How Can We Help?

Search for working dog conference recordings, articles on health, breeding, socialization and organizational management, webinar recordings and more.

Old Conference Archive Page | View all IWDC 2023 Recordings | View all Breeder's Workshop Colorado Recordings

Print

What are EBVs and how do they help?

What are EBVs?

Estimated Breeding values (EBVs) are a genetic selection tool we can use to identify which breeding dog has better genetic merit for certain desirable traits, compared to other dogs of its breed.  Dogs don’t always produce themselves (“breed true”) because what you see (phenotype) is a combination of the dog’s environment and genetics.

Most behavior and health traits are controlled by many genes (polygenic) and with each mating, the mix of genes is not exactly the same as either parent. Instead the mix of the two parents creates a unique combination of genes for that individual which changes the genetic potential for that individual. EBVs combine all of the data on all dogs and uses the pedigree relationships to predict the genetic merit much more accurately than what is observable.

The interesting thing about EBVs is that they incorporate data not just from the dog in question, but also its relatives and progeny. Nearly complete data on all or most dogs from multiple generations for a trait provides an accurate understanding of the real genetic merit of the dog and some idea about other related dogs (depending on the depth of data on the dogs and all close relatives). EBVs enable us to make more informed decisions that can result in real changes in the health and behavior of our dogs when used correctly over multiple generations, and produce puppies more ideal for your program (based on the traits you’re looking for – like improved hip health or certain temperamental traits). Here are some examples of how EBVs assist with this:

  • Selecting your next generation of breeding dogs: Which young adult offspring from litters available to you are most likely to produce improvement in certain traits. Keep as breeders the offspring that are more highly ranked than your current breeders.
  • Which breeding dogs to pair: Which mate maintains genetic diversity (e.g. limits the  rise of inbreeding) but also compliments the improved traits you are looking for.
  • Help guide you so you don’t “throw the baby out with the bath water”.  No dog is perfect.  Start with what you have, store the data and use it to select genetically better dogs in the next generation.  Establish breeding goals with like-minded breeders and together make data driven decisions to breed dogs that are healthy and meet your other breeding objectives through improvement with each generation. 

Estimated Breeding Values have been used in cattle since the 1950’s with great success, and have been implemented by a few guide dog organizations since the 1980s, For most others, EBVs for working dogs are a comparatively new concept but one made available to working dog organizations thanks to the International Working Dog Registry, which combines EBV specific data behind the scenes to calculate EBVs while keeping user and organizational private data secure.

Throughout this website and article you will find many more resources on EBVs, genetics and breeding, targeted specifically to dog breeders to learn about EBVs and understand how to use EBVs effectively.

Concepts of using data to determine risk for a disease is very common in human medicine. The terms used in the below video (“Polygenic Score”) are different, but the it is still helpful in explaining how data can identify genetic risk (which in humans is used to help identify risk of developing a disease and implement preventative care and early screening).

Video from the Broad Institute, available on Youtube: https://www.youtube.com/watch?v=HwEAjZWPw2k  

An EBV is a number

Just like the “Polygenic score” in the video above, an EBV is a number that is a prediction of an individual dog’s genetic merit for that trait. EBVs are calculated from whatever health and behavior data are available about that dog and all dogs of that breed which have data on that trait. In the International Working Dog Registry, EBVs are currently calculated in IWDR for Labrador Retrievers, Golden Retrievers and German Shepherds as those are the breeds most commonly entered into the database and therefore have sufficient health and behavior data available for EBV calculations.

The actual EBV is a number from 1 to -1 as seen in the example below. However, to make EBVs more user-friendly, the IWDR provides a percentile ranking of the dog, based on its raw EBV data, so you know at a glance where that dog stands in comparison to your other candidates and the other dogs of its breed. Breeding managers use the percentile ranking to identify which of the young dogs are among the top 20-25% of dogs available to be among the pool of dogs which are further evaluated as replacement breeders. Dogs with the higher percentile ranking are more likely to produce the desired trait than dogs with a lower percentile ranking. Genetic improvement only occurs when the next generation is better than the one before.

Below is an image showing some example EBV data.  You can see corresponding percentage rankings for each EBV. For example, dog Curtis has a Noise EBV of 0.35 which when compared with all other Labradors at that time, Curtis has one of the highest EBV values and is in the 95th percentile (e.g. top 5%) of his breed for the “Noise” trait.

Genetic Selection Methods

With that in mind, how do EBVs compare to other typical selection methods?

 

Phenotype Selection

The most common type of genetic selection is phenotype selection. This method involves studying phenotype data on the dogs under consideration, including the data you can gather on their relatives. Most of that data comes from:

  • searching online registries for health clearances and performance titles
  • speaking with other breeders/organizations
  • and observing dogs performing in behavior tests, work, shows or competitions.
  • You may also use tools to assess and preserve genetic diversity while practicing line-breeding to intensify desired phenotypes (for example, calculating inbreeding coefficients to maintain a level of genetic diversity during line-breeding).

There are significant risks when relying on phenotype selection: unfortunately, not all relatives are represented and often, registries only report information on unaffected (“Normal”) dogs. You may be forced to rely on second or third-hand subjective opinions on a dog’s performance or behavior. You may be subject to selection bias, where only some pieces of the whole story are made available for you. 

The most frustrating aspect of phenotype selection is that despite all of the cost and effort invested, breeders are still at the mercy of chance to systematically improve their dogs’ health, behavior, working ability and conformation simultaneously while maintaining genetic diversity in the population. With the ever-increasing number of genotypes being reported and other tools now available, even the most adept breeder studying the data of dogs and their relatives cannot adequately estimate the relative genetic value of the underlying genes and how they will interact across loci in the next generation.

For traits or breeds where EBVs are not available, phenotype data on all or most relatives can be useful.  However, EBVs are the most effective in systematically making genetic improvement.

 

Genetic Tests

The expression of some diseases and physical traits such as coat color are controlled by one or a few genes for which there are commercially available tests. Test results from a reliable laboratory are successfully used by breeders to avoid matings that have a high risk of producing affected progeny. Genetic test results are taken into account whether EBVs are used or not.

 

The 3 Principles of Genetic Selection

Breeders using EBVs must also consistently implement the 3 principles of genetic selection in order to maintain diversity and direct genetic improvement over generations of selection. By understanding the principles of genetic selection and by consistently applying these principles over generations, you can achieve a measurable increase in the quality of dogs in your breeding program.

Successfully Using EBVs

When used properly, EBVs have been proven to work   Key points to reliably improve the traits selected for are to:

  • Collaborate with others to collect data measured in the same way on all or most dogs for the most important traits
  • Enter your data and pedigrees through at least 5 generations to tie the dog’s together.
  • Consistently use the 3 principles of genetic selection by:
    • Choosing the next generation of breeders that are genetically better than the generation before.
    • Choosing mates which limit the rise in inbreeding (which maintains genetic diversity)
    • Limiting the number of progeny any individual dog produces

You will see improvement after 2-3 generations and can reliably improve the traits selected for. Breeding dogs no longer has to be the “heartbreak hobby” where breeders work so hard to make good decisions but are at the mercy of chance.   With EBVs, breeders can greatly increase their odds in the next generation being better than the one before.   Breeders still use all of their knowledge of the lines but have a powerful tool to really make a difference.

As we mentioned earlier, current EBV data in the IWDR is available for Labrador Retrievers, Golden Retrievers and German Shepherds. Over time, organizations have worked with the IWDR to define which traits they need to improve through genetic selection, and assessments have been made to confirm which traits are most heritable and therefore more useful to include in EBV calculations.   Below are some examples of genetic improvement through the use of EBVs. 

Genetic Improvement Using EBVs

The example below is from a breeding colony established in 2013 where many different organizations own the breeding dogs and all progeny have the same data collected – in the same way, therefore, consistently – for the desired traits.  The principles of genetic selection have been used with IWDR EBVs for 10 years (3 generations).  

The graph below shows the reduction in the percent incidence of dogs with the undesirable skin quality scores which include diagnoses such as atopy and chronic ear infections.  The first generation consisted mostly of breeders which had the more desired skin quality.  They produced some offspring that were like them, but also produced a higher percent of offspring with undesirable skin diagnoses.  Over the next 10 years (3 generations) of being able to use EBVs to identify the replacement breeders that were better than the existing breeders, the percent of dogs with the undesirable skin diagnoses declined.  Because they also used a selection index to “breed for the whole dog”, this collaborative breeding group also reduced the incidence of undesirable diagnoses in heart, elbow, hip, other skeletal, neurologic and eye problems while also making some progress or at least maintaining a low incidence of undesirable working behaviors. 

The box and whisker plot below displays the average EBVs by generation for this same colony for the skin quality EBV.  The blue box represents the middle 50% of the EBV’s for the dogs born in that generation class.  The line above represents the top 25% of the EBVs for dogs with the best genetic merit.  The line under the blue box is the 25% of EBVs for dogs with the lowest, least desirable genetic merit.  It is obvious to see that with each generation the bottom blue line gets shorter and also higher.  Selective breeding using EBVs and the principles of selection improved the overall skin quality in this breeding colony for all dogs, not just the parents.  

At the same time, the rate of increase of inbreeding was well controlled.  The last 2 generations are represented by only 73 and 33 dogs which may or may not accurately reflect the actual rate of rise of inbreeding in these recent generations.  However, it is important to check at least quarterly so adjustments in mate selection can be made to choose mates where the progeny will have lower inbreeding coefficients.

Additional Resources

Was this article helpful?
0 out of 5 stars
5 Stars 0%
4 Stars 0%
3 Stars 0%
2 Stars 0%
1 Stars 0%
5
Please Share Your Feedback
How Can We Improve This Article?
Table of Contents