Do horses enjoy interaction with humans? – Design of a choice experiment

Horseback riding is under increasing critical scrutiny, as clearly shown by recent media reports followed by public discussion, and a critical report from the Danish Animal Ethics Council. It is unclear where the critique stops and one may wonder whether our use of horses in general can be defended. If it can, a possible argument in favor of using horses for riding and other purposes, could be that positive experiences of the horses may override negative ones. However, so far there has been a lack of scientific study of positive welfare in animals and thus there is no evidence for arguing that human-horse interaction actually does include positive experiences for horses that potentially could outweigh any negative experiences. A focus on the assessment of positive horse welfare in human-horse contact is therefore necessary.

Researchers in a large EU project (EU COST Action LIFT) have recently defined positive animal welfare as “animals flourishing—both physically and mentally—in their environment and having rewarding experiences”. An important aspect of positive welfare is positive engagement (LIFT, 2024). While a horse in a human-horse interaction will be engaged to a certain degree, concerns have been raised about whether that is actually perceived as positive by the horses. The big question here is whether human contact is perceived as truly positive by the horses (apart from the long-term evolutionary benefit following from feeding and caring for them). The closest we get to assessing positive welfare so far is by conducting choice experiments where it is tested whether an animal prefers one situation over another. In horses, a preference test was used by e.g. von Borstel and Keil, (2012), in which the horse could choose between shorter and longer riding bouts (however, results remained inconclusive which the authors explained with a suboptimal experimental setup).

Von Borstel et al., (2009) found that horses prefer being ridden in a normal head-neck position over being ridden in hyperflexion in a choice-test. Lee et al., (2011) tested the preference of horses to exercise on a treadmill, being turned out to a paddock alone or with a companion. However, Lee et al., (2011) neglected human-horse interactions. Hence, although preference tests can be a relevant tool for welfare tests in horses, whether horses would choose an interaction with a human has not been tested so far, although this seems critical to answer the question of whether horses perceive human contact as positive. The limitation of the approach is of course that even if a horse prefers ‘option a’ over ‘option b’ it does not directly imply that ‘option a’ is positive rather than just less negative than ‘option b’. Including additional aspects of horse behaviour, e.g. anticipatory behaviour and body language, may help to overcome this limitation.

In this project, we will design a choice experiment for horses in which horses can freely choose to engage in a human-horse interaction or leave the experimental situation. Therewith, a first but important step towards answering the question whether horses find something positive in interaction with humans can be made which will help guide the public debate scientifically and contribute towards maintenance of social license to operate for the horse sector.

This experiment is relevant for everyone involved in the horse sector. It is important to stress that if the outcomes of this study should be negative, i.e. that horses do not choose interaction with a human voluntarily, this does not mean that social license to operate no longer exists, but instead, that we need to consider how we interact with and train horses, so the interaction becomes more desirable to them. This project is meant to be a basis for a larger focus area around positive welfare in horses within the Danish Centre for Companion Animal Welfare, i.e. the outcomes here will help shift the focus towards positive animal welfare and thereby be beneficial to everyone involved in the horse sector, as the expected knowledge gain is fundamental.

The main aim of this project is knowledge gain to fill a significant research gap which is currently one of the reasons for a polarised and opinionated debate that is strongly in need for scientific input. However, to be able to do this we need the knowledge here sought for. As this research has not been carried out before, the project is an important basic element for the further progression of a study of the human-horse relationship and the strengthening of the horse sector in the future.

Apart from the described deliverables which include outreach to other scientists, the horse sector and the wider public, the direct collaboration with the Danish Center for Companion Animal Welfare secures the continuation of communication of the new knowledge gained from this project as well as further research projects that will build on the scientific findings obtained from this project.

Therewith, the project is perfectly positioned to guide the current debate in a sensible, science-based way and improve the status, and, depending on the outcomes, potentially our interactions with horses in the future. A focus around positive animal welfare will strengthen the horse sector and a guided dialogue will ensure the social license to operate for the horse sector in the future. The knowledge gain aimed for in this project is thus essential not only for scientists and stakeholders in the horse sector but also to the society in general. A research focus around positive animal welfare in human-horse interactions is, at current, strongly needed and this project has the potential to be an important starting point therein.

Overall description of activities

This project will set up a test situation where horses are free to interact with a human handler performing one of four different treatments in a balanced order (a) no interaction (control), (b) human-horse interaction simulating grooming (scratching and petting), (c) training of a simple task using positive reinforcement, or (d) human-horse interaction including feeding food rewards by hand. Each treatment is performed across three subsequent days with 10 min pr horse/day, i.e. across 12 days pr horse with 4 days break between treatments (4 weeks in total with one treatment pr week). The interaction will be carried out in a test arena (approx. 20 x 40 m) with a known companion horse eating hay at one end of the arena (behind a fence). The test horse is free to stay with the human handler or leave the interaction doing something else such as standing by the companion horse.

Twenty young warmblood horses from a private stud will be included. The horses will be known to each other (group housed) and be used to occasional, yet not intensive handling by humans. During an initial habituation period of two weeks, the horses will be made familiar with the test arena and test procedure. The test person will lead the horse into the middle of the arena and start the interaction. A second person will take time and help ensure a standardized behaviour of the companion horse. A video camera will record the sessions to enable a subsequent, detailed behavioural analysis.

The whole experimental setup can be divided into five work packages, which are described here: 1) pilot testing, 2) practical data collection, 3) video analysis, 4) statistical data analysis, 5) dissemination of research results. 

WP1: stud and student recruitment, pilot testing

The first WP comprises agreements to be made with 1-2 private studs to ensure a sufficient number of horses for the study. Only then can the exact details of the setup be decided upon, as those will depend on the structure of the private studs involved. The set-up is tested through pilot studies and a student from either KU or AU is recruited to help out with the actual experiment. These kinds of projects are very popular among students, and we do not anticipate problems recruiting talented and devoted students.

WP2: habituation, training, practical data collection

WP 2 comprises the practical data collection, starting with habituation of the horses to the test arena and handlers. A habituation criterion will be used to ensure that the test horses are ready for the experiment, i.e.each horse can stay calmly in the test arena for 10 min. The actual tests are carried out across a 4-week period with 3 days of testing pr week (one treatment pr horse per week). It is expected that 10 horses can be included pr day, i.e. the data collection period will last 8 weeks for the 20 horses. The end of the practical data collection poses Milestone (M1).

WP3: video analysis

All tests will be video recorded. This material will be analysed using a pre-defined ethogram to analyse the behaviour of the horses in each test situation. Besides the evaluation of the behaviour by the ethogram, the horses’ position within the test arena will be determined every 10s. Therefore, the test arena will be separated into 10m2 grids and the grid in which the horse is positioned will be noted down. The observer will be thoroughly trained beforehand and inter- and intra-observer reliability checked regularly to prevent observer drift by repeated evaluation of a training data set that is not part of the analysed data. Frequencies and duration (when relevant) of different behaviours will thus be obtained and used for further analysis. The end of the video analysis poses Milestone (M2).

WP4: statistical data analysis

Data will be analysed with appropriate statistical tests. The exact models to be built will depend on the observed frequencies and duration of behaviours, but most likely, some of the behaviours will be grouped to enhance the data distribution. In general, the dependent variables will be: choosing human interaction yes/no, showing conflict behaviours yes/no, showing positive anticipatory behaviours yes/no, showing behaviours linked to positive emotions yes/no. Independent variables will be tested with regard to model quality (as determined by Akaike Information Criterion (AIC) but most likely include age and sex as fixed effects and horse (as random effect). The connections between the different variables obtained will be checked with a correlation matrix (in case of numerical distribution) and Chi2 tests (in case of categorical distribution).

WP5: dissemination of research results

The results will be published in at least one scientific publication, published gold open access to ensure broad accessibility worldwide (M3). The project group always puts a high emphasis on scientific outreach, i.e., making scientific results and their implications available and understandable to the broader public. All project partners have a broad outreach, for example they organize regular meetings and panel debates, both for scientists and the broader public, on relevant animal welfare topics and have good contacts to journalists from horse magazines such as Ridehesten, Cavallo and The Horse. The project group will use newsletters to reach out to stakeholders and actors within the horse industry (e.g., horse breeding organizations, studbook keepers, horse sports organizers, animal welfare organizations). In addition, information about the results will be disseminated via blogs aimed at horse owners via social media. Also, an effort will be made to reach a wider public through newspapers, radio and TV. Within the Danish Centre for Companion Animal Welfare, the aim is to establish a research focus to explore to what extent positive animal welfare in human-horse interactions exists and, if so, can be promoted, and this study aims to be the starting point for this initiative.

The project will therefore receive great attention at internal meetings and beyond due to the involvement of scientists and stakeholders. This study, naturally, has its limitations,and further studies are planned to build on the yielded insights so that outreach can well be expected for many years to come, even after the project has officially ended. For example, it is expected that the project results will be presented at two conferences in 2026, i.e. after the project has officially ended. It is important to stress that no matter what the scientific findings of this study are, this project will be beneficial for the horse industry as the project will help with the public dialogue as well as with improving our horse-human relationship and enhancing positive animal welfare in the horse sector, i.e. even if the horses should not choose human contact voluntarily, this project will help to guide steps into improving the horse-human interaction in determining what the best possible human-horse interaction looks like (as perceived by the horses).

The horses will be housed in their usual home environment. The experimental procedure described is not expected to cause any harm, suffering or damage to the animals and is not to be considered an animal experiment according to national law requirements. Nevertheless, ethical allowance will be applied for with the Animal Ethics Institutional Review Board at the University of Copenhagen upon acceptance of the project.

The project is a collaboration between University of Copenhagen and Aarhus University. The project leader (Irena Czycholl, University of Copenhagen) is in charge of taking care that everything can be carried out as described. Janne Winther Christensen will be in charge of the practical data collection and the contact to the private stud(s). Irena and Janne will co-supervise the participating student and Peter Sandøe will ensure outreach to the general public.

Irena Czycholl: Irena has started in 2022 as Associate Professor for Animal Welfare and Behaviour at the University of Copenhagen. She is an educated veterinarian and has obtained her PhD about welfare assessment in 2015 at the University of Veterinary Medicine Hannover, Germany. Thereafter, she has worked as Postdoc at Kiel University (farm animal science) and focused amongst others on relevant challenges in horse husbandry (e.g. possibilities and challenges in group housing systems (e.g. Hildebrandt et al., (2020; 2021; 2022); Weidling et al., (2024)) and riding (e.g. evaluation of stress in riding sessions (Stange et al., 2023) as well as measuring animal welfare in horses (Czycholl et al., (2017; 2018; 2019; 2021). Irena Czycholl is involved in the EU Cost action on defining and finding indicators for the reliable, valid and feasible assessment of positive animal welfare. Furthermore, she is vice president of the Danish Centre for Companion Animal Welfare and will thus be responsible for development and maintaining a research focus of positive animal welfare in horses.

Janne Winther Christensen: Janne has a background in biology, a PhD in ethology and has been Associate Professor in animal science at Aarhus University since 2015. She has a strong research focus in equitation science and thus extensive experience in evaluating training methods in horses (e.g., Christensen et al., 2005; 2006; 2012; 2021b; Hartmann et al., 2011; McLean and Christensen, 2017), the human-horse relationship (Hartmann et al., 2017), as well as conflict and other stress-related behaviors in the ridden horse in different contexts and breeds (Christensen et al., 2014; 2021a; 2024). She therefore has extensive experience in designing ethograms to evaluate human-horse situations. She focuses not only on the riding situation, but also on the situations before (e.g., tacking up, mounting) in a current ongoing project. Her role as Senior Vice President of the International Society for Equitation Science gives her a large international network through which she can also disseminate the results of the project. Janne Winther Christensen has several contacts to private studs in which the experimental setup will take place. Like Irena and Peter, she is senior member of the Danish Centre for Companion Animal Welfare and will support development and maintenance of a research focus around positive animal welfare in horses.

Peter Sandøe: Peter is Professor of Bioethics at the University of Copenhagen. He studied philosophy and also obtained his PhD in this field at Oxford University in 1988. Since 1990 a major part of his research has been within bioethics with particular emphasis on ethical issues in the human-animal-relationship. He has largely committed to interdisciplinary work combining perspectives from natural science, social science and philosophy. For the last 20 years he has collaborated with sociologists to study ethical issues and animal welfare challenges from an empirical perspective, originally mostly through qualitative studies, recently through mixed methods and surveys. Peter Sandøe is involved in the EU Cost action on defining and finding ethical standards for positive animal welfare. Likewise, his former work on positive animal welfare (Lawrence et al., 2019; Vigors et al., 2021; Rault et al., 2022) can be seen as preliminary studies for the research envisaged here. Furthermore, he has been the driving force in establishing and ensuring funding for a highly visible and productive Centre for Companion Animal Welfare, which will be involved in co-funding this study and, with the present study as a starting point, developing a larger research focus around the topic of positive animal welfare in the ridden horse/horses in horse-human contact in the years to come. Finally, he is very skilled in and devoted to ensuring that scientific findings achieve a wide public dissemination, nationally and internationally, as documented by a couple of communication prizes). Altogether, this project group does not only have the scientific background needed to carry out this project, but also the perfect basis for dissemination of research results and maintenance of a larger research focus in positive animal welfare in horses in the future with this project as starting point.

The measurable project deliverables will be:

• One scientific peer-reviewed publication (gold open access) in a relevant journal (e.g., Applied Animal Behaviour Science)
• Information about the project and project results on the homepage of the Danish Centre for Companion Animal Welfare
• Dissemination via blogs and social media
• At least two popular scientific contributions in horse magazines
• At least two cases of dissemination via newspapers, radio or TV or as podcast (e.g. via Videnslyd)

One of the main deliverables of the project is the publication in a scientific peer-reviewed journal as gold open access to ensure broad and general accessibility for everyone. This is also important to inform other scientists about this project and guide future projects in the area of positive animal welfare within the human-horse interaction. Apart from the scientific publication, the project group will inform about the project and project results in at least two popular scientific contributions to classical horse magazines such as Ridehesten or Cavallo to inform especially interested horse owners about this project.

Also, we will ensure that a wider public hears about the project via traditional news media. The group has a very good track record for this kind of dissemination, both nationally and internationally. Likewise, an audible deliverable will be given due to the known wide outreach of podcasts and radio interviews. On a homepage, as well as on social media channels commonly used by the project group for science communication, the public will further be informed about this project and the new insights gained. Science communication and the broad spread of the expected important findings of this project will further be ensured by the fact that this project is meant to be the starting point for a research focus in the area of positive animal welfare of the Danish Centre for Companion Animal Welfare in the years to come.

The project group offers the perfect frame for this and the establishment of such a research focus offers the unique opportunity for this project to serve as a beacon in this area. It is important to stress that even if the outcomes should be that the horses in this experimental setup do not voluntarily choose human interaction, this would not mean that the horse sector would lose its social license to operate but in contrast strengthen the horse sector as guidance on how human-horse interaction might need to be changed to be perceived as positive by the horses could be provided. Research as proposed in this project does not yet exist but is strongly needed to guide the often highly polarised public debate scientifically and secure the social license to operate for the horse sector in the years to come.

The large outreach of the project group secures implementation of the scientific findings and thus guiding the direction of the debate in a sensible, science-based way. As all members of the project group are involved in teaching activities, of course, the results of this important project will also be implemented in those, i.e. education within the larger area of ethics in human-animal-interactions, animal and veterinary science.

Christensen, J.W., Ahrendt, L.P., Lintrup, R., Gaillard, C., Palme, R., Malmkvist, J., 2012. Does learning performance in horses relate to fearfulness, baseline stress hormone, and social rank? Applied Animal Behaviour Science 140, 44-52.

Christensen, J.W., Ahrendt, L.P., Malmkvist, J., Nicol, C., 2021a. Exploratory behaviour towards novel objects is associated with enhanced learning in young horses. Scientific reports 11, 1428.

Christensen, J.W., Beekmans, M., Van Dalum, M., VanDierendonck, M., 2014. Effects of hyperflexion on acute stress responses in ridden dressage horses. Physiology & behavior 128, 39-45.

Christensen, J.W., Jensen, D., von Borstel, U.U.K., 2024. Conflict behaviour in Icelandic horses during elite competition. Applied Animal Behaviour Science 271, 106166.

Christensen, J.W., Keeling, L.J., Nielsen, B.L., 2005. Responses of horses to novel visual, olfactory and auditory stimuli. Applied Animal Behaviour Science 93, 53-65.

Christensen, J.W., Munk, R., Hawson, L., Palme, R., Larsen, T., Egenvall, A., von Borstel, U.U.K., Rørvang, M.V., 2021b. Rider effects on horses’ conflict behaviour, rein tension, physiological measures and rideability scores. Applied Animal Behaviour Science 234, 105184.

Christensen, J.W., Rundgren, M., Olsson, K., 2006. Training methods for horses: habituation to a frightening stimulus. Equine Veterinary Journal 38, 439-443.

Czycholl, I., Büttner, K., Klingbeil, P., Krieter, J., 2018. An indication of reliability of the two-level approach of the AWIN welfare assessment protocol for horses. Animals 8, 7.

Czycholl, I., Büttner, K., Klingbeil, P., Krieter, J., 2021. Evaluation of consistency over time of the use of the animal welfare indicators protocol for horses. Animal Welfare 30, 81-90.

Czycholl, I., Klingbeil, P., Krieter, J., 2017. Suitability of animal-based indicators for the detection of animal welfare in horses. Tieraerztliche Umschau 72, 209-217.

Czycholl, I., Klingbeil, P., Krieter, J., 2019. Interobserver reliability of the animal welfare indicators welfare assessment protocol for horses. Journal of equine veterinary science 75, 112-121.

Dyson, S., Pollard, D., 2020. Application of a Ridden Horse Pain Ethogram and its relationship with gait in a convenience sample of 60 riding horses. Animals 10, 1044.

Hartmann, E., Christensen, J.W., Keeling, L., 2011. Training young horses to social separation: Effect of a companion horse on training efficiency. Equine veterinary journal 43, 580-584.

Hildebrandt, F., Büttner, K., Krieter, J., Czycholl, I., 2022. The behavior of horses stabled in a large group at essential resources (watering point and lying halls). Journal of Equine Veterinary Science 118, 104106.

Hildebrandt, F., Büttner, K., Salau, J., Krieter, J., Czycholl, I., 2021. Area and resource utilization of group-housed horses in an active stable. Animals 11, 2777.

Hildebrandt, F., Krieter, J., Büttner, K., Salau, J., Czycholl, I., 2020. Distances walked by long established and newcomer horses in an open stable system in Northern Germany. Journal of Equine Veterinary Science 95, 103282.

Holzer, S., Herholz, C., Tanadini, L.G., Ineichen, S., Julliand, S., 2022. Hay preferences in horses versus selection by their owners. Livestock Science 258, 104896.

Kieson, E., Felix, C., Webb, S., Abramson, C.I., 2020. The effects of a choice test between food rewards and human interaction in a herd of domestic horses of varying breeds and experiences. Applied Animal Behaviour Science 231, 105075.

Lawrence, A.B., Vigors, B., Sandøe, P., 2019. What Is so Positive about Positive Animal Welfare?—A Critical Review of the Literature. Animals 9, 783.

Lee, J., Floyd, T., Erb, H., Houpt, K., 2011. Preference and demand for exercise in stabled horses. Applied Animal Behaviour Science 130, 91-100.

LIFT, 2024. A definition of positive animal welfare.

McDonnell, S.M., 2003. The equid ethogram: a practical field guide to horse behavior. Eclipse Press.

McLean, A.N., Christensen, J.W., 2017. The application of learning theory in horse training. Applied Animal Behaviour Science 190, 18-27.

Mejdell, C.M., Jørgensen, G.H., Buvik, T., Torp, T., Bøe, K.E., 2019. The effect of weather conditions on the preference in horses for wearing blankets. Applied Animal Behaviour Science 212, 52-57.

Phelipon, R., Ascione, A., Ruet, A., Bertin, A., Biau, S., Arnould, C., Boichot, L., Lansade, L., 2024. Differences in behaviour, facial expressions and locomotion between positive anticipation and frustration in horses. Applied Animal Behaviour Science 274, 106262.

Rault, J.-L., Sandøe, P., Sonntag, Q., Stuardo, L., 2022. Positive Animal Welfare: Bridging the Gap or Raising Inequalities Worldwide? Frontiers in Animal Science 3.

Stange, L.M., Wilder, T., Siebler, D., Krieter, J., Czycholl, I., 2023. Comparison of physiological and ethological indicators and rein tension in headshakers and control horses in riding tests. Applied Animal Behaviour Science 263, 105943.

Vigors, B., Sandøe, P., Lawrence, A.B., 2021. Positive Welfare in Science and Society: Differences, Similarities and Synergies. Frontiers in Animal Science 2.

von Borstel, U.K., Keil, J., 2012. Horses' behavior and heart rate in a preference test for shorter and longer riding bouts. Journal of Veterinary Behavior 7, 362-374.

von Borstel, U.U., Duncan, I.J.H., Shoveller, A.K., Merkies, K., Keeling, L.J., Millman, S.T., 2009. Impact of riding in a coercively obtained Rollkur posture on welfare and fear of performance horses. Applied Animal Behaviour Science 116, 228-236.

Weidling, G., Krieter, J., Lübben, R., Czycholl, I., 2024. Dominance hierarchy does not influence distances travelled and area utilization in a large group of ponies. Applied Animal Behaviour Science 271, 106178.