Population Genetics

by Lisa Van Loo

Charles Darwin, often called the "Father of Evolution", described how natural selection causes changes in populations over time, which creates evolution of species and types within a species. By selecting breeding partners in our domestic dogs, we are forcing evolution to follow the paths we dictate. This evolution by artificial selection has proceeded much more rapidly than natural evolution; most of the breeds of dog we know today were really only developed in the last 150 years or so.

Knowledge of population genetics and how it applies to purebreds, is a key to understanding the limitations of dog breeding. Population genetics is simply the study of how genes work within a population. First of all, we need to define the population we are working with. All domesticated dogs are of the same species. Chihuahua to Saint Bernard, they are all dogs!

The canine species has been sub-divided, so to speak, into many different breeds of dogs.

Breaking the population into smaller parts limits the gene pool (total number of genes available in a population) within each breed. This limiting happens when breeders select against certain traits — in effect, throwing away certain alleles (strands of DNA that code for a specific trait) while retaining other desirable alleles. In Chesapeakes the allele for the color black was selected against. Even though the allele for black color exists in the dog species as a whole, there are no black Chesapeakes because that particular allele has been eliminated from the breed’s gene pool.

This selection for some traits and against others has also unintentionally selected for some alleles that cause hereditary problems. This happens because genes do not occur alone. They are attached to chromosomes (long strands of DNA that contain many genes). These chromosomes can hold many hundreds of genes. Some of the genes code for good traits, others for undesirable ones. While selecting for a specific positive virtue, we may have been selecting for a negative trait that rides along on the same chromosome copy.

Founder Effect is the term used to describe what happens genetically within a population that is descended from one or a few common ancestors. These few ancestors appear many times in the background of the breed, usually because they had many of the traits breeders wanted and passed them on to their offspring. Unfortunately, there may have been bad genes associated with the desired ones. As the doubling-up on these few ancestors continues for many generations, the chance that these bad genes will find each other and cause hereditary problems to crop up increases.

Founder Effect is easily studied in populations of animals that occur on islands. Because they are physically cut off from others of their species, island populations become more inbred over time. Eventually, they can take on a whole new look and/or behavior than the original species. Some island populations become completely new species because of this difference and narrowing of the DNA. Darwin’s observation of island populations was what led him to begin formulating his ideas on evolution.

The different breeds of dogs can be thought of as separate islands in the ocean of canine DNA. Genes cannot be created; each breed can only use the genes that were present in the foundation animals. Because of the large number of different alleles available at most of the loci in the dog species (the "plastic" genotype of the species) the breeders were able to select for many different types of dogs. By keeping the traits they wanted and not breeding from dogs that had undesirable traits, breeders were able to create a vast array of different breeds.

Most breeds of dog are traceable to one or a few founding animals. Much crossing of the breeds was allowed in the early days of breed development but, through selection, many of the genes brought in through these crosses were lost. Many times, an individual dog is identified that holds many of the traits desired by breeders. This outstanding dog may be recognized early on and be bred from extensively. As time goes on, the exceptional qualities of the offspring and further descendants are recognized, and these then are bred to each other in an effort to concentrate the genes of the outstanding dog. This is known as popular sire effect, although many "popular dams" exist as well.

A dog can only pass on to its offspring the alleles that it has. Any stud dog will only have at most two different alleles at any locus, one from his sire and one from his dam. Popular sires frequently become popular because they consistently pass their traits on to their offspring. This is often because the popular sire is homozygous (both copies of an allele are the same, either dominant, or recessive) for alleles at many loci in his genome (total of all the alleles the dog has). This animal is sometimes called prepotent. In essence, he doesn’t have two different copies of the allele to pass on to his offspring, he only has one — the same allele was passed to him from both his father and his mother.

If this stud dog is used frequently, it will automatically reduce the number of alleles available in the population. Breeding "like to like" is a time-honored system; a number of the females bred to a popular male will be similar to him, both in looks and genetically, thus many of their alleles would be the same as the sire’s. In this way, the number of different alleles available in the gene pool becomes smaller still. As generations come down from the popular sire, much line breeding of his descendants may take place in an effort to "concentrate" his genes. This is how bottlenecks are created and diversity decreases in a population.

We often hear the term bottlenecking. A bottleneck occurs any time the gene pool for a population becomes very small. A bottleneck is a result of reducing the number of alleles available at each locus. These bottlenecks can occur for a number of reasons. The use of popular sires is only one factor diminishing a gene pool.

Another cause of bottlenecking is the reduction in total numbers of individuals of a breed. The few remaining members are bred together in an effort to keep the breed going. Much inbreeding is necessary in these small populations, and again, diversity is lost. We saw this with the Chesapeake breed immediately after the Civil War. Many kennels and individual dogs were lost in this period. There were very few purebred specimens of this breed left in the place of their origin. Most breeders resorted to crossing with other breeds to keep the Chesapeake alive. This resulted in a loss of type, as the "foreign" genes came into the breed. Did this lead to increased diversity in the breed? Not in the long term.

Breeders, wanting to regain the original type of the breed, began weeding out dogs that appeared to have foreign blood. Any dogs that showed evidence of cross breeding (hound or spaniel type, for instance) were excluded from the breeding population. By avoiding these undesirable genes, breeders gradually removed the "new" genes that had been brought in by cross breeding. Many of the genes that may have increased genetic diversity in the breed were lost in the process.

The Great Depression followed immediately by World War II led to further depletion of the breed, as many of the larger kennels were disbanded. The number of dogs available for breeding during this time was very small. Then, in the 1950’s America had a boom. Breeders could afford to use top quality stud dogs on their bitches. Airline travel became more reliable, and breeders could ship bitches. This meant that many bitches, from everywhere, were bred to a relatively few stud dogs.

We see with these examples that the Chesapeake population has bottlenecked on several occasions. Genes in a bottlenecked population do not flow. There are fewer genes to "pick" from, if you will. A very small gene pool will have very few genes to select from. A trait cannot be selected against if there is no substitute gene available in the population. In breeds with a small gene pool and limited choices at certain loci, this can create an ethical dilemma for the breeder. We are taught not to ever breed a dog that has a health problem, or a soundness or type fault. Reality, however, shows that in the breeds with restricted gene pools, the breeder must make some very hard decisions. Very few dogs would be able to pass all of the genetic tests, and of those, some would be carriers of the recessive defects. Not affected themselves, they can still pass the genes on to their offspring.

In an effort to keep track of recessive genes, as well as traits controlled by dominant genes, breeders will often use pedigree analysis. Pedigree analysis is simply studying the dog’s pedigree for known animals with shared traits, both good and bad, then deciding whether the dog has a chance of inheriting those traits.

Pedigree analysis can take many forms. Most people simply look at the animals actually listed on the pedigree. Three, four or five generation pedigrees are most often used. In a breed as geographically spread out as ours, with much mixing of the bloodlines, it is often difficult to analyze a pedigree in this way, as there may be many dogs in the pedigree that are not known to the breeder. Pedigree analysis only works to the extent that the breeder has useful knowledge about the dogs in the pedigree. This, in turn, depends on open communication among breeders. Without complete information, use of the pedigree analysis method to select breeding stock is faulty.

In the case of a recessive defect, like the prcd form of PRA, pedigree analysis alone is not successful, as nobody knows how many dogs have had PRA. Many Chesapeakes are gun dogs or pets. If they develop PRA, it may not be noticed until the dog is very old. Their Vet may just assume it is blindness due to old age, and never make a true diagnosis. Others may be diagnosed correctly, but the owners do not let anyone (the dog’s breeder or the ACC) know. Others may adopt the "shoot, shovel and shut up" policy, never telling anybody if a problem crops up in their line. Without complete information, it is nearly impossible to tell how many dogs have had this devastating disease, or which lines may be affected.

The only way in which OptiGen results may be used in analysis is on an individual basis. For instance, one bitch was not tested, but a dog she was bred to was a Pattern A. A puppy from this mating was tested and found to be a Pattern B. This meant that the dam, had she been tested, would be either Pattern B or Pattern C. Another puppy from this mating had a Pattern A result. This narrowed the possible results for the dam. If she were tested, she would test Pattern B, as a Pattern C cannot have Pattern A offspring. Now, it is not really necessary to test this particular bitch, because her result is assumed to be Pattern B. Using this type of analysis, a breeder can decide which animals in a kennel will need to be tested.


Population Genetics and Genetic Testing

The question has been raised about the frequency of the prcd gene in the breed. This question is one that cannot be answered at this time. First of all, the false allele situation exists. It is not possible at this time to determine which dogs have a false allele. Even when the gene-specific test is developed, however, the data generated would not apply to the breed as a whole, unless many more dogs were tested. Here’s why.

At present, approximately 250 Chesapeakes have been tested. This seems like a lot, but in looking at the sires and dams of tested dogs whose information has been publicly released, a pattern emerges. Most of the dogs tested up to this date can be grouped into four breeder groups. These would include all dogs bred by a particular breeder that are related to each other, as well as dogs sired by that breeder’s stud dogs. Among these four groups, there is some crossing over; for example, Breeder A may use Breeder B’s stud dog, while Breeder C has sold puppies to Breeders A & D. A fifth group, although bred by several different breeders, consists of many closely-related individuals. When the dogs from these groups are removed from the data set, only about 20 dogs remain which are not closely related to dogs in the other five groups.

What causes this? Whenever a new genetic test comes out, there usually is a small group of dedicated individuals who will test their dogs. When the tests come back with good results, other breeders with related dogs will then test theirs, as the chance exists that their dogs will also clear. Thus, whenever a new test is brought out, we find that many of the early dogs that are tested are related to each other. This was true for early OFA and CERF numbers in Chesapeake Bay Retrievers as well.

These test results do not reflect frequencies in the Chesapeake gene pool as a whole, as the demographics of the tested dogs (sires, dams, half-siblings, etc) do not reflect the demographics for the breed as a whole. For instance, only four breeder groups account for over 90% of dogs with OptiGen results. When we look at AKC Stud Books, however, we note that these four breeder groups are not responsible for 90% of all Chesapeakes!

Rather than try to interpret genetic test results for the whole Chesapeake population, breeders should use the results to determine breeding partners on an individual basis. Selecting of mates based on a whole-dog approach, rather than one trait or test result is the best approach. With the new genetic testing being developed today, we now have an opportunity to shape canine evolution in newer, more positive ways than ever before.


© 2000 Lisa Van Loo