for decades there was an apparent paradox in horse racing. The sport is lucrative (Mage, this year’s Kentucky Derby winner, earned his owner $1.9 million) and the fastest horse wins. Horses with good results and a good pedigree are used as breeding stock for the next generation. Horse breeders were armed with a lot of data, a single trait to optimize, and strong incentives to do so. Yet several studies have suggested that despite their best efforts, race times were not improving.
The most common explanation was that, physiologically speaking, it was always more difficult to breed a horse that could run faster than existing horses already do. The modern Thoroughbred racehorse dates back at least three centuries. Perhaps the years of selective breeding had already discovered and exploited almost all the genetic potential of the breeds.
This made no sense to Patrick Sharman, a racing enthusiast and geneticist at the University of Exeter, England. After all, cattle breeding has been going on for hundreds of years, yet it continues to create cows that produce more milk. Artificial selection applied to chickens is still breeding plumper birds. It would be strange, he thought, if racehorses were the one domesticated animal that humans could no longer improve. Then, along with Alastair Wilson, who had once been his Phd supervisor, started digging.
Their first paper was published in 2015 and looked at a much larger dataset of British breeds dating back to the 1800s than other papers. He found that, contrary to accepted wisdom, horses actually got faster. In sprint races, those covered five to seven furlongs (1-1.4 km), the average speed needed to win has increased by about 0.1 percent every year since 1997. Their latest paper, published May 27 in Inheritance, try to gauge how much of that improvement is attributable to genetics. In other words, is the time-, energy- and money-intensive profession of horse breeding worth it?
The answer seems to be yes, although less so than farmers might like. By linking a large performance database, containing nearly 700,000 race times recorded in Britain between 1995 and 2014, to a family tree of over 76,000 horsepower, they found that speed is heritable, albeit weakly, and that breeding he is improving it, but slowly.
The drive is more pronounced for sprints and middle-distance races (812 furlongs). Drs Sharman and Wilson conclude that about 12% of the variation in horse speed at these distances comes down to genetics. (This is roughly the same heritability as neuroticism, extraversion, or lifespan in humans.) And they found that the improvements to that genetics accounted for more than half of the speed increase observed over that time period. . The rest, Dr. Sharman says, is likely due to non-genetic factors like better nutrition or veterinary care for better maneuvering technique.
When it comes to long-distance racing, it’s not clear if the times are improving. One reason, says Dr. Sharman, it could be that genes that are good for sprinting don’t necessarily make good endurance athletes. It appears that breeders select for sprint performance because it offers faster commercial returns. Sprinters tend to start racing around the age of two, long-distance horses at three.
Horse breeders can also face other trade-offs. Selecting solely for speed can increase your risk of injury. (Churchill Downs Racecourse, where the Kentucky Derby is run, suspended racing for a month from June 7, after more than a dozen horses had died of injuries in the past six weeks). Temperament also matters: a fast horse is of little use if it is not rideable.
Despite the difficulties, there is also evidence that breeders may be leaving some potency in the gene pool. At least in Britain, says Dr. Sharman, breeders still rely, to some extent, on their professional judgment when evaluating horses. Less intuitive, more objective statistical techniques have transformed other sports, most famously baseball, over the past two decades. Even horse racing may be ripe for its Moneyball moment.
#study #asks #racehorses #reached #genetic #peak