Abstract: Previous empirical work on animal social learning has found that many species lack the ability to learn entire action sequences solely through reliance on social information. Conversely, acquiring action sequences through asocial learning can be difficult due to the large number of potential sequences arising from even a small number of base actions. In spite of this, several studies report that some primates use action sequences in the wild. We investigate how social information can be integrated with asocial learning to facilitate the learning of action sequences. We formalize this problem by examining how learners using temporal difference learning, a widely applicable model of reinforcement learning, can combine social cues with their own experiences to acquire action sequences. The learning problem is modeled as a Markov decision process. The learning of nettle processing by mountain gorillas serves as a focal example. Through simulations, we find that the social facilitation of component actions can combine with individual learning to facilitate the acquisition of action sequences. Our analysis illustrates that how even simple forms of social learning, combined with asocial learning, generate substantially faster learning of action sequences compared to asocial processes alone, and that the benefits of social information increase with the length of the action sequence and the number of base actions.

Abstract: Innovations and problem-solving abilities can provide animals with important ecological advantages as they allow individuals to deal with novel social and ecological challenges. Innovation is a solution to a novel problem or a novel solution to an old problem, with the latter being especially difficult. Finding a new solution to an old problem requires individuals to inhibit previously applied solutions to invent new strategies and to behave flexibly. We examined the role of experience on cognitive flexibility to innovate and to find new problem-solving solutions with an artificial feeding task in wild redfronted lemurs (Eulemur rufifrons). Four groups of lemurs were tested with feeding boxes, each offering three different techniques to extract food, with only one technique being available at a time. After the subjects learned a technique, this solution was no longer successful and subjects had to invent a new technique. For the first transition between task 1 and 2, subjects had to rely on their experience of the previous technique to solve task 2. For the second transition, subjects had to inhibit the previously learned technique to learn the new task 3. Tasks 1 and 2 were solved by most subjects, whereas task 3 was solved by only a few subjects. In this task, besides behavioral flexibility, especially persistence, i.e., constant trying, was important for individual success during innovation. Thus, wild strepsirrhine primates are able to innovate flexibly, suggesting a general ecological relevance of behavioral flexibility and persistence during innovation and problem solving across all primates.

Abstract: Social animals should have plenty of opportunities to learn from conspecifics, but most studies have failed to document social learning in horses. This study investigates whether young Icelandic horses can learn a spatial detour task through observation of a trained demonstrator horse of either the same age (Experiments 1 and 2, n = 22) or older (Experiment 3, n = 24). Observer horses were allowed to observe the demonstrator being led three times through the detour route immediately before being given the opportunity to solve the task themselves. Controls were allowed only to observe the demonstrator horse eating at the final position, but not the demonstration of the route. Although we found a tendency towards better performance by observer horses in the second experiment, we were unable to repeat this result in a similar set-up with a new group of horses and older, dominant demonstrator horses. We conclude that horses exposed to prior demonstration did not perform better than control horses in solving spatial detour tasks.

Abstract: Highlights
�Horses remember the location of food hidden by the experimenter after a delay.
�They understand the communicative meaning of a human positioned close to the target.
�The same horses are capable of changing their decision-making strategy.
�They are able to shift from accuracy inferred from human given cues to speed.
�Horses can use human cues or not depending on time, cost, experience and reward.
Abstract
To date, horses have seemed capable of using human local enhancement cues only when the experimenter remains close to the reward, since they fail to understand the communicative meaning of the human as momentary local enhancement cue (when the human is not present at the moment of the animal's choice). This study was designed to analyse the ability of horses to understand, remember and use human-given cues in a delayed (10 s) three-choice task. Twelve horses (experimental group) had to find a piece of carrot hidden under one of three overturned buckets after seeing the experimenter hide it. The results were then compared with those of a control group (twelve horses) that had to find the carrot using only the sense of smell or random attempts. At the beginning, the experimental horses made more correct choices at the first attempt, although they took more time to find the carrot. Later the same horses were less accurate but found the carrot in less time. This suggests that the value of the proximal momentary local enhancement cues became less critical. It seemed, in fact, that the experimental and control group had aligned their behaviour as the trials proceeded. Despite this similarity, in the second half of the trials, the experimental group tended to first approach the bucket where they had found the carrot in the immediately preceding trial. Our findings indicate that horses are capable of remembering the location of food hidden by the experimenter after a delay, by using the human positioned close to the target as valuable information. The same horses are also capable of changing their decision-making strategy by shifting from the accuracy inferred from human given cues to speed. Therefore, horses are able to decide whether or not to use human given-cues, depending on a speed-accuracy trade-off.

Abstract: Most horses are trained by negative reinforcement. Currently, however, no standardised test for evaluating horses' negative reinforcement learning ability is available. The aim of this study was to develop an objective test to investigate negative reinforcement learning in horses. Twenty-four Icelandic horses (3 years old) were included in this study. The horses were tested in a pressure-release task on three separate days with 10, 7 and 5 trials on each side, respectively. Each trial consisted of pressure being applied on the hindquarter with an algometer. The force of the pressure was increased until the horse moved laterally away from the point of pressure. There was a significant decrease in required force over trials on the first test day (P<0.001), but not the second and third day. The intercepts on days 2 and 3 differed significantly from day 1 (P<0.001), but not each other. Significantly stronger force was required on the right side compared to the left (P<0.001), but there was no difference between first and second side tested (P=0.56). Individual performance was evaluated by median-force and the change in force over trials on the first test day. These two measures may explain different characteristics of negative reinforcement learning. In conclusion, this study presents a novel, standardised test for evaluating negative reinforcement learning ability in horses.