Temporal-spatial associations of large groups of laying hens in a quasi-commercial barn

Abstract

Social behaviour is established in animal systems with clear benefits, including improved welfare, where conspecifics can attenuate the effects of stress. Previous work from our group demonstrated that laying hens moved in a coordinated manner through the external areas of a barn. The current work sought to expand on this initial finding by focusing on movements within a barn. For this purpose, from 17 weeks of age the location of 1125 hens within five pens (225 hens/pen) were tracked. We used radio frequency identification (RFID) transponders in a commercially-relevant barn consisting of thirty-two antennae placed throughout a multi-tier aviary with access to an external wintergarden. After processing transponder registrations, our final dataset comprised location data for 1122 individuals along 214 days and was used to determine spatio-temporal associations with the Gaussian Mixture Model method. We explored associations among individuals within different areas, whether weekly social networks changed over time, and if multiple community structures were present within pens. Social networks were more structured than by chance (mean p-value across weeks < 0.001, range: 0–0.0006), though associations were generally weak with hens affiliated with all other hens. Association indices decreased with age (t = -120.20, df = 877894, p < 0.001), though approximately 10.3 % of hens had at least one relatively strong and durable tie that appeared to last for the entire 36-weeks observation period. Node strength was generally low (12.44 ± 2.24, range: 1.57–19.28), though strength values among hens within the same week varied greatly. Social networks were dynamic across time but also showed a temporal correlation such that association matrices were more correlated closer in time (range R: 0.12–0.90, mean: 0.55, corrected p-values < 0.01). We found no evidence of community structure using any algorithm. Our study indicated that hens within a relatively large group did not seem to manifest strong movement associations. Yet, a proportion of individuals maintained ties for the entire observation period, which could be associated with social motivations. We believe our work establishes a foundation to consider benefits of social connections on hen welfare within commercial laying hen systems.