For some time now we have been of the opinion that there has been over selection of the paternal side of the pedigree and under selection of the maternal side of the pedigree in thoroughbred breeding.
That is, with large books and better veterinary practice, the bar to become a stallion has probably moved up a little too high (you are just about required to be a GI winner now to have any chance) while because of the same quest for large stallion books, the bar for a mare to enter the breeding shed has slipped to the extraordinary low (basically two functional ovaries gets you in).
We have done a lot of our own statistical research on this and found that there is a significant advantage in buying yearlings out of mares that had some race performance over others that had none and we have been advising our clients/buying for ourselves for some time now to buy yearlings out of mares that at least showed some talent on the racetrack.
Why would we do this? Shouldn't the Sales Select test that we undertake to select yearlings be able to tell if the foal has inherited the characteristics of elite performance regardless of the performance of the dam?
Yes, but no....
Yes we can certainly measure some of the variants that are associated with performance but no, we cannot measure all of the variants and we certainly can't measure epigenetic effects (the environmental impact on the genes), although measuring the cardio does pick some of this effect up. This inability to measure all variation results in what is known as missing heritability, that is, there is a portion of performance that is passed on from the parents to the foal that we miss. Put another way, we are able to measure the genetic changes in the foreground that effect performance, but the combined effect of the genetic changes in the background we miss and they have an impact on the outcome.
By selecting yearlings out of mares that could run themselves we believed that we were somewhat mitigating the effect of missing heritability. A new paper released this week on the Importance of Maternal Line in Breeding from Dr Allan Davie and his group at Southern Cross University and in China has come to some similar conclusions with our previous research.
Using a database of 675 racehorses the group established that the heritability of performance between dams and their foals (r=0.141) was significantly higher than between sires and their foals (r=0.035). They further found that a good stallion could generally not negate the heredity effect of a poor race mare but that a good race mare could negate the hereditary effect of a poor sire.
This finding isn't unexpected. Given that the performance level in terms of race ability of the stallions becomes more homozygous, the effect of that performance reduces. They note in the paper that in the group of horses they used, 89.5% of the stallions performed in Group races while only 29.9% of the dams did.
One thing that must be made clear is that the paper is specifically talking about the racing class of the dam of the yearling itself. One of the more frustrating aspects of the way that a catalog is designed is that it overly values the more distant maternal line. Genetically speaking, there is not much difference between the broodmare sire of the yearling and the second dam of the yearling, yet the race performance and offspring of the second dam is detailed for all to see and value and the broodmare sire of the yearling is just a single name on the page. There is very little, if any, value in considering the performance of the third dam of a yearling. While it is often a case of convenience (they can't afford the 'good race mare') many breeders and yearling buyers significantly over estimate the importance of the 'family' while discounting the importance of the race performance of the mare herself.
The interesting aspect in our research on this is that the mares themselves don't actually have to be a stakes performer to offer significant value in terms of their potential to produce a superior runner. We use the Class Performance Index (CPI) as a proxy for race performance as it is measured across all racing jurisdictions and allows us to look at catalogues were the dams may have performed in different countries in a more standardized fashion. After completing a study on some 400,000 CPI values and establishing a normal distribution and 'frequency bins', we believe that buying yearlings out of mares with a CPI of 2.4 or better offers significant advantage. You could use Timeform, Beyer, or any other rating to work out what level of performance is needed, but it invariably points to horses that are of 'good' standard, being those able to win a race at Saratoga, Flemington or Newmarket is enough of an advantage.
Another interesting aspect of the paper was that they found an effect of age of mare on the subsequent outcome of the foals with foals born out of mares that were ages 8-11 having the largest effect on the racing outcome. This is somewhat parallel to our research where we believe that the first 6 foals out of the mare are significantly the best and offer a good selection technique for yearling sales.