Smart horses may not necessarily be the easiest to train, research in Poland suggests.
Scientists from the University of Life Sciences in Lublin set about a series of tests on 120 horses to determine whether their ability to open the wooden lid on a box to access food inside indicated how they would handle training.
Their results revealed that shorter times taken to learn how to open the box was associated with a longer time needed before being mounted for the first time in a round pen.
“It seems that horses characterized by a high learning skill in this aspect do not cooperate well with the trainer during mounting,” Witold Kędzierski and his colleagues reported in the journal, Applied Animal Behaviour Science.
“Professional riders often complain that high-performance horses are very difficult to train because they do not want to be submissive,” they reported, pointing to 2012 research showing that horses who were more attentive to humans had lower cognitive skills than horses which were not dependent on humans.
“Our results correspond with that finding.”
Kędzierski, joined in the research by Izabela Wilka, Anna Stachurska and Iwona Janczareka, based their findings on experiments involving 120 horses undergoing training.
“Training the horse is based on human-horse relationship, which requires that the horse shows active, conscious behaviour as well as an ability to learn,” the researchers said.
Their study involved 40 Thoroughbreds, 40 purebred Arabians, and 40 Anglo-arabians (thoroughbred-arabian cross). There were 20 colts and 20 fillies in each breed group. All the horses were intended for racing.
The times needed by the horse to open the food box, in the first phase of the study, were assumed as a measure of an aspect of each individual horse’s learning skill.
In this phase, each horse witnessed a woman placing several handfuls of oats into the box. She then demonstrated the operation of the lid.
Each horse was then given a maximum of 15 minutes to access the food. If a horse was successful, the lid was replaced and the horse monitored to see if it opened the lid successfully again. (This was to ensure the first successful access was not a fluke.)
The horses’ box-opening abilities were assessed again the following day, with the times taken to open it measured with a stopwatch. They were, in general, much faster on the second day.
The researchers then evaluated the times needed to the first saddling and the first mounting of each horse, as well as each animal’s reactions. Three licensed trainers were used, employing what the authors called a sympathetic method based on a freestyle training technique applied in Europe.
Heart rate and heart-rate variability were monitored during the saddling and mounting tests.
They then assessed these findings with the results of the food-box test in a bid to identify patterns.
“It took the Thoroughbred horses significantly longer to open the crib than it took the Anglo-arabian horses – 306 seconds versus 191 seconds, respectively,” they reported.
The researchers said they found many statistically significant correlations.
“It seems that the horses which quickly succeed in the crib opening test are calm during the first phase of the initial training until saddling, but they do not willingly accept mounting.
“However, the relationship between the time of opening the crib and response to the training should be analysed separately in particular horse breeds and sexes.”
The researchers said it was known that individual horses differed in their learning ability and ability to memorise.
“Individual learning abilities in horses are generally consistent over a time, especially when aversive stimuli are used.
“Therefore, it appears that the ability to learn depends on stable individual characteristics. The learning ability seems to be related to the horse breed rather than to age or sex.”
They noted earlier research which showed that individual task-solving skills by horses correlated negatively with the interest shown by the horses towards humans. “To the best of our knowledge, the influence of the horse’s learning ability on the course and effectiveness of natural training is still unknown.”
Kędzierski and his colleagues noted that none of the horses showed extreme negative emotions which prevented them from continuing the tests.
During the tests, of 120 primary tested horses, 15 did not open the crib, and of the 105 remaining horses, 22 did not manage to be mounted.
In the saddling test, all the horses were successfully saddled. It took the Anglo-arabian foals more time to be saddled than the Thoroughbred and purebred Arabian horses.
“Taking into account the sex within a breed, only the colts and Anglo-arabian fillies differed significantly: the fillies were quicker. When compared to all the other groups, it took the Anglo-arabian colts longer to be saddled.
“The percentage of horses successfully mounted was the highest for the Thoroughbred,” they reported. “The purebred Arabians showed a lower percentage and the Anglo-arabians had the lowest percentage.”
There were no significant differences in the groups regarding the time it took to be mounted, they reported.
Discussing their results, the scientists said the influence of the sex and breed on learning ability had been described in previous studies. “In our study, however, the high individual variability in crib opening test times, shows that the learning ability is, to a certain extent, an individual feature.”
The breed and, partly, the sex of the horse only affected the time it took to successfully perform the saddling test, they noted, with the Anglo-arabian colts needing a very long time to be saddle-trained – yet they achieved the shortest times in the crib opening test.
However, when comparing how long it took to be successfully mounted, the horse groups did not show any breed or sex-related differences, they said.
They noted that the objectives of the three tests performed in this study were different.
The first test, involving the box, involved a prize of food, whereas in the saddling and mounting tests the horse had to control its flight reaction in response to novel stimuli.
“The response to reinforcement depends on the horse’s temperament. Some horses learn better with negative reinforcement, others with positive reinforcement.
“This may explain the opposite relationships found between the results of the crib opening test and training tests.
“The cause of the different reactions may be the horse’s limited capacity to generalise during habituation.”
The findings suggested that horses of higher learning ability were not as willing to be mounted, whereas horses of lower learning ability cooperated better with the trainer and accepted more easily the presence of a human on their back.
This phenomenon confirmed the observation described by Clémence Lesimple and her colleagues in earlier research indicating that horses of low learning skills showed more of a dependency on humans to complete a task.
“In light of the present results, the Anglo-arabian horses differ from other breeds as they are more independent and not submissive during training.”
They concluded: “The response to the training is affected by the horse breed and sex. Therefore, the relationship between the learning skill and response to the training should be analysed separately in particular horse breeds and sexes.
“The physiological reactions and behaviour of horses are also affected by the type of reinforcement: positive or negative.”
The study was supported by the Polish Committee for Scientific Research.
Izabela Wilka, Witold Kędzierskib, Anna Stachurskaa and Iwona Janczareka.
Are results of Crib Opening Test connected with efficacy of training horses in a round-pen?
The abstract can be read here.