Search

Despite the domestication of sheep and goats by humans for several millennia, we still lack comparative data on their cognitive capacity. Comparing the cognitive skills of farm animals can help understand the evolution of cognition. In this study, we compared the performances of sheep and goats in inference by exclusion tasks. We implemented two tasks, namely a cup task and a tube task, to identify whether success in solving the task could be attributed to either low-level mechanisms (avoiding the empty location strategy) or to deductive reasoning (if two possibilities A and B, but not A, then it must be B). In contrast to a previous study comparing goats and sheep in a cup task, we showed that both species solved the inferential condition with high success rates. In the tube task, performances could not be explained by alternative strategies such as avoiding the empty tube or preferring the bent tube. When applying a strict set of criteria concerning responses in all conditions and controlling for the potential effects of experience, we demonstrate that two individuals, a goat and a sheep, fulfil these criteria. This suggests that sheep and goats are able to make inferences based on deductive reasoning., (© 2022. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.)

Recruitment is a process by which animals can initiate collective movements: the action of an individual prompts conspecifics to follow. Although it has been hypothesized that animals may be able to intentionally recruit others, there is no experimental evidence of this to date. We tested this hypothesis in two pairs of Tonkean macaques in a situation requiring the subjects to find a food site in a 2800 m 2 area, and approach the site together to release rewards. Each subject was informed of the location of either highly or little-valued rewards. We recorded attention-action sequences in which an individual checked that his partner was attending to him before moving, and also simple departures (i.e., not preceded by eye contact). Analyses showed that sequences were more often followed by recruitment and leading the partner to a baited site than simple departures were. Moreover, subjects used attention-action sequences more frequently when informed of the location of the highly valued reward. This may be explained by the fact that the more motivated they were by the expected rewards, the more likely they were to actively recruit their partner. No such effect was found when subjects performed simple departures. We conclude that Tonkean macaques are capable of intentional recruitment because the subjects voluntarily behaved with the goal of influencing their partner's movement: they checked that the partner was paying attention to them and prompted him to follow by moving. Such performances can be accounted for either by associative learning or by intentional communication., (© 2022. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.)

In species that move in cohesive groups, animals generally reach decisions through socially distributed processes, and individual knowledge is expected to influence collective decision making. Pooling of information should not be considered a general rule, however, since conflicts of interest may occur between group members. When resources are limited or highly attractive, higher-ranking individuals can prevent others from accessing food, and subordinates may have an interest in withholding information about its location. We investigated the role individual knowledge may play in recruitment processes in four groups of horses (Equus caballus). Animals were repeatedly released in a food search situation, in which one individual had been informed about the location of a preferred food, while another was a naïve control subject. Horses that were informed about the location of the feeding site were seen to approach the food source more steadily and were followed by a higher number of group members than their uninformed counterparts. Recruitment processes appeared mostly passive. Among the informed subjects, lower-ranking individuals were overall less followed than the higher-ranking ones. Most lower-ranking horses arrived alone at the feeding site. Non-followed informed subjects spent less time in active walk and used direct paths less frequently than followed animals, and they were joined by fewer partners at the attractive food source and spent more time feeding alone. This indicates that the influence of informed individuals on the behavior of other group members was a mixed process. Some horses brought nutritional benefits to their conspecifics by leading them to food supplies, whereas the behavior of others might be functionally deceptive.

By distinguishing the attentional cues of their mates, animals can learn what part of their environment is of potential interest. However, recognizing the attentional states of others through auditory perception appears to be difficult, since these states are generally not accompanied by ostensive signals liable to reveal them. In this context, one study found that rhesus macaques withhold their action when unobserved, then concluding that they know what others can hear. We replicated this study by testing 18 Tonkean macaques in an experimental setting where subjects had to choose between two boxes containing a food reward. One box made a noise when opened, and the other opened silently. An experimenter was present and could either have her back to subjects or be facing them. If subjects aimed to avoid attracting the experimenter's attention, they were expected to select the silent box while the experimenter's back was turned. Results showed that subjects did not discriminate between boxes, whatever the experimenter's position. We thus found no evidence that Tonkean macaques are able to take the auditory attentional states of a human into account. It is therefore premature to conclude that monkeys can establish a link between hearing and knowing.

Whether or not non-human animals can plan for the future is a hotly debated issue. We investigate this question further and use a planning-to-exchange task to study future planning in the cooperative domain in two species of monkeys: the brown capuchin (Cebus apella) and the Tonkean macaque (Macaca tonkeana). The rationale required subjects to plan for a future opportunity to exchange tokens for food by collecting tokens several minutes in advance. Subjects who successfully planned for the exchange task were expected to select suitable tokens during a collection period (5/10 min), save them for a fixed period of time (20/30 min), then take them into an adjacent compartment and exchange them for food with an experimenter. Monkeys mostly failed to transport tokens when entering the testing compartment; hence, they do not seem able to plan for a future exchange with a human partner. Three subjects did however manage to solve the task several times, albeit at very low rates. They brought the correct version of three possible token types, but rarely transported more than one suitable token at a time. Given that the frequency of token manipulation predicted transport, success might have occurred by chance. This was not the case, however, since in most cases subjects were not already holding the token in their hands before they entered the testing compartment. Instead, these results may reflect subjects' strengths and weaknesses in their time-related comprehension of the task.

Exchanges form the basis of human economies. Animals too can engage in reciprocal interactions but they do not barter goods like humans, which raises the question of the abilities necessary for trading to occur. Previous studies have shown that non-human primates can exchange food with human partners. Here, we tested the ability of brown capuchin monkeys and Tonkean macaques to reciprocate in a task requiring two conspecifics to exchange tokens in order to obtain rewards from an experimenter. We recorded 56 transfers between subjects in capuchin monkeys and 10 in Tonkean macaques. All transfers were passive in both species. Capuchins preferentially picked up tokens valuable for them in the partner's compartment. They tended to manipulate the partner-valued tokens more often than the no-value ones, leading to more opportunities for these tokens to end up within reach of the partner. Despite optimal conditions where values of goods were defined and known by partners, however, none of the pairs tested engaged in short-term reciprocal interactions. These results indicate that calculated reciprocity was difficult if not impossible in the animals tested.

The ability of animals to delay gratification is crucial for complex goal-directed action. It may help them in making effective decisions when facing a choice. We tested the ability of nine long-tailed macaques (Macaca fascicularis) to delay gratification in several experiments. In exchange tasks, subjects had to keep a small piece of cookie before returning it to an experimenter in order to get a larger food amount. Results showed that animals could wait between 10 s and 10 min depending on individual and sizes of reward. In another experiment, subjects could immediately give back the initial piece of cookie then wait for the return. Their performances more than doubled, demonstrating the role of consumption inhibition in postponing gratification. Such achievements underscore delays of gratification which until now were not thought possible in monkeys. Finally, subjects were presented with an accumulation of food pieces added at short intervals until they seized them. They mostly waited between 30 s and 1 min, which points at the consistency of our data, compared to those of other studies. Our results indicate that long-tailed macaques anticipated the duration of delays. We may account for their remarkable performances by their achievements in the social context.

The transmission of tool use is a rare event in monkeys. Such an event arose in a group of semi-free-ranging Tonkean macaques (Macaca tonkeana) in which leaning a pole against the park's fence (branch leaning) appeared and spread to several males. This prompted us to test individual and social learning of this behavior in seven young males. In the first experiment, three males learned individually to obtain a food reward using a wooden pole as a climbing tool. They began using the pole to retrieve the reward only when they could alternatively experience acting on the object and reaching the target. In a second experiment, we first tested whether four other subjects could learn branch leaning after having observed a group-mate performing the task. Despite repeated opportunities to observe the demonstrator, they did not learn to use the pole as a tool. Hence we exposed the latter subjects to individual learning trials and they succeeded in the task. Tool use was not transmitted in the experimental situation, which contrasts with observations in the park. We can conclude that the subjects were not able to recognize the target as such. It is possible that they recognized it and learned the task individually when we alternated the opportunity to act upon the object and to reach the reward. This suggests that these macaques could then have associated the action they exercised upon the pole and the use of the pole as a means to reach the reward.