A two-stage method to approach weaning stress in horses using a physical barrier to prevent nursing

Many domestic horses are weaned through abrupt physical separation of the foal and mare. Deviations from normal behaviours testify that this abrupt method has an adverse effect on the well-being of both mares and foals. To reduce negative changes in behaviour and physiology induced by weaning, a two-stage approach was tested: the first stage achieved nutritional separation by means of an udder cover to prevent nursing, and the second stage of physical separation occurred four days later. Behaviours (standing, walking, running, eating, drinking, nursing, aggressing, vocalizing) of mares and foals weaned either by abrupt physical separation (control; n=8), or two-stage (TS; n=7) were recorded in real time for 8h/d, and triaxial accelerometers recorded lying time of mares and foals 24h/d. Heart rates and global position were recorded on foals during live observations, and fecal samples from mares and foals were collected for cortisol analysis. Behavioural and physiological data were analyzed using a mixed model for repeated measures. No treatment effect was observed in foals, however physical separation induced an increase in vocalizations (0.3±0.13 versus 56.9±39.3 vocalizations/foal/d±SEM; p<0.0001), observations of running (0.31±0.08 versus 3.2±0.25 observations/foal/d±SEM; p<0.0001), observations of aggression (2.4±0.27 versus 7.6±0.52 observations/foal/d±SEM; p<0.0001), and distance travelled (1275±150 versus 4679±497m/d±SEM; p<0.0001). Prior to treatment, foals lay 34.91±12.4min/bout in 8±3.4 bouts/d. During nutritional separation, control foals lay down more than TS foals (p<0.0031). There was no treatment effect on lying duration (p>0.75) or number of lying bouts (p>0.17) upon physical separation, but lying time was significantly less in all foals (p<0.01). Prior to treatment, mares lay an average of 14.16±11.21min/bout in 2±1.7 bouts/d. Mares displayed no difference in lying duration or number of lying bouts between treatments (p>0.99) or upon physical separation (p>0.37). Two-stage mares vocalized less frequently than control mares upon physical separation (7.6±0.6 versus 12.6±0.9 vocalizations/mare/d±SEM; p<0.0001). Fecal cortisol concentrations increased in both mares and foals upon physical separation (146±13.5 versus 193±21.9ng/ml±SEM; p<0.0032). Although some minor reductions in stress response were achieved by use of a two-stage approach to weaning, these results more clearly show the effect on mares and foals to physical separation. Large variation in both behavioural and physiological responses to stress may indicate a role of personality and attachment theory in effective approaches to weaning.