Inbreeding and genetic diversity in horses
Two horses can look very different and still share ancestors. Their foal may receive two copies of a genetic version descended from the same ancestral copy. Understanding inbreeding helps assess a planned mating; maintaining genetic diversity also means considering which families contribute to future generations.
How relatedness, inbreeding and genetic diversity differ
Relatedness describes the connections between two horses that share ancestors. Inbreeding concerns an individual whose parents are related. At a given DNA location, a foal receives one version from its dam and another from its sire. These versions, called alleles, may descend from the same copy in a shared ancestor.
Genetic diversity describes the range of genetic versions and combinations within a population. It cannot be inferred simply from the number of coat colors or breeds represented. Horses of the same color can be distantly related, while horses that look different can come from closely related families.
Why inbreeding can increase risks associated with recessive variants
Related parents are more likely to carry the same variant inherited from a shared ancestor. Inbreeding therefore increases the probability that their foal receives two copies. When a harmful variant causes a recessive disease in horses with two copies, this combination can produce disease. Inbreeding does not create that variant.
Increased homozygosity, meaning more DNA locations with two identical versions, can also expose the effects of several harmful variants. Inbreeding depression is a population-level decline in the average value of certain traits, such as reproduction or the ability to perform an activity. A Thoroughbred study, for example, linked genomic inbreeding to the probability of racing. This finding does not predict an individual horse’s health or fertility, or establish the same effect in every breed.
What the inbreeding coefficient F measures
In a pedigree, F estimates the probability that the two copies at a genetic location are identical by descent: both trace to the same copy in an ancestor shared by the parents. The estimate relates to the ancestry and reference population used in the calculation. An F of 12.5% does not mean 12.5% faulty genes or a 12.5% chance of disease.
A future foal’s F depends on the relationship between its parents. Two breeding horses whose own calculated F is zero can be full or half siblings and produce an inbred foal. A parent’s coefficient is not passed to the foal like an allele. The mating must be assessed, rather than adding or averaging the two parental coefficients.
Example: a half-sister and half-brother can each have an F of zero
The mare and stallion in this example share a sire but have different dams. Their own F is zero in the pedigree considered, yet their relationship gives their foal a nonzero coefficient.
Mare, daughter of this stallion
Each parent’s F
0%
Stallion, son of this stallion
Each parent’s F
0%
This half-sister and half-brother have a foal together.
Foal from this mating
Expected F of the foal
12.5%
Assumptions: the shared ancestor and both dams are not inbred; these three founders are unrelated to each other, and there are no additional relationships in this simplified pedigree.
The 12.5% describes expected identity by descent in this example. It is not a probability of disease, and this diagram is not a mating recommendation.
Sources [1]
Why an incomplete pedigree can underestimate inbreeding
Pedigree calculations can identify only the connections recorded in the data. A missing branch may hide a shared ancestor. A zero coefficient then means that no inbreeding was detected within that pedigree and its assumptions; it does not prove the absence of older relatedness.
To compare results, check the number of generations considered, the proportion of known ancestors, possible parentage errors and the reference population. A five-generation calculation and one using a much deeper pedigree can give different results for the same horse. A lower value obtained from less information does not demonstrate a better breeding choice.
Sources [1]
What genomic analyses can add to a pedigree
Analyses of many DNA markers can identify regions where the two chromosomes contain long stretches of identical versions. These are called runs of homozygosity, or ROH. The proportion of the genome covered by these stretches provides a measure of genomic inbreeding; long segments can, in particular, indicate recent relatedness.
This measure and pedigree F use different information and should not be treated as interchangeable. The markers studied, thresholds and method affect the result. A small panel of disease or color tests does not measure diversity across the whole genome. Even a genomic analysis cannot diagnose every disease.
How heavy use of a few stallions can reduce population diversity
A population can contain many horses yet depend on a small number of families. When a few stallions produce a large share of offspring, their genetic versions become heavily represented and less-used origins can disappear. In a global Thoroughbred study, researchers observed increasing genomic inbreeding and declining diversity associated with influential sire lines.
Maintaining diversity therefore requires considering family contributions and changing relatedness over several generations, beyond one mating. Among horses whose abilities and health fit the breeding program, using a wider range of origins can limit concentration. An imported horse or a different bloodline name does not necessarily mean low relatedness to horses already in the population.
Managing a recessive variant while maintaining genetic diversity
A low inbreeding coefficient does not prevent two horses from carrying the same recessive variant. Equally, a high coefficient does not identify which variant is present. Targeted tests on both breeding horses and analysis of their relatedness therefore answer complementary questions.
For HERDA, an unaffected carrier of the tested variant mated to a horse confirmed without that variant cannot produce a foal with two copies of it, but can pass on its copy. In a program suited to this disease, this option can avoid the affected combination while retaining a genetic origin. It requires follow-up of offspring and does not automatically apply to dominant diseases. Excluding all carriers of every variant without considering the population can also remove diversity.
Assessing a mating for both the foal and the wider population
Start with information on both breeding horses: health, family history, identity, DNA results and pedigree. Compare possible matings with the same method and a similar depth of information. The foal’s expected coefficient helps assess parental relatedness, while disease tests clarify some identified risks.
The decision also involves breeding goals, welfare and the representation of families. A veterinarian and, where appropriate, a population genetics specialist can help interpret this information. The site’s coat calculator calculates neither inbreeding nor an overall health risk for the foal.
Questions about inbreeding and diversity
Does a zero inbreeding coefficient guarantee a healthy foal?
No. It means that no identity by descent is estimated within the data and assumptions used. Ancestors may be missing, and two horses can carry the same disease-causing variant without detected relatedness. The coefficient cannot replace targeted tests or a health assessment.
Does mating horses from two different breeds eliminate all genetic risks?
No. Different breed names do not prove the absence of shared ancestors or variants. A cross can change diversity, but its value and risks need to be assessed using both horses’ data and the program’s goals.
Does a negative DNA panel prove high genetic diversity?
No. A panel looks for a limited list of variants. By itself, it does not measure relatedness between two horses, runs of homozygosity across their genomes or diversity within their population.
Is there a universally risk-free inbreeding percentage?
A single number cannot guarantee a horse’s or foal’s health. Its interpretation depends on the method, data, population and variants involved. Management goals need to suit the breeding program and also monitor how inbreeding changes over time.
Sources on inbreeding and genetic diversity
The explanations and examples draw on IFCE and university resources and research in Thoroughbreds. Findings from these studies are not individual predictions or values that apply to every breed.
- IFCE — Inbreeding coefficients, pedigree depth and mating examples
- IFCE — Genetic diversity and contributions of different origins
- University of Minnesota — Mendelian inheritance and recessive variants
- Hill et al. (2022) — Genomic inbreeding and the probability of racing in Thoroughbreds
- McGivney et al. (2020) — Genomic inbreeding and influential lines in the global Thoroughbred population
- UC Davis VGL — HERDA and interpretation of genetic results