Can behavioural observations give an evaluation of welfare? A study of time budget and social behaviour of arab mares at paddock
Résumé
A unstable herd of 44 Arab mares was observed in a 2500m2 paddock in summer in Tunisia. Fifteen-minute focal animal samples and scan sampling were used to determine the time budget of the mares during the period from 9am to 3pm and to study their social behaviour. A total of 42 hours of observations were performed. The behavioural states recorded were feeding, drinking, alert-standing, stand-resting, self-grooming, locomoting, rolling and social interactions. Locomotion was the most frequent activity at 27.9 ± 19.47% (mean ± SD), followed by feeding, and standing resting. The mares spent 14.88 ± 18.83% of their time alertstanding, 1.03 ± 3.51% self-grooming, 4.75 ± 7.19% drinking and 0.83% ± 2.64% interacting socially. 269 agonistic interactions were noticed. Retreat and avoidance represented the most frequent agonistic interactions (65% of total interactions). Almost 60% of all aggression consisted of threats to bite. Threats to kick (27%) were next in frequency of occurrence. Rolling was observed only once and mares were never observed allogrooming, lying down, urinating, defecating nor vocalizing. Social behaviour was restricted to agonistic interactions and preferential associations that could be determined were very rare. The tendency to stay away from other mares (percentage of time spent at more than 3 horse's body length) differed largely between mares and was unexpectedly high given the high density of animals (8.28 ± 15.37%). The data obtained in this high density situation therefore reveals restricted behavioural repertoires, unusual time budgets with a high frequency of active walk that constitutes the most frequent activity, and a low level of social interactions that were almost entirely restricted to agonistic interactions. These results are discussed in relation to other similar studies on domestic horses and questions of welfare assessment.