Estimated breeding values (EBV) describe the relative genetic value of each member in a breeding programme. They can be used for objectively choosing which animals are the best candidates for breeding to produce a better next generation. Livestock breeders have used EBVs for decades, so a great deal of research has gone into proving their efficacy, but dog breeders are fairly new to the game and are only just beginning to see their value.
Having said that, in the 1980's EBVs were put in place reduce the frequency of canine hip dysplasia (CHD) in several breeds. A breeding program against CHD in the Hovawart. introduced severe breeding restrictions in 1984 that banned all affected dogs from breeding. This actually made things worse, reducing the number of unaffected dogs and increasing the number of dogs classed as borderline.
In 1989, the restrictions were scrapped and selection based on EBVs was instituted, which produced immediate, significant improvement. In only 5 years more than 80% of all dogs had normal hips and severely afflicted animals were nearly eliminated (1).
Another amazing example happened with Boxers in Germany(1). Historically, German Boxers had a high incidence of cryptorchidism (one or both testicles fail to descend) because apparently, most living Boxers can trace their pedigrees to four German stud dogs - Sigurd von Dom and his three grandsons, Utz von Dom, Dorian von Marienhof, and Lustig von Dom, all of whom consistently produced cryptorchid offspring.
First efforts to reduce the frequency of cryptorchidism began as far back as 1942, with a total ban on breeding them (interesting that the Germans had time to think of doing this smack bang in the middle of, you know, a world war!). Nevertheless, the incidence of cryptorchidism increased over the next 40 years, The unification of Germany in 1984 widened the gene pool, but cryptorchidism increased unabated for the next 10 years.
In 1996, even stricter regulations on breeding were imposed. These measures reduced the frequency of cryptorchids for a time, but it also removed 84% of the reproductive dogs from the breeding pool and improvement tailed off after a few years.
Finally in 2000, Germany removed all restrictions related to cryptorchidism and instituted the use of EBVs. Within only 3 years they saw marked improvement according to a report published in 2003.
You can use EBVs for any trait you have information for. It's been used for hips, elbows, heart defects, cataracts, cancer, etc, as well as behaviour, intelligence, size, hunting traits... anything, really that has a genetic component but is too complex and multi-factorial to screen through a simple DNA test because EBVs account for environmental factors as well as heritability (2).
The key issue for Chihuahua breeders, however, is there are no official EBVs available for our breed.
They will come eventually, but in the meantime, we have to rely on the information in our own database. Although we can trace the lines of some of our Kitaco dogs back to the 1930s, in many cases, all we have is a name, and no further information.
These old breeders, and sadly, many active breeders today, consider breeding more of an art than a science. Many scoff at the notion of using science or mathematics to improve the health of our breed. We have struggled to get reliable health information about the dogs further up the line from our dogs and even been actively blocked from learning anything about the potential health issues that might exist in our lines that we know nothing about.
This is endlessly frustrating, but probably the inevitable by-product of "show breeding" where the purpose, in a competitive environment, is to keep up the illusion your breeding is flawless, and every dog you breed is a winner - hence the oft-heard and highly flawed "quality over quantity" boast - which one hears bandied about. This noble-sounding quote flies in the face of science-based breeding which encourages widening the gene pool, understands that in order to improve, you need to breed enough dogs so you can select to improve, and that sharing vital, health related information would help all breeders eliminate a recurring problem in the breed.
As an aside, any breeder who insists they have no health, temperament or construction problems in their lines is lying. If they are breeding perfect Chihuahuas, we'd have all heard about it by now, LOL
But back to the science. Bear with us, for a moment, while we get the mathematics out of the way.
In 2003, Eldin A. Leighton, Ph.D. the Jane H. Booker Director of Canine Genetics, presented a paper to the The Seeing Eye, Inc, in New Jersey, titled How To Use Estimated Breeding Values to Genetically Improve Dog Guides
It's an academic paper and pretty dry, also filled with wads of incomprehensible algebra and fun statistical models, but there is some excellent information that explains what it is that an EBV calculates.
Dr Leighton says "Phenotype is what can be seen and measured with respect to the physical trait... Each of these phenotypes is conceptually composed of two parts: a genetic component and a non-genetic component. The genetic component was passed down to each individual dog by its parents when a sperm and an egg united to form a new zygote. The non-genetic component includes the environment in which the dog was raised and all other aspects of each dog's own existence...
"... Very simply, then, an observation on a single animal can be conceptualised as :
Phenotypic observation = Environmental effects + Genetic effects + Residual effects
or, in terms of a statistical model, this can be written as:
So, thanks to Dr Leighton (and many other geneticists, it has to be said) we have our formula (try putting that into an Excel spreadsheet, BTW, without your brain exploding!) but what happens when you have no data to start calculating EBVs?
Well, we are fortunate that we have been able to confirm the health status for most of our dogs back at least 5 generations, either because we bred the dogs ourselves, have DNA evidence, know the breeders well enough that they have trusted us with the information, and in some cases, the owners are Chihuahua club members or regulars at Playgroup, and have provided us with the information, themselves. In this way, we have been able to painstakingly compile a list of health and phenotype information on our lines so we can build our very own Kitaco EBV database and start putting this powerful tool to use for the betterment of our breeding.
It's not enough to have just a score, though, and this is where it becomes so important that you know that your bitch's Great-Great-Great Grandfather had a terrible under-bite, for example (we're looking at you, Carlos!) that might pop up any old time in the future. If you don't know about it, you might inadvertently put her to another dog with a similar fault lurking hidden in his line, or worse, "line-breed" the fault back into existence.
So the score must also have a confidence factor, that tells us how sure we are of this information. In our calculations, we factor in up to 30 ancestors, eliminating one point for each ancestor's health and phenotype we are certain of (lower scores are better). There will come a time in the future when that score will be zero for all our dogs, but until then, the scores will be far higher than we'd like.
If you'd like to know more, Dr Carol Beuchat PhD has created a simple summary of how to use estimated breeding values, based on the interface used on Cornell University's EBV website. EBVs are now being used by kennel clubs around the world` - for example, for hips and elbows by the Finnish Kennel Club and UK KC, and for syringomyelia in Cavalier King Charles Spaniels by the UK KC.
To quote Dr Beauchat, "EBVs ARE coming to dog breeding, and they will be the most useful and powerful tool breeders have ever had to improve dogs through selection. Start learning about them so you can make the best use of them in your breeding program."