Learning of safety by a social fish: applications for studying post-traumatic stress in humans

Abstract

As animal behaviour theory has developed over the past 70 years, much attention has gone towards social behaviour. While our basic knowledge of social systems has grown substantially, it has rarely been applied to human issues. Here, we attempted to bridge the gap between animal behaviour theory and human psychology by conducting social experiments involving fish. As in many species, minnows (family: Cyprinidae) repeatedly exposed to risky situations can develop a behavioural phenotype characterized by neophobic tendencies, pacing and stereotypic behaviours. Here, we tested whether the simple presence of calm (or un-calm) conspecific models could lead to a weakening of the high-risk phenotype in minnows that acquired fear either in isolation or within a group. We first documented that the social context of risk exposure impacted the intensity of the high-risk phenotype, with minnows exposed to risk in isolation showing stronger high-risk traits compared to those that were exposed to risk in groups. However, individuals exposed to risk in isolation were more influenced by calm models, despite their more pronounced phenotype. We argue that group exposure led to social reinforcement of risk, which in turn decreased the information transfer about safety in these individuals. We also demonstrate that a group of calm models, and not un-calm models, was required to weaken the high-risk phenotype. These findings highlight the interplay between social reinforcement of risk and safety in social groups and the impact of groups on information transmission. Moreover, our results parallel anecdotal reports of successes or failures of social therapies for post-traumatic stress (PTS) patients based on the social context of symptom acquisition, suggesting that understanding the transfer of information in social animals could prove fruitful in understanding and modelling PTS recovery.