Contemplating the Welfare and Training of Horses

Putting the science back into equine behaviour – Equine Behaviour Symposium 2011 by Suzanne Rogers

There are many websites, books and people (including me) claiming expertise in equine behaviour. They give advice to horse owners, demonstrate practical work with horses and often sell specialized tack or other tools that claim to fix a myriad of problems. However, many of the messages are confusing and conflict with each other – it can be difficult for people who want to learn about behavior to identify credible sources of information and find training methods that will not only work but also be enjoyable for owner and horse.

Horse sense/horse nonsense?

Many of the methods of horsemanship that are becoming increasingly popular in the UK simply do not make sense. For example, some claim to be based on how horses communicate in a herd but then upon investigation equine behaviour is actually being interpreted through the eyes of humans, ignoring the findings of people who have studied horses and published research. Some claim that their methods are kinder than other methods but again these claims are often made by those who have little understanding of how animals learn, or how horses communicate that they are stressed, anxious and frightened. Thus, there is a gap between what we know about horses and what is being marketed and sadly many of the methods people use in an attempt to improve their relationship with their horse do the opposite – the horse becomes frightened of the owner.

Learning from dog training

The world of dog training has faced the same problems. In the past, much dog behavior was interpreted in terms of ‘dominance’. This was due to early studies on wild wolves and their social structure. People were quick to apply this to pet dogs with disastrous results. However, scientists later realized that it is wrong to apply what they had seen in some wolf social structures to that of the domestic dog – and certainly wrong to think of the human as part of a canine social structure. Theories of training based on dominance were thus disproved. However, because of the simple message ‘you need to dominate your dog!’ some owners and trainers are still basing their training on this flawed reasoning. For a useful guide to this topic see the information on the Welfare in Dog Training campaign website (

Using science as a basis for how you treat your animals therefore seems a much better idea than believing what people tell you because that is what they have come up with. However, as more is found out about behavior we might need to leave some of our theories behind – it is important to think as scientists and make decisions using the information available. The Equine Behavior Forum aims to help people interested in equine behavior to find out about the latest research and do just that.

Time to stop keeping horses in stables?

This year’s symposium started with Dr Emma Crieghton (Newcastle University) presenting her work on group housing for horses. Although intuitively many of us would suggest that keeping a social prey species in a small wooden box for many hours a day is not going to meet their behavioural needs, stabling horses is an embedded part of equestrian culture in the UK. Emma first compared horses living in single stables to humans living in jail cells. Keeping horses in stables results in well-documented behavior problems such as stereotypical behavior (crib biting, weaving, box walking etc.). Research shows that the more time horses spend in stables the more behavioural problems are reported by the owners and that stabled horses have disrupted feeding behavior and are more likely to have an abnormal level of activity upon release.

But is it practical to keep horses in social housing such as large barn areas with more than one horse? Emma described her work with the Blue Cross comparing horses kept in social housing with those in individual stables. They found that:

  1. Horses kept in social housing take less time to manage than those kept in traditional stables
  2. Horses kept in social housing require less bedding than when kept in traditional stables
  3. Horses in social housing are less ‘dirty’ than those kept in traditional stables
  4. Horses in social housing are a little more difficult to catch but this might be because the carers were not used to catching horses kept in groups.

In addition, little aggressive behavior between horses was seen in social housing. Therefore it seems clear that social housing for horses is preferable than traditional stables – both for the welfare of the horses and for ease of management. Emma reported some encouraging changes made by welfare organizations and colleges moving towards social housing. Those of us working with horses and their owners now need to ensure that this information is more widely available and to provide support for people considering a change of management system.

Comparing (aversive) training methods

The second speaker was Dr Veronica Fowler (Institute for Animal Health, Guildford). Veronica presented the preliminary results of a long-awaited study. There has been considerable discussion in the equine training community about whether or not some natural horsemanship methods, such as those by Monty Roberts, are more ethical (as claimed) than other approaches. Equine scientists argue that a key component of these methods known as ‘join up’ is based on flawed interpretation of equine ethology and without understanding of how horses learn. They argue that join up and the subsequent training cause stress to horses, are based on motivating the horse through fear and involve erroneous interpretation of body language and behavior. It was thought that this study was about researching scientifically if these methods are actually stressful to horses through recording the heart rate (a controversial method of measuring stress in itself) during such a training process. However, the study presented was actually a comparison between the Monty Roberts technique and a ‘conventional’ technique for the initial training of riding horses. The methods were not described apart from join up being used by Monty Roberts and ‘conventional/traditional’ training being used by the other trainer. The hypothesis was that the former would show less conflict behavior, decrease psychological and physiological stress and not have any compromise on performance.

The study was conducted at Sparsholt College with 14 untrained horses; 4 mares and 10 geldings. The horses arrived 2 days before the study started and were stabled with ad lib hay apart from two, ten minute sessions a day turnout in the school. The horses were trained for 30 minutes per day each and then on days 21 and 22 a series of tests were completed (ridden flatwork and an obstacle course) and judged by a panel of multi-disciplinary experts.

So what did they find? The horses trained by Monty Roberts were saddled quicker than by the conventional trainer and the horses had a lower heart rate during the procedure. The horses trained by Monty Roberts had a lower maximum heart rate when a rider was first introduced. The horses trained by Monty Roberts performed 30% better on the tests at the end.

What does this tell us? Firstly, is speed important in training horses? Does it matter if one method takes a little longer than the other? Secondly, for many people interested in the ethics of training this study compared one aversive method with a different aversive method. As such the results aren’t that interesting but this study has perhaps been the push needed towards more research of this kind. Hopefully future studies will compare aversive methods with those that apply what we know about the science of learning and equine behaviour (those based primarily on the use of positive reinforcement). Surely this study can’t be far behind so we must try to curb our frustration that this method might be cited for years to come to show that Monty’s method is less stressful than another method, and consider it as an early step in successive approximation (shaping) towards research that will give us the whole picture comparing different methods on criteria that are important to the horse as well as the human.

From theory to practice?

The final presentation was by Dr Andrew McLean (Australian Equine Behaviour Centre) and was entitled ‘equine learning theory and its effect on training and behaviour’. Dr McLean introduced the subject of cognitive ethology and its importance and asked the audience to not think of the horse as human but to instead train properly – not punish the horse for not knowing. He explained that humans have previously thought that horses must be submissive but that we don’t need the flawed concept of needing to dominate our horses, just to be a good trainer. I must admit to finding some of the presentation confusing as many well-established learning theory terms were described in a way that doesn’t match what is published elsewhere. Also Dr McClean is known for publishing and promoting positive reinforcement but appeared to describe training that focused mainly on negative reinforcement and positive punishment. Although I don’t doubt that his application of learning theory to training is a vast improvement on some traditional methods, this left the audience a little confused at the gap that seemed to exist between the theory and practice, and lead to a lively discussion in the questions and answer sessions.

Final thoughts

Overall the symposium was an excellent reminder that science provides evidence for the application of ethology. The challenge now lies in helping the results reach the horse owners, and therefore horses, who need it. Only then can we change the equestrian culture surrounding management and training to one based on what we actually know about horses, not assumptions. The symposium highlighted for me that perhaps the biggest challenge lies in ensuring that research undertaken provides answers to questions that reflect the modern equestrian world and that it is not so far behind as to answer questions no longer being asked.