Search

This project received funding from the European Union's 8th Framework Programme, Horizon 2020 (grant agreement number: 802719) and the St Andrews Restarting Research Funding Scheme (2020). 1. Studying animal behaviour allows us to understand how different species and individuals navigate their physical and social worlds. Video coding of behaviour is considered a gold standard: allowing researchers to extract rich nuanced behavioural datasets, validate their reliability, and for research to be replicated. However, in practice, videos are only useful if data can be efficiently extracted. Manually locating relevant footage in 10,000 s of hours is extremely time-consuming, as is the manual coding of animal behaviour, which requires extensive training to achieve reliability. 2. Machine learning approaches are used to automate the recognition of patterns within data, considerably reducing the time taken to extract data and improving reliability. However, tracking visual information to recognise nuanced behaviour is a challenging problem and, to date, the tracking and pose-estimation tools used to detect behaviour are typically applied where the visual environment is highly controlled. 3. Animal behaviour researchers are interested in applying these tools to the study of wild animals, but it is not clear to what extent doing so is currently possible, or which tools are most suited to particular problems. To address this gap in knowledge, we describe the new tools available in this rapidly evolving landscape, suggest guidance for tool selection, provide a worked demonstration of the use of machine learning to track movement in video data of wild apes, and make our base models available for use. 4. We use a pose-estimation tool, DeepLabCut, to demonstrate successful training of two pilot models of an extremely challenging pose estimate and tracking problem: multi-animal wild forest-living chimpanzees and bonobos across behavioural contexts from hand-held video footage. 5. With DeepWild we show that, without requiring specific expertise in machine learning, pose estimation and movement tracking of free-living wild primates in visually complex environments is an attainable goal for behavioural researchers. Publisher PDF

Animal communication has long been thought to be subject to pressures and constraints associated with social relationships. However, our understanding of how the nature and quality of social relationships relates to the use and evolution of communication is limited by a lack of directly comparable methods across multiple levels of analysis. Here, we analysed observational data from 111 wild groups belonging to 26 non-human primate species, to test how vocal communication relates to dominance style (the strictness with which a dominance hierarchy is enforced, ranging from 'despotic' to 'tolerant'). At the individual-level, we found that dominant individuals who were more tolerant vocalized at a higher rate than their despotic counterparts. This indicates that tolerance within a relationship may place pressure on the dominant partner to communicate more during social interactions. At the species-level, however, despotic species exhibited a larger repertoire of hierarchy-related vocalizations than their tolerant counterparts. Findings suggest primate signals are used and evolve in tandem with the nature of interactions that characterize individuals' social relationships. Publisher PDF

Despite important similarities having been found between human and animal communication systems, surprisingly little research effort has focussed on whether the cognitive mechanisms underpinning these behaviours are also similar. In particular, it is highly debated whether signal production is the result of reflexive processes, or can be characterized as intentional. Here, we critically evaluate the criteria that are used to identify signals produced with different degrees of intentionality, and discuss recent attempts to apply these criteria to the vocal, gestural and multimodal communicative signals of great apes and more distantly related species. Finally, we outline the necessary research tools, such as physiologically validated measures of arousal, and empirical evidence that we believe would propel this debate forward and help unravel the evolutionary origins of human intentional communication. This article is part of the theme issue ‘What can animal communication teach us about human language?’

Cross-species comparison of great ape gesturing has so far been limited to the physical form of gestures in the repertoire, without questioning whether gestures share the same meanings. Researchers have recently catalogued the meanings of chimpanzee gestures, but little is known about the gesture meanings of our other closest living relative, the bonobo. The bonobo gestural repertoire overlaps by approximately 90% with that of the chimpanzee, but such overlap might not extend to meanings. Here, we first determine the meanings of bonobo gestures by analysing the outcomes of gesturing that apparently satisfy the signaller. Around half of bonobo gestures have a single meaning, while half are more ambiguous. Moreover, all but 1 gesture type have distinct meanings, achieving a different distribution of intended meanings to the average distribution for all gesture types. We then employ a randomisation procedure in a novel way to test the likelihood that the observed between-species overlap in the assignment of meanings to gestures would arise by chance under a set of different constraints. We compare a matrix of the meanings of bonobo gestures with a matrix for those of chimpanzees against 10,000 randomised iterations of matrices constrained to the original data at 4 different levels. We find that the similarity between the 2 species is much greater than would be expected by chance. Bonobos and chimpanzees share not only the physical form of the gestures but also many gesture meanings., Author summary Bonobos and chimpanzees are closely related members of the great ape family, and both species use gestures to communicate. We are able to deduce the meaning of great ape gestures by looking at the ‘Apparently Satisfactory Outcome’ (ASO), which reflects how the recipient of the gesture reacts and whether their reaction satisfies the signaller; satisfaction is shown by the signaller ceasing to produce more gestures. Here, we use ASOs to define the meaning of bonobo gestures, most of which are used to start or stop social interactions such as grooming, travelling, or sex. We then compare the meanings of bonobo gestures with those of chimpanzees and find that many of the gestures share the same meanings. Bonobos and chimpanzees could, in principle, understand one another’s gestures; however, more research is necessary to determine how such gestures and gesture meanings are acquired.

Research on great ape gestural communication has hitherto prioritised the signaller, focusing on the expressed repertoire, the set of gesture types that an individual deploys. In this study of wild bonobos (Pan paniscus), we also examine the understood repertoire, the set of gesture types that a recipient understands: a recipient understands a gesture if they respond to the signaller with an Apparently Satisfactory Outcome (ASO). Combining the understood and expressed repertoires substantially expands an individual’s overall repertoire. Hence, we propose that including the understood repertoire better represents an individual’s actual communicative repertoire, and may for the first time reveal gesture types received exclusively by specific subsets of individuals. Significant differences between expressed and understood repertoires might result from two things: (1) gestures are ritualised between signallers and recipients and may therefore be unidirectional in some cases, or (2) the circumstances of use differ among age-sex groups, limiting the need for an individual to use a certain gesture type. In the latter, we expect consistency in particular age-sex groups whereas unidirectional comprehension as a result of ritualization would predict random distribution among individuals. We further compare our dataset with and without gesture instances occurring in play, as play may be a context in which individuals deploy gestures that they rarely or never use otherwise. Assessing the understood repertoire is a useful tool for animal communication researchers, not only allowing shorter study-periods, but also to discover gesture types specific to subsets of individuals, which may reveal much about a species’ social behaviour.

Funding was provided by the Wenner-Gren Foundation Dissertation Fieldwork Grant, Japan Society for the Promotion of Science (JSPS) Grant in Aid for Scientific Research, JSPS Core-to-Core Program, and University of St Andrews 600th Anniversary Scholarship. In animal communication, signallers and recipients are typically different: each signal is given by one subset of individuals (members of the same age, sex, or social rank) and directed towards another. However, there is scope for signaller-recipient interchangeability in systems where most signals are potentially relevant for all age-sex groups, such as great ape gestural communication. In this study of wild bonobos (Pan paniscus), we aimed to discover whether their gestural communication is indeed a mutually understood communicative repertoire, in which all individuals can act as both signallers and recipients. While past studies have only examined the expressed repertoire, the set of gesture types that a signaller deploys, we also examined the understood repertoire, the set of gestures to which a recipient reacts in a way that satisfies the signaller. We found that most of the gestural repertoire was both expressed and understood by all age and sex groups, with few exceptions, suggesting that during their lifetimes all individuals may use and understand all gesture types. Indeed, as the number of overall gesture instances increased, so did the proportion of individuals estimated to both express and understand a gesture type. We compared the community repertoire of bonobos to that of chimpanzees, finding an 88% overlap. Observed differences are consistent with sampling effects generated by the species’ different social systems, and it is thus possible that the repertoire of gesture types available to Pan is determined biologically. Publisher PDF