One of the most interesting things about trying to classify different dogs in terms of their ancestry and relationship to other extant breeds is how much cultural prejudices come into play.
I can’t think of a better example than the Rottweiler.
In North America, Rottweilers are “tough guy dogs.” In the 1990’s, they were the second most popular breed in terms of AKC registrations. And there were plenty of nasty Rottweilers running about attacking people– even though uncontrollable aggression is actually a major fault in this breed.
If you were to ask they average North American what the Rottweiler’s closest relatives are, I’m sure you’d get someone saying that they are definitely related to “pit bulls.” Others might say boxers.
Still others– who are bit more informed– would say Dobermanns. Of course, this is kind of right. Dobermanns are thought to have a tiny bit of Rottweiler ancestry.
But that doesn’t mean that the Rottweiler is most closely related to the Dobermann.
If Dobermanns have just at bit of Rottweiler ancestry, the bulk of their ancestry is actually non-Rottweiler.
Furthermore, classifying Dobermanns is very difficult. As a modern breed with a wide variety of ancestors, it is virtually impossible to get them to fit into any sort of grouping of related breeds. Some claim that they are molossers, but others claim they are the biggest of all pinschers.
So let’s leave the Dobermann out of our analysis, and look at what the genetic literature says about the ancestry of Rottweilers.
Rottweilers were included in the 48,000 SNP study that revealed that dogs were most closely related to Middle Eastern wolves.
It also looked at dog breed relationships, and the results were a bit surprising. Retrievers and Newfoundland were closely related, but they didn’t fit with the other gun dog breeds. They fit with what might be called a “mountain dog clade” that includes the St. Bernard, the Great Dane, and the Bernese mountain dog.
And according to this study, the closest relative of the Rottweiler is the Great Dane.
It is also closely related to two Swiss “mountain dogs”: the Bernese mountain dog and the St. Bernard.
The relationship with the Great Dane suggests that the ancestral German boarhound was drawn out of the common working “mastiff” of Europe.
Throughout Europe, there were always mastiff-type dogs hauling carts. The ancestral type of this sort of dog would have probably looked like a trekhond or “Belgian mastiff.” These dogs varied greatly in appearance, but even in Belgium, some looked more like Great Danes and others like Rottweilers or Swiss mountain dogs.
My guess is that if the Rottweiler’s genetic material were compared to the other Swiss mountain dogs– of which there are four breeds– the Rottweiler would fit nicely in that family of dogs.
Two of these mountain dogs (Sennenhund) were primarily used as cattle-herding dogs, which was one of the original functions of the Rottweiler. These two dogs are smaller than the Rottweiler, but they originated in areas of Switzerland that are adjacent to Germany.
Furthermore, Rotweil was actually part of the Swiss Confederacy. It was never accepted into the full nation of Switzerland, but the city remained closely linked to Switzerland, which lies just to the south.
My guess is that trade between Rottweil and Switzerland meant that the Sennenhund-type of dog would eventually become established in Rottweil. The dogs may have driven herds of Swiss cattle into the slaughterhouses of Rottweil.
Rottweilers became associated with the butcher shops of their hometown, for the butchers needed dogs that could herd cattle.
Yes, I mean herd. Rottweilers are herding dogs. They are actually a well-known droving breed.
This differs very much from boxers and dogs of the bulldog family. These dogs may have started out as herders, but as time progressed, their main function was to grip and bait cattle.
And you’ll note on that phylogenetic tree that the bulldog and English mastiff family is very different from this mountain dog clade.
This should tell us that mastiff-type dogs have been derived from different stocks and at different times. The old claim that they all descend from the Tibetan mastiff has yet to be proved and is very unlikely.
People need to be careful when trying to classify dog breeds. There is actually quite a bit of convergent evolution through artificial selection in dogs, which is why we have rose-eared greyhounds having virtually no common ancestry with tazi/Saluki/Afghan hound family. Dachshunds are not closely related to terriers from the British Isles, even though they have a similar function.
If one didn’t know Rottweil’s history, it would be easy to make the claim that Rottweilers were German pit bulls or giant black and tan boxers.
They are actually derived from the working mastiff-dogs of that region.
They are not in the bulldog family.
Can anyone explain me the the figure shown? I am unable to get it.
The middle of the circle represents the common ancestor. Each branching off represents a clade of related breeds.
Basenjis are off on their own. They aren’t closely related to any of the breed dogs or those from East Asia.
Dingoes are nested within the East Asian dogs.
Thanks for pointing that out. Pet peeve here – lumping the basenji, east asian dogs, south asian dogs, etc, all together as “primitive” or “pariah”. As you have emphasized, they are not related. Spread the word!
That convergence may be the key to saving breeds since it would allow for crossbreeding for a type while broadening the genepool.
I totally agree Massagu and believe this likely reason Retrieverman shares his research. Saving individual breeds seems the focus and key to the secrets of crossbreeding. As Species and subspecies ofwolf.coyote and jackals apparently have crossbred producing viable hybrids. Taxomony has help to reveal some of these anchestoral cross breedings. The other element that is vital to consider as is pointed out in this post…the convergent evolution due to artifical selection plays a critical part in our need knowledge.
Any two-dimensional representation is almost certainly over-simplifying something as complex as the relationships of the various sorts/types/kinds/breeds of dogs. but it can be helpful, at least as a beginning.
I find nothing surprising about placing the Rottweiler with the “mountain dogs”, particularly the Sennenhund. And quite possibly they may also have a dash of mastiff-type or something else. Reality is seldom as simple as researchers might like it to be.
What do you think the placement of the Chihuahua on that chart reveals?
It’s no longer predominantly an indigenous breed.
Kate like Retrieverman states it is no longer predominately an indigenous breed. There is speculaition ,theory and myth surround the Chinuahua. When indigenous back in Mayan times the dog was long haired and mute. It seems even speculated they ate them. Believed it was used in religious ceremonies of the Mayans. Then when they were conquered became pets. However, it seems taccepted they were crossed with A dog the Explorers brought with them. Thus the species of today.
Does that chart use only mtdna? Is there any native dog dna to put on that chart? If there really aren’t any native dogs, do any of the ones we have still have, have some native dna?
I am a theoretical breeder, nowadays; I do not actually breed anymore and when I did breed, it was dogs I thought at the time were Indian dogs. Long before any of the current breeders were in operation. So many people I knew that had them were Indians- from WA state to the northern plains states, and down to Tucson, I never questioned their origin back then, except they had no wolf -though people thought they may have had wolf way back when.
But how do you know if the phenotype is like the old dogs, that the dna isn’t also very similar if you have nothing to compare it to?
What impresses me about all the breeding programs of dogs that look wolfy is how easy it is to achieve the look without using wolf. I had no trouble maintaining a gently wolfy type when I did breed. I recently found some photos from those years and will scan and post them soon, in their own blog.
Hi C-mae
I have written several posts on the history of the Chi. There were actually several distinct types of dogs when the Spaniards got here, including the techichi, the alco, the Talthan bear dog- and the tiny 7 Nations dogs of the upper East Coast which type name I do not know, but I have seen in paintings of the time and place. The Aztecs definitely did a big food trade in small, fat, dogs. I await further DNA studies to see where chis might fit, though I have an alternate theory about that.
Has anyone read this site on the Rottie? I found some of the comments quite interesting and would like to check out their resource information.
http://www.rottweiler.net/forums/breeding/35976-genetics-wolf-dog.html
Please accept apology if this material has been prediscussed on this Blog,but found this materiall. I found this information so in tune to Retrieverman’s discussions on the Red Wolf.in biological research. http://search.aol.com/aol/search?enabled_terms=&s_it=comsearch51&q=Genetic+Anchestry+of+Coyote
I don’t really like Rottweilers. Don’t get me wrong, I think they are beautiful animals. It’s just that I fear them very much. They could ripp my arm off if they got the chance!
Those who assert a “Roman herding dog” origin for both the Rottie and the Swiss herding dogs will find this supportive of their views. Be interesting to see a more through comparison — Pyrenian shepherd, Picard, Briard, Belgians (and Dutchies, who ought to show very close relationship to the Belgians), Bouvier, even some of the Neapolitan and Dogue de Boudreaux could be useful inclusions in a study like this. I will note that virtually every such DNA study has some breeds “moved” from relationships – I saw one that put Tervuren next to Beagles (whatever gene or genes they were looking at, I don’t think that the conclusion was correct).
But it’s not so very surprising. A “mastiff” that was a good dog for protection, hauling carts and also to help manage cattle is probably the foundation for all mastiff types. Cattle herding dogs tend to take two directions (related or not) – selection for a dog that can “duck under” the kick and selection for a dog big enough to boss more docile cattle. But given the GSD origin that included some of the “Altdeucher” that were more mastiff in nature, I’m surprised that NO relationship was found. As someone else posted, I don’t think a two dimensional study is ever going to be accurate. Most breeds are probably “mix / match” and have genes from several ancestral “landraces” to obtain size, color, various desired attributes like scenting ability, bite, bark, etc.
I lived in a dorf (village) outside of Leipzig for a couple years. Rotties were common as farm dogs, and seem right at home in the walled farm compound. History hasn’t been polite in that part of the world . . . invasions, wars, hostilities between ethnic groups . . . guardian dogs have been important. It makes sense that they group with other Central European farm dogs.
I wouldn’t infer a close relationship between the rottie and the retrievers from that diagram. You need to go up 7 or 8 branches on the cladogram before you hit a common ancestor. And I wouldn’t be surprised if the classification algorithm has some instabilities . . . change a few assumptions or throw in a different data set and you get different linkages.
I’d love to see the full scale version of the diagram. It would be good to see all the breeds included in the dataset.
What you’d actually need is a bigger sample with more breeds and more individuals per breed. The technology behind doing genomic analyses such as this one is getting better all the time, and the cost is getting cheaper. A more complete cladogram will be drawn in the very near future.
I’m not saying that they are closely related. What I’m saying is a golden retriever shared a more recent common ancestor than the rottweiler and the boxer. The purpose was to show that mastiffs are not all derived from the same stock, as is the common claim.
Golden retrievers aren’t even remotely like Molosser dogs.
This cladogram is based upon what is still a very small sample of the genome, but it won’t be long before we can see how they all fit together more completely.
Phylogenetic trees have their limitations. They are just meant to illustrate common ancestry and decent, but they have a harder time if organisms that are not that closely related. And in domestic dogs, that’s likely at play all over.
Please post if you see an updated version of Osterander (sp?) and Wayne, 2005.
A paper published in June 2012 doing cladistics for 80 breeds ends up with a real hash job . . . individual dogs of the same breed don’t end up in the same group . . . the Labradoodle doesn’t come out close to the Labrador . . . the Akita and corgi end out closely related, etc, etc., etc.. (see http://dnaresearch.oxfordjournals.org/content/19/3/275.full ). Their methods sound impressive but their results don’t pass the test of face validity.
Of course, for all I know, Osterander and Wayne cheated by forcing all the dogs from one group into a composite and putting the composite through the classification algorythm. (That is one hard to read paper and the press releases don’t go into methods). Who knows, they might have ended out with dogs in the same breed getting split between different groups if they hadn’t forced them into the same category.
That phylogenetic tree is based upon mtDNA only.
I don’t trust phylogenetic trees on dogs with just mtDNA.
My last post still needs moderation, maybe because of the change of address.
I thought it might just be mtdna,
Is that the kind of testing Mars lab uses for breed testing?
I’ve been without internet all weekend. I’m sorry it was in moderation for so long.
No.
http://www.wisdompanel.com/why_test_your_dog/faqs/#33
That IS the Mars lab http://www.wisdompanel.com/why_test_your_dog/faqs/#33
I think in many cases a given gene may be shared without the dogs being particularly related. For example, take the crested /queensland dog posted about elsewhere on this blog. Even several generations down the line, dogs from this animal might well show up “Queensland” related even though it is due to a single cross. It all depends on what genes get selected for or get inherited by sheer chance. Yes, one only inherits 50% from each parent, but that doesn’t mean that any given gene is or isn’t inherited. Since Mix/match was common prior to the 1840s and really, for some time thereafter, asserting that a given gene being shared means that the dogs are “closely related” is naive. What it means is that there was a shared ancestor — but that doesn’t make the dogs “close”.
excellent theory, and with a lot of documentation. I have to re-read your post, in order to understand everything.
thank you