Companionship counts
Investigating social housing conditions and welfare in privately owned rabbits
Cecilie Ravn Skovlund, Niels Munch Corneliussen, Laura Benoni Kofod, Isabella Thybo Mejlstrup, Peter Sandøe, Björn Forkman
Applied Animal Behaviour Science, 2026
Abstract
Rabbits are among the most popular companion animals. Despite being a social species, most companion rabbits are solitary housed due to challenges with pairing and cohabitation, which may compromise their welfare. This study therefore aimed to investigate the effects of social housing (solitary vs. social) and dyadic relationships (using a Friendship Index derived from rates of allogrooming and contact) on rabbit welfare. Moreover, dyad-related factors (sex combination, age difference, life stage combination, early socialisation, and neuter status) and resource-related factors (group size, spatial availability, enrichment, raised platforms, and visual barriers) previously linked to social dynamics in farmed and laboratory rabbits, were examined for their impact on dyadic relationships. This was assessed through behavioural observations based on home-pen videos collected by owners (in the morning, 5–10 a.m., and evening, 5–10 p.m.), using behavioural welfare indicators for rabbits. Linear and generalized linear mixed effects models were used to assess welfare outcomes and relationship indices, with ‘owner’ as a random effect. Video material of 122 rabbits from 74 owners were analysed, with approximately five hours per rabbit. Socially housed rabbits spent 21 % of scans in contact and 75 % in proximity to conspecifics. Agonistic interactions were rare. Solitary rabbits expressed significantly less behavioural diversity (P = 0.02) and more awake inactivity (P = 0.04) than socially housed rabbits, indicating improved welfare for the latter group. However, solitary rabbits spent significantly more time on environmental interaction (P = 0.03), plausibly due to their inability to socialise. A higher age difference was found to be significantly associated with a reduced Friendship Index (P = 0.002). Minimising the age difference may therefore be considered for future recommendations regarding pairing of companion rabbits. No significant effects of dyadic relationships on welfare were found; however the sample was biased towards positive relationships. The large housing space available to the rabbits may have contributed to the low level of agonistic behaviour and positive dyadic relationships. In summary, statistical inferences were hindered by data homogeneity and zero-inflation, likely due to convenience sampling, along with a lack of evidence-based welfare indicators for rabbits. Future studies should investigate optimal observation times and durations, as well as assess validity, reliability, and feasibility of existing behavioural welfare indicators for rabbits. Despite methodological limitations, this study provides insights into the social environment and welfare of companion rabbits, emphasizing the benefits of social housing and the potential advantages of minimizing age differences in rabbit dyads.