Abstract: Investigations were conducted on four horses from a group of 12 Przewalski mares raised in different zoos and kept in a 44-ha enclosure under semi-natural conditions. Activity and feeding were continuously measured every second and were saved every 15 min by the storage-telemetry system ETHOSYS, from June 1995 to July 1996. Body mass of the horses was regularly recorded. Daily and monthly mean values, power spectra and DFC (as a measure for stability of rhythms synchronised with circadiurnal period) for activity and feeding were calculated. The general pattern of activity and feeding over the year was closely related to sunrise and sunset. Feeding accounted for 40% of total activity in summer and 62% in spring (all-year average being 52%). The level of activity was lowest in winter; whereas feeding was lowest in summer. The time budget for feeding reflected both feeding conditions and the annual pattern of body condition. Greatest activity occurred during daylight hours. Only on hot summer days, activity at night was higher than during daylight hours. Spectral analysis of activity and feeding in Przewalski horse showed a time pattern which was characterised by 24-h rhythmicity, but also by ultradian components with period lengths between 4.8 and 12 h, i.e., an activity pattern of up to five strong bouts per day. Annual variation in the pattern of power spectra was not high during the year. Results are discussed in connection with horse feeding strategy. Analysing the time structure of long-term and continuously measured activity and feeding could be a useful method to follow the general living conditions, especially the nutritional situation and to detect stressful conditions.

Abstract: The responsiveness of 10 horses and 10 ponies to environmental challenge (represented by an open field test) was assessed using a qualitative approach based on free choice profiling methodology (FCP), which gives observers complete freedom to choose their own descriptive terms. Data were analysed with generalised Procrustes analysis (GPA), a multivariate statistical technique associated with FCP. A cross-validation of the outcomes of this approach to data recorded through quantitative behaviour analysis, and through a questionnaire given to the animals' owner/riding instructor, was also performed using principal component analysis (PCA). Twelve undergraduate students generated their own descriptive vocabularies, by watching 20 horse/pony video clips lasting 2.5 min each. GPA showed that the consensus profile explained a high percentage of variation among the 12 observers, and differed significantly from the mean randomised profile (p < 0.001). Two main dimensions of the consensus profile were identified, explaining 60% and 5.2% of the variation between animals, respectively. The 12 observer word charts interpreting these dimensions were semantically consistent, as they all converged towards the same meaning, albeit using different terms. The most used term to describe the positive end of axis 1 was “quiet”, whereas “attentive” was the best positive descriptor of axis 2. The most frequently used descriptors for the negative ends of axes 1 and 2 were “nervous” and “bored”, respectively. Thus, axis 1 was labelled as “quiet/nervous” and axis 2 was named as “attentive/bored”. A marked effect of animal category was observed on the scores of the animals on the first dimension (p < 0.001). Horses received significantly higher scores, and were thus assessed as more quiet and calm, than ponies. Conversely, ponies tended to receive lower scores on the second dimension (p < 0.12), therefore they appeared less curious and attentive. The results of the PCA showed that the variables from different types of measurement clearly had meaningful relationships. For instance, the variables with the highest loading on the positive end of axis 1 were all indicative of tractable and docile animals, whereas axis 2 showed high loadings on the positive end for variables indicating attentive animals. Qualitative behaviour assessment proved to be an appropriate methodology for the study of horse behavioural responsiveness, in that it provided a multifaceted characterisation of horse behavioural expression that was in agreement with other quantitative and subjective assessments of the animals' behaviour.

Abstract: To achieve optimal performance in equine sports as well as in leisure not only the physical abilities of the horse should be considered, but also the horse's personality. Besides temperamental aspects, like emotionality, or the horse's reactivity towards humans in handling situations, the learning ability of the horse is another relevant personality trait. To study whether differences in learning performance are consistent over time and whether individual learning performance differs between learning tests or is affected by emotionality, 39 young horses (Dutch Warmblood) were tested repeatedly in two learning tests. An aversive stimulus (AS) was used in one learning test (the avoidance learning test) and a reward was used in the other learning test (the reward learning test). During both learning tests behaviour as well as heart rate were measured. Each test was executed four times, twice when horses were 1 year of age, and twice when they were 2 years of age. Half of the horses received additional physical training from 6 months onwards. In both tests horses could be classified as either performers, i.e. completing the daily session, or as non-performers, i.e. returning to the home environment without having completed the daily session. There were some indications that emotionality might have caused non-performing behaviour, but these indications are not convincing enough to exclude other causes. Furthermore, there seem to be no simple relationships between measures of heart rate, behavioural responses putatively related to emotionality and learning performance. Horses revealed consistent individual learning performances within years in both tests, and in the avoidance learning test also between years. There was no significant correlation between learning performances in the avoidance learning test and the learning performances in the reward learning test. It is concluded that individual learning abilities are consistent over a short time interval for an avoidance learning test and a reward learning test and over a longer time for the avoidance learning test. Furthermore, results indicate that some horses perform better when they have to learn to avoid an aversive stimulus while others perform better when they are rewarded after a correct response. It is suggested that these differences may be relevant to design optimal individual training programmes and methods.

Abstract: Current definitions of horse personality traits are rather vague, lacking clear, universally accepted guidelines for evaluation in performance tests. Therefore, the aim of the present study was to screen behavioural and physiological measurements taken during riding for potential links with scores the same horses received in the official stallion performance test for rideability and personality traits. Behaviour, heart rate (HR) and HR variability from thirty-six stallions participating in a performance test were recorded repeatedly during their performance test training. Using the coefficient of determination, regression analysis revealed that about 1/3 of variation (ranging between r = 0.26 (“constitution” (i.e. fitness, health)) and r = 0.46 (rideability)) in the personality trait scores could be explained by selecting the three most influential behaviour patterns per trait. These behaviour patterns included stumbling (with all traits except character), head-tossing (temperament, rideability), tail-swishing (willingness to work), involuntary change in gait (character) and the rider's use of her/his hands (constitution, rideability), voice (temperament) or whip (constitution). Subsequent mixed model analysis revealed a significant (P < 0.05) influence of the behaviour pattern “horse-induced change in gait” on character (-0.98 ± 0.31 scores per additional occurrence of change in gaits), of head-tossing (-0.25 ± 0.08 scores) and rider's use of voice (-0.51 ± 0.25; P = 0.0594) on temperament, and of stumbling on each of the following: willingness to work (-2.5 ± 1.2), constitution (-2.5 ± 1.2 scores; P = 0.0516) and rideability scores (-3.3 ± 1.4). In addition, constitution scores tended (P = 0.0889) to increase with higher low frequency/high frequency heart rate variation ratios (LF/HF), indicating a shift towards sympathetic dominance and thus a higher stress load in horses with higher scores for constitution. Rideability scores from the training phase were also significantly influenced by head-tossing (-0.5 ± 0.1), and in addition rideability scores from the final test were influenced by the training rider, ranging between average estimated rideability scores of 6.8 ± 0.4 for one training rider and 8.36 ± 0.3 scores for another training rider. Horses ridden with their nose-line predominantly behind the vertical received higher scores for rideability (8.3 ± 0.3) than horses ridden with their nose-line at the vertical (7.7 ± 0.2). These findings indicate that either judges perceive horses to have a better rideability when they readily offer a more extreme poll flexion, or that riders make use of horses’ better rideability by imposing a more extreme poll flexion. Several of the above described associations, but also of the non-existing links (e.g. no association between shying or heart rate and temperament) between behaviour patterns and scores for personality traits are rather surprising, warranting further investigation regarding the underlying causes of these relationships. Some of these behaviour patterns should be considered when redesigning the current guidelines for evaluation of personality traits during breeding horse performance tests, ultimately leading to improved genetic selection for equine personality traits. However, ethical implication of defining aversive behaviour such as head-tossing as an indicator of, for example, poor temperament, should not be neglected when devising new guidelines: such aversive behaviour may in fact be an indication of inadequate training techniques rather than poor horse personality.

Abstract: Behavioural observations suggest that smell is important in social discriminations between horses but balanced studies of this capacity are lacking. We used a habituation-discrimination procedure to investigate the ability of horses to distinguish between pairs of odour samples from different individuals. In Study 1, separate tests were conducted for urine, faeces or fleece fabric previously rubbed on the coat (to pick up body odour samples (BOS)) and donor pairs differed in sex, and age. 10 pregnant mares each underwent three tests, one per sample type. A test consisted of three successive 2-min presentations of a sample from Individual A with a simultaneous presentation of a sample from Individual B during the final presentation. Doubly repeated measures ANOVA indicated a main effect of sample type on investigative response (df = 2, f = 7.98, P = 0.004): durations were longer for BOS than for urine or faeces but habituation across trials was most consistent for urine. In the final presentation, mares demonstrated discrimination by investigating the novel urine sample (B) more than the repeated sample (novel: median 8.0 s, IQR = 10; repeated: median 2.5 s, IQR = 6; z = -2.558, P = 0.008). In Study 2, urine samples from castrated male donors were used and neither mares nor their 4-month-old foals discriminated between samples from different individuals in the final presentation. The findings suggest that urine odour may contain some information that horses can use to discriminate between conspecifics. This may be limited to the level of broad categories such as sex or reproductive status; further investigation is needed to reveal what functional information can be transmitted and what compounds are involved.