03 58 299 566  gvequine facebook

Reproduction | Efficiency


Reproductive Efficiency of horses in Australia


 What is a feasible outcome of yearly breeding of horses? Is it reasonable to expect a foal very year?

To answer those questions we need to look at some historical perspectives of breeding efficiency and some physiological parameters of horses.

Firstly, horses have adapted over millions and millions of years to create a reproductive pattern that ensures their survival in the wild. We have then domesticated them and imposed our own constraints on their breeding performance. We artificially impose the time of the breeding season wherein most Thoroughbred farms have stopped breeding at the time of summer solstice (Dec 21 in the Southern Hemisphere) which happens to be the time that the highest % of mares are cycling (Osborne 1966) . We have no or very little selection pressure on fertility; rather primarily it is the horses’ performance that dictates the breeding pressures. As well, the high economic value of many horses dictates breeding with older animals that may have accumulated a variety of negative influences on breeding potential.

Secondly, the horses gestation is ~340 days. This only gives us the ability to maintain a yearly inter-foaling when mares are back in foal with a successful breeding occurring no later than 25 days after foaling. If mares are not bred on the foal heat, then there is only one opportunity to maintain a yearly inter-foaling interval. If the mares’ gestation was ~280 days as with most breeds of cattle, then there would be plenty of time to breed her and maintain a yearly inter-foaling interval. Sheep have a gestation of ~148 days, so there is even a chance for two pregnancies per year in those breeds that cycle all year round (but who do you know that can produce lambs twice a year?). It would be a much easier decision of whether or not to try and maintain a yearly inter-foaling interval of one year if the gestation of a horse was the same as a Rhinoceros (540 days), Giraffe (440 days) or Elephant (22 months).

 What then should we be aiming at with regard to annual foaling percentages? Hans Merkt and colleagues made an interesting observation on records from German Hanoverian horses over a 158-year period. ‘An evaluation of the foaling rate achieved in the Hanoverian breed in Germany between 1815 and 1973 showed that during this period no improvement of the reproductive rate was achieved. The decennial foaling percentage remained within 50-60% except for those decades which included the two World Wars and their aftermaths. The annual foaling percentage remained maximal until the number of mares covered/stallion rose above 80 and it also remained high throughout the reproductive life of the stallion. Only in the oldest stallion (32 years) was there a significant lowering of fertility’ (Merkt et al., 1979).

A more recent report suggested that breeding efficiency had improved in the UK (Ricketts and Young 1990) . They reported seasonal pregnancy rates of 63% in 1971 and 72% in 1989. The live foaling rates were 54% and 66% respectively. This improvement in number of mares pregnant and number of foals born corresponds nicely to the gradual increase in research interest and establishment of a number of laboratories devoted entirely to the study of equine reproduction. However a survey of well (intensively) managed Thoroughbred farms around Newmarket (Morris and Allen 2001) suggested that the main improvement in efficiency on those farms was an improvement in the number of mares that did not loose their pregnancy compared to a previous study by one of the authours (Sanderson and Allen 1987) . The number of mares foaling was 82.7% in 1998 and 77% in 1983.

Presented to the left is some interesting information from our Thoroughbred breeding data (The Australian Stud Book C/O Dr. John Digby).

Figure 1 shows that the number of stallions peaked in 1985-6 and has been gradually dropping since.

When we look at the number of mares versus stallions (at left) we can see that both mare and stallion numbers dropped off quickly from the heights of the mid 1980’s however the mare numbers seem to have stabilised somewhat.

When we compare the number of available mares per stallion we can see that is steadily rising. A figure that ultimately should put pressure on breeding efficiency.

 Figure 3. The number of available mares per stallion


Perhaps the most interesting statistic to be noted was the live foal per mare bred percentages (at left). Back in 1981 we were only recording around 40% live foals per mare covered. That figure was almost 76% by the end of 2000. Apart from the afore mentioned UK experiences it would stand to reason that the improvement in fertility could be related to a multi-factorial influence. Firstly, the value of the individual has risen dramatically and coupled with that has been an information explosion, together with an improvement in our equipment to manage reproductive processes. This increased understanding has been passed on to all levels of farm management. The net result has been a stabilising in the numbers of foals born despite still steadily falling mare numbers.

However, there are some other issues that should be examined. Firstly, does the Thoroughbred breeding record presented above reflect the other breed associations and secondly, is the live foal rate consistent between farms or is it affected by management?

Before looking at those issues in depth we should try to standardize our terms of reference.

Is the number of foals born an accurate estimate of the real fertility of a breed of horse? Probably not we think. The most accurate estimate of a stallions’ fertility data is the number of mares pregnant each cycle. At the end of the season the number of mares pregnant may reflect the number of cycles that the stallion had an opportunity to breed rather than the true fertility. For instance, if one farm has a 75 % pregnancy rate per cycle then after two cycles for each mare (eliminating mares pregnant), 93% of mares are pregnant. If you have another farm that has a pregnancy rate per cycle of only 40%, then they can achieve a 93% pregnancy rate after 5 cycles. The difference is huge in relation to agistment and veterinary costs, not to mention lost opportunity costs because the foals are born so much later. This later discussion may in part explain the English report that quotes Weatherbys general Stud Book as reporting the percentage of live foals to be 79% in 1998 compared to our Stud Book records that officially list percentage of foaling mares as 74% in 1998 and 76% in 1999 and the year 2000. The more recent English report (Morris and Allen 2001) list almost 20% of breedings after the first 3.5 months of the breeding season. Our experiences would indicate that in Victoria and NSW not many mares are bred in December (the fourth month). The average pregnancy per cycle in the UK study was 60%, which is lower than the 74% reported for intensively managed Thoroughbred studs in Australia (McKinnon 1998a) .

The goal of any breeding farm should be to get mares in foal as soon as possible whilst maintaining their ability to deliver a live foal. There is little point in having a mare become pregnant if she has a high probability of loosing it. If we examine breeding efficiency (not necessarily the same as fertility) it could probably be better defined as % of foals born per service or per cycle bred. Well-managed Thoroughbred breeding farms will maintain foaling % of around 70% per covering. Despite all the figures and preceding discussion the goal of most breeding farms is actually to maximise the number of foals born. Just because the efficiency figures look good does not imply that the return from the farm is maximal. It stands to reason that the farm has the same result financially when a foaling percentage of 70% is obtained with 60 mares occurs versus a foaling percentage of 60% for 70 mares, when the service fees are identical. So the goals may be defined from the point of farm management as ‘obtaining the highest number of live foals’.

 Current RIRDC Study

In 1999 RIRDC funded a study on the breeding efficiency of horses in Australia. The project aims are listed below.

Appended at the back of this article is the Access database that we have developed to analyse breeding efficiency data.

The project is still years from completion however some very useful results have already been forthcoming. Currently we are validating the database and running trial ‘queries’, the term used in Access for asking questions of databases.

Project Synopsis


Outcomes and deliverables of the proposed research

Outcome: A thorough analysis of the efficiency of breeding practices in the Thoroughbred and Standardbred industries in Australia.

Objective 1): Analysis of breeding records via the Stud Books. This is only expected to provide an overview of industry practices. For a variety of reasons, records submitted to the Stud Book (both racing codes) are only a superficial view of breeding practices, however they do represent the true and relevant fertility’s of the different breeds and even different stallions on the same breeding farms.

Objective 2): Analysis of individual mare records from selected breeding farms.

This is an ambitious attempt to collate data from a wide variety of sources.

It is estimated that  individual mare records of each breed from a variety of studs that have been chosen to represent three levels of management will be used to compare factors such as fertility per cycle, effect of fluid pre and post service, effect of scanning or not, rate and time period of early embryonic death, etc.

Answers to these questions may provide RIRDC with a balanced view of problems in this area of the industry and areas that should be targeted for further research. For example we estimate that early embryonic death accounts for as much as 15% of losses on Thoroughbred farms, however this data comes from small populations that may be skewed or not representative of the total population. Establishing an accurate figure would be useful in targeting funding for further research.


Background, relevance and potential benefits

Research into breeding practices is hampered by lack of a large data set to draw conclusions from. For instance we published on the incidence in Thoroughbred versus Standardbred mares of twins (15% versus 5%) and yet few studies have been able to demonstrate these differences due to inadequate sample size.

Establishing a data base on the breeding performance of the Thoroughbred and Standardbred horses in Australia is of particular use due to identifying trends (i.e. have we improved efficiency on well managed breeding farms?) and identifying areas that need research (i.e. how much is lost to the industry each year from foals that are either lost as early embryonic death or aborted and should those areas be targeted for serious and expensive research?).


Preliminary results from data entered into the Access database.

Some specific factors that effect breeding efficiency have been recorded. Where possible we have tried to contrast the differences according to breed (Thoroughbred (TB) versus Standardbred (SB)).

Numbers of mares examined.

The database we have chosen to work with for this presentation has 1833 TB mares and 1330 Standardbred mares. The number of cycles examined was 2436 and 2086 respectively. All Thoroughbred mares were bred by natural service and all Standardbred mares by artificial insemination (AI). The data was retrieved from the years 1994-5.

Some summary data is listed below.

Table 1:

 Fertility statistics:






Mare numbers



Cycle numbers



Single pregnancies:



Twin pregnancies:



Twins per cycle:


9.7 %


2.2 %

Twins per pregnancy


13.8 %


3.5 %

Triplet pregnancies:


0.35 %





Total mares pregnant at 15 days


93.6 %


97.3 %

Total mares pregnant at 45 days


86.7 %


88.6 %

Fertility per cycle:


70.4 %


62.0 %

Cycles per pregnancy:





Services per cycle





Table 2:

Twins and Early Embryonic Death (EED)






Twins (multiple) present at ~15 days


14.3 %


3.5 %

Loss to 25 days


4.0 %



Loss between 25 and 45 days


3.5 %



Total EED (15-45 days)


7.3 %



EED (15-25 days)


3.7 %


7.1 %

EED (26-30 days)


2.1 %


1.5 %

EED (31-45 days)





Table 3:

Foal Heat






Mares bred on foal heat



Mares pregnant at 15 days


64.5 %


51.0 %

Loss between 15 and 45 days


17.5 %


8.0 %

Table 4:

Ovulation induction: Thoroughbred


Number of mares

% Pregnant/cycle

% Twins/pregnancy



72.0 % (147/204)

19.7 % (29/147)



72.0 % (1042/1447)

14.8 % (154/1042)

No drugs


64.5 % (527/817)

11.8 % (62/527)

Ovulation induction: Standardbred


Number of mares

% pregnant/cycle

% Twins/pregnancy



50 %



 63.4% (121/191)

 4.1 % (5/121)

No drugs


61.8% (1172/1896)

3.5 % (41/1172)

Table 5

Effect of fluid and treatment on pregnancy data and EED:


No. pregnant/No detected (%)

EED (%)

Intraluminal fluid pre service

271/388 (69.9 %)

24/271 (8.9%)

Intraluminal fluid post service

114/222 (51.4%)

10/114 (8.8 %)

Single treatment pre service

266/358 (74.3 %)

24/266 (9.0 %)

More than one treatment pre

134/185 (72.4%)

11/134 (8.2 %)

Single treatment post service

291/381 (76.4 %)

19/291 (6.5 %)

More than one treatment post

393/578 (68.0 %)

33/393 (8.4 %)

Effect of fluid and treatment on pregnancy data and EED:


No. pregnant/No detected (%)

EED (%)

Intraluminal fluid pre service

88/186 (47.3 %)

10/88 (11.4%)

Intraluminal fluid post service

97/234 (41.6%)

10/97 (10.3 %)

Single treatment pre service

53/99 (53.5 %)

4/53 (7.5 %)

More than one treatment pre

3/9 (33.3 %)


Single treatment post service

129/252 (51.2%)

13/129 (10.1 %)

More than one treatment post

11/36 (30.6%)

2/11 (18.2 %)

Conclusions on preliminary data:

Fertility statistics:

Pregnancy rate per cycle:

The higher pregnancy rate per cycle in the TB population compared to the SB (70.4 % versus 62.0 % P < 0.0001) was not expected but likely reflects economics rather than the population of mares. The differences become even clearer when examining the average services per cycle, which was 1.04 for the TB population (on average four mares got re-bred from every 100 cycles) and 2.21 for the SB. Interesting enough, the SB seasonal pregnancy rate  (number pregnant at 45 days) was slightly higher (88.6%) than the corresponding TB figure (86.7%).

Although not examined at this time, we have long identified SB breeders as having less regard for time of foaling than the TB breeder. Possible reasons for this are 1) The SB breeder is more likely to retain the offspring to train themselves and thus not as concerned as to time of foaling 2) Most TB breeding farms stop breeding by Christmas (< 4 month breeding season), whereas SB farms continue to breed on well into February and even March and 3) SB farms are less likely to involve veterinarians in routine breeding farm management, due primarily to economics. This is the reason that the average service per cycle for TB mares is 1.04 versus 2.21 for SB. So SB mares are bred multiply and can have increased chances for intrauterine infections and are more likely to be bred at inopportune times such as post ovulation, especially if teasing isn’t integral to management. In addition to the above, the staff levels tend to be much higher per number of horses on TB farms than the SB.

The fact that we have shown the SB seasonal pregnancy rates are identical to the TB (despite a significant difference in fertility per cycle), clearly illustrates that they breed longer into the breeding season. The SB regulatory authorities should also be congratulated on delaying the date of first breeding (~October 1st to coincide with an official first foaling date of September 1st - exactly one month behind the TB breeders) and clearly helpful in utilising the effect on increasing cyclicity from increasing daylight and warmth).

Twins and EED

Many more twins were identified both per cycle and per pregnancy (9.7% and 13.8%) for TB compared to the SB (2.2% and 3.5%) respectively (P <0.001) at the (~) day 15 pregnancy test. These differences between breeds have been reported before (Ginther 1992) and relate mostly to the frequency of ovulations between breeds. Especially interesting was the outcome of twin pregnancies. The reduction of the twins to singletons (success of veterinary intervention, i.e. to crush one) was high in the TB population, as only 10 of 245 twins (4.0%) were not detected as single pregnancies at day 25. During the same period 8 of 46 SB twin pregnancies were both lost (17.4%), which is significantly more (P<0.006). We believe that this difference is related to fundamental management practices on breeding farms. Because the incidence of twins is recognised as low on SB farms, in general the first pregnancy test is delayed until the mare should have returned to oestrous. This delay until day 17-20 often interferes with our ability to manipulate twins apart for a successful crush as fixation (the beginning of implantation) has occurred and twins together can be difficult to move part for crushing (McKinnon and Rantanen 1998b) .

The incidence of EED was similar between mare populations with more (P< 0.01) pregnancies being lost in the SB group (7.1%) between day 15-25 than the TB group (3.7%). It is interesting to speculate as to the differences being due to a difference in the number of mares with post breeding uterine inflammation as a result of less examination and treatments.

Foal heat breeding

  In general we try to avoid foal heat breeding as there are reduced pregnancy rates per cycle (McKinnon et al., 1988) and an increase in the amount of EED in those mares that do become pregnant (Loy 1980)   (see later). Management can play a part in increasing foal heat fertility (McKinnon et al., 1988) . In this study the fertility of the foal heat mares in the TB group was not different to the total population of TB (64.5% versus 70.6%, P<0.3). However pregnancy per cycle was reduced for the SB group and the summation of both groups revealed a significant lowering of fertility in foal heat bred mares (165/307, 53.7% compared to 2845/4205, 67.7% P <0.0001). However there was no increase in EED for the summed data of both breeds. For mares bred on foal heat the losses by day 45 were similar (17/165, 10.3%) compared to all other heat cycles for both breeds 224/2845 (7.9%) (P<.23).  

Induction of ovulation

  A greater percentage of mares had ovulation induced by either hCG or Ovuplant in TB mares (1651/2468, 66.9%) compared to the SB population (193/2089, 9.2%). An interesting observation was a slight increase (P<0.012) in the number of TB mares with twins (per pregnancy) detected when treated with Ovuplant (29/147, 19.7%) versus no treatment (62/527, 11.8%).

The low number of mares treated with Ovuplant was only because the drug was not commercially available at that time (1994-5) and we were doing some trial work for the company. Data collected over the last few years has indicated closer to 50% of mare cycles are associated with Ovuplant administration. The low number of Standardbred mares treated with ovulation induction drugs reflects the economics of breeding as well as the reduced necessity to use them with AI programs.

Another interesting finding was the increased fertility (P< 0.0001) (pregnancy rate per cycle) of TB mares with induction of ovulation (1189/1651, 72.0%) versus not treated (527/817, 64.5%).

Effect of intraluminal fluid and or treatment on fertility.

As previously reported (McKinnon et al., 1987; Adams et al., 1987)  intrauterine fluid post ovulation can have a negative effect on pregnancy results. In the current study fluid post service significantly effected (P<0.0001) pregnancy in TB mares (114/222, 51.4%) compared to non-detection of intrauterine fluid (1602/2216, 72.3%). The same was true for the SB mares (97/234, 41.6% compared to 1197/1852, 64.6%, P< 0.0001).

In the TB population treatments pre service or a single treatment post service was associated with normal pregnancy rates, however more than one treatment post service was associated with a reduction (not significant P<0.15) in pregnancy rates (465/681, 68.3% compared 1251/1755, 71.3%).

  Factors affecting breeding efficiency

Decreased time to conception

            For mares entering breeding farms, the time to the first breeding has a big impact on when mares will foal. Early foals consistently occur from early breeding. Many mares are not cycling naturally at the time of year when we wish them to become pregnant (early September), this is especially true in the southern latitudes.

How can we get mares to cycle earlier?

1)   Provide an increased amount of light beginning around 60-90 days before the first anticipated breed. 16 hours per day is necessary. French workers have determined the amount of light is not as important as previously thought. Another interesting finding from the same people was that the light regime does not have to continue for the whole 90 days. It is just as beneficial to start the program and then stop the lights after 35 days. Beware!!!, this does not mean that the lights can be started late. All light does is send signals to the brain (through a negative melatonin feedback). If the program starts late the mares will cycle late. Another commonly forgotten part of the early cycling equation is warmth and nutrition. Mares not fed well or stressed will not cycle as early as well fed and non-stressed mares. This may mean that providing light in large paddocks where mares are fed together is not as useful as previously thought.

2)   Drugs.

            Stimulatory drugs. These drugs start mares cycling when they do not appear to have real ovarian activity. There is lots of new information coming through on these types of drugs and unfortunately some of it is not as conclusive as previously thought. Currently it appears that Dopamine antagonists (Domperidone and Sulpiride) hold the most promise in shifting the balance on ovarian activity. Please recognise that not all these drugs work all of the time, in fact some work only 60% of the time. In Australia, Domperidone is marketed for humans as ‘Motilium’.

            Ovulatory drugs. Drugs that consistently ovulate follicles that previously were thought to be too immature to respond are now being used clinically. Our laboratory (Goulburn Valley Equine Hospital) was first to report of the use on Ovuplant in treating mares to have them cycle early at a predetermined time. A detailed discussion of research work supporting the clinical use of Ovuplant is provided below. Breeding farms that have used Ovuplant for forcing mares to cycle can attest to its success, but many forget that the mare has to be close to ovulating anyway and we are merely advancing her time of the first ovulation in a controlled manner. This drug will not work when mares have little ovarian activity.

Increase pregnancy rate per cycle

            Decreasing the number of breeds per cycle

                        Decreasing number of breeds per cycle provides less contamination of the reproductive tract and prevents depletion of sperm numbers from the stallion. Our recommended breeding objectives are a breed per cycle figure of 1.05. In this scenario every 100-mare cycles would account for 105 breeds. This has important ramifications for management. To achieve this objective mares need to be managed intensively. When a suitable sized follicle is detected then a drug to induce ovulation is administered. Ovulation must be confirmed at 24 or 48 hr after breeding because some stallions do not have sperm that lasts well in the reproductive tract.

It has only been readily accepted by breeding farms relatively recently that this practice will decrease labour and increase pregnancy rates per cycle and it has become absolutely paramount for management of breeding stallions with large books (>100 mares) or restricted services such as some shuttle stallions.

One of the most useful drugs for controlling ovulation time and thus keeping the number of breeds per cycle down to ~1.05 is Ovuplant. The properties and results of our experiments with this drug are listed in the short articles towards the end of this paper.

            Management of mare cycles

There is no doubt the single biggest factor in improved breeding efficiency has been the introduction of ultrasonography. This has made veterinary involvement mandatory on large breeding farms and has dramatically improved all our knowledge of the mares cycle and overall fertility, not to mention improvement in the management of twins. Equine reproductive management can now be practiced as a science instead of an art.

Few people predicted the impact that ultrasonography would have on equine reproductive management and understanding of reproductive physiology. The ability to examine a mare's reproductive tract non-invasively with ultrasonography provided the opportunity to diagnose pregnancy earlier than by rectal palpation, effectively manage twins and detect impending early embryonic death (EED). However, ultrasonography should not be limited to these areas. It can be used to diagnose uterine pathology, such as intrauterine fluid, air, debris, cysts and occasionally abscessation and neoplasia. In addition, ultrasonographic examination of the ovaries may aid in determining stage of oestrous cycle, status of preovulatory follicles, development and morphologic assessment of the corpus luteum (CL) and in interpreting ovarian irregularities, such as anovulatory or haemorrhagic follicles, neoplasia and peri-ovarian cysts. The costs of equipment initially resulted in a rather limited application of reproductive ultrasonography. Clients enthusiastically support use of ultrasonography to detect pregnancy. However, the same fee schedules for routine examination before and/or after breeding are not as easily accepted. An approach that allows us to scan multiply while still keeping clients and farm managers happy is something all of us strive to organise each year. A more logical and thus practical approach to diagnosis and treatment of physiological and anatomical abnormalities of the mare's reproductive tract would be forthcoming if we can continue to develop a means to use the equipment more routinely. In addition, valuable information would be available from correlation of fertility data with normal and abnormal ultrasonographic observations. Regardless, informed clientele prefer routine ultrasonography and its use results in a more interactive approach to farm management with an increased awareness of the events associated with breeding, ovulation and early foetal development.

            Stallion fertility and management

The stallion is forgotten in the fertility equation despite being responsible for half of the probability outcome in each breeding. If a group of matings results in an outcome of 80 % of mares becoming pregnant per cycle then the mare and stallion fertility contributions are most likely 0.9 and 0.9. When the fertility of stallion drops then the outcome is affected. If his fertility is halved then only 40% of mares will go in foal per cycle. However because people have the mistaken belief that there is nothing we do to influence stallion fertility, they ignore his contribution. It is disappointing to recognise that within Australia, only a few people are adequately trained to evaluate stallion fertility parameters. Perhaps that is another reason why the stallion is too often forgotten.

Management has a huge impact on fertility. Well-managed farms have good pregnancy and foaling rates. However, when pregnancy rates per cycle on the same farm are compared between stallions quite often there are major differences in the figures. In those cases, providing the percentages of mares in the barren, maiden and foaling groups are similar; the stallion is demonstrating his “inherent” fertility. That is something he is born with. I am sure many of us have seen the stallion that just seems to have to sniff his mares and they go pregnant. Compare him with the stallion that takes two, three or even four cycles to get mares pregnant. They can end up with the same number or percentage of mares pregnant at the end of the breeding season, but it is a lot more work with the less fertile stallion and his progeny may have an average foaling date that is later than acceptable for commercial foal sales. Nobody knows what the real differences in fertility are caused by. However, quite frequently we see a difference in the characteristics of progressive motility and good tense sperm producing testicles between the stallions. Despite this, there are sometimes no obvious differences and the only explanation is that one stallion is able to produce more fertile sperm than another.

Mare fertility and management

                        Infectious fertility.

Ability to recognise that some mares have poor clearance of uterine fluids was proposed many, many years ago, however only recently has a technique been able to measure the lack of evacuation of uterine contents from some mares. This technique from Florida has resulted in the knowledge that some treatments are extremely useful such as oxytocin and certain forms of prostaglandin. In general ultrasonography tells us how bad the fluid is (usually infection) and also how to treat it. I.e. treat locally with antibiotics or treat with agents to evacuate the fluid from the uterus or a combination.

                        Foal Heat breeding.

Many of our clients wish to breed mares on foal heat. The reasons are generally obvious. The gestation of the mare averages 340 days. However some are longer. Most of us involved with commercial breeding farms recognise a financial penalty for foals born later than the normal accepted times. Without breeding at foal heat it may be hard to maintain a foaling interval of 365 days in some cases. In addition clients with brood mares on farms sometimes are paying high agistment costs and place pressure on managers to get their mares pregnant and send them home as soon as possible. Sometimes farm mangers cannot resist the temptation to breed at foal heat despite the potential problems, just because the mare is showing heat and their knowledge that foal heat mares ovulate quite quickly.

Put simply the positive aspects of foal heat breeding is a decrease of the inter-foaling interval and all the attendant benefits. The negative aspect that many fail to recognise is the increase in early embryonic death. The interplay of these opposing outcomes should be factored into any decision-making analysis.

GVEH Breeding approach

1)  All mares are examined at day 2-4 post foaling. By that time if problems have occurred during foaling then they will be identified. We believe that this exam is critical if we are to prevent mares become “problem broodmares”. Each year we identify mares that if inappropriately treated will become long-term disasters.

2)  All mares are treated with an infusion of Lactated Ringers Saline with antibiotics. We believe that this helps only the occasional mare, but more importantly we have invaded the mares’ reproductive tract and that may need to be treated in case we have introduced something.

3)  Mares looking good are scheduled to be re-examined on Day 9-10 post foaling. We will not breed mares any earlier than Day10. On day 9-10 if the mare has ovulated she is scheduled for PGF2µ administration in 6 days. If she looks good (no fluid and uterine tone is good) she is scheduled to breed if possible. If there is a question then she is treated and or re-examined according to follicle size and presence or absence of uterine fluid.

4)  Great care is taken to avoid breeding mares with intraluminal fluid accumulations.

Lastly there are two more rules that we adhere to. A) Only breed mares that are young and reproductively healthy (i.e. <12 YO) and B) Do not breed on the foal heat if the mare has been boxed and not forcibly lunged.

Lastly, old mares that are bred on foal heat and become pregnant have such a high rate of early embryonic death or abortion we strongly recommend against the practice.

Decrease pregnancy losses

Pregnancy losses can broadly be divided into early embryonic death (EED) (< day 45 of gestation) and abortion.

There are well-documented probabilities of foetal loss for different time periods of pregnancy. The recognition that these losses have repeatable frequencies reported in different studies should point to a common denominator that is present in multiple populations of mares.

My view is that management is the most important factor affecting the well being of horses. I would go one step further and say, “all disease is caused by management”. This is of course a dramatic statement but many times true if you ponder the myriad of diseases seen on stud farms. In particular, overcrowding, concentration of disease and the handling of horses in systems designed for efficiency rather than safety are all factors in breeding wastage.

Below are figures on the number of mares bred and the resultant number of foals born. Clearly the fertility of horses is quite low when compared to other species. However there is an encouraging trend in the overall fertility.

 Fertility of mares bred in Australia (1992-1996).

Why are these figures so low? To start with there is no selection for fertility at all. We are selecting for performance (mostly through economics) and the conformation associated with good performance seems to favour a long sloping vagina, which is definitely a poor conformation for breeding. Can we defend the lack of selection for fertility? Most think so, as surely our aim is to produce the ultimate athlete rather than a series of fertile but slow or similarly performing animals. Given the lack of selection for fertility it seems difficult to defend the ban on AI on technological grounds. A restriction on the number of mares bred and their overall fertility makes sense on commercial grounds and is more than likely driving the current sentiments on retaining a ban on AI. Why the difference in Thoroughbred versus Standardbred fertility? The Standardbreds used artificial insemination (AI) and that should have the helped the pregnancy rate. When we looked at the number of cycles and the time of breeding cessation we found that Standardbred breeders bred longer and for more cycles and thus the overall fertility was not different. The reason for the differences in pregnancy rate per cycle became more apparent when we examined the number of veterinary examinations per mare. Significantly more examinations were performed in Thoroughbred mares versus Standardbred mares. Why? Economics!!! Many of the Standardbred mares were bred on multiple occasions per cycle and probably with little or no relation to ovulation, but certainly with no data on fluid accumulation in the uterus (infection) that could only be obtained with ultrasonography. The Standardbred mares had a much higher percentage of non-cycling mares, which was actually due to failure to identify teasing mares rather than not cycling. Mostly, the lack of veterinary management was false economics because at the end of the season there were fewer mares in foal and the dates of conception were much later in the Standardbreds.

My point in the previous discussion is that it is the improvement in management (both veterinary and farm) that results in more mares in foal at the end of the season, but that because there are no selection procedures for fertility perhaps we are getting mares in foal that normally would not have. In this scenario there are mares with uterine abnormalities that need treatment and that with therapy they become pregnant and then later have problems, as they were unable to sustain the good uterine environment that is necessary to maintain pregnancy. A classic example of this is the older mare bred on foal heat. As many as 35% of mares over 15 years of age can be expected to loss their pregnancy by 45 days (after being positively diagnosed as pregnant on day 15).

We are pushing these mares and they are not always standing up to the pressure.

In the early 1990’s we reviewed the veterinary literature on pregnancy loss and found that most losses were associated with infection and that most were bacterial in origin. Many people are aware that infectious viral abortion caused by a Herpes virus can cause devastating outbreaks of abortion in intensively managed horses (management again), yet this is insignificant as a cause of wastage compared to the losses of early embryonic death and routine abortion losses. Another interesting observation we noted was that most mares undergoing early embryonic death had uterine fluid (inflammation) present when they returned to heat (yet it wasn’t present when losses were first diagnosed) and the cells types in the uterus demonstrated by biopsy were of the chronic inflammatory type (lymphocytes). These findings led us to believe that probably the most important cause of pregnancy losses was associated with uterine inflammation/infection and that the reason theses mares were susceptible was related to our pushing them to maintain an inter-foaling interval of one year. In other words we are pushing mares too hard to get pregnant early.

It was these philosophies that led us to the thoughts on antibiotics during pregnancy that are outlined below.

We decided to perform a study and treat potential cases of chronic uterine infection with antibiotics, monthly, during all of thepregnancy. Because not all EED and abortion is caused by bacterial infection/inflammation it was not expected that all mares would maintain pregnancy. The population we chose to treat had all undergone either EED or abortion at least once in the last season prior to entering the trial.

At the conclusion of the study 43 mares had entered the program, they had a mean age of 14.2 years and had EED or abortion an average of 2.2 times in the preceding years.

The results of the study were that 36 mares became pregnant (84%). Of the 36 pregnant mares 3 had EED (8.3%), 4 aborted (11.1%) and 29 had foals (80.6%).

A study such as this, although encouraging, must be interpreted with caution as there are no controls (i.e. a similar population of 43 mares that had no treatment). Despite this we are confidant that we are beginning to make some headway into preventing pregnancy losses.  

Another area of recent improvement is recognition and management of the high-risk pregnancy. These mares may have placental problems and present with premature lactation or problems may be inferred from mares with stresses such as colic surgery.

Commitment to research has resulted in many exciting discoveries that should continue and help us get more live foals each year and in addition more foals from difficult breeders.