Monday, June 21, 2021

A look at the genetic relations in groups of dogs

What do we really know about free living groups of dogs? Published work by the Swiss animal behaviourist, Rudolph Schenkel, of his wolf studies in the 1930s and 1940s gave rise to the idea of pack leaders and this idea is still referred to all too often today. His work was based on captive wolves in a zoo and has since been recognized, including by Schenkel, as not giving a picture of free-living wolves or dogs. More recently claims have been made that a “pack” consists of the mother and her pups and that only the “alpha female and alpha male breed”. This is then also picked up and claimed to be the same with free living dogs. Free living dogs do not have rules they follow and there has not been any long-term study of many dog groups, individuals or their movements done over several generations to give a clear picture and it is unlikely that one size fits all. From genetic studies we do know, however, that in general freeborn dogs have a lower coefficient of inbreeding (COI) than do human controlled breeds suggesting movement between groups is common. There is of course an occasional exception when a parent and sibling or 2 siblings have mated but his seems to be rare and may be a result of movements having been restricted when dogs find themselves caught in a built-up city area. One of the dogs in the group initially tested, Aswad, was later shown by Embark to have a close relative, Ramy. On checking it was confirmed that they had both came into the same shelter together and that the size difference suggested they might be a father and son. From the genetics they share 28% similarity so it seems they were half siblings from separate litters, with their haplotypes suggesting they probably had the same father but different mothers. Genetic results of other dogs living in a group or close by groups show interesting connections as follows. Of the dogs tested none of the dogs in the same group as Pack leader, so named because of the role the people feeding these dogs observed him to be, showed as his pups. He was related to Figa, Hannah and Finn at a level of half siblings, aunts/uncles or grandparents, and to Jimmy and Jenny as first cousins and yet Hannah and Figa were living together in a separate area to Pack leader. Hannah and Figa shared their genetics (40%) at a level where they were probably half siblings or possibly aunts/uncles or grandparents. They have the same maternal haplotype so likely had the same mother. Hannah and Finn could also be half siblings of Pack leader but did not have the same mother as Pack leader since Pack leaders maternal haplotype was different. Jimmy and Jenny could be as related to Pack leader as first cousins and share 43% of their genetic make up and have the same maternal haplotype that none of the other dogs had, so probably had the same mother. They were in a separate group. Most closely related were Jenny and Jimmy at 43% and with the same mother as they both have the same maternal haplotype. Hannah and Figa at 40% and again with shared maternal haplotype so same mother but different to Jenny and Jimmy. Dasha and Dave with 31% and again the same mother as they have the same maternal haplotypes but different to the other siblings. Lara and Luna at 22% and again share the same maternal haplotypes so had the same mother that none of the others had. Figures shown are the percentage of DNA shared between connected dogs.
Hannah and Figa share 40% of their genes so are probably half siblings and share the same maternal haplotype A437 so probably the same mother.
Jenny and Jimmy share 43% of their genetic make up so are probably full siblings and share the same maternal haplotype C34 so probably the same mother. They were from a different group to Hannah Figa and Pack leader.
Lara and Luna share 22% of their genes as well as the same maternal haplotype so probably have the same mother but different fathers and were not family relatives of other dogs tested. Savvy and Sandy from Dahran were thought to be siblings or half sibling but only share 17% DNA, a level equivalent to first cousins. They do share the same maternal haplotype suggesting their mothers may have been siblings.

Wednesday, July 1, 2020

Where did all the German Shepherd genes come from?

Almost every “village dog” shows a trace of German Shepherd in its DNA. A range of other “breeds” also show up infrequently including Saluki, Basenji, Doberman Pinscher, Pomeranian, Boxer, Collie, Harrier, Labrador Retriever, Dalmatian,  and others. Considering that all of todays human established breeds originally come from the early natural dogs or by mixing other breeds to get a type wanted it is not surprising to find some shared DNA and the more pariahs that are tested the more are likely to be found. Some also have the same haplogroups and haplotypes  that are found in some “pedigree Canaan dogs”. It would be surprising if that was not the case since the Canaan breed is simply a small group, or family, established by taking in some 30 of these pariahs and breeding them for sale then excluding the rest. “The “breed” can be distinguished in much the same way as a human family that has bred in a closed community could be identified from other humans in the same geographic area. Most dogs in todays breeds are as closely related as second cousins. It is rare to find a dog with no genetic structure linking it to other breeds.

When “pedigree” dogs or unknown dogs that turn out to fit the genetic profile of the “pure” breed are tested then any trace breed found does not get reported. When a new “pariah” is added to the “Canaan breed” kennel clubs only require that the offspring breed true for 3 generations with existing registered dogs. That is equivalent to going back as far as great grandparents, each of  which would show up as 12.5% of the dogs genetic makeup. Once a new pariah is introduced to the breeding program and its offspring “breed true” to and is registered and its DNA added to the breed profile then all the related dogs will of course also test as being “pure Canaan”. If the relative finder on Embark tested 100% pure Canaan dogs is used GSD’s do show up as here:
So how do we explain the presence of this connection.
From Grace Gartel: “Since we first started testing dogs at Embark (and even before then, during our village dog research), the Embark team noticed that most village dogs across the globe share some German Shepherd Dog DNA (when it is less than 5% of their DNA, they are a "trace breed"). To be honest, we're not yet sure exactly why this is, but it definitely is a real biological phenomenon and has been noted by other labs (see for example
The average amount of relatedness varies across the globe, and the exact amount varies between village dogs from the same region. This could relate to German Shepherd Dogs traveling around the world in the World Wars and other times, but we are not certain of that at this time. As we work to understand village dog genetics and history better, we hope to unravel this mystery!”
The suggestion that GSD’s were introduced by military into the area seems unlikely to explain this. Many other dog breeds have also been introduced and there is no reason to believe GSD’s could have bred so widely with pariahs worldwide that close to all pariahs even in remote areas now have some GSD in them. A more likely explanation would be that it is simply a continuation of genetic strains dating back to ancient times and has remained as a stable pattern. 

4 of 67 village dogs from this region had a trace of Gray wolf in their genetic makeup, 2 from Israel, 1 from Jordan and 1 from Kuwait. No doubt there are others not tested and again this is further evidence of the ancient origin of our desert dogs.
Some potential genetic health issues have been found in pariahs. Apart from the more common degenerative myelopathy (DM), which has become a major issue in the pedigree breed due to inbreeding, others already spotted in lower numbers in the pariahs include von Willebrands type II , factor VII deficiency (a blood clotting deficiency) and “Collie eye” Without fully testing any new pariah intended to be added to the breed BEFORE using them breeders could well be creating future problems. Thousands of desert dog pariahs still survive but, sadly, efforts to eliminate them by “rescuing” them or by Trap Neuter and Release (TNR) or worse,  to shoot or poison them if “successful” could lead to the end of these natural dogs while we still have much to learn from them.

Sunday, January 26, 2020

Dogs of ancient Egypt

Dogs have lived in Egypt for thousands of years and there are many artistic depictions of them. These include both carvings and paintings. Thousands of mummified dogs were placed in one tomb that has been known for some time but only recently properly studied. Many of these dogs may have been killed as part of a ritual but some had received better and more careful treatment and were likely to have been loved pets.

A number of distinctly different breeds were recognised. Rosellini attempted to put together a collection of the various dog types found and produced the following 2 plates. Most modern popular breeds have existed for a mere few hundred years but many people try to link them to Egyptian dogs for prestigious reasons but without factual proof. Today’s Pharaoh hound for example has been proven to be a modern mix designed to resemble the original. Africa’s Basenji is more likely to have descended from Egyptian types. Sighthounds were reserved for the elite people and their descendants have continued until today. Today’s Baladi dog in Egypt is another likely survivor from ancient times and is probably related to the desert pariah found throughout the middle-east and which has probably followed Bedouin and earlier hunter gatherers where ever they moved, throughout the Arabian peninsula and perhaps the ancient fertile crescent where man settled and grew crops. These desert dogs have survived in harsh conditions often without help from man and are the natural stock of today’s Canaan breed selected from these pariahs. The popular and intelligent (one has been shown to know over a thousand words) border collie, with known origins in the English/Scottish border area has also been claimed by some to have desert pariah connections.

Monday, January 6, 2020

Genetics of registered Canaan dogs versus pariahs.

Here is an interesting illustration of just what the difference is between the pure free living pariah dogs and the pure Canaan “breed” established by simply capturing pariahs and breeding them for sale.  Kiva is a dog I saw as a puppy in 2013 in Israel. Her parents were Janey, a freeborn bitch from Jordan, and Tsuk, a freeborn from Israel. She now lives in America. Based on that information only her gDNA results could be expected to show her as a “village dog”. However she comes out as a Canaan. Looking deeper gives the explanation. Unlike most established breeds that have long been totally closed, the Canaan breed still adds new dogs after they have gone through a kennel club process. In simple terms this is based on the new dog coming from the area i.e. not from other continents where similar looking pariahs exist, then by breeding through 3 generations with the offspring breeding true to the look. The 3rd generation can then be registered as a Canaan with the kennel club. When it comes to DNA identification of any breed this is of course based on what kennel clubs deem to be the breed.  At least two maybe three dogs related to Kiva have been incorporated into the breed and have had their DNA checked by Embark so are therefore now part of the Canaan breed profile. At least one of these was descended from a brother of Janey.  Therefore Kiva now fits the updated Canaan profile.  Sadly many of Janey’s pups from her 3 litters were considered “aggressive” and were killed rather than spend the time needed to work on this. This is NOT the way to go if breeders honestly want to preserve this type of dog.  Starting at bringing in more new dogs they should be fully tested with state of the art methods (Embark being the only available one with pariahs in their database). This would firstly ensure that the dog is not a mixture. It would also ensure that no potential genetically associated health issue is added and could become a problem in future.  There is no simple “marker” that says a dog is a Canaan. The only “Canaan” that may be found living free is one related to the few that have been used to create the “breed” or have been added to them.   There is no reason why a pariah or village dog that tests as a pure Arabian or Middle Eastern village dog should not be added to the Canaan breed if it also has no genetic health issue. Ideally it would be preferable to concentrate on allowing these free living dogs to live where they are and as they have done for thousands of years without human interference, provided they are not in unsafe areas where they are at risk of being killed.

Of the pariahs I tested the close relatives were all from the same pack of dogs living together.

Tuesday, December 17, 2019

Coefficient of in breeding of Arabian desert dogs

The mean coefficient of inbreeding (COI) of the 47 dogs tested was 6.255. An inbreeding coefficient of 0 indicates a dog that comes from two unrelated parents; 6.25 is equivalent to mating of first cousins; 12.5 equates to the genetic equivalent of a dog produced from a grandfather to granddaughter mating; 25 would equate to the genetic equivalent of a dog produced from a father to daughter mating.


Although some inbreeding is probably inevitable this does not show that it is common and in fact is less than in many pedigree dogs.

Wolfiness as measured by Embark:

Embark explains this as:

Your dog’s Wolfiness Score is not a measure of recent dog-wolf hybridization and does not necessarily indicate that your dog has some recent wolf ancestors. (If your dog has recent wolf ancestors, you will see that in the breed mix report.) Instead, the Wolfiness Score is based on the number of ancient genetic variants your dog has in our unique Wolfiness marker panel. Wolfiness scores up to 10% are almost always due to ancient wolf genes that survived many generations, rather than any recent wolf ancestors. These ancient genes may be a few thousand years old, or may even date back to the original domestication event 15,000 years ago. They are bits of a wild past that survive in your dog!

Your dog’s Wolfiness Score is based on hundreds of markers across the genome where dogs (or almost all of them) are the same, but wolves tend to be different. These markers are thought to be related to "domestication gene sweeps" where early dogs were selected for some trait. Scientists have known about “domestication gene sweeps” for years, but do not yet know why each sweep occurred. By finding rare dogs carrying an ancient variant at a certain marker, we can make associations with behavior, size, metabolism, and development that likely caused these unique signatures of “doggyness” in the genome.

Most dogs have wolfiness scores of 1% or less. We find populations and breeds with higher scores of 2-4% occasionally, and unique dogs with scores of 5% or above more rarely.

Our results:

Mean:  2.83; Low 0.6; High 11.7.

3 were less than 1% (2% of the total); 35 of the rest were less than 4% (72.9% of the total);  6 were above 5% (12.5% of the total).

Friday, October 25, 2019

The importance of interpreting published work in an unbiased manner.

Recently  (22 August 2019) Dr. Tamas Jakkel of the Federation Cynologique Internationale (FCI) gave a talk “Preserving breeds, debunking myths”. (available on their FB page) In it he referred to a publication that he claimed showed that pedigree dogs are healthier than mixed dogs. In reality the exact opposite was true if anyone looks at the original publication. The publication did find that mixed dogs have more gDNA alleles that are associated with known health issues. This can be expected since the mixed dog may well pick up such a gene from both parents who may have different issues. However what he did not say or recognise or at worst deliberately hid was that the mixed dogs were nonetheless healthy because they only had one copy of these mutant genes so were “carriers” rather than “at risk”.

There have also been talks given and blogs written giving as an example that while boxers have the highest numbers of dogs “at risk” for DM they are not the highest to develop DM. There is of course a simple reason if we look a little bit deeper. DM effects dogs as they mature and sadly many boxers simply die too early for DM to show up. "With a median life span of 10.5 years and an expected range of 9 to 12 years, one does not expect a Boxer dog to live far into his teens. Reasons for this include the quite high cancer rates with this breed and heart issues. "  You can find the factual explanation and link to the actual publication here.

DM is a big problem in many dogs with Canaan pedigree dogs being among the most effected – certainly in the USA. I have yet to see figures for Canaan dogs in Europe but do know that a number of breeders have experienced this problem.  We have to remember that being at risk ie. 2 copies of the gene,  to quote Embark DNA testing (run by Boyko of Cornell University), “Testing positive is predictive of your dog being affected by this condition, but it is not a final diagnosis nor does it predict when symptoms may occur or the severity of a condition in your dog.”  DM has no cure and is a nasty way for a dog to die so breeders really need to avoid producing at risk dogs. We also need to realise that having only one copy nor indeed none, of some of these health  issues does not mean a dog may not develop a problem since outside factors also have a role to play.

The ONLY way to know the health status of a dog to be used for breeding is to have it tested and not only for issues known to exist but for all, unless of course both parents are known to have been affected, or all dogs in a breed have been previously tested clear. For most breeds of dogs breeders have to work around this. Simply not using any dog with any of the known genetic health problems could lead to the loss of other valuable (to breeders) genes.  So what to do?

If a dog is at risk (2 copies) if bred with a dog known to be clear (0 copies) then ALL the pups will be carriers, fine if they are never bred but important to know if down the line an owner decides to breed. If an at risk dog is bred with a carrier (1 copy) then for each individual puppy born there is a 50/50 chance it will be at risk or a carrier. It is possible all will be at risk or all will be carriers. Not worth the risk.

If a both dogs bred are carriers (1 copy of the gene) then each pup may have a 25% chance of being clear, a 50% chance of being a carrier, or a 25% chance of being at risk. Yes once more it is possible all will be clear, all will be carriers or all will be at risk. There is no way to predict.  Again not worth the gamble if we don’t want to be guilty of producing dogs that will suffer.

If a dog that is a carrier is bred to a dog that is clear then each pup has a 50/50 chance of being clear or a carrier. Again not a problem but needs to be understood and tested if used for breeding.

For most breeds that’s about all that can be done without some form of out breeding.

In the case of the Canaan pedigree dog there remains a chance of bringing in more freeborn dogs, however it is vital that in doing so they should be genetically tested FIRST and not used if they have any genetic health issue. Why go to the trouble and cost of bringing in more problems.  Embark them first!

Breeders also need to accept that the type of pariah dogs that were used to establish the human controlled breed are wide spread and plentiful as genetic evidence shows. Do they test as “Canaan dogs”? No of course not because the pedigree dogs come from little over 30 from a relatively small area, with a few added more recently, so are all now more closely related and we cannot expect  to find those close relatives in the wild, although not impossible. Some of these dogs end up in shelters where they are invariably neutered but if the desire to preserve an ancient type was properly explained I believe there could be co-operation set up to introduce some to the pedigree breeders. This is a unique opportunity for Canaan breeders and could even be used if handled properly in the long term to total phase out even carriers of DM and other genetic issues.

Genetic health issues in free born  “village dogs” versus “Canaan dogs”.

Looking at a greater number of these freeborn dogs than was used to establish the Canaan pedigree breed, none of these dogs were at risk for any of the genetic health problems tested by Embark.

9 of 47 (19%) dogs carried one copy of the allele coding for Degenerative Myelopathy (DM) so 81% were normal. The Orthopaedic Foundation for Animals (OFA) in USA figures for 1918 show Canaan dogs, a pedigree breed established in the 1930s from village dogs show an alarming rate of 6.4% abnormal (at risk) for DM with 38.2% carriers with only 55.4% normal. Compared to other breeds Canaan dogs ranked a high 22nd. No figures were available for Canaan dogs in Europe but it is known that DM has been a problem in them.

Canaan dogs
Village dogs
Degenerative Myelopathy

At risk

Choroidal hypoplasia, (Collie eye anomaly) was carried by 2 “village dogs”. These 2 dogs from Umm al Quwain, UAE were closely related, sharing 40% of their DNA (half siblings). This does not seem to have been reported in Canaan dogs but surely any new freeborn dogs added should be tested for all known health issues so as not to inadvertently create a problem.

von Willebrand type 1 a bleeding disorder was carried in 1 dog from Sohar, Oman with 1 copy of the allele. Again a reason to fully test all new freeborn dogs before adding them to the pedigree Canaan dogs.

Low normal ALT 25 of 47 dogs (53%) had one copy of the allele coding for low normal ALT and 1 had 2 copies. This does not effect the health of dogs but is important to know since a dog having this, when suspected of having liver problems, could give an ALT result generally considered normal but could in fact be raised for that dog. With a high percentage of this being present at a carrier level. Without testing pedigree Canaan dogs we simply do not know if it is present in the breed or not.

Saturday, September 21, 2019

Some interesting haplotypes found in Arabian village/desert dogs.

Some of the interesting haplotypes found in these dogs with notes from Embark.

H8.1 found in Max from al Khobar and Pavlova from Dubai. H8.1 - Part of the E haplogroup, this haplotype has been spotted in Basenjis, Canaan Dogs, and village dogs spanning from Africa into the Middle East.

H8.4 found in Jimmy from Um al Quwain. H8.4 is found in dogs from the south and east of India.

H9 found in Wiley from Egypt and in Basenjis and African village dogs. Part of the F haplogroup F is the odd duck in the family of domestic dog male lineages. This paternal lineage is genetically closer to wolves, foxes, and jackals than to other dogs. This indicates that it came into the dog population after dogs were originally domesticated, when one particularly attractive male wolf mated with a female dog, over 6,000 years ago. Since then, these dogs found their way into Africa and Mongolia. It hasn't been found outside those areas except in Basenjis.  Basenjis are an iconic African breed, that first made its way to the USA in the early 20th century when a handful of individuals were imported from the Congo. The Basenji is an ancient breed which is distantly related to other dog breeds (most of which are European or Asian), and it has the earliest separation date from all other breed populations. Unsurprisingly, the F lineage has also been found in African village dogs, as well as, surprisingly, some samples from Mongolia. The fact the lineage is found in two very distant places is evidence that it entered the dog population many thousands of years ago.

C1 found in Pack leader, Jenny and Jimmy from um al Quwain. Jenny and Jimmy are siblings and are related to Pack leader.  C1 - Congratulations, C1 is a very exotic female lineage! It is more closely associated with maternal lineages found in wolves, foxes and jackals than with other dog lineages. So it seems dogs in this are group have a common male dog ancestor who, many thousands of years ago, mated with a female wolf!

A158 found in Robin from Dubai. A158 - Part of the large A1e haplogroup, we have detected this haplotype in village dogs in India.

A264 found in Cookie from Jeddah. A264 - Part of the large A1e haplogroup, this haplotype occurs in Irish Wolfhounds and village dogs from Iraq and Lebanon.

B105 found in Honey from Fujairah B105 - Part of the B2 haplogroup, the B105 haplotype occurs most commonly in Middle Eastern Village Dogs. It's a rare find!

B77 found in Tess from Fujairah.  B77 - Part of the B1 haplogroup, this haplotype occurs most frequently in Japanese Chins.