The Thoroughbred gene pool: the natural tendency to select Thoroughbreds based on racing performance can have unfortunate consequences for the breed as a whole

Citation metadata

Author: Deb Bennett
Date: May 2015
From: EQUUS(Issue 452)
Publisher: Active Interest Media
Document Type: Article
Length: 4,351 words

Document controls

Main content

Full Text: 

By numerous measures, the Thoroughbred is the world's most important kind of horse: by the huge amount of money involved in Thoroughbred racing and breeding; by the fact that Thoroughbreds form part of the foundation stock for numerous derivative breeds; and because historically they have been the premier "extreme athletes" of the equine world. Thoroughbreds function as "sport horses" in the diminishing number of "marathon" steeplechase, timber, and point-to-point races that are the last remaining vestiges of the "stayer" contests upon which the breed was founded more than three centuries ago. Thoroughbreds are the majority breed in the hunter-jumper world and in three-day eventing and are highly influential in dressage--Olympic disciplines that engender enormous popular and media interest.

About 23,000 Thoroughbred foals are registered annually in North America. Australia ranks second with just under 14,000 foals a year. Add to that the tens of thousands of foals produced elsewhere--the United Kingdom, Japan, New Zealand, Saudi Arabia, Germany, France, Argentina--and it's obvious that only a fraction of the Thoroughbred population will run in the 2,000 races open to the breed held each year. And of those only a tiny number ever win, place or show. In previous installments of this series, we highlighted the contribution of King Charles II of England, who invented standardized performance testing, which is, in biologist's language, what the rules for racing actually are. Although it is true that between about 1750 and 1875 the rules, and thus the conditions for winning, changed greatly, flat-track horse racing remains a highly effective form of performance testing.

Such testing is, however, a two-edged sword. On the one hand, it efficiently identifies the horse best suited--in terms of both physique and physiology --to succeed under the mandated conditions. On the other hand, if winners are bred only to winners--and remember that there are very few winners--it creates what biologists call a "genetic bottleneck." In racing, speedy horses are by definition the winners. Thus horses less endowed with speed but replete with other worthy and necessary qualities--particularly stamina, balanced conformation, disease resistance, and soundness--will inevitably be weeded out. If the lion's share of breedings are awarded to stallion "A," then the particular genome of "A" will spread like wildfire through the population. "A" thus becomes the bottleneck, so that all subsequent generations begin with the same genome as "A."

In Thoroughbred racing, raw speed rather than "bottom" or stayer capability has become the supreme goal of the breeder. Whereas flat-track racing in the time of Charles II began by requiring horses to run multiple heats totaling up to 16 miles in a single day, within the first century most races were only three to four miles long. In the 1770s, "futurities" open to horses that had not yet reached physical maturity were inaugurated at distances of 1.5 to 2.5 miles. By 1850, the concept of "classic distance" racing was firmly embedded at from one to 1.5 miles so that today, less than 5 percent of all races worldwide require Thoroughbred horses to cover distances greater than 1.5 miles (12 furlongs). The most common racing distance worldwide is 1.25 miles (10 furlongs), but in the United States its only one mile (eight furlongs). In America after World War II, "sprint distance" races burgeoned; contests of less than one mile--a mere four to six furlongs--are now a close second in popularity to the classic-distance races.

In nature, selection for a single characteristic such as speed never occurs. Even under the strongest "selection pressure," i.e., extremely high death rates such as in albino animals who are more visible to predators and thus easily picked off, there are always mitigating factors that allow some albinos to survive. It is entirely possible for an albino animal to possess greater agility, intelligence or disease resistance than its normally pigmented relatives, for example. The same applies in racing: The horse with a high number of lifetime starts, who does not break down from congenital weakness, who retires sound, and yet who (most of the time) finishes somewhere back in the pack has much higher value than his win/ loss record signifies.

These facts all derive from a branch of biology called "population studies," which aim to clarify the effects of natural selection not upon the individual animal--who either succumbs to or escapes disease or predation--but upon the population it belongs to. Although every individual possesses a genome, it is the population as a whole that possesses the gene pool--that is, the entire range of alleles or different genetic "options" for any given characteristic. Thus it is to the population that we must look in order to know how broad a range of genetic options is available. Very briefly, here are concepts from population biology that impact horse breeding:

* Founder effect. The smaller the population at the time a wild species becomes isolated upon an island--or at the time for a horse breed when its registry book becomes "closed"--the less total genetic variability and the fewer the allelic options will be available from that point forward. The Thoroughbred has the advantage of having been created through crossbreeding unrelated individuals who were primarily Hobbies, Barbs or Turkmenes. However, the total number of founding individuals was relatively small, almost certainly including fewer than 500 unrelated broodmares.

A study done by researcher E.P. Cunningham and colleagues at Trinity College in Dublin, Ireland, examined relatedness in Thoroughbred horses both by pedigree study and by genetic sampling. They compared Thoroughbreds (TB) with a population of Arabian-related horses from Egypt (EG) and with Middle Eastern Turkmenes (TU). TBs surprised the research team because their diversity score was actually higher than that from either EG or TU--but this is because the Hobby contribution to the foundation of the Thoroughbred was not considered.

Population biologists usually begin by assuming that populations of wild animals isolated upon an island are panmictic, which means that statistically every animal of breeding age has equal access to mates and equal likelihood of mating with any member of the opposite sex. In domestic breeding this is rarely the case, and the unequal contribution of founding animals has enormous consequences in subsequent generations. The Trinity College researchers found that the 10 TB founders with the highest contribution were responsible for 45 percent of the variability in the 1990 sample. The top 20 founders contributed 65 percent of genes, and the top 30 founders 78 percent. All other identified founders taken together accounted for only 3 percent additional variability. The most influential founding sires were the Godolphin "Arabian" (13.8 percent), the Darley "Arabian" (6.5 percent), the Curwen Bay Barb (4.2 percent), the Ruby Mare (4.2 percent) and the ByerleyTurk (3.3 percent). Significantly, the total and relative contributions of the top 10 founders have been practically identical since 1790. This exemplifies the principle that relative founder contributions quickly become stabilized in closed populations and subsequently tend to remain stable. The effective number of founders becomes a constant after only a few generations.

* Selection pressure. In the wild, selection pressure is the total combined effect of disease, predation or any other factor that causes individuals to be less able to breed. Individuals less well adapted to the environment in which they live--weaker or slower, for example--will die younger and at higher rates, and over time their unique allelic makeup will drop in frequency within the population to near-zero or even become entirely extinct. Among domestic racehorses, selection pressure is entirely artificial, but it is not simply the result of whatever the racing rulebook requires; it is also the product of human ethics. "In earlier eras," writes Thoroughbred researcher Ellen Parker, "most people bred horses to race them, and they had a stake in the animals' soundness. By contrast, modern commercial producers breed horses in order to sell them, and if those horses are unsound, they become somebody else's problem."

Today the individual who breaks down after several successful races--or one who looks likely to break down--is often quietly whisked off to stud or nursery, there to perpetuate speed but at the same time whatever weaknesses he or she may possess. The lesson we must derive from this is that, severe as the "law of nature" may sound, selection pressure is a good thing because it acts to strengthen a breed--but only so long as it is allowed to come into force. When drugs that mask conformational weaknesses or that artificially bolster athletic prowess are allowed at the track, and when mares or stallions with known defects are bred for profit alone, the overall effect is to weaken rather than strengthen the Thoroughbred horse.

* Directional selective pressure. This occurs when extremes of phenotype" or athletic ability become the ideal. In horse as well as dog breeding, the ideal or "adapted" individual has high economic value, and thus among domestic animals, selective pressure is often directed by economic factors, including greed. Have modern Thoroughbred breeders found the "perfect" phenotype, the horse perfectly suited to win at classic distances, or the one perfectly suited to win at five furlongs? Is such a goal actually achievable, or is it instead a dangerous chimera?

What characterizes the "adapted" flat-track racer? Pedigree researcher Anne Peters notes that selection for the single characteristic of raw speed is unwise. Besides speed, winners possess other crucial characteristics, the first being congenital soundness or "toughness." Such horses "recover quickly from a race and run back to form without excess coddling." Champions also exhibit the desire to win, and a kind of regal presence; as longtime Canadian trainer Bill Phillips expressed it to me, "Winners have the attitude that it offends them when another horse tries to pass." Peters notes, "Gameness and courage have won as many races as sheer physical ability."

The ability to "quicken" is another crucial characteristic that is not the same as either raw speed or stamina, but the ability to accelerate when called upon at the key strategic moment in a race. "In longer races," Peters observes, "efficiency of stride becomes more important, but [even in longer races] the ability to quicken remains a more desirable asset than the ability of a horse to gallop others into the ground. [Such] tireless but non-brilliant horses ... are referred to as 'plodders.'"

Ellen Parker writes, "Speed is absolutely necessary to the Thoroughbred horse ... but speed alone with no thought of bottom or soundness or good bone can lead to only one place. Whatever quality the horse possesses is of little use if it ends in a Ruffian-like tragedy."

The production of great horses--and the very survival of the Thoroughbred as a breed--requires that they be valued for many qualities other than raw speed. Although the extreme athlete may become a goal in breeding domestic animals, in the wild it never occurs because selective conditions always change over time. Under shifting selective conditions, it is the generalist rather than the overspecialized bloodline that survives. While environmental change may be either slow or abrupt, Nature is almost infinitely variable--as contrasted to the unchanging, two-dimensional world of the racing rulebook.

* Genetic bottlenecks. In nature these are typically the result of a disease epidemic, such as tularemia among rabbits or West Nile virus among red-tailed hawks, which kills off a very high number of susceptible individuals, leaving only a few from which population numbers can rebound. Among domestic animals, one of the most famous examples of bottlenecking concerns the Irish Wolfhound, which had declined to only eight purebred dogs by the early 19th century and was thereafter reconstituted from those dogs plus some part-related Scottish Deerhounds and Greyhounds. The Wolfhound today remains a valuable and interesting type of dog, but it is most decidedly not the same animal known from antiquity, which had such stamina that it could outpace a stag in flight and such fierceness that it could bring it down by eviscerating it on the run.

The decline in the Wolfhound was due to exactly the same factor that promotes genetic bottlenecking in Thoroughbred horses: fashion--which is

nothing other than the very human tendency to try to imitate success. In the last 200 years of "modern" classic-distance racing, in actuality there have been dozens of bloodlines capable of winning, but pedigree studies clearly show that breeders have repeatedly flocked to only certain sires. Ellen Parker notes, "It's no secret to those who examine pedigrees that sire lines rise and fall. There are a number of reasons for this....However, when this swing occurs--say to Bold Ruler, then to Northern Dancer, then to Mr. Prospector and so forth--we should not be so enthusiastic in embracing the new line that there is nothing left of the valuable old ones."

According to the Trinity College study, thanks to repeated swings in fashion that have created sequential bottlenecks, 78 percent of alleles in the current Thoroughbred population can be traced to just 30 foundation animals, 27 of them male. Ten foundation mares account for 72 percent of maternal (tail-female) lineages, and a single stallion (Eclipse) appears in 95 percent of tail-male lineages. While numerous stallions were involved in founding the Thoroughbred, by the mid-1800s only three remained, each linked to the modern population by a single male ancestor (in parentheses): the Byerley Turk (Herod), the Godolphin "Arabian" (Matchem) and the Darley "Arabian" (Eclipse). The Wolfhound nearly became extinct because small, soft-coated, gentle little dogs had become fashionable. In a similar way, Thoroughbreds descended from great sires--horses with both speed and bottom, such as Princequillo, Ben Brush, Man o'War, Citation, Hindoo, Hyperion and Son-in-Law--are now all but extinct upon the racetrack, whereas those descended from a single early 20th century sire, Phalaris, now hold complete world domination with--as the Jockey Club Review notes --"no end in sight."

Line breeding and inbreeding, in short, lead directly to loss of genetic variability. The current deleterious trend would have proceeded much faster were it not for the live-cover rule that governs Thoroughbred breeding. A Thoroughbred foal can be registered only if it is conceived by live cover. Originally, this rule was intended to help guarantee accuracy, so that the foal's parentage would be correctly recorded. Today, it has the highly beneficial side effect of reducing the number of foals that any one stallion can produce. With modern management techniques, a successful racehorse can sire up to about 250 foals per year, and while this is a much higher number than in earlier times, it is nothing compared to what can be done with dog or horse breeds that allow artificial insemination (AI); a single popular dog sire can in a breeding career of perhaps three years produce more than 1,000 litters averaging six pups apiece. The most popular sires in horse breeds that allow AI can have an equally overwhelming influence since, although foals are usually produced as singletons, a stallion's breeding career may extend for more than 15 years.

* Genetic load. While inbreeding "doubles up" on desirable alleles, it also multiplies the number of undesirable ones. Inbreeding among mammal species is rare in nature, occurring only where the population is geographically confined, for example, isolated upon an island. Among island populations of horses, or herds confined to a given range, every stallion capable of breeding covers as many undefended mares as he can find and commonly tries to further broaden his reach by stealing mares from weaker herd sires. Feral stallions very rarely cover their own daughters, and as their sons reach reproductive age, they drive them away. For these reasons we do not see hydrocephalus[??], "wobbler[??] syndrome," offset knees, fractured cannon bones, chronic splints, fractures of the sesamoids, ruptured tendons or breakdowns due to small feet or thin, shelly hoof wall in 2-year-old mustangs, despite the fact that these immature animals periodically run miles over rocky, hilly ground.

Likewise, among the Turkmene tribes that still breed the sire-line ancestors of the Thoroughbred--the Yomuds, Karabairs, Akhals and Tekes--these problems are not seen. Neither are they seen in the often-overworked equine youngstock of other Middle Eastern nations, including Egypt. In the Trinity College study, the TB population showed significantly lower diversity than EG or TU. The number of shared alleles within EG and TU was less than would be expected under random breeding, a finding which suggests that breeders have avoided inbreeding.

The TB also showed some avoidance of inbreeding, but less than EG or TU. The bottom line is that practical-minded, often illiterate and impoverished Middle Eastern tribesmen certainly are not blind to conformational weaknesses in their herds and are aware that inbreeding weakens a bloodline. They eliminate rather than perpetuate weak individuals, and they do this through their own forms of performance testing, i.e., informal race meets and other contests requiring speed, stamina and soundness. Congenital unsoundness is thus NOT an unavoidable side effect of breeding for any single desirable factor, whether that be speed or beauty, size or color, but rather comes from breeding blindly, carelessly or for the sole purpose of economic gain.

This article features more than a dozen historically important stallions descended from Eclipse through the Irish horse Birdcatcher, born in 1833, comprising only the older bloodlines in this family. As noted in previous installments in this series, horses descended in sire line from Herod and Matchem are now seldom seen on the flat track, and a similar situation has evolved with the horses surveyed here. Next month, we will complete our study of Thoroughbred families with a look at the single largest and currently wildly popular branch of Thoroughbreds--horses descended from Phalaris.

Deb Bennett, PhD

Coming next: Preserving the "Classic Thoroughbred"


A simple survey of only three top-class classic distance races is indicative of an alarming trend in flat-track racing worldwide. Although, as the chart shows, the longer the race, the higher the number of career starts (the San Juan Capistrano Invitational Handicap is currently the longest flat-track race in the United States), the average number of career starts has declined.


Data in these charts comes from the Trinity College genetic study led by E. P. Cunningham ("Microsatellite diversity, pedigree relatedness and the contributions of founder lineages to Thoroughbred horses," Animal Genetics, December 2001). When change in selective conditions is overlaid (gray and yellow zones), the reason for the switch in popularity of Eclipse- vs. Matchem- and Herod-line sires becomes clear. Note the 30- to 50-year lag time. Fashions for mares are much less clear-cut than those for stallions, and even the most popular dam (Tregonwell's Natural Barb mare) holds a frequency of only 16 percent. This indicates that allelic variation is highest in mare lines, perhaps because breeders have continued to try to "outcross" by using a wide variety of mares.


The Birdcatcher families or sire-lines are charted here; the following sidebars expand the Isinglass and Teddy families. (The huge Cyllene/Phalaris family will be covered next month.) Those who own Thoroughbreds bred for steeplechasing, jumping or other "sport horse" divisions will find many of their horses' ancestors here. Note the thickly-muscled, "rounded" look of most of these individuals. They tend to be somewhat long and soft in the back and "stand over a lot of ground," measuring longer from breast to point of buttock than their withers height. Isinglass is the exception, and even from this very early example we learn the importance of having something other than Eclipse blood in the pedigree. Isinglass takes after his dam's dams, who are descended (i.e., in "tail female") from Herod.

Isinglass 1890 (by Isonomy, out of Dead Lock, a Touchstoneline mare). The Touchstone line was highlighted in our last installment.

Bend Or 1877 (by Doncaster out of Rouge Rose, by Thormanby, a Herod-line horse). Note the broad, rather soft back.

Teddy 1913 (by Ajax out of Rondeau, by Bay Ronald). This thickly-muscled horse stands over a little less bone than would be ideal. The tendency toward loss of bone, as well as loss of visible withers height, grows with increasing Eclipse blood.


This is the last family that typically produces Thoroughbreds with angular, "framey" conformation. They are muscular yet balanced. This is also the first Birdcatcher family to contain stallions with chronic soundness problems who were nonetheless bred because of their capability for speed.

Star Shoot 1898 (by Isinglass out of Astrology, by a Touchstone-bred sire but carrying a good deal of Herod). This horse represents a nice blend of sprinter muscularity while yet retaining level balance, visible "frame" and excellent bone.

Sir Barton 1916 (by Star Shoot out of Lady Sterling, a granddaughter of the great Herod-line champion Hindoo, and tracing to Matchem in tail-female). The first winner of the Triple Crown and possessed of a wonderful pedigree, this horse should have epitomized the ideal of speed combined with toughness. Unfortunately, he suffered from chronic hoof cracks, perhaps a factor in his 1920 loss in a famous match race against Man o' War. Forgotten thereafter by the public, he spent most of the next decade at Fort Robinson, Nebraska, as a U.S. Army remount sire, and near the end of his life on a Wyoming ranch where he sired "usin' hosses."

Swynford 1907 (by John O'Gaunt out of Canterbury Pilgrim, a Touchstone-bred mare tracing to Herod in tail-female). Though a lovely horse with a pleasant temperament, and like Sir Barton possessed of a goodly amount of non-Eclipse blood through his dam, Swynford was retired to stud when he broke a fetlock at 4 years of age, having started only 12 times.

Gallorette 1942 (by Challenger II out of Gallette, a Teddy-family mare with close-up St. Simon, tracing directly to Regulus and Flying Childers several times without going through Eclipse in tail-female). I present this mare as the best representative of her bloodline. Gallorette beat almost all the colts who challenged her, including the brilliant Stymie. She exemplifies ideal Thoroughbred conformation and raced sound through the end of her 6-year-old year, compiling an "iron horse" record of 72 starts in which she won or placed 54 times.


Blenheim II 1927 (a grandson of Swynford by Blandford, out of Malva, double-bred to St. Simon and tracing to Herod and Tregonwell's Natural Barb mare in tail-female). There are some very nice points to this horse, but like many close relatives, he suffered from weak ankles and was retired to stud after rupturing a tendon as a 4 year old with only 10 career starts. At stud he was successful primarily because his wealthy owners provided him the best mares, including the fabulous Mumtaz Mahal, upon whom he got Mumtaz Begum. Other progeny include Whirlaway (out of a Ben Brush mare) and Jet Pilot (out of a mare tracing to Teddy and Ben Brush).

Mahmoud 1933 (by Blenheim II out of Mah Mahal, a daughter of Mumtaz Mahal by Gainsborough, a Bay Ronald grandson). Mahmoud had a somewhat checkered career yet set a speed record when he won the Epsom Derby that was not broken for another 59 years. He retired sound as a 4 year old after only 11 starts. A succession of wealthy breeders and owners guaranteed him access to very strong mares, and thus Mahmoud's progeny include many winners.

Mumtaz Mahal 1921 (by The Tetrarch, a Herod-line stallion, out of Lady Josephine by Sundridge, a Vedette-line stallion, but with multiple crosses to Lexington). This mare is definitely not of the Birdcatcher family but is presented as an example of the balanced, flat-muscled, "framey" phenotype highly beneficial to them in terms of soundness. Compare her conformation to that of Gallorette. Though possessed of an archetypal "stayer" pedigree, this mare was offered only 10 starts, most at "sprint" distances. She was indeed very fast and is an ancestress of both Bold Ruler and Secretariat, who will be discussed in our next installment.


In the Teddy family we begin to observe some individuals who are, in terms of both conformation and athletic capabilities, closely similar to Quarter Horses.

Sir Gallahad III 1920 (by Teddy out of Plucky Liege, a Touchstone-line mare with both Matchem and Herod in tail-female). This stallion shows the "rounded" conformation typical of Teddy-line horses but with enough bone to support the mass above. A full brother to Bull Dog, he sired no less than three Kentucky Derby winners: Gallant Fox (1927), Gallahadion (1937) and Hoop Jr. (1942).

Bull Lea 1935 (by Bull Dog out of Rose Leaves, a Touchstone-line mare that also traces to Eclipse in tail-female In this horse we get the first taste of what "Eclipse on Eclipse" looks like: a massively muscular body above, fortunately supported by excellent bone below. These horses resemble Quarter Horses not only in muscularity but because they stand relatively short on the leg and have noticeable "downhill" balance. Retired sound to stud at age 5 after a respectable 27 starts, Bull Lea went on to sire 58 stakes winners.

Citation 1945 (by Bull Lea out of Hydroplane, a Hyperion-Bay Ronald line mare with multiple crosses to Lexington and Herod in the tail-female). This is the smart way to breed Eclipse--by guaranteeing that there is enough "stayer" blood in the mare to preserve a balance between the weight of the body above and the construction and toughness of the limbs below. Enough positives simply cannot be uttered about this horse, who has been voted one of the top 10 flat-track racers of all time. Citation compiled 32 wins in 45 career starts. A Triple Crown winner, he continued racing until age 6 and was one of only three horses to win 16 straight races. He was the first racehorse to earn $1 million.

Damascus 1964 (by Sword Dancer out of Kerala by My Babu, going back to Herod in tail-female). Closely related to Citation, Damascus is reminiscent of him both conformationally and in the fact that he too had a highly successful race career, earning more than $1 million. He won 21 of 32 career starts including the Preakness and Belmont Stakes. Retired to stud at age 4 after bowing a tendon, he did better than Citation at stud, siring 71 stakes winners and many champion-producing mares.

Source Citation

Source Citation   

Gale Document Number: GALE|A411614797