Ti

EEa

b

c

a

ARRAA

KCPFFES

1

hsMhatDTsipap

h0

Behavioural Processes 105 (2014) 71–78

Contents lists available at ScienceDirect

Behavioural Processes

jo ur nal homep ag e: www.elsev ier .com/ locate /behavproc

he interplay between individual, social, and environmentalnfluences on chimpanzee food choices

mma Finestonea, Kristin E. Bonniea,b, Lydia M. Hoppera, Vivian M. Vreemana,lizabeth V. Lonsdorfa,c, Stephen R. Rossa,∗

The Lester E. Fisher Center for the Study and Conservation of Apes, Lincoln Park Zoo, Chicago, USADepartment of Psychology, Beloit College, Beloit, USADepartment of Psychology, Franklin and Marshall College, Lancaster, USA

r t i c l e i n f o

rticle history:eceived 26 July 2013eceived in revised form 21 January 2014ccepted 18 March 2014vailable online 26 March 2014

eywords:himpanzeean troglodytesood preference

a b s t r a c t

The foraging activity of chimpanzees requires individuals to balance personal preferences with nutri-ent requirements, food availability, and interactions with members of their social group. To determinewhether chimpanzee food preferences are fixed or malleable across varying socio-ecological contexts,we presented six zoo-housed chimpanzees with pairwise combinations of four different foods under twoexperimental conditions. First, we individually tested each chimpanzee’s choices for the four foods toascertain individual preferences. Second, we tested the chimpanzees in a situation which more-closelymimicked the foraging pressures experienced by wild chimpanzees. In this second condition, the chim-panzees were tested in a group setting and the food availability was less predictable, such as in a patchyforaging environment. Subjects expressed significant variation in their selection of which foods to con-

ood choicenvironmental predictabilityocial facilitation

sume in the two different contexts and also appeared more willing to consume less-preferred foods inthe unpredictable, social environment. These results suggest that chimpanzees’ food preferences are notfixed, but change with context and are likely mediated by social facilitation. This is not only importantto understand chimpanzees’ foraging patterns and dietary requirements, but also has implications forexperimental paradigms that rely on food preferences.

© 2014 Elsevier B.V. All rights reserved.

. Introduction

Does what we choose to eat, reflect what we prefer to eat? Forumans, such choices are influenced as much from a nutritionaltandpoint, as from a social–psychological perspective (Pliner andann, 2004). In contrast, the food choices of nonhuman animals

ave classically been studied within an ecological framework withn aim to understand how animals forage most effectively givenhe environmental pressures they face (e.g., Bates and Byrne, 2009;uffy and Hay, 1991; Janmaat et al., 2006; Janson, 1998; Sih, 1982;orres-Contreras and Bozinovic, 1997). Less import is given to howocial and environmental pressures interact, and whether theympact food preferences and food choices. This is particularly sur-rising when considering chimpanzees (Pan troglodytes), which are

highly gregarious nonhuman primate species that have a com-lex, omnivorous diet.

∗ Corresponding author. Tel.: +1 312 742 7263.E-mail addresses: sross@lpzoo.org, chimpfreq@yahoo.com (S.R. Ross).

ttp://dx.doi.org/10.1016/j.beproc.2014.03.006376-6357/© 2014 Elsevier B.V. All rights reserved.

In the wild, chimpanzees spend 40–60% of their active timefeeding (Wrangham, 1977) yet they consume only a small portionof available food items (Sugiyama and Koman, 1987; Wrangham,1977). Chimpanzees are constantly making choices about whatfoods to eat, but are their foraging choices dictated primarily bypersonal preference? Models of primate socioecology suggest thatchimpanzee foraging behavior is guided by a complex interplaybetween food availability, food choices, group composition, andlocation (Wrangham, 1980; Isbell, 1991). Furthermore, primatebehavior and foraging is heavily shaped by periods of food short-ages, during which rare but preferred foods and low-value ‘fallback’foods become especially important (Marshall and Wrangham,2007; Rosenberger, 2013). Thus, it is presumed that social and envi-ronmental influence interact to affect food choices, but there is littledata from controlled experiments to validate these predictions.

It might be assumed that animals preferentially select thosefoods that are nutritionally beneficial for them (e.g., Carlson et al.,

2013), or that they would select those foods that are easy to obtain,the proverbial ‘low hanging fruit’ (Koops et al., 2013). However, anumber of species will travel considerable distances when forag-ing (e.g., Apis mellifera, Beekman and Ratnieks, 2000;Ateles geoffroyi

7 ural P

yndMaCeaaa

fPmffscsshrtfe

twlD2tp2dfmftpt2aeeetIwmocts

ofierpdtc2w

2 E. Finestone et al. / Behavio

ucantanensis, Valero and Byrne, 2007). Furthermore, even whenutritionally-appropriate foods are readily available, animals stillo not always select them (Forbes and Kyriazakis, 1995; Sousa andatsuzawa, 2006; Yeomans et al., 2004). Potentially ‘healthy’ foods

re overlooked in preference for foods that are more palatable.himpanzees, for example, appear to use sugar content (Wranghamt al., 1998; Remis, 2002, 2006) and energy yields (Matsumoto-Odand Hayashi, 1999) to guide their food selection, but both texturend taste have also been shown to be factors that primates use tossess food quality (Remis, 2002).

In addition to the physical properties of the food itself, primateoraging patterns may also be affected by social influences (e.g.,apio cynocephalus, Alberts et al., 1996). Chimpanzees forage withembers of their social group (Nishida, 1968) and rarely encounter

ood sources alone. Thus, social environment may affect individualood choice (Hopper et al., 2011; Lonsdorf and Bonnie, 2010, butee van Leeuwen et al., 2013). Synchronization of feeding activitiesan result in competition for food (Oates, 1987) and chimpanzeeshow contest-type competition (Wittig and Boesch, 2003). Con-equently, low-ranking females, and other subordinates, tend toave a lower quality diet, and forage less efficiently, than higher-anking female chimpanzees (Murray et al., 2006; Parish, 1994). Inhis way, the pressures of group living can directly impact whichoods an individual chooses to eat, regardless of personal prefer-nce.

The impact of a chimpanzee’s social group on foraging pat-erns may also arise through social facilitation. Like for humans,ho have been shown to develop preferences for foods that are

iked by their family, peers, or esteemed individuals (Birch, 1980;uncker, 1938; Pliner and Pelchat, 1986; but see Pliner and Mann,004), there is emerging evidence that chimpanzees may chooseo eat foods that a dominant animal selects, even if it is not theirreferred food (Hopper et al., 2011, see also Galef and Whiskin,008; Sherwin et al., 2002; van de Waal et al., 2013, for comparableata with other species). But social facilitation does not only mani-est through copying the food choices of others; observing a group

ember eating can also encourage the consumption of a differentood (i.e., eating behaviors generally increase). There is evidencehat social facilitation can drive capuchins to select novel or lessreferred foods when observing a group member eat, regardless ofhe food type eaten by the demonstrator (Visalberghi and Addessi,000; Addessi and Visalberghi, 2001; Dindo and de Waal, 2007)nd similar responses have been reported for chimpanzees (Hoppert al., 2011, but see Addessi and Visalberghi, 2006; van Leeuwent al., 2013). Such results suggest that although an animals’ pref-rence may be constant (Visalberghi et al., 2003), it is still possiblehat their choices may vary according to their social environment.mportantly, and depending on the relationship an individual has

ith those animals present at a feeding site, the social environmentay inhibit the individual’s consumption (i.e., due to competition)

r induce greater consumption (i.e., due to social facilitation) ofertain food items. When testing food choices, therefore, it is impor-ant to consider the interaction between the individual and theocial environment in which it feeds (Galef, 1996).

In addition to being of theoretical interest, an understandingf what drives individual preferences and choice is importantor a number of applied reasons. For example, an understand-ng of what dictates animal food choices and preferences cannable us to provide better husbandry and care in a captive envi-onment (e.g., Clay et al., 2009; Gaalema et al., 2011) and mayrovide insight on food availability requirements when reintro-ucing chimpanzees into the wild. Additionally, food preference

ests form the basis of many behavioral and cognitive tests withhimpanzees (e.g., Brosnan et al., 2005; Slocombe and Zuberbühler,006) and understanding whether ‘choice’ reflects ‘preference’ould provide greater context for such research.

rocesses 105 (2014) 71–78

Our aim was to distinguish between chimpanzees’ food preferen-ces and food choices in differing social environments. To determineindividual food preferences for four different foods we tested thechimpanzees individually (‘individual’ condition). In this condi-tion, the foods that the chimpanzees had to choose between wereclearly visible and the chimpanzee was afforded time to make theirselection with no social pressures. In the wild, in a patchy environ-ment, food sources may be less predictable (Houle et al., 2007),foods may require more processing, cognition, and behavioral flex-ibility to obtain (Milton, 1981; Lonsdorf, 2005), and individualchimpanzee’s foraging strategies may be dictated by their group’s(Lehmann and Boesch, 2002). To reflect this naturalistic foragingcontext, the second condition increased both the social and envi-ronmental pressures. We gave the same chimpanzees the same fourfoods, also through a series of pair-wise choices, but presented viaan artificial termite mound and the chimpanzees were tested intheir familiar social group (‘social’ condition).

Investigating food selection in these two contexts is impor-tant for detecting factors that influence group-level food choicethat may be diluted or absent when individuals are tested in soli-tary conditions. We predicted that individual food choices wouldchange across experimental conditions (the individual, stable, con-dition versus the social, unpredictable, condition) however wemade no directional predictions as to how these choices wouldbe expressed (accordingly, all analysis was two-tailed). As socialpressure might affect the foraging choices of animals differen-tially, we also wished to determine if any differences in food choicebetween the conditions reflected a group-wide shift or merely achange in choices by only certain group members. To test this,we considered the chimpanzees’ selections at both the individual-and group-level. This allowed us to evaluate the extent to whichsocial, individual and environmental factors drive the chimpanzees’choices in each condition. Furthermore, in the social condition, wealso analyzed whether an individual’s food selection was related tothe number of group mates concurrently at the mound.

2. Methods

2.1. Ethical note

This study was approved by the Lincoln Park Zoo Research Com-mittee, which is the governing body for all animal research at theinstitution. No social group manipulations occurred as the resultof this project and animal separations were always voluntary onthe part of the apes. Food substances, amount and frequency werereviewed and approved by veterinary and nutrition staff prior tothe start of the project. No modifications were made to standardanimal care routines. This research adhered to legal requirementsin the United States of America and to the American Society ofPrimatologists Principles for the Ethical Treatment of Non-humanPrimates.

2.2. Subjects and housing

The subjects were six captive-born chimpanzees housedtogether at the Regenstein Center for African Apes (RCAA) at theLincoln Park Zoo (Chicago, Illinois, U.S.A). The group included twomales and four females (average age: 17.3 years, range: 11–26years). These chimpanzees inhabited an expansive indoor/outdoorexhibit that included climbing structures and deep-mulch beddingand an off-exhibit holding area (details below). Throughout the

study, the chimpanzees had outdoor access when weather condi-tions were appropriate (>4.5 ◦C). Fresh produce and primate chowwere scattered twice daily throughout their exhibits. The exhibitspace housing these subjects features an artificial termite mound

ural P

psse2

2

iekis2mthweepocc

2

2

orbmAbttlwwc

2

apfepsm

oi3tawcct1ttd

E. Finestone et al. / Behavio

reviously described by Lonsdorf et al. (2009). Prior to this study,ubjects participated in extractive tool studies and learned to fishuccessfully, and as such, are considered proficient at the task. Thexperiment was conducted between March 2009 and November010.

.3. General procedure

We tested chimpanzee food preferences in two conditions:ndividual and social. We conducted individual testing in an off-xhibit holding area composed of six enclosures, around a centraleeper area, which measured 10.8 m2. We conducted social test-ng in a group setting in the publically viewed, naturalistic exhibitpace. This exhibit measures 408.4 m2 for the indoor exhibit and011.7 m2 for the adjacent outdoor yard with ceilings of approxi-ately nine meters. In both conditions we ran pair-wise preference

ests with four foods selected to represent different basic tastes:oisin sauce, ketchup, peanut butter, and mustard. The subjectsere familiar with peanut butter, ketchup, and mustard. How-

ver, hoisin was a novel food item, and prior to each phase of thexperiment additional trials were run to ensure that the chim-anzees were familiar with the palatability and visual propertiesf each substance. As we used four foods, a total of six pair-wiseombinations were possible and all combinations were tested in aounterbalanced manner.

.4. Individual food preference test

.4.1. ApparatusEvery pair combination of the four foods was presented in turn

n a portable tray. This tray was constructed from wood, PVC andecycled plastic boards (Fig. 1) and was designed so that it coulde slid through the gap between the floor and the bottom of theesh door in the holding area and be presented to the chimpanzee.

deep bumper at the back of the tray prevented the tray fromeing pulled completely into the animal area. Handles attached tohe back of the apparatus allowed experimenters to safely retracthe tray after each trial. On the tray’s top surface were two shal-ow wells–approximately 4 cm in diameter and 60 cm apart—into

hich the test foods were placed. The placement of these wellsas such that only one food could be retrieved at a time by the

himpanzee.

.4.2. ProcedureTo test the food preferences of every chimpanzee, we tested each

nimal individually in their holding area. We presented only oneair of foods in any one test session and each session consisted of 10orced-choice trials. Each chimpanzee received three sessions forach food pair combination before proceeding on to the next foodair combination (30 trials for each combination). As there wereix possible food pair combinations, each chimpanzee received ainimum of 18 preference test sessions (6 × 3).For each trial, we placed a standard amount (2.5 mL) of each food

n paper squares (3 cm2), which we presented to the chimpanzeen the shallow wells on the test tray. All six subjects completed0 trials of each of the six food combinations for a total of 180rials per individual (540 trials total). In each trial, subjects werellowed access to the tray for up to 10 s. A chimpanzee’s ‘choice’as indicated by the food that they selected first. As soon as the

himpanzee made a selection, we withdrew the tray so that thehimpanzee could not obtain the alternative food and loaded theray with the next trial. There was an interval of approximately

0 s between each trial to allow time to load the tray. However, ifhe subject did not make a choice within 10 s of presentation, theray was withdrawn and re-presented a second time after a shortelay. If the chimpanzee did not make a selection during this second

rocesses 105 (2014) 71–78 73

presentation, we removed the tray again and that trial was aborted,and we then prepared the tray for the next trial.

These food preference test sessions were terminated when all10 trials had been completed or if the subject did not select a sub-stance in two consecutive trials. If the latter occurred, we re-ranthese ‘missing’ trials and completed them at a later date. Betweentrials, we cleaned the apparatus with a low concentration cleaningsolution in order to remove any remaining food residue from thewells.

To control for side biases, the test foods were randomly andevenly distributed between the left and the right side of the traythroughout the session with both foods being presented on theleft and right of the tray an equal number of times. Individu-ally, four of the six subjects exhibited significant individual sidebiases. Two were biased toward selecting the food on the right,�2(1) = 9.80, P = 0.002; �2(1) = 8.89, P = 0.003; and two toward theleft �2(1) = 4.36, P = 0.040; �2(1) = 5.69, P = 0.020. Because trialswere counterbalanced for side, if the chimpanzees showed a foodpreference for a particular trial, they would have had to overridetheir side bias to select a specific food. Thus, food preferences wouldneed to be especially strong to be significant because selecting thesame food many times in a pair-wise combination would requiresubjects to override their side biases repeatedly.

2.5. Social food preference test

2.5.1. ApparatusFor this condition, foods were presented to the chimpanzees via

an artificial termite mound, designed to replicate a natural termitemound found in wild ape habitats (Fig. 1, see Lonsdorf et al., 2009for details). The mound measured 274 cm wide by 205 cm tall andprotruded 104 cm into the animal exhibit. The structure was hol-low with a 15 cm thick concrete crust. The placement of the moundwas such that keepers could access the inside of the mound throughan access door without entering the chimpanzees’ enclosure. Themound had eight holes bored through its surface that terminatedin the interior of the mound with a screw attachment for polyvinylchloride (PVC) tubes. PVC tubes (4.5 cm diameter, 13 cm length)were then filled and attached from the inside of the mound. Oncebaited in this manner, the chimpanzees could access the food inthese tubes from their enclosure by reaching into each hole withstick tool. We did not provide tools for the chimpanzees for thispurpose, but individuals were able to make tools from natural veg-etation (hay, trees, branches, etc.) available in their indoor andoutdoor exhibits.

2.5.2. ProcedureWe tested the chimpanzees in their social group in their home

enclosure, as described above. During each test session the chim-panzees had access to the mound, which was always baited withtwo of the four test foods. The mound was baited four days aweek (Tuesday, Wednesday, Thursday and Friday) for 1 h (begin-ning at 14:15) using a pseudo–random, balanced schedule. For eachsession, the mound was baited with 75 mL of two of the test foodsand the foods were distributed evenly across the eight numberedPVC tubes (four tubes per food). As for the individual preferencetests, all six possible combinations of the four foods were run andthese pair-wise combinations were counterbalanced for location.In total, we ran twelve sessions (two for each combination). Eachsession lasted for 20 min and commenced as soon as the first chim-

panzee in the social group examined the mound. In order to controlfor any observer bias, the researcher observing the chimpanzeesduring these test sessions was blind to which holes contained thetwo foods in the mound.

74 E. Finestone et al. / Behavioural Processes 105 (2014) 71–78

Fig. 1. Test apparatus. (a) The tray used for the food preference tests in the individual condition, shown from the chimpanzees’ perspective. A view of the artificial termitem d (c) fk

3

tttrsptsasFe

ound used in the social condition (b) from within the chimpanzee enclosure aneeper.

. Coding and data analysis

In the individual preference tests, a food choice was defined ashe first food that the chimpanzee selected within a trial. A “selec-ion” was defined as a chimpanzee removing the food sample fromhe testing tray with either their hand or mouth. For each trial, theesearcher recorded which food the chimpanzee selected first, theide of the tray that food was on, and the latency for the chim-anzee to choose the food. For each of the six food combinations,he researcher recorded the number of trials in which a subjectelected each food. To examine the data, choice matrices were cre-

ted for each subject, and pairwise comparisons for each of theix food combinations were made using Chi-square Goodness ofit tests (expected values were set at 50% of trials completed forach food pair). Exact P values are reported. To correct for multiple

rom the inside from where the baited tubes could be attached by a researcher or

comparisons, we used a false discovery rate control, as described byBenjamini and Hochberg (1995) see also, Storey (2002) and Hopperet al. (in press).

In the social condition, a food selection was defined as thefood that the chimpanzee extracted from the mound duringscan-samples over a 20 min period. In this social condition, thechimpanzees could thus make multiple choices and so we wereinterested in the proportion choice of the two foods selected withina given session. We used scan-sampling with a 30 s inter-sampleinterval to record the behavior of each chimpanzee that was within1 m of the mound. When the subject’s behavior was “extract”

(i.e., when they were actively getting food from the mound), werecorded the hole number. When the subject extracted from a hole,that scan-sample was considered a trial and the food that the holecontained was considered the food choice.

E. Finestone et al. / Behavioural Processes 105 (2014) 71–78 75

Table 1Proportion of forced choice trials substance (Hoisin, Ketchup, Mustard or Peanut Butter) was chosen by each subject in individual and social conditions.

Subject Condition Hoisin vsketchup

Hoisin vsmustard

Hoisin vspeanut butter

Ketchup vsmustard

Ketchup vspeanut butter

Mustard vspeanut butter

Rankorder

Ha:M,20 Individual H = 50.0 H = 96.7 H = 50.0 K = 96.7 K = 50.0 PB = 96.7 H2.5 = K2.5 > PB1 > M0

P =1.000 <.00001 =1.000 <.00001 =1.000 <.00001Social H = 58.8 H = 80.6 H = 51.1 K = 79.4 PB = 53.7 PB = 71.1 H2 = K2 = PB2 > M0

P =.303 =.0006 =.8848 =.0006 =.6390 =.0094Op:M,11 Individual H = 53.3 H = 93.3 H = 66.7 K = 86.7 K = 96.7 PB = 86.7 K2.5 > H2 > PB1.5 > M0

P =.7153 <.00001 =.0679 =.00005 <.00001 =.00005Social H = 52.4 H = 85.7 H = 76.9 K = 68.2 K = 75.0 M = 79.2 H2 = PB2 > K1.5 > M.5

P =.8266 =.0075 =.0522 =.0881 =.0455 =.0043Ch:F,11 Individual K = 90.0 H = 56.7 PB = 93.3 K = 100.0 PB = 70.0 PB = 100.0 PB3 > K2 > H.5 = M.5

P =.00001 =.4654 <.00001 <.00001 =.0285 <.00001Social K = 55.3 H = 65.2 PB = 69.2 K = 52.0 PB = 44.4 PB = 81.0 PB2 = K1.5 = H1.5 > M1

P =.5164 =.1444 =.0499 =.8415 =.5052 =.00006Ka:F,20 Individual K = 53.3 H = 83.3 PB = 80.0 K = 96.7 K = 56.7 PB = 100.0 PB2.5 > K2 > H1.5 > M0

P =.7153 =.0003 =.0010 <.00001 =.4653 <.00001Social H = 51.4 H = 94.1 H = 52.4 K = 77.8 PB = 81.6 PB = 75.6 PB2.5 > H2 > K1.5 > M0

P =.8648 <.00001 =.7579 =.0184 <.00001 =.0010Ca:F,26 Individual K = 76.7 H = 83.3 PB = 93.3 K = 93.3 PB = 73.3 PB = 53.3 PB3 > K2 > H1 > M0

P =.0035 =.0003 <.00001 <.00001 =.0106 <.00001Social H = 51.5 H = 68.4 PB = 81.5 K = 65.7 PB = 60.0 PB = 62.9 PB2 > K1.5 = H1.5 > M1

P =.8625 =.0231 <.00001 =.0630 =.2059 =.1282Na:F,16 Individual H = 56.7 H = 100.0 H = 53.3 K = 83.3 K = 63.3 PB = 90.0 H2 = K2 = PB2 > M0

P =.4653 <.00001 =.7153 =.0003 =.1442 =.00001Social H = 61.9 H = 69.2 PB = 59.5 K = 65.7 PB = 60.0 PB = 63.9 H2 > K1.5 = PB1.5 > M1

P =.1228 =.0163 =.1036 =.0630 =.2059 =.1282

While data are presented here as proportions for ease of comparison, statistical analyses were run on raw choice frequencies. Bold indicates a significant preference asd rials, dw and ti

oocp(Ew

uwSsbtts

4

4

cji(mmc(asspT

etermined by Chi Square Goodness of Fit analyses (expected values were 50% of tith rank order indicate frequency of preference for each food; with preference = 1

Data were collected by two observers, who had achieved inter-bserver reliability rates of at least 85% prior to the commencementf the study. As for the individual condition, choice matrices werereated for each subject and pairwise comparisons for each foodairing were made using exact Chi-square Goodness of Fit testsexpected values were set at 50% of trials completed for each pair).xact P values are reported. To correct for multiple comparisons,e again used a false discovery rate control.

To examine how food choice changed between the individ-al and social conditions, we compared the proportion of trials inhich each food was selected across conditions using a Wilcoxon

igned Rank analysis. Finally, to examine how food choice in theocial condition was driven by the presence of other group mem-ers, we calculated the average number of individuals extracting athe mound during each trial. Friedman’s tests were used to comparehe average number of chimpanzees fishing when each subjectselected each of the four food items.

. Results

.1. Food choices in the individual condition

All six subjects completed 30 trials of each of the six foodombinations. While there was some individual variation, all sub-ects chose one food on significantly more trials than expectedn at least three food pairings, indicating individual preferencesTable 1). In addition, subjects demonstrated a clear aversion to

ustard: all six subjects chose both ketchup and peanut butter overustard significantly more than expected, and five of six subjects

hose hoisin over mustard significantly more often than expectedTable 1). Using the results of these pairwise comparisons we cre-ted a count matrix of the number of times each food was chosen

ignificantly more than expected. When food choice did not differignificantly than expected each food was given a .5 rating. Rankedreferences, determined by these count matrices, are presented inable 1.

f = 1 in all cases) when P (exact) is corrected for multiple comparisons. Subscriptse = .5. Each chimpanzee is identified with: ID: sex, age (at time of testing).

4.2. Food choices in the social condition

In the social condition, all six subjects completed at least 13 tri-als for each of the six food combinations (range 13–74; M = 34.3;SD = 11.6). While there continued to be some individual variation,subjects tended to choose foods with more equal frequency in eachfood pairing (Table 1). One subject (Op) who showed a prefer-ence for peanut butter over mustard in the individual condition,showed the opposite, choosing mustard significantly more oftenthan peanut butter, in the social condition. As with the data fromthe individual condition, we created a count matrix of the numberof times each food was chosen significantly more than expected.When food choice did not differ significantly than an expected ascore of .5 was given to both foods. Ranked preferences for individ-ual and group choices are presented in Table 1.

4.3. Change across conditions

Subjects altered their overall food choices when in the socialcondition compared to when tested in the individual condition(Fig. 2). Subjects chose mustard on a significantly higher propor-tion of trials in the social condition than in the individual condition(Wilcoxon Signed Rank, Z = 2.20, P = .028). In contrast, subjectschose ketchup on significantly fewer proportion of trials in thesocial condition than they had in the individual condition (Z = 2.20,P = .028). Subjects chose both hoisin and peanut butter equally inthe individual and social conditions (hoisin: Z = .42, P = .674, peanutbutter: Z = 1.99, P = .063).

4.4. Effect of other group members on food choice in the socialcondition

The mean number, and range, of chimpanzees present alongsidea focal animal for each selected substance were: ketchup (mean:2, range: 1.9–2.1), hoisin (mean: 1.9, range: 1.9–2.0), peanut but-ter (mean: 2.0, range: 1.4–2.2), and mustard (mean: 1.8, range:

76 E. Finestone et al. / Behavioural P

Fig. 2. Change from individual to social conditions, in proportion of trials each of fourfoods were selected by each of six chimpanzees. Positive values indicate that foodwvc

1swdPn

5

bcaapctuftrw1ScccwuIc

phccltenp2pbtp

as selected on a higher proportion of trials in the individual condition. Negativealues indicated that food was selected on a higher proportion of trials in the socialondition.

.5–2.1). Therefore, although the chimpanzees showed preferentialelection of certain foods over others in the social condition, thereas no significant difference in the number of chimpanzees presenturing each subject’s food selection (Friedman’s test) �2 (3) = 7.20,

= 0.066. Thus, food choice did not seem to be impacted by theumber of groupmates present at the mound during feeding bouts.

. Discussion

When the chimpanzees were provided opportunities to chooseetween foods, individuals demonstrated strong food preferen-es (and aversions), but the foods they selected to eat variedccording to the experimental setting. First, when tested individu-lly, the chimpanzees variably selected food substances such thateanut butter and ketchup were selected most often over the otherhoices (hoisin and mustard), and mustard was avoided more oftenhan all other substances. Therefore, although there was individ-al variation, they each showed a strong preference for one of theour foods, and all six chimpanzees universally disregarded mus-ard. These results are consistent with previous studies that haveeported the preferential selection of different food items by bothild and captive chimpanzees and other apes (Wrangham et al.,

998; Remis, 2002; Sousa and Matsuzawa, 2006; Clay et al., 2009).econd, chimpanzees altered their substance choices in the moreomplex, naturalistic social environment. Strikingly, although thehimpanzees still universally did not prefer mustard, they signifi-antly selected mustard more often in this condition than they hadhen tested individually. This not only held true for some individ-als; all six chimpanzees shifted in their food selection patterns.

n this way, although the chimpanzees’ preferences may not havehanged across conditions, their choices did.

In the individual condition, chimpanzees may have selectedeanut butter and ketchup for nutritional reasons. Compared tooisin and mustard, peanut butter has high-energy yields and fatontent, while ketchup has high caloric (i.e., sugar) and proteinontent. Of the four foods, the least-selected mustard offered veryittle calorific or protein yields. However, mustard and peanut but-er both provided comparably high levels of iron and calcium,specially in comparison to hoisin and ketchup, which containeither. In accordance with previous food preference data for chim-anzees (e.g., Matsumoto-Oda and Hayashi, 1999; Remis, 2002,006; Wrangham et al., 1998), these data suggest that the chim-

anzees’ preferences (for peanut butter and ketchup) were driveny a desire for sweet, calorific foods. Note, however, that these par-icular chimpanzees are provided with a comprehensive diet thatrovides them with the requisite daily nutrient content and thus,

rocesses 105 (2014) 71–78

unlike wild chimpanzees (e.g., Hohmann et al., 2010; Tennie et al.,2009), these captive animals did not need to seek out foods basedon their nutritional needs. While it is possible that taste or nutri-tional motives contributed to the chimpanzees’ food choices, suchmotivations cannot explain the shift in choices between the twoexperimental conditions. Rather, we propose that this shift is bestexplained by ecological and social factors introduced in the socialcondition.

While primates are motivated to obtain the most desired food intheir environment (Hopper et al., 2013; van Leeuwen et al., 2013;Sheskin et al., 2013), there is also substantial evidence that socialinfluences affect their food choices (e.g., van de Waal et al., 2013).Specifically, chimpanzees appear to monitor the quality of foodsthat others eat (Brosnan, 2011; Price and Brosnan, 2012), whichmay also explain why chimpanzees have also been shown to con-form to food choices made by other group members (Hopper et al.,2011). Although the pressure to conform may also co-occur withthe opposing pressure of competition over desirable food resources(Wittig and Boesch, 2003; Murray et al., 2006) as the number ofchimpanzees that fed at the mound did not increase when moredesirable foods were presented, we propose that levels of compe-tition did not vary across sessions either. Furthermore, althoughdetailed data on the specific hole choices made by each chim-panzee according to their relative rank to other group members alsopresent at the mound was not available, the shift in food choices forincreased selection of mustard was shown by all six group mem-bers (i.e., not just by subordinate individuals) also suggesting thatrelative-rank did not drive food choices.

Rather than competition or conformity, however, we suggestthat the chimpanzees’ shift in food selection between conditionswas likely driven by social facilitation and local, or stimulus,enhancement; the absolute number of chimpanzees at the mounddid not impact an individual’s choice, but simply being in the pres-ence of their groupmates (whatever the number) stimulated themto select their less preferred food more often. It is well documentedthat social facilitation encourages individuals to eat greater quan-tities in the presence of companions (e.g., Harlow and Yudin, 1933;Bell and Pliner, 2003; Hoppitt et al., 2007). Indeed, primates willeven eat less preferred foods after seeing a social partner eating(Visalberghi and Addessi, 2000; Dindo and de Waal, 2007; Dubreuilet al., 2006; Hopper et al., 2011). In addition to the potential influ-ences of social facilitation, which arose from the observation ofgroup mates eating, simply seeing other chimpanzees present atthe mound may have encouraged greater exploration of the moundand the foods therein (i.e., stimulus enhancement). In the socialcondition, when a chimpanzee approached the mound and beganfeeding, this may have elicited the same behavior from other groupmembers and encouraged more consumption across the group. Thisdoes not imply that social facilitation actively shifted a subject’schoice from one substance to another, simply that it encouragedgreater eating. Furthermore, this shift in choices would have beenmost apparent for food items that were consumed in low qualitiesin the individual condition (in this case mustard).

Considering ecological and environmental pressures on foodchoices, obtaining food from the mound was more complex andfood choices were less transparent compared to in the individualtrials. Even though this may lead one to question if chimpanzeeswere even able to preferentially select some foods over others, wepropose that they were. Although the chimpanzees ate more of theleast-preferred mustard in the social condition, the chimpanzeesstill showed strong preferences in this condition, suggesting thatthey were not just randomly sampling all the baited holes in the

mound. Given that chimpanzees evidence robust memory for thelocation of specific foods in their environment (e.g., Sayers andMenzel, 2012; Janmaat et al., 2013), even if a chimpanzee ‘acciden-tally’ selected a food they did not prefer, they could have simply

ural P

accmaswlaflf

ioe2ambiapresttsctaavbdntipspue

A

fiCshsfr

R

A

A

A

E. Finestone et al. / Behavio

voided that hole when dipping again and thus still expressed ahoice for a preferred food, and this is what we observed. Becausehimpanzees rely on multiple sensory and environmental cues toake initial food discoveries (Janmaat et al., 2013) they would be

ble to effectively sample the available foods, with taste, smell andight, and then make their choices, as we have previously shownith this group of chimpanzees (Bonnie et al., 2012). This flexibility

ikely enabled them to navigate the current experimental set-upnd avoid certain holes/foods. In sum, the cryptic nature of theood in the mound cannot fully explain increased selection of theess-preferred mustard and we suggest instead that it likely aroserom an interplay between environmental and social factors.

Previous studies with captive chimpanzees have providednsights to the ways that nutrition, palatability, taste and physi-logical needs shape food preference by controlling for social andnvironmental factors (Benz et al., 1992; Sousa and Matsuzawa,006; Remis, 2002, 2006). In contrast, we incorporated socialnd environmental pressures into a more naturalistic design. Thisethod revealed that the chimpanzees altered their food choices

etween environments, indicating that the context in which foodtems are presented can significantly affect choice. When testingn animal alone, it is highly likely that they are expressing foodreferences, however, in a social setting, food choice is not alwayseflective of preference and is likely to reflect a combination of sev-ral factors. This is especially pertinent when we consider thosetudies that have tested food preferences individually and usedhose data to determine experimental design within a social con-ext (e.g., Bräuer et al., 2006, 2009; Brosnan et al., 2010). Futuretudies are needed to evaluate inter-subject variability in foodhoice and determine how specific social dynamics may influencehe choices of some subjects differently than others. Age, sex, rank,nd genetics are relevant factors to consider when evaluating for-ging strategies in a social environment (e.g., Hopper et al., 2011;an de Waal et al., 2012). For example, young individuals maye more inclined to copy the food choices of others, even if theyo not correspond with personal preference. In addition, subordi-ate individuals’ food choices may differ depending on whetherhe testing situation allows them to copy the choice of dominantndividuals without directly competing (i.e., enough items of thereferred food present such that competition is not necessary). Inummary, we suggest that a holistic approach to the study of foodreference and choice is the most fruitful way to gain a completenderstanding of chimpanzee behavior whether in a captive or wildnvironment.

cknowledgements

The authors wish to thank the Leo S. Guthman Fund for theirnancial support of the Lester E. Fisher Center for the Study andonservation of Apes, Lincoln Park Zoo. We also thank Marissa Mil-tein, Sarah Calcutt, Kathy Wagner and Karen Anderson for theirelp with data collection, and Maureen Leahy and the animal caretaff at the Regenstein Center for African Apes, Lincoln Park Zoo,or their assistance with animal management and cooperation withesearch.

eferences

ddessi, E., Visalberghi, E., 2001. Social facilitation of eating novel food in tuftedcapuchin monkeys (Cebus apella): input provided by group members andresponses affected in the observer. Anim. Cognit. 4, 297–303.

ddessi, E., Visalberghi, E., 2006. How social influences affect food neophobia in

captive chimpanzees: a comparative approach. In: Matsuzawa, T., Tomonaga,M., Tanaka, M. (Eds.), Cognitive Development in Chimpanzees. Springer, Tokyo,pp. 246–264.

lberts, S.C., Altmann, J., Wilson, M.L., 1996. Mate guarding constrains foragingactivity of male baboons. Anim. Behav. 51, 1269–1277.

rocesses 105 (2014) 71–78 77

Bates, L.A., Byrne, R.W., 2009. Sex differences in the movement of free-ranging chim-panzees (Pan troglodytes schweinfurthii). Behav. Ecol. Sociobiol. 64, 247–255.

Beekman, M., Ratnieks, F.L.W., 2000. Long-range foraging by the honey-bee, Apismellifera L. Funct. Ecol. 14, 490–496.

Bell, R., Pliner, P.L., 2003. Time to eat: the relationship between the number of peopleeating and meal duration in three lunch settings. Appetite 41, 215–218.

Benjamini, Y., Hochberg, Y., 1995. Controlling the false discovery rate: a practicaland powerful approach to multiple testing. J. R. Stat. Soc. B 57, 289–300.

Benz, J.J., Leger, D.W., French, J.A., 1992. Relation between food preference and food-elicited vocalizations in golden lion tamarins (Leontopithecus rosalia). J. Comp.Psychol. 106, 142–149.

Birch, L.L., 1980. Effects of peer models’ food choices and eating behavior onpreschoolers’ food preferences. Child Dev. 51, 489–496.

Bonnie, K.E., Milstein, M.S., Calcutt, S., Ross, S.R., Wagner, K., Lonsdrof, E.V., 2012.Flexibility and persistence of chimpanzee foraging behavior in a captive envi-ronment. Am. J. Primatol. 74, 661–668.

Bräuer, J., Call, J., Tomasello, M., 2006. Are apes really inequity averse? Proc. R. Soc.B 273, 3123–3128.

Bräuer, J., Call, J., Tomasello, M., 2009. Are apes inequity averse? New data on thetoken-exchange paradigm. Am. J. Primatol. 71, 175–181.

Brosnan, S.F., 2011. A hypothesis of the co-evolution of cooperation and responsesto inequity. Front. Neurosci. 5, 43, http://dx.doi.org/10.3389/fnins.2011.00043.

Brosnan, S.F., Schiff, H.C., de Waal, F.B.M., 2005. Tolerance for inequity may increasewith social closeness in chimpanzees. Proc. R. Soc. B 1560, 253–258.

Brosnan, S.F., Talbot, C., Algren, M., Lambeth, S.P., Schapiro, S.J., 2010. Mechanismsunderlying responses to inequitable outcomes in chimpanzees, Pan troglodytes.Anim. Behav. 79, 1229–1237.

Carlson, B.A., Rothman, J.M., Mitani, J.C., 2013. Diurnal variation in nutrients andchimpanzee foraging behavior. Am. J. Primatol. 75, 342–349.

Clay, A.W., Bloomsmith, M.A., Jackson Marr, M., Maple, T.L., 2009. Systematic inves-tigation of the stability of food preferences in captive orangutans: implicationsfor positive reinforcement training. J. Appl. Anim. Welfare Sci. 12, 306–313.

Dindo, M., de Waal, F.B.M., 2007. Partner effects on food consumption in browncapuchin monkeys. Am. J. Primatol. 69, 1–9.

Dubreuil, D., Gentile, M.S., Visalberghi, E., 2006. Are capuchin monkeys (Cebus paella)inequity averse? Proc. R. Soc. Lond. B 273, 1223–1228.

Duffy, J.E., Hay, M.E., 1991. Food and shelter as determinants of food choice by anherbivorous marine amphipod. Ecology 72, 1286–1298.

Duncker, K., 1938. Experimental modification of children’s food preferences throughsocial suggestion. J. Abnorm. Soc. Psychol. 33, 489–507.

Forbes, L.M., Kyriazakis, I., 1995. Food preferences in farm animals: why don’t theyalways choose wisely? Proc. Nutrit. Soc. 54, 429–440.

Gaalema, D.E., Perdue, B.M., Kelling, A.S., 2011. Food preference, keeper ratings, andreinforce effectiveness in exotic animals: the value of systematic testing. J. Appl.Anim. Welfare Sci. 14, 33–41.

Galef, B.G., 1996. Food selection: problems in understanding how we choose foodsto eat. Neurosci. Biobehav. Views 20, 67–73.

Galef, B.G., Whiskin, E.E., 2008. ‘Conformity’ in Norway rats? Anim. Behav. 75,2035–2039.

Harlow, H.F., Yudin, H.C., 1933. Social behavior of primates: social facilitation of feed-ing in the monkey and its relation to attitudes of ascendance and submission. J.Comp. Psychol. 16, 171–185.

Hohmann, G., Potts, K., N’Guessan, A., Fowler, A., Mundry, R., Ganzhorn, J.U., Ort-mann, S., 2010. Plant foods consumed by Pan: exploring the variation ofnutritional ecology across Africa. Am. J. Phys. Anthropol. 141, 476–485.

Hopper, L.M., Lambeth, S.P., Schapiro, S.J., Brosnan, S.F., 2013. When given the oppor-tunity, chimpanzees maximize personal gain rather than “level the playingfield”. Peer J. 1, e165.

Hopper, L.M., Price, S.A., Freeman, H.D., Lambeth, S.P., Schapiro, S.J., Kendal, R.L. 522Influence of personality, age, sex, and estrous state on chimpanzee problem-solving 523 success. Anim. Cogn. DOI 10.1007/s10071-013-0715-y.

Hopper, L.M., Schapiro, S.J., Lambeth, S.P., Brosnan, S.F., 2011. Chimpanzees’ sociallymaintained food preferences indicate both conservatism and conformity. Anim.Behav. 81, 1195–1202.

Hoppitt, W., Blackburn, L., Laland, K.N., 2007. Response facilitation in the domesticfowl. Anim. Behav. 73, 229–238.

Houle, A., Chapman, C.A., Vickery, W.L., 2007. Intratree variation in fruit productionand implications for primate foraging. Int. J. Primatol. 28, 1197–1217.

Isbell, L.A., 1991. Contest and scramble competition: patterns of female aggressionand ranging behavior among primates. Behav. Ecol. 2, 143–155.

Janmaat, K.R.L., Byrne, R.W., Zuberbühler, K., 2006. Primates take weather intoaccount when searching for fruits. Curr. Biol. 16, 1232–1237.

Janmaat, K.R.L., Ban, S.D., Boesch, C., 2013. Taï chimpanzees use botanical skillsto discover fruit: what we can learn from their mistakes. Anim. Cognit.,101007/s10071-013-0617-z.

Janson, C.H., 1998. Experimental evidence for spatial memory in foraging wildcapuchin monkeys, Cebus apella. Anim. Behav. 55, 1229–1243.

Koops, K., McGrew, W.C., Matsuzawa, T., 2013. Ecology of culture: do environmentalfactors influence foraging tool use in wild chimpanzees, Pan troglodytes verus?Anim. Behav. 85, 175–185.

Lehmann, J., Boesch, C., 2002. Social influences on ranging patterns among chim-

panzees (Pan troglodytes verus) in the Taï National Park, Côte d’Ivoire. Behav.Ecol. 14, 642–649.

Lonsdorf, E.V., 2005. Sex differences in the development of termite-fishing skillsin wild chimpanzees (Pan troglodytes schweinfurthii) of Gombe National Park,Tanzania. Anim. Behav. 70, 673–683.

7 ural P

L

L

M

M

M

M

N

O

P

P

P

P

R

R

R

S

S

S

S

8 E. Finestone et al. / Behavio

onsdorf, E.V., Bonnie, K.E., 2010. Opportunities and constraints when studyingsocial learning: developmental approaches and social factors. Learn. Behav. 38,195–205.

onsdorf, E.V., Ross, S.R., Linick, S.A., Milstein, M.S., Melber, T.N., 2009. An experi-mental, comparative investigation of tool use in chimpanzees and gorillas. Anim.Behav. 77, 1119–1126.

arshall, A.J., Wrangham, R.W., 2007. Evolutionary consequences of fallback foods.Int. J. Primatol. 28, 1219–1235.

atsumoto-Oda, A., Hayashi, Y., 1999. Nutritional aspects of fruit choice by chim-panzees. Folia Primatol. 70, 154–162.

ilton, K., 1981. Diversity of plant foods in tropical forests as a stimulus to mentaldevelopment in primates. Am. Anthropol. 83, 534–548.

urray, C.M., Eberly, E.E., Pusey, A.E., 2006. Foraging strategies as a function of sea-son and rank among wild female chimpanzees (Pan troglodytes). Behav. Ecol. 17,1020–1028.

ishida, T., 1968. The social group of wild chimpanzees in the Mahali mountains.Primates 9, 167–224.

ates, J.F., 1987. Food distribution and foraging behavior. In: Smuts, B., Cheney, D.,Seyfarth, R., Wrangham, R., Struhsaker, T. (Eds.), Primate Societies. The Univer-sity of Chicago Press, Chicago, IL, pp. 197–209.

liner, P., Mann, N., 2004. Influence of social norms and palatability on amountconsumed and food choice. Appetite 42, 227–237.

liner, P., Pelchat, M.L., 1986. Similarities in food preferences between children andtheir siblings and parents. Appetite 7, 333–342.

arish, A.R., 1994. Sex and food control in the “uncommon chimpanzee”: howbonobo females overcome a phylogenetic legacy of male dominance? Ethol.Sociobiol. 15, 157–179.

rice, S.A., Brosnan, S.F., 2012. To each their own: variability in the responses toinequity in nonhuman primates. Soc. Justice Res. 25, 140–169.

emis, M.J., 2002. Food preferences among captive western gorillas (Gorilla gorillagorilla) and chimpanzees (Pan troglodytes). Int. J. Primatol. 23, 231–249.

emis, M.J., 2006. The role of taste in food selection by African apes: implicationsfor niche separation and overlap in tropical forests. Primates 47, 56–64.

osenberger, A.L., 2013. Fallback foods, preferred foods, adaptive zones, and primateorigins. Am. J. Primatol. 75, 883–890.

ayers, K., Menzel, C.R., 2012. Memory and foraging theory: chimpanzee utilizationof optimality heuristics in the rank-order recovery of hidden foods. Anim. Behav.84, 795–803.

herwin, C.M., Heyes, C.M., Nicol, C.J., 2002. Social learning influences the preferen-ces of domestic hens for novel food. Anim. Behav. 63, 933–942.

heskin, M., Ashayeri, K., Skerry, A., Santos, L.R., 2013. Capuchin monkeys (Cebusapella) fail to show inequality aversion in a no-cost situation. Evol. Hum. Behav.,http://dx.doi.org/10.1016/j.evolhumbehav.2013.10.004.

ih, A., 1982. Foraging strategies and the avoidance of predation by an aquatic insect,Notonecta hoffmanni. Ecology 63, 786–796.

rocesses 105 (2014) 71–78

Slocombe, K.E., Zuberbühler, K., 2006. Food-associated calls in chimpanzees:responses to food types or food preferences? Anim. Behav. 72, 989–999.

Storey, J.D., 2002. A direct approach to false discovery rates. J. R. Stat. Soc. B 63,479–498.

Sousa, C., Matsuzawa, T., 2006. Token use by chimpanzees: choice, metatool, and cost(Pan troglodytes). In: Matsuzawa, T., Tomonaga, M., Tanaka, M. (Eds.), CognitiveDevelopment in Chimpanzees. Springer, Tokyo, pp. 411–438.

Sugiyama, Y., Koman, J., 1987. A preliminary list of chimpanzees’ alimentation atBossou, Guinea. Primates 28, 133–147.

Tennie, C., Gilby, I.C., Mundry, R., 2009. The meat-scrap hypothesis: small quan-tities of meat may promote cooperative hunting in wild chimpanzees (Pantroglodytes). Behav. Ecol. Sociobiol. 63, 421–431.

Torres-Contreras, H., Bozinovic, F., 1997. Food selection in an herbivo-rous rodent: balancing nutrition with thermoregulation. Ecology 78,2230–2237.

Valero, A., Byrne, R.W., 2007. Spider monkey ranging patterns in Mexican subtropicalforest: do travel routes reflect planning? Anim. Cognit. 10, 305–315.

van de Waal, E., Borgeaud, C., Whiten, A., 2013. Potent social learning and conformityshape a wild primate’s foraging decisions. Science 340, 483–485.

van de Waal, E., Krützen, M., Hula, J., Goudet, J., Bshary, R., 2012. Similarity in foodcleaning techniques within matrilines in wild vervet monkeys. PLoS One 7,e35694.

van Leeuwen, E.J.C., Cronin, K.A., Schütte, S., Call, J., Haun, D.B.M., 2013. Chimpanzees(Pan troglodytes) flexibly adjust their behaviour in order to maximize payoffs,not to conform to majorities. PLoS One 8 (11), e80945.

Visalberghi, E., Addessi, E., 2000. Seeing group members eating a familiar foodenhances the acceptance of novel foods in capuchin monkeys. Anim. Behav. 60(1), 69–76.

Visalberghi, E., Sabbatini, G., Stammati, M., Addessi, E., 2003. Preferences towardsnovel foods in Cebus apella: the role of nutrients and social influences. Physiol.Behav. 80, 341–349.

Wittig, R.M., Boesch, C., 2003. Food competition and linear dominance hierar-chy among female chimpanzees of the Taï National Park. Int. J. Primatol. 24,847–867.

Wrangham, R.W., 1977. Feeding behavior of chimpanzees in Gombe National ParkTanzania. In: Clutten-Brock, T.H. (Ed.), Primate Ecology: Studies of Feeding andRanging Behavior in Lemurs, Monkeys and Apes. Academic Press, London, pp.504–538.

Wrangham, R.W., 1980. An ecological model of the evolution of female-bondedgroups of primates. Behaviour 75, 262–300.

Wrangham, R.W., Conklin-Brittain, N., Hunt, K.D., 1998. Dietary response ofchimpanzees and cercopithecines to seasonal variation in fruit abundance. I.Antifeedants. Int. J. Primatol. 19, 949–970.

Yeomans, M.R., Blundell, J.E., Leshem, M., 2004. Palatability: response to nutritionalneed or need-free stimulation of appetite. Br. J. Nutr. 92, S3–S14.