Dog Breeds and Where They Actually Come From

Dogs have a word for this that cats mostly do not, and the word is the whole argument. Landrace means a population that was never closed off from others, that breeds freely within the area it lives in, and whose characteristics developed because the animals were surviving there rather than because anyone selected them toward a written standard. The alternative, a breed, is a sub-population with a standard and a studbook.

The reason that distinction is worth an article is that both sides of the “adopt, don’t shop from a breeder’s list” argument tend to blur it. Enthusiasts list breeds by country of origin as though geography were a breed characteristic, and shelter advocates call everything with ancient ancestry a breed. Both mistakes come from the same place, which is treating an administrative category as a natural one.

What a registry actually requires

The Fédération Cynologique Internationale runs the most explicit recognition procedure of the major registries, and reading it is instructive because every criterion is countable and none of them is a map.

Recognition requires a DNA test, membership of one of eight groups, a minimum of fifteen years of prior national registration, and an effective population size above a stated threshold. The application must come from the national canine organisation of the country of origin of the breed.

That last one is the phrase people misread, so it is worth being precise about. The anchor country is the body that owns and writes the standard, not the place the dogs live. The cleanest disproof is the Chow Chow’s own FCI standard, which reads origin China, patronage Great Britain. Two different countries in one document. The anchor country is whoever maintains the standard.

So when a registry says a breed is of a given origin, it means a national body in that country administered the standard. It does not mean the breed is found there and nowhere else, and it certainly does not mean the dogs are still there.

What “primitive” and “aboriginal” actually mean, and where the words come from

The American Kennel Club’s own account is the most honest available, and it starts by disowning the vocabulary. There are varied terms used to refer to ancient dogs, including primitive, aboriginal, village and landrace, and these words have different implications regarding ancestry and evolution. The page then concedes that what we call primitive dogs might be more accurately referred to as aboriginal.

It is worth being clear about what that page does and does not argue, because it is often cited as though it were a rejection of breeds. It does not reject the recognised-breed frame. It defends it, defining the purebred as a closed population and pointing to a dedicated society for definitions. What it concedes is narrower and more interesting: that today’s breeds are no longer exactly the same as the aboriginal village dogs their ancestors were hand-picked from, and it gives the Samoyed as a case where today’s standard is narrower than the original population.

That is a real concession with a real consequence. A “primitive breed” is a modern reconstruction of something older, and the reconstruction is the part with a studbook.

A note on sourcing, because this topic is worse served by sources than most. The Animal Health Foundation page that circulates on this subject is a reprint of a syndicated newspaper column, credited on the page to a syndication agency, with the Foundation as the reposting site. It contains no foundation-authored definitions at all. The definitions themselves come from the Primitive and Aboriginal Dog Society, whose site was down when checked. If you want the actual definitions rather than a newspaper’s summary of them, that body is the place, and the AKC cites it.

What the genetics actually shows, and it is not what the breed lists assume

This is the part that should change how you read a “native dog breeds of ” list, and it comes from a peer-reviewed study of native East Asian breeds rather than from anyone’s summary.

The study sequenced newly obtained genomes from five named Korean and Chinese breeds alongside public reference data, and its findings point consistently away from the closed-breed picture:

East Asian indigenous dogs are not genetically isolated. That is the study’s own phrasing.

Differentiation between them is very low. The F statistic between southern Chinese and Vietnamese indigenous dogs is 0.016, and between the two Korean breeds and southern Chinese indigenous dogs 0.051 and 0.090. The authors read the low differentiation among the native Korean breeds as suggesting gene flow between them, or that they were never isolated.

The inbreeding coefficients run backwards from what a closed breed predicts. The lowest median inbreeding coefficients of any breed studied were for the Korean breeds, at 0.080 and 0.095, where Chinese and European breed dogs ran between 0.289 and 0.519. A closed studbook produces high inbreeding. These populations show the opposite.

One of them carries the most European admixture of any breed in the study. The Sapsaree shows 40.4% West Eurasian ancestry, indicating ancient admixture of European ancestry, and divided by the measured haplotype sharing, the highest of any breed in the dataset.

The authors’ own caveat should travel with the finding: per-population sample sizes were small, between 3 and 22 animals, averaging under six, and they flag the lack of ancient dog DNA and limited sample size as limitations. So this is evidence of gene flow, not a map of ancestry.

What it does mean is that a named “native breed” with a closed studbook, developed from a population that never stopped interbreeding with its neighbours, is a description of the last two hundred years rather than the last ten thousand. That is not a criticism of the breed. It is a correction about what the word is describing.

One specific claim, checked

“The Siberian Husky is native to Siberia and found nowhere else.” The provenance half is misleading and the exclusivity half is not supported.

The modern breed was developed in Alaska in the early 1900s from dogs imported from Siberian peoples, and it is registered and standard-administered by American bodies. A 2025 genetics study sampled 237 Siberian Huskies that were all AKC-registered in the United States, and found substantial structure within the breed separating sled, show and pet lines. The breed as it exists today is an American population, in the specific sense that the sampled dogs are.

The ancestry claim is the interesting half and it survives. The Siberian Husky forms its own clade in the East Asian genetic structure, separate from other Asian breeds, and it shares a substantial steppe-related ancestry signal with the Sapsaree. So the animals do descend from a distinct Arctic lineage. “Found nowhere else” is contradicted by the breed’s own existence as a recognised breed in a country that is not Siberia.

What a landrace is, and why it is not a lesser thing

The descriptions in circulation are approximate rather than formal, and the one that carries the most weight is the second. A pariah type is a free-breeding dog foraging in human settlement, described across a wide band from India through Southeast Asia. A landrace is a population adapted to a particular locale or culture, whose characteristics developed more in response to survival in a particular environment than to human design.

That last clause is the distinction and it is not a ranking. These animals are doing the thing a standard is trying to preserve, without a standard, and they have been doing it for far longer than any registry has existed. The genetics above is consistent with exactly that: gene flow, low differentiation, and lower measured inbreeding than the closed populations.

The practical consequence is that the phrase people sometimes use, “what breeds are not man made”, has an honest answer that is more interesting than the one usually given. It is not that some dogs are natural and some are artificial. It is that the artificial ones are all recent, and the recentness is a choice made by people, and the ancestry underneath it is not.

What this has nothing to do with

Nothing here is an argument about which dog to have, and no line of it should be read as one.

Breed origin does not establish temperament. No registry document, and no study cited above, connects a breed’s country of origin to behaviour, health in an individual dog, or suitability for a household. A landrace dog from a particular region is a population with a history, not a temperament specification.

Do not use any of this to justify a purchase or a refusal. The most common misuse of regional breed content is as a shopping filter, in either direction.

What we still do not know

How much any named breed diverged from its source population, and when. For breeds with recent written standards this is documented and recent. For breeds claimed to be ancient, nobody can date the narrowing, and “ancient” in breed writing usually means nineteenth century.

What most of the world’s dogs actually are. The genetic work above is regional and well-sampled by regional standards. The global picture of free-breeding dog populations is poorly described, and the absence of a study is not evidence of a similarity.

How much local adaptation is genetic and how much is learned. A landrace characteristic can be inherited, can be passed on culturally by the people who keep the dogs, or both, and separating those is hard.

Every factual claim above traces to one of the sources listed below. I have flagged the two places where the widely cited source is weaker than its reputation, the syndicated column and the dog-society website that was unavailable, rather than passing them on as though they were the registry’s own words. The framings I have added are the emphasis on the anchor country being about who writes the standard, and the point that these are all recent constructions.

I am not a veterinarian and I have never kept a landrace dog. This is general information about registries, genetics and history, and the authoritative version of any of it is the registry’s own current rulebook and the original study.