Abstract: In recent years, Przewalski’s horses have been increasingly kept in semi-reserves. However,there areonly few studies ontheir behaviour and their ability to adaptto management interventions.In the main part of my dissertation, I focus on investigatingthe animals’ behaviour in different semi-reserves with varyinghabitats and living spaces. In addition, I investigate the horses’ behaviour during various management interventionsand analysetheensuing changes instress levels. Another aspect of my dissertation is the studyof social behaviour inPrzewalski’s horses. I investigate theparameters that should be used to demonstrate social bonds between individualsandassess whichdata provide the most meaningful results.In the commentary tochapter 1,several studies investigatingsocial bonds in horsesare discussed. Comparing the various studies, it is strikingthat no homogeneous analyses orevaluation criteria exist. While some authors only considersocial grooming, others include data onthe spatial proximity of the individuals in their evaluations, and various definitionsof proximity can also be found in the literature. Additionally, someauthors use friendly approaches between individuals asa furtherparameter wheninvestigating the social bonds.Continuing with this theme, in chapter 2I investigate the social behaviour of the horses and compare various analysis methods. I show that proactive behaviour, such as friendly approaches, is a good alternative to spatial proximity when investigating social bonds between group members, andis also useful for expanding the often very small data sets of mutual grooming in horses. Comparing Przewalski’s horses with wild living horses, I found no significant differences in the social behavior and the frequency of social interactions, regardlessof group size, group composition, habitat, and individual parameters such as age and gender.Inchapter 3,I investigate the behaviour of a Przewalski’s horse group when exploring a new area of their enclosure. Their behaviour changed, showing less resting and more feeding. Furthermore, the animals maintained greater distances from each other, and the alpha male, instead of herding the group from behind, led the group around the new area and walked in front of the other group members. Moreover, he showed a substantial increase in stress level during the first day.A general comparison of the behaviour of the Przewalski’s horses in different semi-reserves is provided in chapter 4. In it, the habitat choice of the animals and their reactions to various management interventions are investigated. It is shown that Przewalski’s horses prefer open grassland to dense woods, although keeping Przewalski’s horses in a pine forest does not influence the animals’ stress level. In contrast to habitat, food range, and changes in the group composition, which do not appear to change stress levels, individual factors, such as the hierarchy, influence the glucocorticoid level of the animals significantly. The largest increases in stress hormones were demonstrated when the horses were temporarily confined in smaller areas.The importance of the available space is also discussed in chapter 5, where it is shown that horses show less aggressive behaviour when more space is provided. In contrast, the husbandry system does not influence the animals’ aggression, but the way of feeding can additionally reduce agonistic behaviour, for example if food is offered ad libitum.In summary, the results of this study provide indications for the optimization of keeping Przewalski’s horses in semi-reserves. The animals can adapt themselves to the environment and thrive in habitats which do not correspond to their original steppe-like home. Nevertheless, the semi-reserves should provide sufficient grassland, as the horses prefer this type of habitat. General speaking, any types of habitat can only offer a suitable living space if the food range is sufficient for the number of horses. Otherwise, and especially during could winter months, supplementary feeding is necessary according to the body condition of the animals. This is particularly important for older, weakened, or very young animals, which are still adapting to life in the semi-reserve. Without sufficient food, stress hormones can increase and negatively influence the well-being of the horses. The same is true for management interventions: restricting the animals to small enclosures, for example, can adversely affect the horses’ well-being and should be only done if absolutely necessary. Targetedbehaviour observations allow the animals that have a special meaning for the group to be identified, and these should not be taken out of the group unless it is unavoidable, as young and unexperienced horses orientate themselves on those animals. This is especially true for the alpha male in a bachelor group, as these groups are often composed of young horses and the alpha-male provides the necessary stability and experience. Social bonds between individuals can be investigated by observing friendly and proactive behaviour, and social grooming and friendly approaches yield suitable data for such analysis.

Abstract: Animal management often involves procedures that, while unlikely to cause physical pain, still cause aversive responses. The domestic horse (Equus caballus) regularly has excessive hair clipped off to facilitate its use as a riding/driving animal and this procedure causes adverse behavioral responses in some animals. The aim of this study was to compare behavioral and physiological measures to assess the aversive effect of this procedure. Ten horses were selected on the basis of being either compliant (C: n=5) or non-compliant (NC: n=5) during this procedure. The horses were subjected to a sham clipping procedure (SC: where the blades had been removed from the clippers) for a period of ten minutes. Measures were taken pre, during and post SC (-10min to +30min) and mean values calculated for ALL horses and for C and NC separately. Behavioral activity was scored (scale 1-5) by twenty students from video footage in (phase/group-blind scoring). Heart rate (HR), salivary cortisol and eye temperature were monitored throughout the procedure. The NC horses were found to be significantly more behaviorally active/less relaxed throughout the trial than C horses (p<0.05) with the greatest difference occurring during the SC procedure (p<0.01). NC horses were more active/less relaxed during, compared with pre or post SC (p<0.05), but showed no behavioral difference pre and post SC. HR of the NC horses was higher than that of the C horses throughout the trial but only significantly so after 10min of SC (p<0.01). ALL horses showed a significant increase in HR between +5 and +10min into the procedure (p<0.05). There was a significant increase in salivary cortisol concentration in ALL horses post procedure (p<0.01) with levels peaking at 20minute post SC. No significant differences in salivary cortisol concentration between C and NC were found at any stage of the trial. Eye temperature increased significantly in ALL horses during SC, peaking at +10min into the procedure (p<0.05) and then decreased substantially when SC had ceased (p<0.01). Although no significant differences were found between C and NC per se, there was a significant interaction between group and phase of trial (p<0.05) with the NC group showing a greater decrease in eye temperature post SC. There was a significant positive correlation between changes in salivary cortisol concentration and eye temperature (p<0.01) but no correlation between any of the other measures. Although the behavioral response of C and NC to this procedure was significantly different the physiological responses indicated that ALL horses found the procedure aversive. Eye temperature could be used as an objective and immediate measure of how an animal is responding to a specific situation in order to evaluate management procedures and adapt them where appropriate to reduce the negative impact on animal health and welfare.

Abstract: Background The increased use of meta-analysis in systematic reviews of healthcare interventions has highlighted several types of bias that can arise during the completion of a randomised controlled trial. Study publication bias has been recognised as a potential threat to the validity of meta-analysis and can make the readily available evidence unreliable for decision making. Until recently, outcome reporting bias has received less attention. Methodology/Principal Findings We review and summarise the evidence from a series of cohort studies that have assessed study publication bias and outcome reporting bias in randomised controlled trials. Sixteen studies were eligible of which only two followed the cohort all the way through from protocol approval to information regarding publication of outcomes. Eleven of the studies investigated study publication bias and five investigated outcome reporting bias. Three studies have found that statistically significant outcomes had a higher odds of being fully reported compared to non-significant outcomes (range of odds ratios: 2.2 to 4.7). In comparing trial publications to protocols, we found that 40-62% of studies had at least one primary outcome that was changed, introduced, or omitted. We decided not to undertake meta-analysis due to the differences between studies. Conclusions Recent work provides direct empirical evidence for the existence of study publication bias and outcome reporting bias. There is strong evidence of an association between significant results and publication; studies that report positive or significant results are more likely to be published and outcomes that are statistically significant have higher odds of being fully reported. Publications have been found to be inconsistent with their protocols. Researchers need to be aware of the problems of both types of bias and efforts should be concentrated on improving the reporting of trials.

Abstract: Wolves returned to Germany in 2000, leading to fear in German horse owners that their horses could be in danger of wolf attacks or panic-like escapes from pastures when sighting wolves. However, reports from southern European countries indicate that wolf predation on horses diminishes with increasing presence of wildlife. Therefore, we conducted a long-term, filed observation between January 2015 and July 2022 on 13 non breeding riding horses, mares and geldings, kept permanently on two pastures within the range of wildlife and a stable wolf pack with annual offspring. Wildlife cameras at the fences of the pastures made 984 times recordings of wolves and 3151 times recordings of wildlife in and around the pastures. Between 1 January 2022 and 23 March 2022 we observed two stable horse groups. Pasture 1 was grazed by five horses of mixed breed, four mares and one gelding, with the median age of 8 years (min. = 6y, max. = 29y). Pasture 2 was grazed by eight heavy warmbloods and draught horses, three mares and five geldings, with the median age of 16 years (min. = 13y, max. = 22y). During this period no wolf was recorded at pasture 2, but wild boar several times, whereas at pasture 1, wolves were recorded 89 times, and for the wildlife mostly hare. Wolves may have avoided pasture 2 because of the presence of wild boar or because the large group of older, heavy breed horses may have formed a stable, protective group. The latter needs to be confirmed in a follow-up field observation, which records anti-predator behavior and welfare indicators in horses. In conclusion, wolves did not attack the mature horses on pastures with plenty of wildlife and the horses did not respond to the presence of wolves with visible signs of reduced welfare or panic. This indicates that wolves may prefer to prey on easily accessible wildlife around and at horse pastures and that Central European horses become accustom to the presence of non-hunting wolves.

Abstract: Outdoor group housing is increasingly recognized as an appropriate housing system for domesticated horses. The objective of this study was therefore to investigate the effect of potential feed shortage in semi-natural horse keeping systems in winter on animal health and welfare. In 10 female Shetland ponies blood concentrations (NEFA, total protein (TP), total bilirubin (TB), beta-hydroxybutyrate (BHB) and thyroxine (T4)), body mass and the body condition score (BCS) were monitored for 7months including a 4months period of feed restriction in five of the 10 ponies. Restrictively fed animals lost 18.4±2.99% of their body mass and the BCS decreased by 2.2±0.8 points (BCS scale: 0=emaciated, 5=obese). Feed restriction led to a continuous increase in TB (P<0.001) and NEFA (P<0.01) concentrations compared to control ponies. The TP and BHB values only differed at the end of the trial with lower concentrations in restricted fed mares (P<0.05). Feed restriction had no effect on thyroxine concentrations. TB concentrations in the feed restricted group were out of the reference range during the entire feeding trial. The increased NEFA concentrations in feed restricted compared to control ponies suggest that fat was mobilized. The BCS, as well as plasma NEFA and TB concentrations were good indicators for a rapid detection of possible health problems caused by undernourishment in horses when kept under semi-natural conditions. In contrast, blood parameters of the control animals were within the reference ranges, suggesting that a year round outdoor housing with additional feed supply is an adequate housing system for a robust horse breed like the Shetland pony.