Artificial insemination of equines has been around a long time and, through the years, has had a profound impact on the horse industry. Most of the impact has been of the positive variety, but there also have been some negatives. Just when artificial insemination (AI) first made its appearance on the equine scene is open to debate. There is the somewhat romantic, and perhaps apocryphal, story of an Arab chieftain who allegedly stole some semen from a prize stallion owned by an adversary. The chieftain then inseminated his prize mare with this semen to obtain a "super" foal. That incident, the story goes, took place in 1322.

Whether that story is fact or fiction is irrelevant, but it does underline the fact that artificial insemination has been around for a good many years. B.W. Pickett, PhD, of Colorado State University, one of the pioneers in the field of research who helped turn the debatable AI approach for achieving pregnancy into a commonplace procedure, tells us that it was in 1776 that a researcher named Spallanzani used stallion semen to observe the effects of cooling spermatozoa. Spallanzani, says Pickett, is the "recognized father of AI."

With that much of a head start, one would think that AI in the equine industry would be the most sophisticated among the species of animals where it is employed. Unfortunately, this isn’t true. The beef industry, for example, routinely uses frozen semen successfully, while the equine industry still struggles to find a way to achieve a satisfactory pregnancy rate with that approach. (One country that has made heavy use of AI for a number of years, says Pickett, is China. There, in 1959, approximately 600,000 mares were inseminated.)

Perhaps the prime reason for the disparity in progress is the role the horse has played in agriculture as compared to beef. While a growing population demanded more and better food on the table at the turn of the century, stimulating research into beef production, horses were on their way out as a utilitarian animal.

Tractors and trucks replaced horses and mules from rural fields to city streets. The number of horses in the United States dropped drastically. After a time, the draft horse, the mainstay of farming in the countryside and dray work in the city, had entered the "rare breed" category.

Then came a switch in attitudes. The horse was newly discovered as a recreational entity. People found that trail riding could be great fun. Still others discovered an exciting world of competition involving horses of all breeds and abilities, and soon horse shows proliferated.

Racing had been around since the country was founded, but it, too, received added impetus. The number of equines involved in racing increased as people with more money and an increasing amount of leisure time found their way to the racetrack.

Horses were on their way back, but in different trappings. It wasn’t long before leaders of the industry decided it was time to play catch-up in the world of research, especially as it related to artificial insemination. Some universities developed ambitious and aggressive research programs to determine the best approach to take in obtaining pregnancies via AI, then added another dimension with embryo transfer.

The research continues today, with the latest breakthrough involving the use of the endoscope to place sperm at the uterotubal junction. More about that later.

The use of AI mushroomed as more knowledge and better techniques became available. Today, the only major equine registry still holding out against the use of AI is The Jockey Club, which registers Thoroughbreds.

It wasn’t always that broadly accepted, says Pickett, who presented many of his views on the subject in a recent article titled "Physiology and Philosophy of Breeding Horses" that appeared in the Journal of Veterinary Science.

"For years," Pickett tells us, "many prominent horse breeders were vehemently opposed to utilization of AI within their breeds. Much opposition was out of fear and ignorance. Many believed that a) foals would be weakened and malformed; b) the procedure was against nature; c) mares would become sterile; d) offspring would be degenerate; e) sex ratio would be altered; f) there would be semen mix-ups; g) diseases would be spread; h) inbreeding would occur, and many other erroneous wives’ tales.

"Actually, none of these things has occurred with any more frequency than observed with natural service. In fact, AI often provides a better solution to problems expressed by these fears than natural service."

The primary objective during the breeding season, Pickett continues, is to impregnate the maximum number of mares in the minimum amount of time.

"This reduces labor costs and other expenses," he notes, "and provides earlier foals, thus shortening the foaling season, again reducing labor and expenses. Artificial insemination can aid in accomplishing these goals."

When discussing the impact that AI has had on the equine industry, one must first understand the advantages it provides as well as the disadvantages.

Pickett outlines a number of advantages for using AI. They can be distilled to 10 key points:

1. Permits disease control. Artificial insemination is used effectively to control venereal disease. This can be accomplished in two ways. First, the amount of semen placed in the mare’s uterus can be reduced sufficiently so that the number of potentially pathogenic organisms is reduced below the number necessary to cause disease. Second, an extender, which dilutes the semen, can be used that contains antibiotics capable of killing the organisms causing venereal disease.

"For example," Pickett says, "the utilization of AI, with a seminal extender containing gentamicin sulfate, probably would have been effective in controlling Taylorella equigenitalis, the causal agent of contagious equine metritis (CEM) in 1977."

2. Reduces the possibility of injury to the mare and/or stallion. Some stallions and some mares react violently during the breeding process. The result can be a stallion which is injured by being kicked or a mare which is savaged by a stallion becoming overly aggressive. Also placed at risk with this type of behavior are the handlers of the stallion and mare.

With AI, a phantom mount or a gentle mare which is acceptable to the stallion can be used for the collection of semen. Once a stallion has been taught to use the mount and ejaculate into an artificial vagina, the risk of injury diminishes significantly.

Using AI also reduces the problems involved when a small stallion is mated to a tall mare or vice versa.

3. Permits use of stallions which have developed poor breeding habits or have been injured. "Stallions have been trained to ejaculate into an artificial vagina (AV) without mounting a mare," Pickett says. "A stallion that cannot be used in natural service because of injury or chronic lameness, etc., can often be trained to ejaculate in this manner and be used in an AI program. In at least two instances in which a stallion was unable to obtain an erection, but showed excellent sex drive, training to an AV was accomplished. Stallions that have been mismanaged, kicked, or otherwise injured during the breeding process occasionally develop poor breeding habits. It is common for these horses to dismount quickly while in the process of ejaculating."

When that happens, of course, much of the ejaculate is lost on the ground. About the only way to circumvent those problems via natural cover is to catch the dismount sample and inseminate with it. However, it quickly becomes apparent that the quality of the semen would be compromised under that scenario.

4. Permits evaluation of semen at each collection and immediate recognition of minor changes in seminal quality. "Formation and maturation of spermatozoa require about 57 days in the stallion," says Pickett. "When AI is used, semen can be evaluated at each collection. In conjunction with an effective herd health program, this will aid in identification of causes of reproductive failure that may otherwise go undetected. When deterioration of seminal quality is observed, steps can be initiated immediately to correct the problem, or mares can be switched to another stallion, preventing disastrous reproductive performance and loss of reputation of the stallion and farm."

5. Aid in identification of reproductive problems. AI can help pinpoint reproductive problems whether they originate with the stallion or with the mare.

6. Prevents overuse of a stallion, particularly early in the breeding season, and enables mares to be bred at the most opportune time for maximum chance of reproduction. Overall, this is perhaps one of the key advantages of AI and an area in which it has had great impact. Some popular stallions are swamped with mares early in the breeding season. In the past, some of them would have been turned down or been bred at a later date. Now, with the use of semen extenders, several mares can be bred with each ejaculate, with the actual number being determined by the stallion’s fertility.

AI is of particular value if a stallion is sub-fertile. Such a horse, when breeding natural cover, might be able to impregnate one mare daily at most, and often fewer. However, if that one ejaculate is collected and extended, the number can be increased.

"When several mares need to be bred on a given day, AI can prevent overuse of the stallion and still enable mares to be bred at the proper time with the optimal number spermatozoa," Pickett said. "Cattle semen can be collected, extended, cooled to 5� Centigrade, and used to breed cows for up to five days without an appreciable loss of fertility. However, too little information is available to recommend this procedure with stallion semen. Our current recommendation for maximum reproductive capacity is to collect, extend, and inseminate fresh semen immediately.

"When numerous mares are booked, AI is the only realistic solution to breeding these mares at the appropriate time for maximum reproductive efficiency. The operative phrase here is ‘reproductive efficiency,’ with special emphasis on maximum, because there are some stallions which can cover effectively 200 or more mares naturally during one breeding season.

"Our current recommendation is to collect the stallion every other day and inseminate all mares that have been in estrus two days or longer, or when it has been determined by ultrasonography and/or rectal palpation that ovulation is imminent."

7. Permits breeding of abnormal mares which could not be used in natural service. Fitting into this category are mares which have been injured or have other disabilities. Also fitting into this category would be the mare which shows no outward signs of estrus at any stage of her cycle. Such a mare might react with violence to the stallion if he attempted to mount. Using palpation and ultrasound to determine when she ovulates allows a mare to be impregnated via AI even though she does not demonstrate the classic signs of being in heat.

8. Permits more effective use of older, more valuable stallions. "As stallions age," Pickett tells us, "degenerative changes occur in the testes that reduce the number of spermatozoa produced per gram of testicular tissue. Also, in many cases the spermatozoa produced by older stallions have a great number of abnormalities and poorer viability, which can further reduce fertility. With a properly spaced seminal-collection schedule, reliable seminal-evaluation procedures, and properly timed inseminations, more pregnancies can be obtained with an older stallion by AI than by natural service. In many cases, this is the only method of obtaining foals from an old, valuable stallion."

9. Increases the genetic pool and permits progeny testing. Breeding horizons have definitely been broadened by the advent of AI. In the past, if a breeder in New York wanted his or her mare bred to a stallion in California, it involved a great deal of planning and forethought. First of all, the trip would be a long one for the mare. Secondly, there would be a good deal of transportation expense involved. Thirdly, the mare would be out of the owner’s care and control for an extended period.

With AI, those constraints have been removed. The mare can remain in her home stable under normal routine while semen is collected from a stallion thousands of miles away and flown to her area.

The above is especially true of mares with foals at side. Added to the cost and inconvenience in shipping a mare and foal cross-country is the threat of injury and/or disease to the foal.

Progeny testing with AI also is a valuable asset. Pickett puts it this way: "There has been very limited progeny testing in horses. There are several conditions in horses assumed to be hereditary, but definite proof is not available. The best method of truly determining the mode of heritability is through AI."

10. Results in higher pregnancy rates than natural service. "Horses have the lowest reproductive rate of all species of farm animals," Pickett maintains. "The major causes of this subfertility appear to be genital tract infections, hormonal dysfunction, the arbitrary January birth date, inadequate management, and lack of selection for fertility."

One of the early indicators that AI was more efficient than natural service in getting mares pregnant came in a study conducted from 1967 to 1970 involving five Quarter Horse and seven Thoroughbred farms. The majority of the Quarter Horse mares were bred with AI while all of the Thoroughbred mares were bred via natural cover.

Of the Quarter Horse mares involved in the study, 51% became pregnant on the first cover compared to 43% for the Thoroughbred mares. The percentage for pregnancies after five services was 84% for the Quarter Horses and 75% for the Thoroughbreds.

The same trend was observed for maiden and barren mares and mares bred on foal heat. With Quarter Horses, 89% of maiden mares became pregnant after five services, compared to 76% for Thoroughbreds. With barren mares, 46% of the Quarter Horses became pregnant on the first service, compared to 34% for the Thoroughbreds. After five services, the spread remained–82% pregnant for the Quarter Horses and 66% pregnant for the Thoroughbreds.

For mares bred on foal heat, 58% of the Quarter Horses became pregnant compared to 41% of the Thoroughbreds.

In commenting on the study, Pickett had this to say:

"We believe these differences were due to utilization of AI by Quarter Horse breeders, since 69% of the Quarter Horse mares were bred with AI compared to none of the Thoroughbred mares. We are not sufficiently naive to believe that AI is not used on some Thoroughbred farms. However, the farms in this study were large, well-known Thoroughbred farms where AI could not have been utilized without detection. Breed difference does not appear to be a logical explanation since many of these mares were 100% Thoroughbred lineage being bred to Quarter Horse stallions and vice versa. Further, many of the Quarter Horse mares had a high percentage of Thoroughbred blood."

Further proof of higher conception rates was provided in a study undertaken by the Arabian Horse Registry, which, in recent years, has been a leader in the utilization of artificial insemination and transported semen, both cooled and frozen. From 1972 to 1976, the Registry was involved in a developmental program involving artificial insemination. In 1977, the Registry formally approved AI. Several years ago, the Registry also approved the use of frozen semen.

As part of the Arabian Registry’s developmental program in the 1970s, data were compiled by a research committee concernng pregnancy rates at a Standardbred farm, comparing natural cover to AI.

During the period when data were collected and analyzed, some 1,466 Standardbred mares classified as being "normal" were bred on the farm. Of that number, 913 were bred AI and 553 were bred via natural cover. The conception rate for the AI bred mares was 83.1% while the conception rate for natural cover was 74.5%. In addition, the AI-bred mares had a live foal rate of 73.8% compared to 68.7% for natural cover.

During that timeframe, the Standardbred farm also bred 905 mares which were classified as being "problem mares." Of that number, 563 were bred AI and 342 were bred by natural service. The AI mares had a conception rate of 44.8% compared to 31.3% for those bred by natural cover. The AI mares had a live foal rate of 35.2% compared to 24.6% for natural service.

"It was determined that there were significant differences in conception rates and live-foal percentages between AI and natural service for both normal and problem mares," said Pickett of the study. "Thus, it is obvious that efficiency of reproduction was higher when AI was utilized compared to natural breeding. This was particularly true with ‘problem mares.’ For example, the percentage difference between AI and natural service for live foals from normal mares was 5.1% compared to 10.6% for problem mares.

"These differences support the contention that problem mares bred by AI are more likely to become pregnant and deliver foals than problem mares bred by natural service."

By perusing the above, one quickly can conclude that there are a number of advantages to using artificial insemination and transported semen, and that these advantages have, and will continue to have, a positive effect on the equine industry.

However, there also are some disadvantages to AI that should be articulated.


Foremost is the technology involved. While the technology is an advantage, its implementation by the small breeder would have to be considered a disadvantage. Pickett estimates that at least 20 mares a year must be bred at a particular farm in order to justify the use of artificial insemination.

In most breeds, farms that breed 20 mares per year are in the minority. The number is far fewer at most farms.

The largest registry in the horse world is the American Quarter Horse Association (AQHA). Debbie Black of the AQHA registration department put it this way: "The backbone of our breed is the small breeder–the person who has five or fewer Quarter Horses."

Also, the backbone of that and other breeds is the person who owns a stallion, but only breeds five or six outside mares per year. It would be difficult for that individual to justify the expenditure involved in purchasing the necessary equipment for an AI program.

The same is true for some breeders in the Rocky Mountain West where horses roam huge pastures. In many instances, owners of these individuals find that pasture breeding is the best way to go.

In the minds of some, the large number of mares that one stallion can breed in a given year by using AI could be construed as being a disadvantage because that narrows the pool of available mares for other stallions. It does not appear to be a valid concern. True, some stallions do breed to a larger book than they would with natural cover, but in checking with a number of registries that approve the use of transported semen, it does not appear that one or two stallions dominate.

For example, the American Quarter Horse Association records show that in 1997, a total of 176,024 mares were reported as being bred. The average number of mares bred per stallion was six. There were only 42 stallions that bred 100 mares or more. By far, the largest number was for stallions which bred 10 mares or fewer–25,336.

The AQHA has permitted the use of artificial insemination on the farm for many years. This means that stallions could be collected and the semen used to inseminate mares that were on the same premises. However, it has only been since 1997 that shipped cooled semen has been approved. The shipped semen must be used within 72 hours. As mentioned earlier, frozen semen has not been approved.

Black said that in 1999, some 30,838 stallions were reported as breeding mares. Of that number, certificates for shipping cooled semen were obtained for 1,100 of them.

Numbers were not available for mares bred on premises via artificial insemination.

While the figures indicate that no one stallion is breeding hundreds and hundreds of mares because of AI and shipped semen, the capability definitely is there.

Pickett provided some eye-opening information when turning his attention to the potential for breeding mares via artificial insemination.

"The first assumption that must be made," he says, "is that management is ideal and that there are sufficient facilities and personnel to perform the inseminations in a reasonable period of time (approximately one hour) after seminal collection. If a daily spermatozoal output (DSO) of 3.5 billion to seven billion is assumed, depending on season of the year, the number of mares that can be bred can be predicted."

He goes on to say that it is recommended that a stallion be collected every other day when the semen is going to be used for AI. This means that the spermatozoal output should be seven to 14 billion, double that of the daily collection.

Also figured into the equation is progressive motility of the spermatozoa. Progressive motility is a characteristic "most commonly used on the farm to predict fertility and estimate the number of mares to inseminate."

In attempting to arrive at numbers of mares which could be inseminated per ejaculate, Pickett and his colleagues used spermatozoal motilities ranging from 25% to 75% and spermatozoal outputs ranging between seven and 14 billion per ejaculate.

They then prepared a table showing what could be done under ideal conditions. If 500 million spermatozoa were used per insemination from an ejaculate of seven billion and the progressive motility was 25%, they reasoned, then three mares could be bred per ejaculate.

However, if there were 14 billion sperm per ejaculate and the progressive motility was 75%, then 105 mares could be bred with that ejaculate.

However, in a normal breeding program, it would be rare that a mare would be inseminated only once.

"In an AI program," Pickett says, "most mares are inseminated more than once for maximum reproductive efficiency. The highest pregnancy rate utilizing the largest number of mares has been obtained with three to four inseminations per cycle, which was without the benefit of ultrasonography.

"Assuming that each mare is inseminated three times, the numbers (in the above referenced table) should be divided by three; thus, the number of mares which can be inseminated with each collection is one to 35.

"Further, if one assumes 70 breeding days per season, then the number of mares that can be inseminated per season in an AI program, based upon these calculations, would range from 70 to 2,450. It is blatantly obvious that more mares can be inseminated, particularly under ideal management, than has been predicted or recommended in the past; particularly if ultrasound is used to maximum advantage and mares are bred only once per cycle.

"In our example, if mares were bred only once per season, the number of mares that could be inseminated would range from 210 to 7,350. It has been reported that in 1960 in China, the two most popular stallions were used to inseminate 4,415 and 3,095 mares for conception rates of 76.9% and 68.1%, respectively. It appeared that 100 million progressively motile spermatozoa were used per insemination.

"One need only compare what was considered a normal book in the early 1900s to what is normal today to recognize that there is a trend toward breeding more mares to popular stallions. Further, current predictions of how many mares can be bred utilizing natural breeding or AI have been grossly underestimated.

"Management is the key to success."

Endoscopic Guidance

While the numbers presented by Pickett are on the verge of being mind boggling to the average breeder, there is more to come. Among the latest in technological advances is the use of the endoscope to inseminate mares artificially.

There have been a number of experiments conducted in this country and England, but little has yet appeared in the literature. What is known and has been reported is that the endoscope makes it possible to deposit spermatozoa at the uterotubal junction–that point where the fallopian tube empties into the uterus.

What this means is that the normal course taken by the spermatozoa through the uterus–often resulting in death to many of the microscopic travelers–can be avoided. Instead, the spermatozoa are placed in a safe haven where they quickly come into contact with the descending egg.

One of the first reports given on the procedure was at the 1998 AAEP meeting in Baltimore. But before you start hounding your veterinarian to use this technique, please note that while it is very promising, it still is in the developmental stage and needs further work and better equipment before it is ready for on-farm use.

Reporting on the procedure was J.J. Vazquez, MVZ, University of California, Davis. Involved in the study were 10 cyclic mares ranging in age from five to 15 years. Their reproductive history was unknown to the researchers.

The mares were sedated, then inseminated with seven million sperm cells. The sperm was from a fertile stallion. Using an endoscope and a polyethylene catheter inserted through the endoscope, the semen was placed onto the surface of the uterotubal junction. Prior to insemination, each mare had received 2,500 IU of human chorionic gonadotropin intravenously when a 35-millimeter follicle had been detected.

Three mares were diagnosed as pregnant with a fetal heartbeat 25 days after ovulation. Of the three mares which were pregnant, one mare was inseminated within 12 hours of detected ovulation and the other two mares were inseminated 24 and 48 hours prior to the detection of ovulation.

Vazquez told his listeners at the convention that one of the reasons for the low success rate was because of the large gauge diameter of the insemination catheter used "and the poor accessibility to the uterotubal lumen in the mare."

He went on to say that the procedure demonstrates that positive results can be realized with a small number of spermatozoa via this method.

"To our knowledge," he said, "this is the first report of successful (non-surgical) uterotubal insemination attempts in the literature. Although the number of mares included in this study (10 mares) is small, this study provides preliminary evidence that a low number of spermatozoa can result in pregnancy if sperm are deposited non-surgically on the surface of the uterotubal joint ipsilaterally (pertaining to the same side) to impending ovulation. This procedure may serve as a possible assisted reproductive procedure for the management of subfertile stallions with sub-optimal number of normal motile spermatozoa or as an alternate procedure for insemination with cryopreserved (frozen) spermatozoa."

There has been a great deal of research work conducted on the procedure since that outlined by Vazquez a little more than a year ago, although much of it still is unreported in scientific literature. In one study, it was found that a 64% pregnancy rate could be obtained on a per cycle basis with as few as one million progressively motile spermatozoa. Researchers have even found that pregnancy can be obtained with as few as one thousand progressively motile sperm.

This compares with normal AI procedures that often use 500 million progressively motile spermatozoa.

As Vazquez pointed out in his conclusion, the procedure will be a major assist for sub-fertile stallions. It also means that if the market were there, it would open the door to breeding even more mares to fertile stallions than the ambitious numbers already discussed by Pickett.

As mentioned, there is no question that artificial insemination in general, and shipped semen specifically, has had an impact on the equine industry. Other than The Jockey Club, there are virtually no registries that do not permit artificial insemination and very few that do not permit transported cooled semen. It is far easier to list the registries that do not permit transported cooled semen than those that do.

Included in the list that do not permit transported cooled semen as of figures available in 1998 are as follows:

American Indian Horse Registry; Colorado Ranger Horse Association; The Jockey Club; Mountain Pleasure Horse Association of North America; Palomino Ponies of America; Pony of the Americas Club; United Quarab Registry; and Universal Perkehner Society.

When it comes to transporting frozen semen, however, there still are some major holdouts, although some of these registries indicate they are studying the matter.

Leading the list of registries that do not permit registration of foals conceived with frozen semen is the American Quarter Horse Association. Following in alphabetical order are the other registries that do not permit the use of frozen semen: American Andalusian Society; American Indian Horse Registry; American Miniature Horse Association; American Paint Horse Association; Appaloosa Horse Club; Belgian Draft Horse Corporation; Colorado Ranger Horse Association; The Jockey Club; Mountain Pleasure Horse Association; National Spotted Saddle Horse Association; Norwegian Fjord Association of North America; Palomino Ponies of America; Pony of the Americas Club; Spanish-Norman Horse Registry; and United Quarab Registry.

(Information concerning registries that do not permit the use of frozen semen also was for 1998.)

So, Who’s Using AI?

While equine registries in general have embraced transported cooled semen, a sampling of registries large and small indicates that the majority of the mares still are bred on premises, either by natural cover or by AI with fresh semen.

In the Arabian breed, for example, Debbie Fuentes of the Registry reports that only about 12% of the mares currently are being bred with transported cooled or frozen semen. By late fall of 1999, she said, the group had registered 9,627 foals born that year. Of that number, only 1,149 were conceived with transported cooled semen–right at 12%.

Current figures demonstrate that there has been growth in the use of transported semen by the Arabian horse industry since its approval in 1995. In 1996, some 9.6% of foals were conceived via transported cooled semen. In 1997, that figure had increased to 10.6%. In 1998, it was 12.3%.

In the Quarter Horse industry the number of mares bred by transported cooled semen is far lower than the number bred on premises by either natural cover or AI. Black says that in 1998 there were 186,458 AQHA mares bred. Of that number, 8,424 were inseminated with transported cooled semen.

She said that breeding reports were received for 30,838 stallions in 1998, and of that number, permits were obtained for 1,110 for shipment of cooled semen.

A registry that is among the smallest–but one in which both transported frozen and cooled