Copying without rewards: socially influenced foraging decisions among brown capuchin monkeys

An individual’s foraging activity can be influenced by the choices made by nearby conspecifics. The interest shown in the location and characteristics of a feeding patch may depend on the feeding success of a conspecific there, a process that needs to be distinguished from choices guided by rewards to the observer itself. We investigated how rewards for both self and others influence the foraging choices of captive capuchin monkeys (Cebus apella). Thirteen adult capuchins observed familiar female conspecific models explore one of three opaque boxes under three conditions. In the first, there were no rewards available to either monkey; in the second, rewards were available to the model only; and in the third, both monkeys could retrieve a reward. Under all conditions, subjects more often explored the same box as the model than was expected by chance. Thus, without ever receiving a reward themselves or without seeing another receive rewards, subjects’ searches were directed at the box explored by another monkey. The tendency to match the model’s choice increased if the subject was rewarded. We compared these results to control conditions in which the model was either absent, or present but not allowed to demonstrate. Subjects’ located the reward less often in control conditions, than in the experimental conditions. We conclude that extrinsic rewards, while helpful, are not required for partners to influence the foraging choices of capuchins, and that the unrewarded copying of foraging choices demonstrated here may provide the basis for additional social influences on learning. This contribution is part of the Special Issue “A Socioecological Perspective on Primate Cognition” (Cunningham and Janson 2007)     

Solving small spaces: investigating the use of landmark cues in brown capuchins (Cebus apella)

Some researchers have recently argued that humans may be unusual among primates in preferring to use landmark information when reasoning about some kinds of spatial problems. Some have explained this phenomenon by positing that our species’ tendency to prefer landmarks stems from a human-unique trait: language. Here, we test this hypothesis—that preferring to use landmarks to solve such tasks is related to language ability—by exploring landmark use in a spatial task in one non-human primate, the brown capuchin monkey (Cebus apella). We presented our subjects with the rotational displacement task, in which subjects attempt to relocate a reward hidden within an array of hiding locations which are subsequently rotated to a new position. Over several experiments, we varied the availability and the salience of a landmark cue within the array. Specifically, we varied (1) visual access to the array during rotation, (2) the type of landmark, (3) the consistency of the landmark qualities, and (4) the amount of exposure to the landmark. Across Experiments 1 through 4, capuchins did not successfully use landmarks cues, suggesting that non-linguistic primates may not spontaneously use landmarks to solve some spatial problems, as in this case of a small-scale dynamic spatial task. Importantly, we also observed that capuchins demonstrated some capacity to learn to use landmarks in Experiment 4, suggesting that non-linguistic creatures may be able to use some landmarks cues in similar spatial tasks with extensive training.     

Experimental evidence for route integration and strategic planning in wild capuchin monkeys

Both in captivity and the wild, primates are found to travel mostly to the nearest available resource, but they may skip over the closest resource and travel to more distant resources, which are often found to be more productive. This study examines the tradeoff between distance and reward in the foraging choices of one group of wild capuchin monkeys (Cebus apella nigritus) using feeding platforms in large-scale foraging experiments conducted over four years. Three feeding sites were arrayed in an oblique triangle, such that once the monkey group had chosen one site to feed, they had a choice between two remaining sites, a close one with less food and the other up to 2.3 times as far away but with more food. Sites were provisioned once per day. The capuchins generally chose the closer feeding site, even when the more distant site offered up to 12 times as much food. The distances to, rewards of, or various profitability measures applied to each alternative site individually did not explain the group’s choices in ways consistent with foraging theory or principles of operant psychology. The group’s site choices were predicted only by comparing efficiency measures of entire foraging pathways: (1) direct travel to the more rewarding distant site, versus (2) the longer ‘detour’ through the closer site on the way to the more distant one. The group chose the detour more often when the reward was larger and the added detour distance shorter. They appeared to be more sensitive to the absolute increase in detour distance than to the relative increase compared to the straight route. The qualitative and quantitative results agree with a simple rule: do not use the detour unless the energy gain from extra food outweighs the energy cost of extra travel. These results suggest that members of this group integrate information on spatial location, reward, and perhaps potential food competition in their choice of multi-site foraging routes, with important implications for social foraging. This contribution is part of the special issue “ A Socioecological Perspective on Primate Cognition” (Cunningham and Janson 2007b).     

Tool choice on the basis of rigidity in capuchin monkeys

Wild capuchin monkeys select stone tools to crack open nuts on the basis of their weight and friability, two non-visual functional properties. Here, we investigated whether they would select new stick-like tools on the basis of their rigidity. In Experiment 1, subjects faced an out-of-reach reward and a choice of three unfamiliar tools differing in color, diameter, material, and rigidity. In order to retrieve the reward, capuchins needed to select the rigid tool exemplar. Capuchins gathered information regarding tools’ pliability either by (1) manipulating the tools themselves (manipulation condition), (2) observing a human demonstrator repeatedly bending the tools (observation condition), or (3) seeing the tools placed on a platform without any manipulation taking place (visual static condition). Subjects selected the rigid tool above chance levels in both the manipulation and observation conditions, but not in the visual static condition. In Experiment 2, subjects needed to select and use a flexible tool to access a liquid reward (as opposed to the rigid tool, as in previous experiment). Again, capuchins selected above chance levels the appropriate tool (i.e., flexible), thus demonstrating a good appreciation of the relation between the tool properties and the task requirements.     

Object permanence in orangutans (Pongo pygmaeus), chimpanzees (Pan troglodytes), and children (Homo sapiens).

Juvenile and adult orangutans (n = 5; Pongo pygmaeus), chimpanzees (n = 7; Pan troglodytes), and 19- and 26-month-old children (n = 24; Homo sapiens) received visible and invisible displacements. Three containers were presented forming a straight line, and a small box was used to displace a reward under them. Subjects received 3 types of displacement: single (the box visited 1 container), double adjacent (the box visited 2 contiguous containers), and double nonadjacent (the box visited 2 noncontiguous containers). All species performed at comparable levels, solving all problems except the invisible nonadjacent displacements. Visible displacements were easier than invisible, and single were easier than double displacements. In a 2nd experiment, subjects saw the baiting of either 2 adjacent or 2 nonadjacent containers with no displacements. All species selected the empty container more often when the baited containers were nonadjacent than when they were adjacent. It is hypothesized that a response bias and inhibition problem were responsible for the poor performance in nonadjacent displacements.     

The operation of an acquired drive in satiated rats

A series of four short experiments indicates that the behaviour of satiated rats in a runway, at the end of which they have previously been rewarded, differs significantly from the behaviour of satiated rats without previous reward in the experimental situation. The former group reach the end box more quickly after having been put in the starting box of the runway and if provided with food in the end box proceed to eat it, although they have just refused similar pieces of food in their home cages. This is shown not to be due to defective satiation or the operation of fear in the control group. When runs and feeding in the end box are separated during the training period, the previously rewarded group still shows a more vigorous response on satiated trials, indicating that it is a reward expectancy about the goal box rather than a running habit which has become “functionally autonomous”, acting as a situation-specific drive.     

Socially learned preferences for differentially rewarded tokens in the brown capuchin monkey (Cebus apella)

Social learning is assumed to underlie traditions, yet evidence indicating social learning in capuchin monkeys (Cebus apella), which exhibit traditions, is sparse. The authors tested capuchins for their ability to learn the value of novel tokens using a previously familiar token-exchange economy. Capuchins change their preferences in favor of a token worth a high-value food reward after watching a conspecific model exchange 2 differentially rewarded tokens, yet they fail to develop a similar preference after watching tokens paired with foods in the absence of a conspecific model. They also fail to learn that the value of familiar tokens has changed. Information about token value is available in all situations, but capuchins seem to pay more attention in a social situation involving novel tokens.     

The order of continuous,partial and nonreward trials and resistance to extinction

In the first experiment rats were given either 16 or 48 nonrewarded or continuously rewarded trials prior to 24 continuously or partially rewarded trials, followed by extinction. Increased resistance to extinction was found for increasing numbers of nonrewarded trials when they were followed by partial reward, but not when followed by continuous reward. Similarly, more continuously rewarded trials followed by partial reward tended to increase resistance to extinction. Because of the theoretical importance of the effect of continuous reward followed by partial, a second experiment was performed where the range of the number of continuously rewarded trials was extended to 0, 48, and 96. Contrary to many theoretical predictions, resistance to extinction increased as a function of increasing amounts of continuous reward.     

Dogs are able to solve a means-end task

Dogs, although very skilled in social-communicative tasks, have shown limited abilities in the domain of physical cognition. Consequently, several researchers hypothesized that domestication enhanced dogs’ cognitive abilities in the social realm, but relaxed selection on the physical one. For instance, dogs failed to demonstrate means-end understanding, an important form of relying on physical causal connection, when tested in a string-pulling task. Here, we tested dogs in an “on/off” task using a novel approach. Thirty-two dogs were confronted with four different conditions in which they could choose between two boards one with a reward “on” and another one with a reward “off” (reward was placed next to the board). The dogs chose the correct board when (1) both rewards were placed at the same distance from the dog, when (2) the reward placed “on” the board was closer to the dog, and (3) even when the reward placed “off” the board was much closer to the dog and was food. Interestingly, in the latter case, dogs did not perform above chance, if instead of a direct reward, the dogs had to retrieve an object placed on the board to get a food reward. In contrast to previous string-pulling studies, our results show that dogs are able to solve a means-end task even if proximity of the unsupported reward is a confounding factor.     

Honeybees (Apis mellifera) Remember Two Near-Target Landmark Constellations

A set of experiments tested whether honeybees can remember two different landmark constellations within the same room, or whether the two constellations are mixed together in a single composite memory. Bees were trained to find a reward location with two different landmark arrays. For example, when the blue landmark was north of the yellow one, the reward was to the east; when the yellow landmark was north of the blue one, the reward was to the west. On occasional unrewarded tests, either one of the training arrays was presented (control tests), or else a training array rotated by 90° (rotated tests). A rotated array consisted of part of one training array added to a part of the other training array. Should honeybees form a single composite memory by combining the two training arrays, they should search as much at the target location on rotated tests as on control tests. Results refute this and suggest that they had two separate memories.     

Can dogs defy gravity? A comparison with the human infant and a non-human primate

We conducted four experiments with 56 adult dogs (Canis lupus familiaris) involving tasks where food was dropped through an opaque tube connected either vertically or diagonally to one of two or three goal boxes. In the first experiment, modelled after studies with children and primates, the dogs first searched significantly more often in the location directly beneath the drop-off point (a gravity bias), although this box was not connected with the tube. These results are comparable to those of human infants and cotton-top tamarins. Experiments 2–4 tested which problem solving strategy the dogs applied to find the food. Results show that they do not understand the physical mechanism of the tube itself, and they apply one of three search strategies: search the gravity box (the one below the drop-off box); search the box in the middle; learn the correct location of the goal box. When the goal box was in the same location the dogs learned to search there over trials, that is, they learned to ‘defy gravity’, but when the location of the goal box changed over trials they showed no learning. These findings are compared with those from human infants and cotton-top tamarins: like these species, the dogs can learn to overcome a gravity bias, but only when the reward is to be found in a consistent location.     

Schedules of reward shift in the double runway

Eighty food deprived rats received 62 trials in a double runway. On Trials 1-30, reward in the first goal box (GB1) was either always two food pellets or always zero pellets. All subjects received two pellets in the second goal box (GB2). On Trials 31-62 subjects in each preshift group (GB1 reward or GB1 nonreward) were shifted to the opposite GB1 reward level on 0, 25, 50, 75 or 100% of occasions. GB2 reward remained unaltered in all cases. For subjects experiencing reward decrease, second runway (A2) run and goal speeds after nonreward were generally enhanced, both within-group and in comparison with never rewarded controls. No such effect was evident on A2 start speed, nor was there any evidence to suggest that A2 performance after decreased reward was a function of the schedule of decrease. Increased GB1 reward resulted in general within-group impairment of A2 start and run speeds, with no effect on A2 goal performance. However, comparisons of speeds after increased reward with those of always rewarded controls revealed no difference on A2 start or run but indicated impairment of A2 goal performance. With the 50% schedule of reward increase, A2 run speeds after nonreward (the training level) exceeded those of never rewarded controls. Results are discussed with reference to McHose's contrast account of double runway phenomena and Amsel's frustration theory.     

Do orangutans (<Emphasis Type="Italic">Pongo pygmaeus</Emphasis>) know when they do not remember?

Metacognition refers to the ability to monitor and control one’s own cognitive activities such as memory. Although recent studies have raised an interesting possibility that some species of nonhuman animals might possess such skills, subjects often required a numerous number of training trials to acquire the effective use of metacognitive responses. Here, five orangutans (Pongo pygmaeus) were tested whether they were able to escape spatial memory tests when they did not remember the location of preferred reward in a relatively small number of trials. The apes were presented with two identical cups, under one of which the experimenter hid a preferred reward (e.g., two grapes). The subjects were then presented with a third container, “escape response”, with which they could receive a less preferred but secure reward (e.g., one grape). The orangutans as a group significantly more likely selected the escape response when the baiting of the preferred reward was invisible (as compared to when it was visible) and when the hiding locations of the preferred reward were switched (as compared to when they remained unchanged). Even when the escape response was presented before the final presentation of the memory test, one orangutan successfully avoided the test in which she would likely err. These findings indicate that some orangutans appear to tell when they do not remember correct answers in memory tests.     

Landmark use by Clark’s nutcrackers (Nucifraga columbiana): influence of disorientation and cue rotation on distance and direction estimates

Many species have been shown to encode multiple sources of information to orient. To examine what kinds of information animals use to locate a goal we manipulated cue rotation, cue availability, and inertial orientation when the food-storing Clark’s nutcracker (Nucifraga columbiana) was searching for a hidden goal in a circular arena. Three groups of birds were used, each with a different goal–landmark distance. As the distance between the goal and the landmark increased, nutcrackers were less accurate in finding the correct direction to the goal than they were at estimating the distance (Experiment 1). To further examine what cues the birds were using to calculate direction, the featural cues within the environment were rotated by 90° and the birds were either oriented when searching (Experiments 2 and 3) or disoriented (Experiment 3). In Experiment 4, all distinctive visual cues were removed (both internal and external to the environment), a novel point of entry was used and the birds were either oriented or disoriented. We found that disorienting the nutcrackers so that they could not use inertial cues did not influence the birds’ total search error. The birds relied heavily but not completely on cues within the environment, as rotating available cues caused them to systematically shift their search behavior. In addition, the birds also relied to some extent on Earth-based cues. These results show the flexible nature of cue use by the Clark’s nutcracker. Our study shows how multiple sources of spatial information may be important for extracting multiple bearings for navigation.     

Invisible displacement understanding in domestic dogs (<Emphasis Type="Italic">Canis familiaris</Emphasis>): the role of visual cues in search behavior

Recently, (Collier-Baker E, Davis JM, Suddendorf T (2004) J Comp Psychol 118:421–433) suggested that domestic dogs do not understand invisible displacements. In the present study, we further investigated the hypothesis that the search behavior of domestic dogs in invisible displacements is guided by various visual cues inherent to the task rather than by mental representation of an object’s past trajectory. Specifically, we examined the role of the experimenter as a function of the final position of the displacement device in the search behavior of domestic dogs. Visible and invisible displacement problems were administered to dogs (N = 11) under two conditions. In the Visible-experimenter condition, the experimenter was visible whereas in the Concealed-experimenter condition, the experimenter was visibly occluded behind a large rigid barrier. Our data supported the conclusion that dogs do not understand invisible displacements but primarily search as a function of the final position of the displacement device and, to a lesser extent, the position of the experimenter.     

奖赏性线索启动和调控视觉搜索额顶网络的神经机制

近期关于奖赏性信息与视觉搜索任务的交互作用研究体现在两个层面, 其一是施加外部奖赏能够影响被试的视觉搜索反应时和正确率, 其二是搜索目标本身具有奖赏特性也能够影响行为反应。但是, 奖赏性信息对负责视觉搜索的额顶注意网络的动态启动和调控作用仍不清楚。本研究结合Posner空间预提示范式和视觉搜索范式, 采用快速事件相关功能磁共振成像技术, 分别操纵线索提示阶段的线索奖赏属性(奖赏性vs.非奖赏性)和线索空间注意属性(有效提示目标位置vs.无效提示目标位置), 并操纵目标搜索阶段的目标属性(奖赏性vs.非奖赏性), 探讨以下两个核心问题：(1)在线索提示阶段, 奖赏性线索产生的自上而下注意准备信号的神经基础, 以及该神经基础与传统的空间注意准备信号的异同; (2)在目标搜索阶段, 奖赏性线索对负责视觉搜索的额顶注意网络产生的启动和调控作用, 重点考察对正确的奖赏性信息的定向过程和对错误的奖赏性信息的重新定向过程。     

Landmark-based spatial search in honeybees. II.

Honeybees were trained to find sugar water at a fixed location in front of a rectangular block or a gap in a row of rectangular blocks aligned east-west. Bees learned to use both blocks and gaps to pinpoint a target place. In approach, the bees tended to head directly towards the block but not the gap. In approaching the gap, they tended to approach the wall, and then fly parallel to the wall until the gap was encountered. In approaching the block, they approached from varying directions. When the width of the block or gap was doubled, bees still searched at the same perpendicular distance from the landmark. When the height of the block was doubled, they searched farther away but not as far as double the distance on control tests, with variations across individuals. Near the target area, the bees tended to face almost parallel to the gap or block, turned slightly towards the landmark by 0–30°. In all setups, bees showed a tendency to search near the landmark, and to search on average closer than the training distance. The results confirm the basic processes identified for landmark-based search in bees, but show variations in how the steps are carried out.     

Do cleaner fish learn to feed against their preference in a reverse reward contingency task?

The ability to control impulsive behaviour has been studied in animals with a standard test in which subjects need to choose the smaller of two food items in order to receive the larger one (reverse reward contingency). As a variety of mammals that have been tested so far (mostly primates) have great difficulties to solve the task, it has been proposed that it is generally cognitively demanding. However, according to an ecological approach to cognition, a species’ ability to solve the task should not depend on its general cognitive abilities but on whether its ecology causes selective pressure on the ability to restrain foraging behaviour. We tested this hypothesis using the cleaner wrasse (Labroides dimidiatus), a fish species that feeds against its preference in nature when engaging in cleaning interactions with so called ‘client fish’. None of the eight tested individuals learned to choose a non-preferred item after 200 trials. In a subsequent test, one subject learned to respond correctly in a large or none contingency task (only the choice of the small food was rewarded). After a short re-experience treatment, this individual learned to solve the reverse reward task after 30 trials. In conclusion, we did not find support for the general idea that interactions with clients prepared cleaners to quickly solve a reverse reward test. However, the results suggest that the potential to solve a reverse reward contingency may not be restricted to mammals but could be present also in a fish species in which the problem of choosing a non-preferred food over a preferred one is an ever present challenge in nature.     

Acquisition and forgetting in monkeys' memory of informational object-reward associations

The first experiment showed the monkeys could recall whether an object had been rewarded with peanuts or with sultanas, two equally preferred foods. The second investigated the effect of rewarded trials with an object on monkeys' ability to recall a nonrewarded trial with the same object. The third demonstrated that monkeys could use the memory of reward to predict nonreward and the memory of nonreward to predict reward, in a Win-Shift Lose-Stay paradigm. The fourth found differences between Win-Shift Lose-Stay and Win-Stay Lose-Shift in the rate at which associations between objects and reward events were forgotten. These results are discussed in relation to D. L. Medin's (In A. M. Schrier, Ed., Behavioral primatology, Hillsdale, N.J.: Erlbaum, 1977, Vol. I, pp. 33–69) distinction between informational and hedonic effects of reward in monkeys. It is argued that the association between an object and a reward event is represented in memory by many independent traces, different traces recording the object's association with different attributes of the reward event.     

Human risky choice in a repeated-gambles procedure: an up-linkage replication of Lakshminarayanan, Chen and Santos (2011)

Lakshminarayanan et al. (J Exp Soc Psychol 47: 689–693, 2011) showed that when choice is between variable (risky) and fixed (safe) food amounts with the same expected values, capuchins prefer the safe alternative if choice is framed as a gain, but the risky alternative if it is framed as a loss. These results seem similar to those seen in human prospect-theory tests in choice between variable and fixed gains or losses. Based on this similarity, they interpreted their results as identifying a between-species commonality in cognitive function. In this report, we repeat their experiment with humans as subjects (an up-linkage replication). Whether choices were rewarded with candy or nickels, choice approximated indifference whether framed as gains or losses. Our data mirror those of others who found that when humans make risky choices within a repeated-trials procedure without verbal instruction about outcome likelihoods, preference biases seen in one-shot, language-guided, prospect-theory tests such as Tversky and Kahneman’s (Science 211:453–458, 1981) reflection effect may not appear. The disparity between our findings and those of Lakshminarayanan et al. suggests their study does not evidence a cognitive process shared by humans and capuchins.