Observational learning in capuchin monkeys: a video deficit effect

Six adult male rhesus monkeys did not learn from the choices of a conspecific shown in videos

There is substantial evidence of group-specific behaviors in wild animals that are thought to be socially transmitted. Yet experimental studies with monkeys have reported conflicting evidence on the extent to which monkeys learn by observing their conspecifics. In this study, we tested the feasibility of using pre-recorded video demonstrations to investigate social learning from conspecifics in rhesus monkeys. With training, monkeys gradually learned to respond correctly following videos of a demonstrator, however, follow-up experiments revealed that this was not due to learning from the demonstrator monkey. In generalization tests with videos that were horizontally reversed, monkeys continued responding to the location they had associated with each video, rather than matching the new choice location shown in the mirrored video. When the task was changed to make location irrelevant, such that monkeys could choose correctly only by selecting the same image selected by the demonstrator in the video, observer monkeys did not exceed chance in 12,000 training trials. Because monkeys readily learn to follow nonsocial visual cues presented on a monitor to guide image choice, their inability to learn from a demonstrator here indicates substantial limitations in the capacity for social learning from videos. Furthermore, these findings encourage deeper consideration of what monkeys perceive when presented with video stimuli on computer screens.

Exploring How White-Faced Sakis Control Digital Visual Enrichment Systems

Simple Summary

Many zoo-housed primates use visual computer systems for enrichment but little is known about how monkeys would choose to control these devices. Here we investigate what visual enrichment white-faced saki monkeys would trigger and what effect these videos have on their behaviour. To study this, we built an interactive screen device that would trigger visual stimuli and track the sakis’ interactions when using the system. Over several weeks, we found that the sakis would trigger underwater and worm videos significantly more than animal, abstract art and forest videos, and the control condition of no-stimuli. Further, the sakis triggered the animal video significantly less often over all other conditions. Yet, viewing their interactions over time, the sakis’ usage of the device followed a bell curve, suggesting novelty and habituation factors. As such, it is unknown if the stumli or devices novelty and habituation caused the changes in the sakis interactions over time. These results also indicated that the different visual stimuli conditions significantly reduced the sakis’ scratching behaviour with the visual stimuli conditions compared to the control condition. Further, the usage of visual stimuli did not increase the sakis’ looking at and sitting in front of the screen behaviours. These results highlight problems in defining interactivity and screen usage with monkeys and screens from looking behaviours and proximity alone.

Abstract

Computer-enabled screen systems containing visual elements have long been employed with captive primates for assessing preference, reactions and for husbandry reasons. These screen systems typically play visual enrichment to primates without them choosing to trigger the system and without their consent. Yet, what videos primates, especially monkeys, would prefer to watch of their own volition and how to design computers and methods that allow choice is an open question. In this study, we designed and tested, over several weeks, an enrichment system that facilitates white-faced saki monkeys to trigger different visual stimuli in their regular zoo habitat while automatically logging and recording their interaction. By analysing this data, we show that the sakis triggered underwater and worm videos over the forest, abstract art, and animal videos, and a control condition of no-stimuli. We also note that the sakis used the device significantly less when playing animal videos compared to other conditions. Yet, plotting the data over time revealed an engagement bell curve suggesting confounding factors of novelty and habituation. As such, it is unknown if the stimuli or device usage curve caused the changes in the sakis interactions over time. Looking at the sakis’ behaviours and working with zoo personnel, we noted that the stimuli conditions resulted in significantly decreasing the sakis’ scratching behaviour. For the research community, this study builds on methods that allow animals to control computers in a zoo environment highlighting problems in quantifying animal interactions with computer devices.

Squirrel monkey responses to information from social demonstration and individual exploration using touchscreen and object choice tasks

We aimed to study whether a non-human primate species responded differently to information acquired socially compared with that acquired individually. To do so, we attempted to train squirrel monkeys to perform binary discriminations. These involved exposure to either social information (human or puppet demonstrator performs an initial 'information trial') or individual exploration (monkey performs information trial as well as subsequent test trials). In Experiment 1, we presented the task on a touchscreen tablet. Only one monkey appeared to learn the significance of the information trial, and across the group there was no improvement in performance over sessions. The proficient individual showed little evidence of successful transfer to three-way discrimination problems, suggesting limited representation of the task structure. In Experiment 2, we used a logically identical task, presented as a physical object choice (inverted cups concealing a food reward). No monkeys learned to use the information trial cues, and success again did not increase over sessions. We concluded that the monkeys' poor performance in Experiment 1 was not attributable to the mode of presentation (touchscreen), but reflected real difficulties with mastering the task structure. For both experiments, we analysed the monkeys' spontaneous responses to the different trial types (social-win, social-lose, individual-win, and individual-lose). We found that monkeys had a tendency to repeat selections made during the information trial, whether these were made by themselves or by a demonstrator. This tendency to repeat was observed even following lose trials (i.e. when incorrect). Apparent 'success' following win trials was probably largely an artefact of behavioural inertia (individual learning conditions) and stimulus enhancement (social learning conditions), rather than sensitivity to the reward cues associated with that stimulus. Although monkeys did respond somewhat differently (more repeats) following win trials, compared with lose trials, this was no more apparent in the object choice task than the touchscreen task, again suggesting that the less ecologically valid presentation medium did not actively disrupt potential for learning the discrimination rule. Both touchscreen and physical object choice tasks appear to be valid methods to study learning in squirrel monkeys, with neither method giving a clear performance advantage over the other. However, this population did not master the contingencies in these tasks.