Abstract: Temperament is an important factor when working with horses. Behavioural tests have already been developed to measure certain dimensions of a horse's temperament (fearfulness, gregariousness, etc.). In order to measure the temperament more precisely, our work aimed to identify a dimension which has already been described in several species but not yet in horses, namely sensory sensitivity. Our study was based on the definition of a temperament dimension as “a behavioural characteristic stable across situations and over time”. We designed several tests for each sense and then determined whether the responses observed were correlated between situations and in time. The principle of the tests was to generate two stimuli of different intensities for each sense (e.g. two different sounds) and to measure the intensity of the horse's response (N = 26). Using Spearman rank correlations, we tested whether the responses to these different stimuli were inter-correlated. We repeated the same tests 5 months later to determine whether the responses were correlated over time. Within each sense, results show that the greater the horses' response to one stimulus, the greater their response to the other. For example, the reaction to the odour of cinnamon (time spent near the source of the odour) was significantly correlated to the reaction to lavender (R = 0.53, p = 0.004). The reactions to two different sounds or to two different tactile stimuli (von Frey filaments, or contact of a brush on the body), were also significantly correlated (R = 0.59, p < 0.0001; R = 0.38, p = 0.029). Finally, the reactions to two different tastes or to two visual stimuli tended to be correlated (R = 0.27, p = 0.09; R = 0.27, p = 0.09). However, there was no significant correlation between the responses to stimuli relating to different senses. Finally, except for the responses to odour, the responses to other sensory stimuli showed stability over a 5-month period (e.g. tactile stimulation: R = 0.71, p < 0.0001). In conclusion, our study revealed characteristics which were stable across situations and over time. The absence of links between the characteristics measured for the different senses suggests that a dimension for each sense exists (e.g. tactile sensitivity) rather than a general sensory sensitivity dimension covering all the senses.

Abstract: Learning performances are influenced by many factors, not only breed, age and sex, but also temperament. The purpose of this study was to understand how different temperamental dimensions affect the learning performance of horses, Equus caballus. First, we carried out a series of behavioural tests on 36 Welsh ponies aged 5-7 years to measure five temperamental dimensions: fearfulness (novel area test and surprise test), gregariousness (social isolation test), reactivity to humans (passive human test), tactile sensitivity (von Frey filament test) and activity level (evaluation of locomotor activity during all the tests). We then presented them with two learning tasks (avoidance and backwards-forwards tasks). In the avoidance task they had to learn to jump over a fence when they heard a sound associated with an aversive stimulus (puff of air). In the backwards-forwards task they had to walk forwards or move backwards in response to a tactile or vocal command to obtain a food reward. There was no correlation between performances on the two learning tasks, indicating that learning ability is task-dependent. However, correlations were found between temperamental data and learning performance (Spearman correlations). The ponies that performed the avoidance task best were the most fearful and the most active ones. For instance, the number of trials required to perform 5 consecutive correct responses (learning criterion) was correlated with the variables aimed at measuring fearfulness (way of crossing a novel area: rs = -0.41, P = 0.01 and time to start eating again after a surprise effect: rs = -0.33, P = 0.05) and activity level (frequency of trotting during all the tests: rs = -0.40, P = 0.02). The animals that performed the backwards-forwards task best were the ones that were the least fearful and the most sensitive. For instance, the learning criterion (corresponding to the number of trials taken to achieve five consecutive correct responses) was correlated with the variables aimed at measuring fearfulness (latency to put one foot on the area: rs = 0.43, P = 0.01; way of crossing a novel area: rs = 0.31, P = 0.06; and time to start eating again after a surprise effect: rs = 0.43, P = 0.009) and tactile sensitivity (response to von Frey filaments: rs = -0.44, P = 0.008). This study revealed significant links between temperament and learning abilities that are highly task-dependent.

Abstract: Managers concerned with shrinking habitats and limited resources for wildlife seek effective tools for limiting population growth in some species. Fertility control is one such tool, yet little is known about its impacts on the behavioral ecology of wild, free-roaming animals. We investigated influences of the immunocontraceptive porcine zona pellucida (PZP) on individual and social behavior in bands of feral horses (Equus caballus) in three discrete populations and used 14 hierarchical mixed effect models to gain insight into the influences of PZP treatment on feral horse behavior. A model of body condition was the strongest predictor of feeding, resting, maintenance, and social behaviors, with treated females allocating their time similarly to control females. Time spent feeding declined 11.4% from low condition to high condition females (F1,154 = 26.427, P < 0.001) and was partially reciprocated by a 6.0% increase in resting (F1,154 = 7.629, P = 0.006), 0.9% increase in maintenance (F1,154 = 7.028, P = 0.009), and 1.8% increase in social behavior (F1,154 = 15.064, P < 0.001). There was no difference detected in body condition of treated versus control females (F1,154 = 0.033, P = 0.856), but females with a dependent foal had lower body condition than those without a foal (F1,154 = 4.512, P = 0.038). Herding behavior was best explained by a model of treatment and the interaction of band fidelity and foal presence (AICc weight = 0.660) which estimated no difference in rate of herding behavior directed toward control versus treated females (F1,102 = 0.196, P = 0.659), but resident females without a dependent foal were herded 50.9% more than resident females with a foal (F3,102 = 8.269, P < 0.001). Treated females received 54.5% more reproductive behaviors from stallions than control mares (F1,105 = 5.155, P = 0.025), with the model containing only treatment being the most-supported (AICc weight = 0.530). Treated and control females received harem-tending behaviors from stallions equally (F1,105 = 0.001, P = 0.969) and agonistic behaviors from stallions equally (F1,105 < 0.001, P = 0.986). Direct effects of PZP treatment on the behavior of feral horses appear to be limited primarily to reproductive behaviors and most other differences detected were attributed to the effects of body condition, band fidelity, or foal presence. PZP is a promising alternative to traditional hormone-based contraceptives and appears to contribute few short-term behavioral modifications in feral horses.

Abstract: Social relationships are important in social species. These relationships, based on repeated interactions, define each partner's expectations during the following encounters. The creation of a relationship implies high social cognitive abilities which require that each partner is able to associate the positive or negative content of an interaction with a specific partner and to recall this association. In this study, we tested the effects of repeated interactions on the memory kept by 23 young horses about humans, after 6 and 8 months of separation. The association of a reward with a learning task in an interactional context induced positive reactions towards humans during training. It also increased contact and interest, not only just after training, but also several months later, despite no further interaction with humans. In addition, this ‘positive memory’ of humans extended to novel persons. Overall, positive reinforcement enhanced learning and memorization of the task itself. These findings suggest remarkable social cognitive abilities that can be transposed from intraspecific to interspecific social contexts.

Abstract: Third-party interventions are defined as the interruption of dyadic interactions by third animals through direct physical contact, interposing or threats. Previous studies focused on the analysis of interventions against agonistic encounters. However, there have been no evaluations of interventions against affiliative behaviours, particularly in relation to the intervening animal&#65533;s social relationships and its social and spatial position. Horses, Equus caballus, are an interesting model species, as interventions against affiliative interactions occur more frequently than against agonistic interactions. In this study, 64 feral horses displayed 67 interventions in affiliative interactions and eight interventions in agonistic interactions within the observation period. We analysed the interventions in affiliative encounters, and found that it was mainly higher-ranking females that intervened in the affiliative interactions of group mates in the stable horse harems. The intervening animals took an active part in affiliative and agonistic encounters within the group, but did not occupy particular social roles or spatial positions. They intervened in affiliative interactions in which group mates with which they had social bonds interacted with other members of the group. They targeted the nonbonded animal and approached the one with which they were socially bonded. We suggest some species use third-party interventions in affiliative interactions to prevent competition for preferred social interaction partners from escalating into more costly agonistic encounters.