hoima town, 12 km from Bulindi, assuming they were of similar quality to ones consumed by the chimpanzees. Where possible, we collected samples from multiple plants of the same species.
We collected food samples in plastic bags and processed them on the same day to include only parts fed on by chimpanzees. For example, we removed outer layers of piths, leaving only the soft inner part. We removed fruit seeds and tough skins, but retained the soft fruit skins if these were normally ingested. Fecal analysis showed that chimpanzees sometimes chewed the soft bean-like seeds of Parkia filicoidea, suggesting they obtained nutrients from them. Occasionally, they ate immature seeds and pods of cultivated beans (Phaseolus vulgaris); thus, we retained a portion of the seed content for these two fruits. We took samples from crops not eaten by the chimpanzees from parts likely to be most palatable, e.g., soft fruit pulp, inner portion of piths.
After processing, we dried samples at 50–55°C using a Shef® food dehydrator. Once dry, we weighed samples, stored them in plastic bags with silica gel, and shipped 5–15 g dry weight per item to University of Hamburg, Germany, for biochemical analyses.
We analyzed samples for macronutrients and antifeedants via standard methods (for reviews of laboratory procedures see Ortmann et al. 2006; Rothman et al. 2012). We ground samples in a Retsch mill to a homogenous powder and dried to 50°C in the laboratory overnight. We estimated nutrient concentrations on a dry matter (DM) basis. We measured total nitrogen (TN) by the Kjeldahl method (Association of Official Analytical Chemists 1990) and determined crude protein (CP) as TN × 6.25. While this conversion factor should be adapted for different food categories, especially tropical fruits (Milton and Dintzis 1981), we use it here to allow for comparison with other studies. As CP does not necessarily reflect protein available for digestion (Rothman et al. 2008; Wallis et al. 2012), we also assessed soluble protein via the photometric BioRad assay after extraction of plant material with 0.1 N NaOH for 15 h at room temperature. A meta-analysis of primate leaf selection found that soluble protein had a greater effect on selection than TN (or CP), suggesting these protein measures differ in ecological relevance (Ganzhorn et al. 2017). Even so, TN and soluble protein correlated highly in our sample of foods (Pearson’s correlation: r = 0.593, N = 78, P < 0.0001). Further, TN in leaves from Uganda correlated well with available protein (Wallis et al. 2012). Therefore, we used CP as our measure of protein in the analysis, but we also report soluble protein in the appendices.
We analyzed neutral detergent fiber (NDF) and acid detergent fiber (ADF) using an ANKOM fiber analyzer (Van Soest et al. 1991). NDF represents the insoluble fiber (hemicellulose, cellulose, and lignin), with ADF representing the cellulose and lignin fractions; hemicellulose (HC) is thus determined by weight difference (NDF – ADF). We determined fat content (lipids) using ether extract and measured ash via combustion (Rothman et al. 2012). We extracted soluble carbohydrates and procyanidin (condensed) tannins with 50% methanol and determined soluble sugars as the equivalent of galactose after acid hydrolyzation of the methanol extract.
We measured concentrations of procyanidin tannins as equivalents of quebracho tannin using the buthanol method, and measured total phenolics (simple phenols and polyphenols) using the Folin–Ciocalteus reagent (Stolter et al. 2006). Tannins inhibit digestion by making some nutrients, e.g., proteins, unavailable for digestion. Simple phenols are small molecules that enter the cell and can act as poisons; these components are volatile and are likely to be lost during the drying process. We based analyses of polyphenols on water extracts. Standard chemical assays of these components represent poor proxies of their actual biological relevance, as both groups of chemicals comprise a plethora of substances with differing properties (Rothman et al. 2009). Nevertheless, we used these analyses to allow comparisons with other studies.
With the exception of ME (expressed as kcal/100 g DM), we present all values as % DM.
We examined differences between food categories in CP, lipids, soluble sugars, TNC, fiber (NDF and ADF), polyphenols and tannins, and ME. Because of unequal samples sizes and nonnormality of some distributions, we used nonparametric statistics. We compared nutritional attributes of major wild food categories (ripe fruits, piths, young leaves) using Kruskal–Wallis ANOVAs followed by Dunn–Bonferroni pairwise comparisons. We compared ripe and unripe samples from fruits that chimpanzees ate in both maturity stages using Wilcoxon signed rank tests. We used Mann–Whitney tests to assess differences between 1) crops eaten and wild food equivalents and 2) cultivated fruits eaten and not eaten; reported z-scores inform about the group with the lowest distribution. We compared wild and cultivated foods for fruit and pith only because the chimpanzees ate leaves from one crop only (yam leaves; not collected for this study). We used one-sample Wilcoxon signed rank tests to assess differences between individual nonfruit crops that were not eaten at Bulindi (but eaten elsewhere) and medians of wild food equivalents.
To control for multiple testing we applied a Holm–Bonferroni sequential adjustment to P-values in all groups of tests. This procedure is considered more powerful than the conventional Bonferroni approach, while still controlling the family-wise type I error (Abdi 2010). Nevertheless, we also report unadjusted P-values in some tests in which the adjustment was likely too conservative given small sample sizes, but these should be interpreted with caution. We performed statistical analyses using SPSS version 23 (SPSS Inc., Chicago, IL, USA) and set statistical significance at P < 0.05; all tests were two tailed.
This research involving wild chimpanzees was noninvasive and adhered strictly to the legal requirements of Uganda and to ethics guidelines detailed by the Association for the Study of Animal Behaviour (UK) and the American Society of Primatologists Principles for the Ethical Treatment of Nonhuman Primates. The study was approved by the Uganda National Council for Science and Technology, the President’s Office, and the Uganda Wildlife Authority.
Wild Foods Compared
Wild and Cultivated Foods Compared
Chemical properties of ripe and unripe fruits of eight species (six crops and two wild species) eaten by chimpanzees at Bulindi in both stages of maturity during this study (2014–2015)
Because few wild unripe fruits were analyzed, we could not compare unripe fruits from wild and cultivated sources. However, no significant differences were apparent between wild ripe fruits and cultivated unripe fruits eaten by the chimpanzees (Electronic Supplementary Material [ESM] Fig. S1).
Crops Eaten and Not Eaten Compared
Papaya leaf, which the chimpanzees did not eat, was higher in CP (29.9% DM) than all 10 wild young leaf species that they did eat (Mdn = 22.7%; one-sample Wilcoxon signed rank test: P = 0.040 with Holm–Bonferroni adjustment). In fact, papaya leaf was highest in protein of all 78 foods analyzed (appendices). Papaya leaves were also low in polyphenols (0.78%) compared to most wild leaf foods (Mdn = 1.48%), though this difference was nonsignificant after adjustment (unadjusted P = 0.028; adjusted P = 0.196). While tannins were not found in papaya leaf, they were present in 7 of 10 wild young leaf foods. Papaya pith, also not eaten, was lower in fiber (NDF = 18.35%, ADF = 13.42%) than all 7 wild piths analyzed (Mdn = 37.64% and 23.37%, respectively), while its ME content was highest (305.14 kcal/100 g vs. 264.70 kcal/100 g [Mdn] for wild piths). A second cultivated pith not eaten at Bulindi (rice) was lower in polyphenols (0.15%) than all wild piths analyzed (Mdn = 0.61%). Only unadjusted P-values were significant (P = 0.018 in each case). Notably, both rice and papaya pith had considerably higher levels of CP (13.4% and 14.2%, respectively) than the four cultivated piths that the chimpanzees did eat (1.8–8.4%; Appendix Tables III and IV). Conversely, sugar concentrations in rice and papaya pith were lower and more similar to those in wild pith foods. The fiber and polyphenol content was overall similar in cultivated piths eaten and not eaten. None of the cultivated piths contained tannins.
Four additional crops analyzed—not eaten by the chimpanzees—are staple foods for local people: cassava and sweet potato (tubers), maize cob (caryopsis), and ground nuts (seed crop). There were no wild food equivalents for these in our sample. These crops were generally low in soluble sugars (Appendix Table IV). However, cassava and maize cob in particular are high in starch (United States Department of Agriculture 2016), which we did not assay. Fiber concentrations in cassava, maize cob, and ground nuts were within the range of other nonfruit items eaten by the chimpanzees. However, sweet potato was high in NDF (58.9%)—almost all hemicellulose. The fiber content of cassava and maize similarly comprised mostly hemicellulose. Ground nuts were rich in protein and contained an exceptionally high lipid concentration. All staple food crops were low in antifeedants.
Our results support the common assertion that crops offer certain nutritional advantages over wild plants for primates in human-modified environments. Chimpanzees within the forest–agricultural mosaic in Bulindi supplement a “natural” diet with various cultivated foods that compared to wild food equivalents, and in accord with our prediction, had higher levels of easily digestible carbohydrates (mainly sugars) coupled with reduced amounts of insoluble fiber and antifeedants. Conversely, however, crops eaten by the chimpanzees were not a good source of protein or lipids relative to some wild foods, which may be true of cultivars generally (Milton 1999). In addition, compared to crops, wild plants may contain higher concentrations of essential micronutrients (vitamins and minerals) that we did not assay here (Milton 1999; cf. Rode et al. 2006). Whether crop feeding primates balance their nutrient intake, e.g., with protein or lipid-rich wild foods, is largely unknown. However, Johnson et al. (2013) demonstrated nutrient balancing in a female Papio ursinus, which included exotic plants and other “human-derived” foods in its diet. Because we did not measure feeding time or food intake by the chimpanzees, we could not estimate nutrient intake. Thus, further research is needed to determine how the chimpanzees prioritize and regulate nutrient intake through their choice of wild and cultivated foods to understand better the role of crops in meeting their nutritional requirements (Felton et al. 2009; Lambert and Rothman 2015).
Our analysis of the chimpanzees’ wild plant diet at Bulindi corresponds with previous chemical analyses of primate plant foods (Lambert and Rothman 2015): ripe fruits provided energy from easily digestible carbohydrates, i.e., sugars; piths were an alternative source of carbohydrate energy, particularly from fiber; and young leaves provided protein, which was low in fruits. Plants eaten by wild primates generally contain low amounts of lipids (Lambert and Rothman 2015; Rothman et al. 2012), as was true of wild plants analyzed here. While previous studies found that ripe fruits provided the majority of lipids in African ape diets (Conklin-Brittain et al. 1998; Reiner et al. 2014), lipids were highest in young leaves in our sample. However, this high “lipid” content likely includes nonnutritive components such as wax and cutin that are also extracted by ether (Palmquist and Jenkins 2003). Nevertheless, individual plants within major food categories—both wild and cultivated—varied considerably in chemical properties (Appendix Tables II and III).
Though unripe fruit contained less digestible carbohydrates and more fiber compared to when fully ripe, it offered a supplementary source of protein and energy. We found no differences in antifeedant content between unripe and ripe samples. However, most fruits sampled in both maturity stages were crops that, relative to wild foods, had small concentrations of polyphenols generally and rarely contained tannins (appendices). Though our sample of unripe fruits was small, the absence of strong differences between unripe cultivated fruits and ripe wild fruits suggests agricultural fruits are relatively nutritious throughout the ripening process. Indeed, chimpanzees often ate unripe fruits of cocoa, mango, jackfruit, and guava when available (McLennan, unpubl. data) (Fig. 6). Again, however, nutrient concentrations in unripe fruits varied considerably. For example, unripe fruit of cocoa, mango, and papaya had sugar levels comparable to those of ripe fruits of many wild species. Conversely, unripe plantain banana contained very little soluble sugar, but may have instead provided energy from hemicellulose (Appendix Table III).
Why Did Chimpanzees Ignore Certain Crops?
Contrary to prediction, we found little evidence that crops ignored by the chimpanzees were less nutritious than those that they did eat. Compared to cultivated fruits eaten, ignored fruits (avocado, pineapple, pumpkin, soursop, tamarillo, and tomato) tended to be lower in nonstructural carbohydrates and more fibrous, which might have influenced whether chimpanzees chose to eat them or not. Conversely, the ignored fruits were a better source of protein and lipids, although chimpanzees probably select ripe fruits primarily for their digestibility and high sugar content. Still, pineapple had among the highest sugar content of all fruits analyzed and should have been highly attractive to chimpanzees. Moreover, all ignored fruits are highly palatable to humans, with the exception of pumpkin, which—although edible raw—is considered too fibrous to eat uncooked by local people, although other primates in Bulindi readily eat it, e.g., Chlorocebus tantalus.
Two cultivated piths not eaten by the chimpanzees (rice stem and papaya leaf petiole) offered a good source of protein with low concentrations of fiber and antifeedants. Still, the greater sugar content of cultivated piths that were eaten (especially sugarcane and yam pith, which had sugar concentrations comparable to crop fruits; Appendix Table III), suggests chimpanzees at Bulindi selected cultivated piths mainly for their sweet taste (or carbohydrate energy), not protein. Young leaves of papaya had the highest amount of crude protein of all foods analyzed. But no evidence suggested the chimpanzees exploit this protein-rich resource (as chimpanzees do at Bossou, for example; Hockings and McLennan 2012), although they often ate papaya fruit.
Nonnutritional factors probably also influence crop selection by primates. In this study, we did not compare availability or abundance of different crops, which might influence whether chimpanzees eat them or not (McLennan and Hockings 2014). With regards to fruits, soursop trees were rare at Bulindi and chimpanzees probably had limited opportunities to encounter the sweet fruits. But other crop fruits not eaten such as pineapple, pumpkin, tamarillo, and tomato were more common than several that were eaten, e.g., lemon, orange, and passion fruit. Other nonfruit crops that were ignored—particularly staple foods for local people such as cassava, maize, sweet potato, and rice—were highly abundant and chimpanzees encountered these foods daily when seasonally available. Thus, availability cannot explain why they did not eat them. In particular, maize cob is among the crops most commonly targeted by chimpanzees across Africa (Hockings and McLennan 2012). Crops that are comparable to wild foods in shape, color, and/or odor, and requiring similar processing, are most likely to be recognized as edible by wildlife (McLennan and Hockings 2014). Chimpanzees probably recognize many fruit crops as palatable from ripeness cues, but some fruits ignored at Bulindi, e.g., avocado and pineapple, are harvested by humans before fully ripe and thus lack a strongly sweet odor, or are encased within a tough exocarp such as pumpkin. However, chimpanzees readily consume cocoa pods, which are similarly tough and not strongly scented. In Bulindi, chimpanzees seem not to have parallels in their natural diet for crops such as cassava tuber, sweet potato, and groundnuts (which are embedded) and maize cob (which is concealed). Such characteristics may help explain why they do not currently exploit them.
A previous study showed that chimpanzees at Bossou, where apes have exploited crops for generations, ate a greater variety of cultivated foods (including staple food crops such as cassava, rice, and maize cob) compared to Bulindi, where major habitat encroachment is more recent (McLennan and Hockings 2014). Fast-changing mosaic landscapes may generate dynamic feeding patterns in wild animals, involving complex interactions between local anthropogenic and environmental factors, e.g., farming practices and the relative availability and nutritional quality of wild and cultivated foods (McLennan and Hockings 2014). Thus, chimpanzees in Bulindi may yet “discover” that certain crops not currently exploited are good to eat in time, as illustrated by Takahata et al. (1986), who described the gradual assimilation of mango, guava, and lemon into the diet of wild chimpanzees at Mahale, Tanzania.
Sustainability of Primate Crop Feeding
Ongoing human settlement and cultivation, especially in the tropics, means that primates should adjust their behavior to survive in modified landscapes, or else go locally extinct (Anderson et al. 2007; Estrada et al. 2012; Nowak and Lee 2013). Supplementing a natural diet with energy-rich crops is one such adjustment, but crop foraging inevitably brings primates into competition with humans (Paterson and Wallis 2005). The relative costs and benefits of eating crops will differ according to species and habitat and, perhaps most importantly, human cultural attitudes and socioeconomic conditions that define tolerance of wildlife, but are subject to change (Hill and Webber 2010; McLennan and Hill 2012; Naughton-Treves and Treves 2005; Riley 2010). Like many primates, chimpanzees show a high level of behavioral and dietary flexibility that enables them to survive in cultivated habitats, providing they are not hunted or persecuted (Hockings et al. 2015; Hockings and McLennan 2016). Despite the tolerance sometimes afforded apes by human cultural beliefs, persistent crop losses and associated problems, i.e., aggression toward people (McLennan and Hockings 2016), can instigate retributive killings and use of lethal control methods (Hyeroba et al. 2011; McLennan et al. 2012; Meijaard et al. 2011). Chimpanzees have slow life histories, and even occasional trappings and killings cause population declines (Hockings and McLennan 2016). Whether agricultural and other matrix habitats can support populations of threatened but flexible primates such as chimpanzees in the long term is uncertain. Ultimately, it hinges on the willingness and capacity of local people to share their landscape and resources with them.
We are grateful to the President’s Office, the Uganda National Council for Science and Technology, and the Uganda Wildlife Authority for permission to study the chimpanzees of Bulindi. Matthew McLennan’s fieldwork was supported by a fellowship from the Leverhulme Trust. We are particularly grateful to Tom Sabiiti for assistance in the field. For help with the chemical analyses we thank Irene Tomaschewski. Mary Namaganda and Olivia Maganyi at Makerere University Herbarium, Uganda, identified the taxonomy of several plants analyzed for this study. We thank Kimberley Hockings, Noemi Spagnoletti, Giuseppe Donati, Joanna Setchell, and the reviewers for helpful comments on the draft manuscript.
Compliance with Ethical Standards
Conflict of Interest
The authors declare no conflicts of interest or competing financial interest.
- Abdi, H. (2010). Holm’s sequential Bonferroni procedure. In N. Salkind (Ed.), Encyclopedia of research design (pp. 573–577). Thousand Oaks: SAGE.Google Scholar
- Association of Official Analytical Chemists. (1990). Official methods of analysis. Arlington: Association of Official Analytical Chemists.Google Scholar
- Conklin-Brittain, N. L., Knott, C. D., & Wrangham, R. W. (2006). Energy intake by wild chimpanzees and orangutans: Methodological considerations and a preliminary comparison. In G. Hohmann, M. M. Robbins, & C. Boesch (Eds.), Feeding ecology in apes and other primates (pp. 445–472). Cambridge: Cambridge University Press.Google Scholar
- Conover, M. R. (2002). Resolving human–wildlife conflicts: The science of wildlife damage management. Boca Raton: Lewis Publishers.Google Scholar
- Eley, R. M., Strum, S .C., Muchemi, G., & Reid, G. D .F. (1989). Nutrition, body condition, activity patterns, and parasitism of free‐ranging troops of olive baboons (Papio anubis) in Kenya. American Journal of Primatology, 18, 209–219.Google Scholar
- Forthman-Quick, D. L., & Demment, M. W. (1988). Dynamics of exploitation: Differential energetic adaptations of two troops of baboons to recent human contact. In J. E. Fa & C. H. Southwick (Eds.), Ecology and behavior of food-enhanced primate groups (pp. 25–51). New York: Alan R. Liss.Google Scholar
- Ghiglieri, M. P. (1984). The chimpanzees of Kibale Forest: A field study of ecology and social structure. New York: Columbia University Press.Google Scholar
- Hockings, K. J., & McLennan, M. R. (2016). Problematic primate behaviour in agricultural landscapes: Chimpanzees as ‘pests’ and ‘predators.’ In M. Waller (Ed.), Ethnoprimatology: Primate conservation in the 21st century (pp. 137–156). Developments in Primatology: Progress and Prospects. Cham, Switzerland: Springer.Google Scholar
- Lorenti, G. (2014). Assessing fragmentation characteristics at Bulindi, western Uganda: Implications for primate conservation in a fragmented landscape. MSc dissertation, Oxford Brookes University.Google Scholar
- McLennan, M. R., & Hill, C. M. (2015). Changing agricultural practices and human–chimpanzee interactions: Tobacco and sugarcane farming in and around Bulindi, Uganda. In Arcus Foundation (Ed.), State of the apes: Industrial agriculture and ape conservation (pp. 29–31). Cambridge: Cambridge University Press.Google Scholar
- Naughton-Treves, L., & Treves, A. (2005). Socio-ecological factors shaping local support for wildlife: Crop-raiding by elephants and other wildlife in Africa. In R. Woodroffe, S. Thirgood, & A. Rabinowitz (Eds.), People and wildlife: Conflict or coexistence? (pp. 252–277). Cambridge: Cambridge University Press.CrossRefGoogle Scholar
- Nowak, K., & Lee, P. C. (2013). “Specialist” primates can be flexible in response to habitat alteration. In L. K. Marsh & C. A. Chapman (Eds.), Primates in fragments: Complexity and resilience (pp. 199–211). Developments in Primatology: Progress and Prospects. New York: Springer Science+Business Media.Google Scholar
- Ortmann, S., Bradley, B. J., Stolter, C., & Ganzhorn, J. U. (2006). Estimating the quality and composition of wild animal diets: A critical survey of methods. In G. Hohmann, M. M. Robbins, & C. Boesch (Eds.), Feeding ecology in apes and other primates (pp. 395–418). Cambridge: Cambridge University Press.Google Scholar
- Paterson, J. D., & Wallis, J. (2005). Commensalism and conflict: The human–primate interface. Norman: American Society of Primatologists.Google Scholar
- Plumptre, A. J., Rose, R., Nangendo, G., Williamson, E. A., Didier, K., et al. (2010). Eastern chimpanzee (Pan troglodytes schweinfurthii) status survey and conservation action plan: 2010–2020. Gland: IUCN.Google Scholar
- Prakash, O., Kumar, R., Mishra, A., & Gupta, R. (2009). Artocarpus heterophyllus (Jackfruit): An overview. Pharmacognosy Reviews, 3, 353–358.Google Scholar
- Priston, N. E., & McLennan, M. R. (2013). Managing humans, managing macaques: Human–macaque conflict in Asia and Africa. In S. Radhakrishna, M. A. Huffman, & A. Sinha (Eds.), The macaque connection: Cooperation and conflict between humans and macaques (pp. 225–250). Developments in Primatology: Progress and Prospects. New York: Springer Science+Business Media.Google Scholar
- Sommer, V., Bauer, J., Fowler, A., & Ortmann, S. (2011). Patriarchal chimpanzees, matriarchal bonobos: Potential ecological causes of a Pan dichotomy. In V. Sommer & C. Ross (Eds.), Primates of Gashaka (pp. 469–501). Developments in Primatology: Progress and Prospects. New York: Springer Science+Business Media.Google Scholar
- Spagnoletti, N., Cardoso, T. C. M., Fragaszy, D., & Izar, P. (2017). Coexistence between humans and capuchins (Sapajus libidinosus): Comparing observational data with farmers’ perceptions of crop losses. International Journal of Primatology. doi:10.1007/s10764-016-9926-9.
- Uganda Bureau of Statistics. (2014). National population and housing census 2014: Provisional results. Kampala: Uganda Bureau of Statistics.Google Scholar
- United States Department of Agriculture. (2016). National Nutrient Database for Standard Reference, Release 28 (slightly revised May 2016). United States Department of Agriculture, Agricultural Research Service, Nutrient Data Laboratory. https://ndb.nal.usda.gov/. Accessed 1 July 2016.
- Warren, Y., Higham, J. P., MacLarnon, A. M., & Ross, C. (2011). Crop-raiding and commensalism in olive baboons: The costs and benefits of living with humans. In V. Sommer & C. Ross (Eds.), Primates of Gashaka (pp. 359–384). Developments in Primatology: Progress and Prospects. New York: Springer Science+Business Media.Google Scholar