Innovations in AI

Over the last 50 years, the equine breeding industry has revolutionised, from wartime and agricultural motivated breeding based on natural covering, to the now predominant artificial insemination (AI) for sport, recreation and leisure breeding. This abstract and accompanying presentation aims to give an overview of the progress and changes in equine breeding over the last half century. Due to the differences between the racing and nonracing sector, each will be commented on separately.

Advances in equine assisted reproduction: nonracing sector

There is no doubt that AI dominates the nonracing sector. While up to 80% of horses were bred by natural cover in the 1980s, up to 90% are now bred worldwide via AI; a complete turnaround in breeding method. So where did AI begin? How did it develop, and come to dominate the nonracing industry?

The earliest possible reference to AI can be found as far back as 1300, to Arabic texts detailing the first possible human intervention in mare insemination. In 1678, microscopic view of sperm cells was reported and later in 1700, L. Spillanzani carried out the first documented scientific research in fertility, most significantly providing direct evidence of the requirement for sperm and egg cell meeting to begin new life. The first officially recognised reference to equine AI can be attributed to French veterinarian Repiquet, who, in 1888, suggested that collection of semen from males could be a helpful way of breeding more mares to one stallion, in addition to aiding in the management of breeding issues, such a thought triggered by the progress and application of AI in research-based fish and canine breeding.

Fast forward to the nineteenth century, and the outbreak of war began a pressure upon equine breeding for military and agricultural needs. Russia in particular invested notably in its studs and, from here, the most significant developments in equine AI as we know it today were made. In 1903, E.I. Ivanoff established the first equine insemination centres, the Imperial Russian Studs. Semen was collected via silk sponges placed in the vagina of mares. Following natural covering, the sponges were removed and placed in a specifically designed press to extract the semen. Systematic insemination of multiple mares using the collected semen was then made. Despite its infancy, pregnancy rates of 76–78% were achieved. Ivanoff also discussed the shipping of semen to neighbouring studs, to improve breeding strategies, prior to further developments including phantom mare collections and the use of condoms for semen collection. The effectiveness of AI of multiple mares from one stallion ejaculate in improving breeding numbers and breeding progress was noted during these wartime efforts and the trend set for continued growth and application of this technique.

Post wartime breeding

At the end of the World Wars, a change in use arose from war and military to sport and recreation. The ongoing development of AI from the wartime period led to the development of techniques for chilling semen and for freezing too, which began to gain in popularity in line with the desire for breeding geographically separated animals. The 1980s saw the development of mare scanning, in addition to artificial hormones to assist with reproductive cycle management and regulation. By 1995, 50% of horses were bred by natural cover and 50% via AI. By 2004, this had increased to 80% via AI and at present, over 90% of sport horses are now bred using assisted reproduction.

Milestones in equine AI

From the primary concept of collecting semen and splitting between mares, the equine AI industry has progressed and continues to progress at a rapid rate of technological development. The development of mare scanning and hormones for cycle management was a notable game changer in the efficiency and success of AI. Pioneered by Professor Twink Allen, this enabled increased accuracy and success of mare insemination and provided the baseline tools for further progress in the industry. 1972 saw the first equine embryo transfer, which then opened the door for other technologies of intracytoplasmic sperm injection (ICSI), embryo freezing and cloning, which all require the input of embryo transfer.

In the 1980s, Martin Boyle began his work on the application of fresh and chilled semen to the commercial breeding environment, progressing to frozen semen in the 1990s. By 1996, West Kington Stud was founded – the first commercial equine AI centre in the UK – with Stallion AI Services founded in the year 2000; centres that are both of note in the development and progression of equine semen processing for AI.

Recent advances in equine AI

The speed of change in equine AI is such that the milestones of scanning, embryo transfer, semen processing, etc., are all considered standard techniques within the field. For the present- day generation, it is technologies such as semen sexing and the increasing focus on genetics that are the latest revolutions within the field.

Equine semen can now be sorted into male and female sperm cells via flow cytometry, in order to aid in rare breed survival, or the production of sport horses of the desired sex. Semen can also be extracted from the testicles post castration in a procedure known as post castration epididymal semen extraction. On the female side, embryo transfer has progressed to oocyte recovery, ovum pick up (OPU) and insemination via ICSI. In 2021, over 6,000 OPU and ICSI procedures were performed by Avantea SRL, the world-leading company in this field of equine conception. Embryo freezing is also progressing, with focus and promise on vitrification to improve results. The ability to combine this with semen sexing could prove revolutionary to the sport horse and rare breed sectors, enabling the banking of sex-determined offspring for future generations.

Moving beyond sperm and egg-based conception, cloning is a technology also gaining pace within the nonracing industry. A method of producing offspring from geldings, increasing the input of mares on the breeding sector, assisting with rare breed conservation and also bringing back the whole genetic set of deceased elite performance horses; merits which the sector is increasingly recognising. Continuing down the genetics route, technologies are developing for increased analysis of mare and stallion breeding potential in terms of their genetic set, karyotyping being one example. Methods are also being developed for semen DNA analysis, to allow for more in-depth assessment of subfertility cases and/or the interaction of external factors on semen fertility.

The application of genomics to the breeding industry is also starting to be seen, with certain stud books beginning the primary stages of linear profile scoring of offspring. 2020 also saw the first application and publication of gene editing to equine breeding – production of myostatin gene-edited embryos by Moro et al. 2020. In the background to this, there is also the ongoing development and application of statistical breed analysis to rare breed breeding strategies. An important note to take forward as the industry continues to progress is the indirect implication of advanced breeding techniques enabling the breeding of subfertile animals, with some preliminary research suggesting raw semen quality as one parameter may have fallen by 10% over the last 20 years.

The application of equine reproductive technology to endangered species preservation is also being seen, with the effectiveness to draw upon every available reproductive element a key trait of benefit in assisting with the survival of threatened species.

The racing sector

While AI has stormed ahead in the nonracing sector, for the Thoroughbred industry, natural cover continues to dominate. Huge advances have been seen with end of season pregnancy rates and live foal rates increasing from 75% and 55% in the 1960s to over 90% and 85% respectively in the present day. Alongside this, stud book size has increased dramatically, from just 40–50 mares in the 1960s, to over 300 mares per season for popular sires.

The effect of popular sires on genetic diversity has, however, attracted attention within the sector. Summarised by HH Aga Khan TBA in 2001, ‘when one stallion can engender up to 300 foals per year, the global inventory of pedigrees, in say 15 years, will present breeders with a totally new problem. In those 15 years a mare will have produced only some 12 foals, whereas a stallion could have sired 4,500 foals’, there are fears that unregulated and unrestricted book sizes could be dramatically depleting diversity within the sector. Such fears led to a pinnacle moment in 2020, where the US Jockey Club introduced a limit on stallion book size. For reasons of a lawsuit, this rule has recently been revoked and its reapplication is under review. Nonetheless, for many, the importance of recognising genetic diversity and its challenges within the racing sector are areas for further revolution within Thoroughbred breeding, to ensure long-term health and performance of the sector.

[...]