Horses: learning and training

.

Visit http://www.soloequestrian.com/ to read about Solo Harness, the loading innovation

.

Learning

Learning is defined as ‘changing behaviour as a result of experience’. The purpose of learning is survival in a changing environment.

Habituation: loss of sensitivity to unimportant stimuli
Habituation stops animals from wasting time or energy reacting to the stimulus e.g. when you first enter a room with a ticking clock you hear every tick, but after a while your brain filters it out so that is no longer heard. Habituation is probably the simplest form of learning.

Classical conditioning: many animals can learn to associate one stimulus with another.
Pavlov, a famous early physiologist, fed meat to dogs, causing them to salivate. Just before the feeding he exposed the dogs to a ringing bell. After a while, the dogs salivated readily in response to the sound alone, which they had learned to associate with the normal stimulus.

Operant conditioning: also called trial and error learning. The animal learns to associate one of its own behaviours with a reward or punishment; the animal then tends to repeat or avoid that behaviour.
Operant conditioning is the basis of most of the animal training done by humans, in which the trainer typically induces a particular behaviour first by rewarding the animal.


Training

Much of the early training of young horses involves habituation – getting the horse used to things, for instance cars.
More advanced training typically follows operant conditioning through negative reinforcement. A correct response to an unpleasant stimulus (relaxing the jaw and backing up in response to pressure from the bit) is rewarded by removal of the stimulus (loosening the reins).
A great deal can be accomplished in training simply by ignoring incorrect responses. A response to a stimulus that is neither rewarded nor punished tends to disappear of its own accord – this is known as extinction.

It is easier to teach a horse to respond to a command when the desired behaviour comes naturally and instinctively – animals may show a ‘biological preparedness’ for learning certain things. At the famous Spanish Riding School of Vienna, stallions are trained to execute airs above the ground. The particular air for which a stallion is trained is the result of observing the young horses at play to see which movements come most naturally to each.

Reinforcement
Appetitive reinforcers increase the likelihood of a behaviour being performed - they are forms of reward. Appetitive reinforcers may act through either positive or negative reinforcement.
Aversive reinforcers, or punishments, have the opposite effect reducing the likelihood of a behaviour being performed.

Note that negative reinforcement is NOT the same as punishment – negative reinforcement comes BEFORE the desired behaviour, punishment comes AFTER it.

Rewards MUST be provided within about half a second of the horse performing the requested behaviour, in order for it to associate the reward with the behaviour.
Consistency of reward is also very important.

Effects of punishment
When punishment is used in training systems, animals may become neurotic.
In other words, they learn faster but are more worried about getting the right answer.

Punishment perceived as reward
Rewards and punishments are defined by their effect not by their intended action. Not uncommonly a handler may end up rewarding a behaviour that they are intending to punish e.g. a horse that kicks the door of it’s stable is often ‘punished’ by being shouted at or having a hand flicked at it. To the horse, however, these bits of attention are more likely to be seen as a reward – if the horse is bored, any attention is better than none.

Lack of focus
Punishment does not help to tell the horse what it should be doing. There are many more ways to get something wrong than to get it right, so one undesired behaviour may be exchanged for another if the appropriate behaviour is not specifically encouraged by reward. There is also the risk of discouraging a behaviour that in later training will be required.

Depending on the method employed, punishment is also potentially an abuse of the animal.

Possible undesirable outcomes of punishment are:
- aversion: the horse becomes fearful of whatever it associates the punishment with – the trainer, the place in which it is trained or some other aspect of the situation.
- habituation: the horse learns to ignore the punishment
- aggression/ learned helplessness: the horse becomes aggressive towards the trainer, or, more damagingly, it ‘gives up’, learning that whatever it does results in punishment so it may as well not try to improve the situation.
- increased timidity: the horse becomes less willing to try anything out of the normal routine.

General guidelines to training new behaviours
Seek opportunities for reward, as these direct the animal’s behaviour towards a precise training goal
Don’t overuse a reward - the horse may become habituated to it and stop responding to it.
Choose an appropriate schedule for the training programme.
Be selective about which behaviours are to be rewarded. This helps you to apply rewards appropriately.
The choice of reward should be as relevant as possible in order to take advantage of the biological biases within the horse. A break from work is often a very good reward.
Be careful of timing of actions - rewards must be given as quickly as possible after the desired behaviour has been performed otherwise the reward may be associates with something totally unrelated to the behaviour.
If an animal is punished when is stops misbehaving, it is being punished for stopping the behaviour NOT for performing the unwanted behaviour, and next time is likely to carry on with the behaviour for longer.
Secondary rewards, including verbal praise and retort, must be classically conditioned to an appetitive or aversive reinforcer before they have any effect on the behaviour. This means that they must be first associated with something the horse likes or doesn’t like, for instance a food snack or a loud noise, before they can have a good effect on their own. This type of classical conditioning is the basis for clicker training – the clicker provides a very recognisable and consistent sound that can be associated with a pleasurable experience.

The vast majority of training can be carried out by rewarding the good and ignoring the bad – done in this way, training is enjoyable for both the horse and the human.

Visit www.soloequestrian.com to read about Solo Harness, the loading innovation

Vice - or self-help?

Visit www.soloequestrian.com to read about Solo Harness, the loading innovation

Stress is defined as 'mental, emotional or physical strain or tension'.
In response to chronic (long-term) stress, the body secretes a family of steroid hormones called corticosteroids. These have undesirable effects on the tissues that are responsible for producing white blood cells, which make up part of the body’s defence system. The number of white blood cells produced is reduced, resulting in an increased susceptibility to infection and disease. This is why people who are stressed are more likely to suffer illnesses such as irritable bowel syndrome, as well as more minor ailments such as colds. Similarly, horses which are subject to excessive stress are likely to develop stomach ulcers and/or be generally unthrifty.

Antidotes to stress

Endorphins are the brains own opiates - they are similar in structure to chemicals such as morphine and heroin. Endorphins are released in response to stress, and appear to have similar calming effects to those of opioid drugs
Certain activities appear to boost or prolong release of endorphins from the brain i.e. they can give an opiate 'hit'. Some repeated, relatively invariant sequences of movements which outwardly appear to have no purpose seem to be both stimulated by the release of endorphins and also prolong that release. These types of movements are known as 'stereotypies'. They are a means by which animals can give themselves the equivalent of a Valium.

Examples of stereotypies in horses are actions such as weaving, box-walking, cribbing and windsucking. Most people are more familiar with the term 'vices' to describe these types of behaviour, however this word is misleading in that it implies that horses are making moral decisions to act in an evil manner. In reality, nothing could be further from the truth - the behaviours are simply the horses response to whatever stressful situation it has been placed in.
Horses are by no means the only animals which exhibit stereotypies in stressful situations - other examples of stereotypies include rocking movements and repetitive vocalisations in institutionalised humans, bar chewing in stalled sows, head swinging in zoo-housed bears and elephants, pacing in captive fennec foxes, eye rolling in veal calves and jumping in caged bank voles.

Why do some horses develop stereotypies while others do not?
There is a strong belief among people who own or care for horses that normal horses will copy the stereotypies of afflicted stable-mates. It is much more likely that a horse who appears to imitate a stereotypy was experiencing the same stressors as the animal it appears to have copied, and so reacted in a similar way. Horses in a yard are likely to be under similar management routines and therefore subject to the same stressors. Equally, a horse which already has a well-developed stereotypy will pose no threat if it is introduced to a yard with a horse-friendly management system.

There is also no evidence to show that stereotypies are inherited. ALL behaviours are the result of the effects of the genes of an individual combined with the effects of their environment, which act in different proportions in different situations. Thus there is probably a small amount of genetic predisposition for horses to perform stereotypies - this would explain why certain horses develop stereotypies though other horses kept in the same environment and managed identically do not. This can be seen in humans too - some individuals are better able to cope with stress than others.

What are the stressors which lead to the development of stereotypies?
Horses have spent the past six million years evolving to suit a particular lifestyle. They are built to spend around 16 hours every day moving around slowly while chewing on fibrous vegitation, and to live in family groups in large, open spaces.
Any factors which remove the horse from its natural environment will cause it stress. These include spending large amounts of time in the stable, not having access to ad lib forage and being deprived of companionship.
Conversely, allowing horses to perform as much of their evolved behavioural repertoire as possible will reduce the risk of the development of stereotypies and may also reduce the amount they are performed in horses which already have established stereotypies.

Development of stereotypies.
Stereotypies tend to develop from the main frustration that the animal faces, for instance a horse which is deprived of the behaviour it would perform in the wild to find food is likely to develop some sort of oral stereotypy such as windsucking.
Feeding-related stereotypies are further explained by evidence which shows that grain-based (‘high sugar’) feeds actually stimulate endorphin release, resulting in enhancement of stereotypic behaviour patterns. Decreasing the amount of grain fed to stereotypic horses, therefore, may reduce the rate of stereotypy performance. A diet based on high quality hay rations may reduce stereotypy directly by minimising endorphin release, as well as indirectly through leading to more time being spent chewing and eating.
Recent evidence suggests a further cause of windsucking and cribbing stereotypies. These types of stereotypy are peculiar to horses - many other animals show weaving or pacing behaviour but no others are known to perform anything resembling cribbing or windsucking. This could be connected with the horses unusual digestive system, which is designed to cope with large amounts of fibrous material and huge volumes of digestive secretions. Horses which are fed mainly on small volumes of concentrate feed may have problems digesting the high levels of sugar which are present in these feeds, which can lead to the digestive tract becoming acidic. In these horses it is thought that cribbing and windsucking may be a means by which they can increase the levels of secretion of digestive fluids and so help dilute the acidity of the lower digestive tract.

Once fully developed, stereotypies will usually be performed in any stressful situation, i.e. they become independent of their original eliciting stimulus. Fully developed stereotypies may not show the current state of the horse, but reflect a past challenge severe enough to have traumatic effects on the central nervous system. This is why horses which have developed a stereotypy will often carry on performing it even when their environment has been improved.

So if stereotypies are actually beneficial to animals, why do we want to treat them?
The moral argument that stereotypies are simply a symptom of stress, and subjecting animals to stress is ethically wrong comes top of the list. Successful treatment of stereotypy involves the removal of the stressor which caused the behaviour in the first place. A secondary factor is that some stereotypies can be damaging to the animal in themselves, for instance cribbing results in damage to the incisor teeth; weaving can result in concussion injuries to the front limbs.
As already discussed, treatment involves returning the animal to as natural an environment as possible i.e. with access to ad lib forage, space and companionship. In cases of weaving and box-walking this alone should be enough to eliminate stereotypy.
In cases of cribbing and windsucking, however, the stereotypy tends to be more persistent, the animal is in effect addicted to its Valium. It is possible that these animals are suffering from chronic low-grade pain from which the stereotypic behaviour may provide relief. This area is currently receiving research interest.

In conclusion, it is evident that stereotypies are much more complex than their common name, vices, would suggest. Stereotypic behaviour gives us an indication that all is not well with particular individuals, and lead us to consider horses which are subject to many stressors but which do not develop stereotypies - are they coping better with the situation or are they simply not coping at all?

Visit www.soloequestrian.com to read about Solo Harness, the loading innovation

Horse coat colour genetics

Visit www.soloequestrian.com to read about Solo Harness, the loading innovation

Horses coat colours are determined by information stored in their DNA. DNA is usually a tangled mass of strands that occupies the nucleus of cells. When cells are about to divide, however, the DNA is tidied up into bundles known as chromosomes. Horses DNA arranges itself as 64 chromosomes (humans = 46). These chromosomes are paired, with each chromosome of the pair carrying the same pattern of genes.
Genes are sections of DNA that affect characteristics of the animal. The pairing of chromosomes means that there are two copies of each gene in animal cells. These two copies may be the same, or they may differ. The different forms that a gene can take are called alleles.
Alleles are usually either dominant or recessive. The dominant allele will be expressed (i.e. will determine the characteristic) when it is paired with another identical dominant allele, and also when it is paired with a different, recessive, allele. Dominant alleles are usually symbolised using a capital letter. Recessive alleles will only be expressed if they are present as two identical copies. Recessive alleles are given small letters.
Egg cells and sperm cells contain half of the chromosomes of the other body cells, one chromosome from each pair. Animals inherit half their chromosomes (and therefore half their genes) from their mother, and the other half from their father. When the maternal and paternal chromosomes come together as one cell, that cell (the embryo) will carry one allele from its mother and one allele from its father for each gene. This results in combinations of alleles occurring in the offspring that may not occur in either the mother or the father individually.


In horses, there are a number of genes that influence coat colour. Each of these genes has two or more different alleles (forms). All horses carry all the coat colour genes, however there is a hierarchy of expression, with some genes masking the effects of others in some animals. The different genes act together to produce the final, visible, coat colour of the horse.

The first gene in the hierarchy of coat colour is the albino gene, ‘W’.
If a horse has just one copy of the dominant W, it will mask to all other coat colours and the horse will be albino (pink eyes and skin, white hair). The horse will carry genes for a variety of other coat colours, but these do not show through in its actual physical colour.
Albinism occurs in many animal species, including humans, and it causes problems in most. Albinos are very sensitive to sunlight and may have poorer vision than their peers. This alone makes albino animals relatively rare – they are not purposely bred very often, however albino horses are extremely rare because of a twist in the equine albino genes tale – horses that inherit two copies of W die just before or just after birth – ‘WW’ is lethal. For a WW horse to be produced, both its parents would have to be albino, so thankfully this situation is usually avoidable.
If a horse is ‘ww’, it will not be albino. Most horses are ww.

The next gene in the hierarchy is the grey gene, ‘G’. Grey, G, is dominant to non-grey, g. Horses that carry the G allele will be grey. Horses that are gg (no pun intended) will be any colour except grey. Horses that carry two copies of G will always breed grey foals. If a horse carries only one copy of G (i.e. it is Gg), each foal will have a 50% chance of being grey (rising to 75% if its other parent is also Gg). A tool called a Punnett Square is a useful way to visualise this:

(N.B Apologies for all the stupid dots, the site keeps trying to rearrange things and get rid of all the spaces, obviously thinks the letters are lonely which is a bit frustrating......)

Parents GG (grey) and gg (non-grey

.................G......... G

g .............Gg ........Gg
..............grey ........grey

g............. Gg......... Gg
...............grey ........grey

All foals have grey coat colour.


Parents Gg (grey) and gg (non-grey)

..............G.......... g

g ..........Gg .........Gg
...........grey ..........grey

g.......... Gg.......... gg
............grey .....non-grey

There is a 50% chance that each foal will be grey, and a 50% chance it will be non-grey.

Parents Gg (grey) and Gg (grey)

.................G ..............g

G............ Gg............ Gg
...............grey .........grey

g ..............Gg ...........gg
................grey ......non-grey

There is a 75% chance that each foal will be grey, and a 25% chance it will be non-grey.

All genetically grey horses begin life with a coloured base coat that turns grey when their foal coat is shed. The coat colour comes in a very dark steel grey which always fades as the horse ages.

The next layer is a little more complex than the first two. This is the bay/black/chestnut layer – the last of the possible base coat colours. These three coat colours result from the combined action of two different genes, extension, ‘E’ and agouti, ‘A’.
The extension gene controls the pigment colour in hairs. The dominant E gives black pigment in the hair; the recessive e gives red pigment.
The agouti gene controls the positioning of black hair around the body. The dominant A restricts black to the points (nose, ears, legs, mane and tail), the recessive allele a gives an even coverage of black all over the horse.

Chestnut horses are all ‘ee’, and the A gene has no effect on their colour. Agouti will ONLY affect black hair, and chestnuts have no black hair to be restricted. Chestnut is the only guaranteed true-breeding colour – if you cross two chestnut horses, you will get a chestnut foal (unless any gene mutation occurs, but that’s another story!).
Black horses must have at least one copy of the ‘E’ allele along with ‘aa’ which means that the black is evenly distributed all over the body.
Bay horses must have at least one copy of the ‘E’ allele along with at least one copy of the ‘A’ allele which restricts the black hair to the points.

Try the Punnett Square for yourself and see what the possible outcomes of mating two bay horses might be!
The first box in the Punnett Square is filled in for you.
Remember
- all chestnuts are ee, with any combination of agouti alleles
- all black horses must have at least one E, but are aa
- all horses that have at least one E along with at least one A are bay.

Can you see why bay horses occur so regularly?

Mating two bay horses that are ‘AaEe’

Parents ....... .........AaEe x AaEe

Sperm/Egg ...AE/ Ae/aE/ ae and AE/Ae/aE/ae

Offspring

............................AE ...................Ae................... aE .....................ae

AE ....................AAEE............... AAEe............................................
............................bay........................................................................
Ae ....................................................................................................
...........................................................................................................
aE ....................................................................................................
.........................................................................................................
ae......................................................................................................
..........................................................................................................

These are the five basic coat colours - albino, grey, bay, black or chestnut.
All other colours are the result of the actions of other genes on these basic colours. These genes are divided into ‘dilution’ genes and ‘pattern’ genes.

Dilution genes

The cream gene, ‘C’ has an effect on bay and chestnut horses.
It shows INCOMPLETE DOMINANCE i.e. unlike the genes we have looked at so far, horses that are Cc are different to those that are CC.
Horses that carry the cc alleles show no dilution – they remain bay or chestnut.
Chestnut horses with one copy of the C allele are diluted to palomino. Chestnut horses with two copies of the C allele are diluted to cremello (pale horse with blue eyes).
Bay horses with one copy of the C allele are diluted to buckskin (although in the UK this is usually (wrongly) called yellow dun – we will get to duns shortly).
Bay horses with two copies of the C allele are diluted to perlino (also pale with blue eyes – difficult to tell the difference between this and a cremello without knowing parentage).
The effect of incomplete dominance is that neither palominos nor buckskin horses breed true. Crossing two palominos (we will stick to palomino for simplicity – buckskins follow exactly the same pattern of inheritance) will give only a 50% chance of producing a palomino foal, with a 25% chance of a chestnut being born and a 25% chance of producing a cremello. The only way to guarantee production of a palomino foal is to cross a chestnut and a cremello.
Punnett Squares should help to visualise this:

Palomino x Palomino

...................C ..............c

C ...............CC ............Cc
..............cremello.... palomino

c ................Cc............. cc
..............palomino... chestnut

There is a 50% chance that each foal bred will be palomino, a 25% chance it will be chestnut and a further 25% chance that it will be chestnut.

Chestnut x Palomino

........................c.................... c

C ...................Cc.................. Cc
.................palomino .........palomino

c ....................cc ...................cc
................chestnut............. chestnut

There is a 50% chance that each foal bred will be palomino, and an equal chance that it will be chestnut.

Cremello x Chestnut

..........................C......................C

c....................... Cc ...................Cc
....................palomino .........palomino

c .......................Cc..................... Cc
...................palomino.......... palomino

All foals will be palomino, bearing in mind that neither of the parents were this colour.

The dun, ‘D’, gene has an effect on all colours. The D gene dilutes body hair but not hair at the points, giving the dun effect e.g. mouse dun is diluted black, yellow dun is diluted bay, red dun is diluted chestnut. Horses diluted are a result of dun, rather than cream, have darker mane, tail, nose, ears and legs (which may have zebra stripings), and also have a dark ‘eelstripe’ along their backbone. Buckskins and yellow duns appear similar, however the eelstripe should distinguish between them. D is a dominant gene.

Pattern genes
The ‘TO’ gene produces tobiano patterning. In the UK this would be referred to as piebald or skewbald, but these terms don’t reflect the genetics of the colour so we will stick to Americanisms. A horse is tobiano if it has white across its backbone. Tobiano is a dominant gene. As with other dominant genes, the TO gene will be expressed in horses that carry TOTO and also in those that are TOto. If a horse carries two identical alleles for one gene (e.g. TOTO), it is termed ‘homozygous’ for that gene. If the alleles are different (e.g. TOto), the horse is termed ‘heterozygous’ for that gene. A test for homozygosity has now become available, due to the increased value that can be commanded by a homozygous TO animal. Previously, and still for many, the only way to distinguish between homozygous and heterozygous was by looking at the offspring of the horse in question. If a coloured stallion has produced a large number of coloured offspring there will be a high probability that he is TOTO, and therefore will always breed coloured horses. If, however, a coloured stallion has produced even one solid coloured foal, he must be heterozygous (TOto) and therefore is not guaranteed to produce a coloured foal even when bred to a coloured mare. This is why breeders advertise their coloured stallions as ‘homozygous’ – they are guaranteeing a coloured foal from a mating with their stallion. Again, Punnett Squares can help to clarify this:

Homozygous tobiano x solid horse

...............................TO ...................TO

to ..........................TOto ................TOto
..............................tobiano............ tobiano

to ..........................TOto................. TOto
...............................tobiano........... tobiano

All foals will be tobiano.

Heterozygous tobiano x solid horse

..............................TO............................. to

to ..........................TOto.......................... toto
..............................tobiano ......................solid

to ..........................TOto........................... toto
.............................tobiano......................... solid

Each pregnancy has a 50% chance of producing a tobiano (coloured) foal and a 50% chance of producing a solid foal. None of the coloured foals produced will breed true for the coat pattern – they will all be heterozygous for TO.

Homozygous tobiano x Heterozygous Tobiano

...............................TO .......................TO

TO ........................TOTO................. TOTO
..............................tobiano ...............tobiano

to ...........................TOto ...................TOto
..............................tobiano................. tobiano

All foals will be tobiano, however only there is only a 50% chance that these foals will breed true for the coat colour.

Heterozygous tobiano x Heterozygous tobiano

...................................TO .......................to

TO ..........................TOTO ..................TOto
.................................tobiano................ tobiano

to ..............................TOto..................... toto
................................tobiano ...................solid

At each breeding, there is a 75% chance that the foal produced will be coloured and a 25% chance that it will be solid, even though this cross is of two coloured horses.


Whether a horse is piebald (white on black) or skewbald (white on brown/ chestnut) will be determined by the base coat colour genes. It is possible to have a tobiano horse that appears grey – in this case the ‘tobiano areas’ will be pink skin, with the rest of the skin being grey. This may only become obvious when the horse is soaking wet!

The ‘O’ gene produces overo patterning. A horse is overo if its white areas do not cross its backbone – it looks as if it has been dipped in paint rather than splashed with it. Overo is also a dominant gene, however it follows the same pattern as albino in that foetuses that inherit two copies of O will die in utero and be reabsorbed by the mother.
The most common breed in the UK to show overo colouring is the Clydesdale.

Tobiano and overo patterning are more common in America than in the UK. Combinations of the two genes may occur, the result being an almost fully white horse with small areas of base colour. These can be as extreme as the ‘medicine hat paint’ – a horse that is white with only the ears and a small patch around the poll area showing the base colour!

The ‘LP’ gene produces leopard pattern, or the appaloosa pattern. It is a dominant gene that shows many different variations.
The ‘RN’ gene produces a mix of white and coloured hair i.e. roan. It is also dominant and also lethal if present in two copies (although there is recent evidence for viable homozygous roan animals, in quarter horses at least). Basic black coat colour with the roan gene produces blue roan, chestnut produces strawberry roan.

Bizarre and wonderful combinations of these base colours and pattern genes are possible – blue roan appaloosa tobiano, red dun overo, palomino medicine hat.

White markings on the face and legs are controlled by a separate set of genes that show a different kind of inheritance to those discussed so far. They can be though of as ‘additive’. If you breed from two horses with very small or no white on their bodies, you will get a foal with virtually no white. Conversely, if you breed from two horses with white faces and a lot of white on their legs, your foal will inherit similar. If once parent is minimally marked and the other has a lot of white, the foal will be somewhere in between. This is termed polygenic inheritance – genes that govern things like height and weight also follow this type of pattern. Most characteristics of animals are controlled by polygenic traits, and are also influenced heavily by the environment. Coat colour is one of the few visible traits on which the environment has little effect – although it’s rare to be certain of the outcome of any breeding project, at least coat colour can be predicted with a reasonable amount of certainty!

Bibliography/ Further Reading

Bowling, A (1996) Horse Genetics, CAB International, Oxon
http://www.equinecolor.com/unusual.html

Visit www.soloequestrian.com to read about Solo Harness, the loading innovation