Effect of meal frequency on outdoor, group-housed horse behavior, social hierarchy, and stress

Social interactions with conspecifics and prolonged foraging behavior promote equid welfare. However, resource availability and feeding method can affect aggression observed in group-housed horses. The objective of this study was to investigate effects of meal frequency on equine behavior, social hierarchies, and welfare. Horses (n = 12) were assigned to one of 2 pens, blocked by body condition score, breed type, age, and sex. Horses were group-housed in 40 × 40 m dry lots with shelters. After a 14-d diet adaptation, pens were randomly assigned an order of meal frequencies in a crossover design and fed once (1 ×, 0800 h), twice (2 ×, 0800 and 1800 h), or 3 times (3 ×, 0800, 1300, and 1800 h) per day for 7 d per treatment. Each pen was fed grass hay at 1.55% of body weight (BW) per day on a dry matter (DM) basis from 7 slow-feed hay nets, approximately 15 min before the pelleted meal. Pelleted, complementary concentrate was fed at 0.45% BW DM per day via individual nosebags. Total rations were divided equally across meals (1 ×, 2 ×, or 3 ×). On d 5 of each treatment, horses were video recorded for continuous observation for 2 h centered around the 0800 h pelleted meal. Behaviors were categorized using an ethogram for frequency (bouts) for all behaviors and duration (%) for select behaviors. Time budgets were calculated for pre-feeding (before hay), post-feeding (after hay), and total time. Dominance ranks within pen were based on displacements given and received. Fecal samples were collected on d 5 for cortisol/corticosterone metabolite (FCM) measurement as an indicator of stress. Data were analyzed using PROC MIXED in SAS with meal, pen, and dominance rank as fixed effects and random effects of period and horse nested within pen. Horses fed 3 × (24 ± 3 bouts) stood alert more frequently (P < 0.0001) than horses fed 1 × (19 ± 3 bouts) or 2 × (12 ± 3 bouts). Pre- but not post-feeding, horses assigned a higher dominance rank showed signs of aggression (e.g., ears back, biting) more frequently (P = 0.03). More signs of aggression were seen post-feeding (P = 0.03) when horses were fed 3 × (11 ± 2 bouts) compared with 1 × (6 ± 2 bouts). Post-feeding, horses fed 3 × walked more frequently (24 ± 3 bouts) compared with those fed 1 × (19 ± 3 bouts) or 2 × (18 ± 3 bouts; P = 0.03), and horses were displaced more frequently when fed 3 × (10 ± 2 bouts) compared with 1 × (6 ± 2; P = 0.05). Meal frequency and dominance rank did not influence FCM concentrations. Overall, increased incidences of aggression occurred when group-housed horses were fed smaller, more frequent meals.