The unique challenges faced by animals living in zoos can lead to the production of anxiety-related behaviours. In this study we aimed to understand what specific factors may cause chimpanzees to display these behaviours. In non-human primates, displacement behaviours, such as self-scratching and yawning, are considered markers of anxiety and stress, and Regurgitation and Reingestion (R/R) is considered an abnormal behaviour with negative consequences for physical health. We examined the possible triggers of R/R, scratching, and yawning in a group of zoo-housed chimpanzees and followed this up with an analysis of long-term data to examine further aspects of R/R behaviour. In the first study we conducted focal observations on 18 adult chimpanzees at Edinburgh Zoo, UK, in addition to all occurrence sampling of visitors using flash photography, screaming and banging on the glass in the exhibit. 158 h of data were analysed and Generalised Linear Mixed Models revealed that yawning was significantly more likely if there was a long period of time since the last feed and when there were moderate numbers of visitors in the zoo. There were trends that yawning was more likely to occur if children screamed and that scratching was more likely to occur if visitors used flash photography. R/R occurred most often within 40 min of a feed, but was not affected by the inter-feed interval preceding that feed, positive or negative social interactions, or visitor numbers or behaviour. As there was no obvious daily trigger for R/R, an analysis of long-term data (2009 to 2015) was conducted to investigate if social or dietary factors affected rates of R/R over a larger timescale. It was found that R/R rates in the months before a significant diet change were not different from R/R rates in the months after, but it was found that R/R rates decreased over the five-year period. Lastly, we found no evidence that the introduction of individuals engaging in R/R lead to resident chimpanzees habitually adopting the behaviour, despite considerable opportunities to observe it. These findings have implications for welfare interventions aimed to reduce R/R and/or anxiety behaviours in captive populations and for the translocation of individuals that are known to engage in R/R between groups.
