Pedigree-based inbreeding: useful, but not the whole picture
The pedigree-based inbreeding coefficient remains a useful tool in purebred dog breeding.
It helps breeders interpret ancestry, compare potential matings and identify close relationships. But it does not measure DNA directly.
That distinction can have major consequences, particularly in breeds with small populations.
Understand its limits. See the bigger picture.
Key points
A single number can make genetic risk seem manageable. That is where the pedigree-based inbreeding coefficient can become misleading: it looks clear, measurable and reassuring, yet it does not show exactly which DNA a dog has inherited.
The problem begins when that number becomes the only basis for a decision. A calculation covering only a few generations can miss much older shared ancestry. Two matings with similar coefficients can also produce different genomic profiles, because each puppy inherits a unique combination of parental DNA segments.
The answer is not to abandon pedigrees. They remain essential for tracing bloodlines, identifying common ancestors and avoiding close matings. But they need to be considered alongside ROH, kinship between potential parents, mean kinship, the diversity available and the health of the bloodlines involved.
The key point is simple: a pedigree-based coefficient starts the assessment; it cannot complete it. Breeding decisions need a broader view.
Why pedigrees have played such a central role in dog breeding
For many years, breeders worked with the information available to them: pedigrees, bloodlines, breed type, observable qualities, the reputations of breeding dogs and the performance of their offspring. That was a reasonable approach, and it helped shape modern breeds.
But the genetic context has changed.
Breeds were not always strictly closed populations. Before standards became established and studbooks were progressively closed, outside dogs were sometimes introduced to strengthen an ability, address a weakness or reinforce a desired trait.
Formal recognition changed that context.
Standards became established and many studbooks closed. Selection then took place within a restricted population. This made ancestry easier to document and helped establish breed type, but it also imposed a constraint: the population largely had to rely on the variation contributed by its founders and early breeding dogs.
That is where difficulties can arise.
Existing variants may persist, become more common or disappear. Outside variation cannot be introduced without a route such as reopening the studbook, admitting eligible dogs through an initial-registration scheme, a managed cross or an outcrossing program. Selection sorts what already exists; it does not create new diversity (Frankham et al., 2010). Variants associated with desired breed type, coat colour, coat texture or other selected traits may be retained. Yet preserving those traits does not preserve all genetic diversity. Other variants can disappear while the breed’s appearance remains familiar.
Genetic drift continues across generations.
Some variants are lost by chance. Others become common because they are carried by popular sires or other heavily used breeding dogs.
Relatedness can increase until inbreeding reflects the population’s history as much as any recent mating.
This is the context in which pedigree-based coefficients are now used.
What the pedigree-based inbreeding coefficient measures
The pedigree inbreeding coefficient estimates the probability that an individual receives two identical copies of the same DNA segment because its parents share one or more common ancestors.
In other words, it starts with the genealogy, not the genome.
For a planned mating, this is the expected value for puppies from that parental combination. All puppies in the same litter therefore have the same pedigree inbreeding coefficient, because they share the same parents and pedigree.
But there is more to it.
Each puppy receives a different mosaic of parental segments. Its actual genomic inbreeding can therefore differ slightly from the pedigree expectation.
The calculation considers common ancestors and the paths of relatedness connecting them through the sire’s and dam’s ancestry. Modern methods allow these principles to be applied to large pedigrees (Meuwissen & Luo, 1992).
An inbreeding coefficient or an inbreeding percentage?
An inbreeding coefficient is a dimensionless value. In genetics, the correct mathematical term is the pedigree inbreeding coefficient: a value without units, often between 0 and 1.
It is often expressed as a percentage to make it easier to read.
For example:
A coefficient of 0.25 is equivalent to 25% inbreeding.
The measure is the same; only its presentation changes.
The important distinction:
The pedigree-based coefficient describes an expectation.
The genomic inbreeding coefficient shows the result.
Why pedigree-based coefficients still matter
Discarding the pedigree-based coefficient would be a mistake. It provides an initial warning of close matings, recurring ancestors, repeated bloodlines and obvious concentrations of ancestry. It also allows breeders to compare potential matings, provided the calculations use the same number of generations.
Pedigrees remain essential.
In a well-documented population, with ancestry traced back to the founders wherever the information is available, this is a valuable starting point.
It should not be the end of the assessment.
Problems arise when the coefficient is taken out of context, calculated over too few generations, or presented as a guarantee of genetic safety.
A pedigree-based coefficient informs a decision; it cannot make the decision on its own.
The built-in limitations of pedigree-based inbreeding estimates
The inbreeding pedigree ratio has its limitations. These limitations do not stem from a breeder’s error but from the very nature of the tool itself.
Here is a non-exhaustive list of the limitations of the pedigree-based inbreeding coefficient:
The result depends on how many generations are included.
A five-generation calculation does not include the same ancestry as one extending back to the founders. Common ancestors beyond the chosen limit are excluded, which can substantially underestimate inbreeding relative to a deeper pedigree.
In a closed population, relationships inherited from earlier generations do not disappear with time. Even when some common ancestors fall outside the pedigree depth being studied, their contribution to homozygosity within the population remains.
By ignoring the oldest common ancestors, a calculation limited to a few generations underestimates the probability that two alleles are identical by descent — precisely what Wright’s inbreeding coefficient measures.
Founders are often assumed to be unrelated and non-inbred.
This is a practical assumption, but it can conceal earlier relatedness in a breed established from a narrow genetic base.
Missing information is not evidence that no relationship existed.
It is an average probability.
A pedigree-based coefficient does not identify the DNA segments a dog actually received.
Recombination means that individuals with similar ancestry can inherit different segments (Knief et al., 2017).
The coefficient alone cannot distinguish the genomic profiles of two dogs with the same pedigree-based value.
It does not tell us which genes become homozygous.
Homozygosity can involve regions with no apparent effect, favorable variants, or harmful variants. The coefficient therefore cannot tell us whether a planned mating will have beneficial, neutral, or adverse consequences.
It is not a diagnosis of health.
A high coefficient does not mean disease is inevitable, and a low coefficient does not guarantee healthy puppies.
All puppies in the same litter have the same pedigree inbreeding coefficient.
Each puppy nevertheless inherits a different combination of parental segments.
The pedigree-based value does not describe each puppy’s realised genomic inbreeding. Differences between siblings can sometimes be substantial (Sams & Boyko, et al., 2019).
How relying on a pedigree coefficient alone can lead to poor breeding decisions
The danger isn't in using the inbreeding pedigree ratio—it's in using it on its own!
A mating can show a moderate rate over 5 generations, while still relying on an already homogenized population. In this case, the figure gives a false sense of security. But the diversity actually available may be low.
That is the first trap.
Two combinations can also display a similar rate, but produce different genomic profiles. One may concentrate long homozygous segments; the other may present more fragmented homozygosity (Kardos et al., 2015).
The overall figure does not show this difference.
Three common mistakes
• Underestimating inbreeding because the pedigree is incomplete.
• Comparing coefficients calculated over different numbers of generations.
• Assuming a low percentage guarantees that two dogs are genetically complementary.
The mistake is not consulting the pedigree coefficient. It is allowing that number to determine the entire decision.
Why small populations face particular difficulties
In a large population, with many available breeders and still abundant diversity, careful selection can more easily limit the increase in inbreeding.
In a small breeding population, The nature of the problem is changing fast.
The risk depends not only on the total number of dogs in the breed, but on its effective population size—that is, the number of breeding animals actually available to produce the next generation.
A breed can consist of several hundred individuals and, in practice, offer few useful options.
Why? Because some dogs are too old, too far away, unavailable, already related to the females intended for breeding, or largely irrelevant regarding health, temperament, and population goals.
This is where the constraint becomes concrete for breeders.
Even when a breeder agrees to look far, compare pedigrees, and avoid obvious combinations, the actual choice can remain very limited.
The process can then accelerate.
When births decline year after year, fewer potential future breeding animals are available. A fall in the number of active breeders also reduces the number of litters. The next generation then has even fewer options.
The consequence is direct: the risk of inbreeding increases, even when breeders sincerely try to avoid it.
In this context, genetic drift acts more strongly. Certain variants disappear because the dogs that carried them did not reproduce. Others become frequent because they are present in a significant proportion of the remaining available breeding dogs.
Diversity is then progressively reduced. Not through extreme choice, but due to a lack of real alternatives.
That is why a inbreeding pedigree rate calculated over just a few generations can be misleading. It may give a false sense of security when the population already shares a significant amount of common ancestry.
Therefore, the subject is not just the rate of a marriage.
It is the genetic structure of the population and the rate at which reproductive options are shrinking.
For a small-scale breed operation, thinking on a "mating pair by mating pair" basis is no longer sufficient.
We must take into account the mean kinship, the available diversity, underrepresented families, the number of truly usable breeding animals, birth trends, and the population’s ability to maintain variation over several generations.
This population-based analysis is all the more important given that inbreeding, morphology, body size, and selection history may interact with the health of canine races (Bannasch et al., 2021).
What a modern approach should bring togethere
A modern approach combines several sources of information.
Pedigrees remain necessary for tracing bloodlines, documenting ancestors and calculating expected kinship. Genomic measures add information about the DNA that is actually present.
One such measure is runs of homozygosity (ROH): homozygous segments showing where a dog’s two copies of DNA are identical along a given length of chromosome. ROH analysis is widely used to investigate patterns of homozygosity and the history of inbreeding (Meyermans et al., 2020; Sumreddee et al., 2020).
In dogs, long ROH deserve particular attention. Research has found substantial overlap between ROH and homozygous recessive disease genotypes, making these patterns useful alongside pedigree information (Sams & Boyko, 2019).
Genomic information also needs context.
A sound breeding decision should consider kinship between the parents, mean kinship within the population, known health in the bloodlines, the dogs’ overall quality, temperament and breed type, and the population’s longer-term needs.
Understanding what each measure tells us
Pedigree-based inbreeding coefficient
Inbreeding calculated from the pedigree, expressed as a coefficient or percentage. It gives the expected level of inbreeding in offspring from a particular mating.
What level of inbreeding does this litter’s pedigree predict?
Kinship coefficient
The probability that two alleles at the same locus, one sampled from each individual, are identical by descent.
How closely related are these two dogs?
Coefficient of relationship
The expected proportion of the genome shared by descent between two individuals.
What proportion of the genome do two dogs share by descent?
Mean kinship
An individual’s average kinship with the reference population.
Are this dog’s genes rare or already widely represented in its population?
Kinship between the parents determines the pedigree-based inbreeding expected in their offspring.
The table below presents some common situations and shows how the parents’ kinship coefficient translates into a coefficient of relationship and an expected inbreeding coefficient for their offspring.
The key point:
A litter’s pedigree-based inbreeding coefficient is equal to the kinship coefficient between its parents. In practice, the pedigree coefficient of genetic relatedness is often about twice as high as kinship coefficient.
Key point:
The genomic coefficient of relationship can nevertheless be higher or lower, because it measures the relationship actually realised in the DNA, influenced by genetic recombination — something the pedigree coefficient of relationship cannot take into account.
For example:
A brother and sister have a kinship coefficient of 0.25 and a pedigree relationship coefficient of approximately 0.50. Their litter’s expected pedigree-based inbreeding coefficient equals the parents’ kinship coefficient: 0.25, or 25% when expressed as a percentage.
What this means for SaarGen
For the Saarlooswolfdog, pedigree-based inbreeding cannot be considered separately from the wider population. The aim extends beyond producing litters that look less inbred on paper: it also involves restoring useful genetic diversity in a breed with a narrow genetic base.
That is a central concern for SaarGen.
An outcrossing program should aim for measurable improvements in available genetic diversity across generations, while remaining consistent with the breed’s identity.
The pedigree coefficient remains part of that monitoring, alongside genomic diversity, ROH, kinship, mean kinship, population structure, health, temperament, breed type and the actual contributions of breeding dogs.
SaarGen does not seek a single score that can summarise everything.
It seeks a more reliable basis for decisions: identifying dogs that contribute useful variation, avoiding rapid reconcentration of the same ancestry, tracking homozygous segments and keeping selection pressure compatible with long-term population viability.
Pedigree-based inbreeding therefore has an important place within a wider population strategy.
What to take away
The pedigree-based inbreeding coefficient remains useful. It helps reconstruct ancestry, compare possible matings, and identify certain risks of relatedness. However, it does not measure DNA directly. Its reliability depends on the number of generations analyzed, the completeness of the pedigree, and the assumptions made about the founders.
In small breeding populations, this limitation becomes critical. A seemingly reassuring coefficient can mask diversity that is already reduced, and two matings with comparable coefficients can produce different genomic profiles.
The benefit for the breeder is clear: no longer deciding on an isolated figure, but reasoning with a more complete vision of the population.
The pedigree provides an estimate.
DNA verifies.
Population management decides.
From ancestry to the genome.
Mini-glossary
Pedigree-based inbreeding coefficient
Inbreeding calculated from the pedigree and often expressed as a percentage. It describes the expected inbreeding in puppies from a particular pair of parents.
Genomic inbreeding coefficient
Inbreeding calculated from DNA and expressed as a percentage. It is generally measured from a saliva sample analyzed by a laboratory.
COI
COI stands for coefficient of inbreeding. The abbreviation is also widely used in French-language discussions of breeding.
Genomic inbreeding
Inbreeding observed or estimated directly from the genome, for example using SNP markers or ROH.
Open studbook
A breed register that allows eligible outside dogs or new breeding animals to be admitted under defined rules.
Closed studbook
A breed register in which recognised offspring must descend from dogs already admitted to the recognised breeding population.
Frequently Asked Questions
Why discuss inbreeding if the SCC prohibits close matings?
This distinction matters. It is one of the most common misunderstandings in discussions of breeding.
Breeders often use inbreeding to mean a close mating, such as father–daughter, full-sibling or half-sibling pairings. In population genetics, however, inbreeding also describes an individual’s genetic ancestry and patterns across a population. A dog can have a high coefficient without an obvious close mating in its recent pedigree, because it comes from a closed population with extensive recurring ancestry.
The Société Centrale Canine prohibits certain matings between close relatives, such as father × daughter, mother × son and brother × sister. These rules aim to prevent a sharp increase in inbreeding in the offspring.
That does not prevent inbreeding from increasing gradually within a closed population. Even when no prohibited mating takes place, shared ancestors may recur across generations, increasing the probability of identity by descent and contributing to homozygosity.
Avoiding the closest matings is therefore not enough to preserve a breed’s genetic diversity. Population management also needs to broaden genetic contributions, limit overrepresentation of particular bloodlines and monitor inbreeding across the whole breed.
“Without inbreeding, we lose breed type.”
That is a dangerous half-truth. Inbreeding has historically been used alongside selection to make certain traits more consistent. But establishing breed type does not mean diversity must continue to decline indefinitely. The aim is a population that remains viable, fertile, healthy and sufficiently diverse to endure.
“Inbreeding improves physical traits and temperament.”
It can fix certain traits more quickly, but it also fixes what accompanies those traits more quickly through linkage disequilibrium. That is the trap. Used without a population strategy, it reduces the scope for selection and increases the risk of concentrating harmful variants. Improving a breed means more than making its dogs more uniform.
“How can a breed be preserved without inbreeding?”
Inbreeding has historically helped breeders make some traits more uniform and predictable. That does not mean deliberately increasing relatedness must remain a permanent objective. Once breed type is established, further increases can reduce diversity without equivalent benefit. Selection should pursue the desired traits while limiting the rise in inbreeding.
“A low pedigree coefficient guarantees a good mating.”
No. A low coefficient may reflect a short or incomplete pedigree. It does not reveal which DNA segments will be homozygous or whether the parents are genetically complementary. A suitable mating also depends on health, breed type, temperament, kinship, available diversity and, where possible, genomic information.
“How can inbreeding exceed 25%? That is already the level expected from parent–offspring or sibling matings.”
Yes, it is entirely possible. The 25% figure corresponds to simple combinations in which the parents are not themselves inbred. In a closed or highly related population, several repeated common ancestors can have cumulative effects. The coefficient can therefore exceed 25%, sometimes by a considerable margin.
“A male with a COI of 20% and a female with a COI of 30% will produce a litter with a COI of 25%.”
No. The litter’s pedigree coefficient is not the average of its parents’ coefficients. It depends on kinship between the sire and dam: their common ancestors and the paths connecting them. Two highly inbred dogs from genuinely distinct lines can produce offspring with a lower coefficient than expected. Conversely, two dogs with low individual coefficients can produce highly inbred offspring if they are closely related to each other.
“Two different bloodlines guarantee low inbreeding.”
No. Lines that appear different in a short pedigree may share extensive older ancestry. Calculations should use the deepest, most complete pedigree available and be complemented by genomic data where possible.
“An inbreeding coefficient is a diagnosis of health.”
No. It indicates a probability of homozygosity and therefore a level of potential genetic risk. It cannot, on its own, predict whether a dog or litter will be healthy or unhealthy.
“An ancestor’s contribution declines over generations, so why talk about inbreeding building up?”
Both statements are correct, but they describe two different processes.
With each generation, the average proportion of DNA passed on by a given ancestor decreases by approximately half. In a closed population, however, descendants gradually accumulate many common ancestors. The homozygosity resulting from all these relationships does not disappear because an ancestor moves further back in the pedigree.
In other words, an individual ancestor’s genetic contribution decreases over time, but the consequences of all the common ancestors continue to shape the population’s genetic structure.
“Why is a five-generation coefficient often lower than one calculated back to the founders?”
Because a five-generation calculation leaves out shared ancestry beyond that boundary.
Yet these ancestors have not disappeared from the population’s genetic history. They still contributed to the formation of DNA segments identical by descent that are passed down through the generations. Excluding them from the calculation leaves some inherited inbreeding uncounted.
“Does that mean all of an old ancestor’s DNA is retained?”
No. A given ancestor’s average contribution decreases over successive generations through genetic recombination. However, if that ancestor appears many times in the pedigree — as is common in a closed population — its overall influence on genetic structure can remain substantial. It is the repetition of the same ancestors, rather than the mere presence of one distant individual, that leads to increasing homozygosity.
Why do people speak of inbreeding “building up”?
Because each generation inherits a population history, rather than starting from scratch.
Each new generation inherits the level of homozygosity already present in the population, to which new instances of identity by descent may be added when related individuals mate. Over time, a closed population therefore tends to accumulate homozygosity and lose genetic diversity, unless management strategies or outside genetic contributions change this dynamic.
Do genomic coefficients make pedigree-based coefficients obsolete?
No. The two approaches provide different kinds of information.
Pedigree-based coefficients estimate inbreeding from documented ancestry. They help researchers and breeders track population trends, compare breeding strategies and anticipate the effects of proposed matings.
Genomic coefficients, such as those provided by Embark, directly measure the homozygosity observed in an individual’s DNA. They reflect the proportion of the genome actually inherited from common ancestors, including those missing from or insufficiently represented in the pedigree.
The two approaches are therefore complementary. The pedigree describes the history of relationships, while genomics measures the biological outcome of that history. Used together, they provide a more complete view of genetic diversity and inbreeding than either method can provide alone.
Sources
Scientific references
The sources below informed this article.
Links to the relevant publications are included in the text.
1- Wright, S. (1922). Coefficients of Inbreeding and Relationship. The American Naturalist, 56(645), 330-338.
2- Meuwissen, T. H. E., & Luo, Z. (1992). Computing inbreeding coefficients in large populations. Genetics Selection Evolution, 24, 305-313. DOI : 10.1186/1297-9686-24-4-305.
3- Kardos, M., Luikart, G., & Allendorf, F. W. (2015). Measuring individual inbreeding in the age of genomics: marker-based measures are better than pedigrees. Heredity, 115(1), 63-72. DOI : 10.1038/hdy.2015.17.
4- Knief, U., Kempenaers, B., & Forstmeier, W. (2017). Meiotic recombination shapes precision of pedigree- and marker-based estimates of inbreeding. Heredity, 118(3), 239-248. DOI : 10.1038/hdy.2016.95.
5- Meyermans, R., Gorssen, W., Buys, N., et al. (2020). How to study runs of homozygosity using PLINK? A guide for analyzing medium density SNP data in livestock and pet species. BMC Genomics, 21, 94. DOI : 10.1186/s12864-020-6463-x.
6- Sams, A. J., & Boyko, A. R. (2019). Fine-Scale Resolution of Runs of Homozygosity Reveal Patterns of Inbreeding and Substantial Overlap with Recessive Disease Genotypes in Domestic Dogs. G3: Genes|Genomes|Genetics, 9(1), 117-123. DOI : 10.1534/g3.118.200836.
7- Sumreddee, P., Toghiani, S., Hay, E. H., Roberts, A. J., Aggrey, S. E., & Rekaya, R. (2020). Runs of homozygosity and analysis of inbreeding depression. Journal of Animal Science, 98(12), skaa361. DOI : 10.1093/jas/skaa361.
8- Bannasch, D., Famula, T., Donner, J., Anderson, H., Honkanen, L., Batcher, K., Safra, N., Thomasy, S., & Rebhun, R. (2021). The effect of inbreeding, body size and morphology on health in dog breeds. Canine Medicine and Genetics, 8(1), 12. DOI : 10.1186/s40575-021-00111-4.
9- Frankham, R., Ballou, J. D., & Briscoe, D. A. (2010). Introduction to Conservation Genetics. Cambridge University Press.