Animal Welfare Information Center

Information Resources on the Care and Welfare of Horses

Return to Contents

Horse Welfare

by Daniel Mills, PhD

Department of Biological Sciences
University of Lincoln
Riseholme Park

The scientific assessment of equine welfare has grown markedly over the last decade, but welfare assessment is not an easy task as it is by its very nature multidisciplinary. It is therefore timely for a review of available resources to help all of us interested in improving horse welfare. It is particularly important as there is still no complete consensus on what “welfare” really is, even among welfare scientists. Some will emphasise the importance of health, others, feelings and yet others the ability for an animal to fulfil its natural potential. For some, the feelings of an individual at a given time should be referred to as its well-being, with welfare considering not just the animal’s current feelings but also the threats to its well-being. Thus it might be argued that horses whose work puts them at risk of particular injury, such as limb injury in race horses, have worse welfare than those whose work does not pose such risks. This approach can be useful when considering welfare within the context of populations, but it is important not to forget the individual. Even if the risk is 1 in 10000 for a catastrophic injury, the consequences for that one individual that is injured are disastrous. Some also consider the effort required by an animal to maintain its state of well-being. Thus Broom (1986) defines welfare as the state of an animal with regards to its attempts to cope with its environment. This would imply that a horse kept in an optimal ambient temperature may be thought to have better welfare than one who must devote additional resources to the maintenance of such a temperature. It is therefore important to clarify what an individual means by the term “welfare” especially when one animal’s welfare is said to be better for another.

However, a publication like this makes no judgement and allows individuals and groups to access the latest information they need for their purposes. It is sad truth that well-meaning intentions do not always translate into well-being of our charges, and horses are particularly a victim of this. Being kept in quarters that look comfortable to carers may not be most appropriate for horses, who have strong social tendencies and a physiology and psychology adapted to a life feeding on open grasslands.

A common criticism of those seeking to assess animal welfare objectively is that we cannot know the mind of another; but this criticism reflects a failure to understand the fundamental principles of the scientific method. Science makes progress through hypotheses which can be tested but which can never be shown to be completely true. In this respect welfare is no different to any other scientific pursuit. All scientific evidence carries with it a degree of uncertainty and we decide what level we are prepared to accept (often a 5% chance level). However these principles are often forgotten by sceptics who, for whatever reason, appear not to want to contemplate what might be happening in the minds of other animals. There is a need for those interested in animal welfare to make clear and defend the scientific basis of their work in order to propose what is possible, realistic and reasonably justifiable.

There are many reasons to be interested in horse welfare. From an ethical perspective it might be argued that we have a responsibility to minimise the suffering we cause to other animals with whom we interact either directly or indirectly. From a practical point of view, animals with good welfare might be expected to perform their work more efficiently. From an academic perspective the assessment of animal welfare is also a challenging intellectual task. However, it is not the aim of this publication to argue why we should measure horse welfare or the ethics of what we do to horses, rather it is hoped that to bring together information to help increase awareness of the methods at our disposal for the assessment and management of the welfare of the horse in a broad range of contexts.

A practical problem for advances in equine welfare concerns the need for funding in this field. Horses are expensive animals to keep and study; and although they are of enormous economic importance, the industry is fragmented and often poorly represented to governments and other funding bodies as a significant concern, except perhaps for the case of the equine athlete. The welfare research field has largely been driven by concern over whether what we do to animals is acceptable and not purely by the intellectual issues involved. As a result research funding has focussed on political priorities. Thus a large proportion of equine welfare research focuses on the problems faced at the extremes of athletic performance, which is largely irrelevant to the vast majority of equids.

More broadly the scientific study of animal welfare has invested enormously in how to assess suffering so that it can be minimised, rather than the evaluation of well-being and positive mental states, which is perhaps the goal of the average carer. Trying to measure “happiness” is not only an enormous intellectual challenge, but also a completely alien concept to many funding bodies and so generally given very little attention despite its central importance to those interested in animal welfare. Scientists also vary in the subjective feelings they are willing to ascribe to a horse (see Price et al., 2002, for an illustration of attitude amongst U.K. veterinary surgeons), and so this field is likely to remain a contentious area of research. Politically speaking laboratory and farm animal welfare have been major areas of concern and so they have been the focus for funding with little money being available to those interested in the well-being of companion animals including the horse. This does not mean that there are any fewer problems in these species, just that they are largely overlooked by both the public and funding bodies. For these reasons there are significant gaps in our scientific knowledge of horse welfare, but we are able to recognise rational approaches to its scientific study. This publication helps to identify what we know and what can be applied, but will also identify the gaps in our knowledge.

Horses have evolved to be adapted to their natural environment and so if the behaviour of a horse resembles this natural state then it might be thought that it is normal and suffering minimal. However, the occurrence of normal behaviour patterns does not necessarily imply good welfare. Some normal behaviours are clearly associated with aversion, such as flight from a predator and so their occurrence is undesirable. In other instances the significance of the behaviour depends on the context. For example, horses may move into water away from dry land in order to avoid the effects of blood-sucking insects which can cause anaemia. Whilst this is obviously beneficial to the horse, this behaviour is not without its costs. They may reduce grazing or reduce other important behaviours as a trade off for escaping the insects. So, whilst standing idle or engaging in social exchanges with others does not appear to be a behaviour of concern, in this case the horses would undoubtedly be better off doing something else if the insects were absent

Some of the adaptations horses have evolved to help them to cope with the challenges they encounter in their natural environment are sufficient to protect them when faced with challenges unique to the domestic environment, but in other cases the horse may not be able to adapt adequately, in which case there must be concern for the animal’s welfare. The problem is identifying when these natural systems are being over-taxed. Behaviour is a form of adaptation to the environment and so its evaluation is dependent upon context. It is important not to make unjustified generalisations. Two horses in different environments might be expected to show different behaviours as well as different patterns of behaviour as part of their healthy adaptation to the different environments. So there may be no norm against which the behaviour of a captive animal can be meaningfully compared. Individuals should avoid the temptation to make arbitrary comparisons when it suits their case, for example the amount of time a horse should spend grazing or alert in a day to be normal or psychological healthy. There is no logic in the assumption that quantitative or qualitative differences in behaviour necessarily imply a difference in welfare. Instead it is important to appreciate the function and regulation of each behaviour in the context in which it occurs, and look for other evidence of the welfare status of the individual. Understanding the natural behaviour of horses is important to the scientific study of equine welfare as it is through this we can appreciate the functional significance of a given behaviour and the mechanisms which may underlie the adaptability of the horse and when these may be strained in an unnatural setting, for example concentrate feeding or the use of raised hay nets.

When trying to evaluate the risks of a given horse management system, we might use one of two approaches and the information in this publication is useful to both processes. We might consider what is within the system which might potentially compromise the welfare of the animal or what signs come out of the system. The former are indirect indicators of areas of potential focus while the latter are more direct measures of what is happening in practice. In both cases the horse, its management and its environment make up the system of interest. Indirect measures can be useful as they can flag up areas of potential concern or interest, for example consider a horse being kept in a livery yard versus one being kept for racing or PMU production. They have different demands put upon them, might be kept in quite different environments by people of differing level of skill in welfare monitoring. One system might be considered lower risk than another, but that does not mean that the welfare is necessarily safe nor the opposite true in a higher risk system, but it does allow prioritisation of concern. The information on the effects of different types of procedures (management, training, veterinary or otherwise), within this particular publication is therefore particularly useful in this regard. The welfare can only be reasonably determined by looking at the direct measures (outputs) of the system, which relate to the behaviour and physiology of the horses and again this publication will help in the evaluation of these measures.

Pain is often the primary concern of most carers and is defined as: “An unpleasant sensory and emotional experience associated with actual or potential tissue damage or described in terms of such damage” (IASP, 1979). The recognition of pain has received most scientific study, perhaps because of the historical importance of the veterinary profession in the study of animal welfare and its obvious association with physical lesions. A variety of techniques have been proposed in the scientific literature for the assessment of pain in the horse based on behavioural and physiological measures e.g. the assessment of activity budgets pre and post surgery (Price et al., 2003), response to analgesia (Dyson 2002), and median frequency within electro-encephalogram recordings (Murrell et al., 2003), but these are generally restricted to use within a specific context. More general measures, such as indicators of sympathetic nervous system activity tend to be non-specific to suffering, unvalidated and / or contradictory. Nonetheless, the ethology of pain is probably worthy of further attention.

It is also important to appreciate that pain is probably not the only aversive feeling experienced by horses. Dawkins (1990) argues that suffering consists of a “a wide range of emotional states that occur when an animal is blocked from carrying out actions that are biologically mandated, normally reduce harm or risk to life or concern reproduction.” This stance has spawned a wealth of work designed to assess the needs of animals in order to determine which are biologically mandated. Simple preference tests may tell us what an animal prefers but they do not tell us if an animal is suffering if it is deprived of the preferred choice. I may prefer Bordeaux wines to Champagnes but I can be happy with either! So scientists have developed techniques where they have started to look at the price an animal is willing to pay for a given commodity. In this way we can have a clearer idea of what is really important to an animal. However, whilst there is nothing to theoretically stop such work being done in the horse, the cost of building the experimental apparatus and housing the number of animals necessary for the time required to obtain sound data has to date largely prevented such work in the horse; although Houpt’s group at Cornell (Lee et al 2001) has conducted preliminary studies to assess the strength of a horse’s motivation for exercise and companionship when confined for 23 hours in the day. A higher price was paid for companionship over exercise and this reinforces much work by our own group which suggests that social isolation is one of the primary problems with many modern housing systems (see Mills and Clarke, 2002 for a review). There is undoubtedly a need for more work in this field and we can only hope that those who have the potential to fund it recognise its importance so we can objectively assess the welfare of the horse in a variety of contexts.

In short this publication is an essential resource for all those who work both directly and indirectly with horses.


Broom DM (1986). Indicators of poor welfare. British Veterinary Journal 142, 524-526.

Dawkins MS (1990). From an animal’s point of view: motivation, fitness and animal welfare. Behaviour and Brain Sciences 13, 1-61.

Dyson S (2002). Subjective and quantitative scintigraphic assessment of the equine foot and its relationship with foot pain. Equine Veterinary Journal 34, 164-170.

International Association for the Study of Pain (IASP) (1979). Pain terms: a list with definitions and notes on usage. Pain 6, 249-252.

Lee J, Floyd T, Houpt K. (2001). Operant and two choice preference applied to equine welfare. In: Proceedings of the 35th International Congress of the ISAE, Garner JP, Mench JA, Heekin SP (eds) The Centre for Animal Welfare, Davis, 110.

Mills DS, Clarke A (2002). Housing Management and Welfare. In: The Welfare of the Horse, Waran N. (ed). Kluwer Academic, Dordrecht, NL. 77-97.

Murrell JC, Johnson CB, White KL, Taylor PM, Haberham ZL, Waterman-Pearson AE (2003). Changes in the EEG during castration in horses and ponies anaesthetized with halothane. Veterinary Anaesthesia and Analgesia 30, 138-146.

Price J, Marques JM, Welsh EM, Waran NK (2002). Pilot epidemiological study of attitudes towards pain in horses. Veterinary Record 151, 570-575.

Price J, Catriona S, Welsh EM, Waran NK (2003). Preliminary evaluation of a behaviour based system for assessment of post-operative pain in horses following arthroscopic surgery. Veterinary Anaesthesia and Analgesia 30, 124-137.

Return to Top

Return to Contents