Benefits of social bonds in domestic horses (Equus caballus): The effect of social context on behaviour and cardiac activity

Horses (Equus spp.) are one of the few species that form close bonds between unrelated and non-reproductive conspecifics. These are characterised by a preference towards specific group members, spatial proximity, and high rates of affiliative interactions and can provide individuals with different benefits. The ultimate benefits are well evidenced, such as increased reproductive success, health, and longevity. However, it is less understood whether these bonds provide benefits on a proximate level, such as social buffering. Thereby, the proximity to, or interactions with, a closely bonded conspecific during or after a stressful situation can provide support and facilitate a decrease of the behavioural and physiological stress response. The aim of this research was to address this knowledge gap by investigating whether domestic horses (E. caballus) benefit from social buffering provided by closely bonded, but unrelated, conspecifics. Through a combination of behavioural observations and heart rate recordings, bond-related buffering effects were investigated in three contexts: (1) during social interactions among group members, (2) during separation from the group, and (3) during novel object exposures. In all contexts, it was assessed whether the interaction with, or the presence of, a closely bonded conspecific would affect the behaviour and cardiac activity of the horses under study. The analysis showed that group-housed horses primarily engaged in low-intensity agonistic interactions. These did not facilitate a pronounced stress response. Social rank but not social bonds between the horses had a small effect on their heart rate during agonistic interactions. The heart rate in receivers of threats was slightly higher the higher the initiator was in rank. Grooming was the only affiliative interaction that corresponded with a lower heart rate. This effect was independent of the bond strength between the horses. During the separation, the horses’ behaviour and heart rate were not affected by the bond relationship to the support provider. Their vigilance during separation was lower when with any conspecific than when alone. Moreover, the rate of affiliative interactions and spatial proximity did not differ between closely bonded and less closely bonded horses and both did not affect their heart rate or heart rate variability. The novel object tests represented a mild stressor. Upon the initial exposure to a novel object, the horses showed proximity seeking to closely bonded conspecifics. Throughout the remainder of the object tests, the horses’ behavioural and physiological stress response and their exploratory behaviour were independent of their social context. These findings make a novel contribution to knowledge as they reveal no strong evidence that close social bonds among unrelated horses provided social buffering in the specified contexts. Rather, the horses’ behavioural and physiological responses were generalised towards closely bonded and less closely bonded members of their group or were independent of their social context. Thus, this research has contributed to the understanding of the role of social bonds between domestic horses in different contexts. Moreover, this research has implications for equine science and can inform housing and husbandry routines from an equine welfare perspective. It identifies that in established groups of horses, agonistic interactions are mostly not associated with a pronounced stress response. Furthermore, grooming and potentially other affiliative behaviours can promote positive welfare. These are valuable findings that support the practice of housing horses in groups. Additionally, in contexts of mild stress, horses seem to benefit from the presence of any familiar conspecific, independently of their bond relationship. These findings suggest that horses can benefit from the presence of familiar conspecifics during potentially stressful husbandry routines.