Historically this occurred through breeding of animals for physical attributes such as size, strength or speed. While selection for desirable behaviours may have played a part, it was undoubtedly of secondary consideration.
Today, horses and ponies are largely bred for equestrian activities rather agricultural or military purposes. Temperament is now at least as important as athletic ability in determining the suitability and value of modern equines.
This begs the question of whether, and to what extent, the temperament of horses is inherited. Or is it that equine personality is shaped by environmental factors, such as diet, workload and interaction with humans and other animals? Can we simply breed better behaved horses or do we need to train them that way?
Evidence for genetic influence comes from the fact that many breeds are described as having characteristic temperaments. Friesians are labelled as loyal, willing, placid and cheerful, whereas the Irish Draught is promoted for having an even temperament.
These stereotypes are supported by the results of scientific research conducted at Bishop Burton, Moulton and Harper Adams agricultural colleges. The personalities of 1223 horses, of eight different breeds, were profiled and this identified clear and characteristic differences.
Thoroughbred horses were found to have high degrees of dominance, anxiousness, excitability, sociability and inquisitiveness compared with other breeds. Irish Draught horses, on the other hand, had relatively low scores for dominance, excitability, sociability and inquisitiveness, but were highly protective in their behaviour.
The results were largely consistent with perceived breed stereotypes, with Thoroughbreds, Arabs and Welsh ponies/cobs being most anxious and excitable. Thoroughbreds were also the most inquisitive and this may account for their aptitude for cross country jumping where quick-wittedness and opportunism help cope with complicated and previously unseen challenges.
In order to further this research, scientists have identified five temperament traits; these include shyness/boldness, exploration/avoidance, activity, sociability and aggression type behaviours. They have investigated the relative contribution of genetic factors to the development and expression of these traits, as well as looking for candidate genes that might explain variation in temperament.
In one study, conducted using Italian Haflinger horses, estimates of the inheritance of temperament traits ranged from 0.02 to 0.06. This suggested that only between 2% and 6% of these horses' temperament was genetic in origin. This compared with 53% for coat colour and pattern.
In a second study, heritability of behavioural responses was studied in Thoroughbred horses in Japan. Here the heritability of behaviour was much higher, at between 23% and 28%, indicating that around one quarter of temperament might be genetic in origin with the remainder resulting from environmental factors.
Further research found that genetic effects appear to influence shyness/boldness and exploration/avoidance behaviours, whereas environmental factors affect sociability and learning abilities. Additive effects were also noted, indicating that the way in which we feed, work and interact with horses can modify inherited behavioural traits.
Efforts to identify genes that might control temperament have focused on those that produce neurotransmitter chemicals and their receptors in the brain. These include dopamine and serotonin, which play key roles in attention, mood, motivation, and learning.
The dopamine receptor gene is known to influence novel-seeking behaviour, or curiosity, in humans. Two distinct variants or polymorphisms of this gene have been identified in horses and their expression appears to correlate with temperament. Thoroughbreds with one of the polymorphisms were found to have higher levels of curiosity and lower vigilance than horse with the other variant.
Similarly, serotonin transport plays a role in the expression of anxiety and four variants of the coding gene have been identified in horses. However, none of these polymorphisms appear to influence equine anxiety, at least in the Thoroughbred horses tested.
This research confirms that temperament is, to some extent, inherited and highlights benefits that will be achieved from breeding horses with good temperaments.
The reality is that, just as most experienced horse folk will verify, between 70% and 98% of a horse's temperament is shaped by his or her experiences. And recent behavioural research has shown that this process starts very early in life.
Gently handling of mares has shown to have a positive effect on the reaction of their foals to humans and strongly influence the development of human-foal relations. Such responses are apparent in foals just a few days old. Similar effects are also seen at six months of age but are, by this age, clouded by the development of foals' own behavioural characteristics.
The importance of starting socialisation and other behavioural training early in life is highlighted by scientific evidence that, once a foal develops a certain temperamental trait, it always has the potential to display this in later life.
Research conducted in France showed that the propensity of foals to react by avoiding humans and novel objects is low at around three weeks but becomes more prevalent as they get older. Once a foal manifests these behaviours, it will always reproduce the same reaction.
Taken together, these studies emphasise the critical importance of early experiences in determining the temperament of horses. Interestingly, later in life, the effects appear less pronounced and more difficult to predict.
Finally, a Dutch study recently examined how personality of riders interacts with the temperament of horses. Horses that were more emotionally reactive were more evasive when ridden and cooperation between horse and rider was greatest with horses that were less emotionally reactive. Interestingly a rider's personality was not a significant factor in determining the relationship between horse and rider.
Fortunately similar research to determine the extent to which the temperament of equine vets influences the behaviour of their patients has not yet been carried out!