Abstract: Although horses can sleep while standing, recumbency is required for REM sleep and since all sleep stages must be completed for an entire sleep cycle, the opportunity for recumbency is essential for animal welfare. Observations on feral horses indicate a minimal lying duration of 30 min, preferably on a deformable and dry ground. In contrast to feral horses, lying behaviour in stabled horses is often affected by the dimensions of the provided lying area and rank.
In Switzerland, minimum requirements (MR) for the littered lying area are established by law to ensure animal welfare (BLV, 2008) (A/N: approximately match German recommendations (BMEL, 2009)). The aim of this study was to assess the adequacy of the dimensions of the minimum requirements for group-housed horses by investigating 38 horses in 8 groups. Further, hard rubber mats were provided supplementary in order to assess their suitability as an alternative to litter. Four treatments were each applied in randomised order:
– 0x MR: no litter + 1.5x MR with rubber mats
– 0.5x MR: 0.5x MR with litter + 1x MR with rubber mats
– 1x MR: 1x MR with litter; 0.5x MR with rubber mats
– 1.5x MR: 1.5x MR with litter + no rubber mats
For each treatment, after a habituation period of 8 days, lying behaviour was recorded (video, accelerometers) continuously for 72 hrs. Statistical analysis was performed using mixed effects models.
Regardless of the ground chosen, the duration of recumbency per 24 hrs was increasing with increasing dimensions of the littered area (F1,93 = 12.9, p = 0.0005; Fig. 1). Whereas the effect flattened from 1x to 1.5x MR, the duration spent on litter – a deformable ground – was increasing continuously (F1,62 = 23.1, p < 0.0001). Further, the proportion of lateral recumbency was increased with increased dimensions of the littered area (F1,79 = 12.3, p = 0.0007). Regarding the number of lying bouts, no differences were apparent between treatments providing litter, but recumbency occurred very seldom if only rubber mats were provided (F1,93 = 14.7, p = 0.0002). Further, low-ranking horses spent more lying bouts on rubber mats than high-ranking horses (F1,29 = 4.4, p = 0.04). Additionally, the larger the dimensions of the littered area the more horses were present in the lying area at the moment of lying down (F1,79 = 6.6, p = 0.01). Moreover, low-ranking horses showed considerably higher percentages of involuntarily terminated lying bouts than high-ranking horses if 0.5x and 1x MR were littered (F1,76 = 8.43, p = 0.005).
Although the indicated minimal lying duration was averagely performed, large individual differences occurred and at least 8% were lying down less than 30 min per 24 hrs in every treatment. Further, the inclusion of social parameters indicated a beneficial effect of an exceedance of the minimum requirements especially for low-ranking horses. Therefore, the minimum requirements established by Swiss law can be stated as adequate but should be perceived as minimum and not optimum dimensions.

Abstract: The complex housing environment and the close contact between humans and horses in equine sports place high demands on the learning capacity of horses. To date only limited information is available on the learning ability of the horse including higher order cognition. A type of higher order cognition is to perceive and discriminate quantities. Several mammals and birds have shown to be capable of discriminating objects due to their quantity (Brannon and Roitman, 2003). With regard to horses, there are only few studies available concerning their numerosity judgment (Uller and Lewis, 2009) and this ability is discussed controversially (Henselek et al, 2012). Possibly, the legacy of ‘‘Clever Hans’’ overshadowed further research on numerical capacity in horses, a horse to whom several psychologists incorrectly attributed the capacity of symbolic calculation (Pfungst, 1907; Rosenthal, 1965). In the present study we wanted to show whether Shetland ponies are able to transfer a previously learned concept of sameness to a numerosity judgment. The base of the test design was a “matching to sample” task, where the ponies had learned to relate abstract symbols to another which were presented on a LCD screen. Three Shetland ponies, which previously solved the matching to sample task, were tested in two test phases. In the first test phase different quantities of dots were presented (1 vs. 2; 2 vs. 3; 3 vs. 4; 4 vs. 5). To exclude discrimination due to the shape of the stimuli, the dots were varied in size and arrangement. The stimuli were presented in a triangular arrangement on the LCD screen; the sample stimulus was presented in the middle above and the discrimination stimuli in the two lower corners (S+ and S-). The pony received a food reward, by choosing the positive stimulus (S+). When the negative stimulus (S-) was chosen, the pony entered the next trial. Each learning session consisted of 20 decision trials. To investigate whether the numerosity judgment was transferable to mixed geometrical symbols (tri-, rectangle, rhomb, dot and cross) a second test phase was designed. All of the three Shetland ponies met the learning criterion of the first test phase (80% correct responses in two consecutive sessions) within the first eight sessions. One pony could transfer all judgments to the mixed symbols (up to 4 vs. 5), another pony up to 3 vs. 5 and the third on the level 2 vs. 3. These are the first reported findings that ponies are able to discriminate up to five objects. The numerosity judgment seemed to be easier for the ponies when homogenous objects were presented, than in the case of heterogeneous symbols. The reaction of the ponies occurred within few seconds, suggesting that the animals used subitizing for their numerosity judgment.
Keywords:
Cognition, numerical capacity, numerosity judgment, Shetland ponys
References:
Brannon E, Roitman J (2003) Nonverbal representations of time and number in animals and human infants. In: Meck WH (ed) Functional and neural mechanisms of interval timing. CRC Press, Boca Raton, pp 143–182
Henselek Y, Fischer J, Schloegl C (2012) Does the stimulus type influence horses’ performance in a quantity discrimination task? Front Psychol 3:504
Pfungst O (1907) Der Kluge Hans. Ein Beitrag zur nicht-verbalen Kommunikation. 3rd edn. (reprint of the original 1983) Frankfurter Fachbuchhandlung für Psychologie, Frankfurt
Rosenthal R (1965) Clever Hans: the horse of Mr. Von Osten. Holt, Rinehart and Winston, New York
Uller C, Lewis J (2009) Horses (Equus caballus) select the greater of two quantities in small numerical sets. Anim Cogn 12:733–738

Abstract: Wrinkles above the eye ball are common in domestic horses but may differ in number and shape both between and within individuals. They are caused by contraction of the inner eye brow raiser, and some people working with horses call them “worry wrinkles”, considering them to reflect emotional states. However, as yet no study has formally investigated the relationship between eye wrinkles and emotional state in horses.
The aim of the present study was to induce states of different emotional valence and to assess whether positive emotional states would reduce the expression of eye wrinkles while negative emotional states would increase it. Sixteen horses were confronted in a balanced order with two presumably positively and two negatively valenced situations each. Positive situations included anticipation of a food reward (FA) and petting (P), negative situations included food competition (FC) and waving a plastic bag (PB). Each situation lasted for 60s (TRT) and was preceded by a 60s control phase (CON). Throughout CON and TRT pictures of the eyes were taken, and for each horse four pictures per situation (FA, P, FC, PB) and phase (CON and TRT) were randomly selected (n = 512) and scored in random order and blind to treatment for six outcome variables: overall impression (qualitative), number, angle and markedness of eye wrinkles, presence of eye white, and shape of eye lid.
Data were analysed separately for the right and left eye using linear mixed effects models (angle, number), generalised linear mixed models (eye white, markedness), and ordered logistic regression (qualitative, shape of eye lid), with “situation” (FA, P, FC, PB), “phase” (CON, TRT) and their two-way interaction as fixed effects.
Expression of eye wrinkles did not vary consistently across “situation” and “phase”. Independent of phase, eye white appeared less frequently during P than during FA (z=-3.15, p=0.009), FC (z=-2.94, p=0.02), and PB (z=4.17, p<0.001) in the left eye and during PB (z=4.10, p 0.001) in the right eye. Similarly, wrinkles were less marked during P compared to the other situations in the left eye (FA: z=3.15, p=0.009; FC: z=-2.94, p=0.017; PB: z=4.17, p<0.001) and compared to PB in the right eye (z=4.10, p=0.001), while no differences between situations occurred in number of wrinkles, overall impression and shape of eye lid for both eyes. Consistent with our hypothesis, P induced relaxation of the underlying muscle in the right eye resulting in a wider angle compared to its control phase (interaction situation*phase: F3,10=3.71, p=0.055; post-hoc comparison: z=-3.57, p=0.009), while FC induced muscle contraction, resulting in a sharper angle in the left eye (interaction situation*phase: F3,11=6.57, p=0.011; z=3.73, p=0.005).
We conclude that emotional valence may affect characteristics of eye winkle expression in horses which might therefore be a promising indicator of horses’ emotional states, but further research is needed to validate the relevant outcome variables.

Abstract: For human athletes, any physical performance classically begins with a warm-up including a phase of cardio-respiratory activation as well as a phase of neuro-muscular mobilization. For the equine athletes similar routines are widely followed despite limited scientific studies addressing its effectiveness. Although physiological bases are of a similar nature in humans and horses, the latter are showing a different “telos” related to their status of prey. Indeed, their survival depends on their ability to detect any predator or potential danger in their environment and to estimate the need to run away. Therefore warm-up techniques should be adapted to this specificity and we added a walk phase preceding the usual warm-up during which the horse was allowed to visually assess its environment without constraint imposed by its rider, reins being long. We assumed that this walk phase would allow a more relaxed mental state and then a decrease in cognitive resources involved in environmental monitoring. Consequently this would release cognitive resources then available for the communication between the horse and the rider.
The autonomous nervous system (ANS) regulates the heart rate via the sympathetic and the parasympathetic nervous systems. The sympathovagal balance and the heart rate variability are considered as good indicators of an acute stress state as well as emotional states. These measures are frequently used to estimate well-being in animals. During the walk phase, we observed a significant decrease of the heart rate (HR), a significant increase of its variability and a significant modification of the sympathovagal balance in favour of a higher contribution of the parasympathetic control. We also measured the non-linear correlation dimension which reflects the degree of freedom of a system. To the best of our knowledge, this method was used until now only in humans to whom it was shown that its decrease is related to stressful events and is associated with a bad prognosis for survival. During the walk phase in horses, the dimension of correlation of cardiac activity was increased.
From a behavioural point of view, we observed a significant decrease of side movements related to head orientation as well as significant changes in ears position, the latter being preferentially directed forwards at the beginning of the walk phase and laterally at the end. As previous research suggests that ears position and head orientation can be involved in attentional mechanisms in horses, our results suggest a decrease of the attention focused on the environment. Furthermore, a significant lowering of the neck was observed as well as a significant decrease of behaviours related to stress.
To conclude our results suggest that a walk phase preceding the usual warm-up of horses would contribute to improve their mental state in favour of more relaxation and of a decrease of the attentional resources invested in environmental monitoring.