Posts Tagged ‘Dog evolution’

pekin duck

One of the more bizarre fights that happens in the world of dogs is how to classify them. Within these debate is always a background of what it actually means to care for a dog “naturally” and what their “natural” life should be.

It actually shouldn’t be this way. Whether one considers a dog to be its own species or a subspecies of Canis lupus, there is no natural way to care for them or natural life for them. That’s because a dog is a domestic animal, the oldest of domestic animals, and there is very little that is natural about them. Over the thousands of years that have been part of human societies, they have adapted very nicely to our needs. Dogs have even developed cognitive short-cuts that have made them better readers of human body language than virtually any other animal, wild or domestic.

The natural way of keeping a dog is that a dog lives with people. It’s an oddball among domestic animals in that is derived from a very old domestication, and it is also the only large carnivoran that has ever been domesticated. Almost all other domestic animals are herbivores, and the other domestic carnivorans–the Near Eastern wildcat, the European ferret, and the red fox– are all small species. Most large carnivorans consider humans to be prey, so there is something very unusual about domestic dogs.

I get that.

But I am not among those who thinks that there is a species called Canis familiaris. I think dogs are most correctly classified as a subspecies of Canis lupus. 

As soon as I say this, people just lose it. That’s because when you say this, it is almost like justifying abusive dog training methods that are based upon dominance. It also might be justifying raw feeding, which is also contentious issue in the world of dogs.

Those are implications that I soundly reject. There is nothing inherent in a classification of an organism that tells you how to feed it or train it, especially when you’re dealing with the real oddball among domestic animals.

Note that it is never controversial to say that a pekin duck is a mallard. In the photo at the top of the page, you can see the old pekin drake that used to rule the pond. His mate was that Rouen-cross hen, which was sold at a feed store as a mallard. They were both of the same species. Her ancestors were some kind of Western Europe mallard, while his were wild mallards living around Nanjing that were later transplanted to the area around Beijing, where they were selected for larger size and white plumage.

Both of these animals were quite different from the true wild mallard that I see swimming in rivers here in West Virginia. She was twice the size of a wild mallard hen and much darker in color, and he was three times the size of wild drake. Neither of them could fly very well. The Khaki Campbell crosses in the photo were actually much better fliers than either of them.

Performance-strain Rouens are not far removed from the wild mallard. Although they are larger and cannot fly, they still produce the large amounts of oil in their plumage that keep them warm and dry even when they swim the coldest water. Female Rouens also retain the brooding instinct and can hatch out their offspring.

Pekins don’t produce as much oil and aren’t as cold tolerant, and if incubators didn’t exist, there would be far fewer pekin ducks in this world. Most pekin hens have no broody instinct. Further, they also grow so much more rapidly the either Rouens or wild mallards that they are prone to growth disorders.

Even though it is so far removed from the wild mallard, the pekin duck is still a mallard. It is as much a modification on a mallard as a St. Bernard is on a wolf.

The only difference is that no one is going to launch into a culture war tirade over the classification of a pekin duck, but if you say a St. Bernard is a wolf, then you will be asking for it.

It is certainly true that dog domestication happened a lot longer than mallard duck domestication. I cannot find any good literature on dates for mallard domestication, but it’s pretty clear that ducks have been kept in Southeast Asia for thousands of years.  Dog domestication dates and locations are still quite contentious, but the best evidence I’ve seen suggests that they were domesticated from an extinct wolf population between 15,000 and 32,000 years ago.

Some will argue that this extinct wolf population actually is a different species from modern Canis lupus, but I’m quite skeptical. Wolves themselves are among the most varied species in the wild. If we were to go back in time  see that ancestral wolf population that gave rise to domestic dogs, I think it would be hard to say that they weren’t within the diversity of phenotype that we see in current Canis lupus populations.

I think the big difference is that these wolves had far fewer reasons to fear people and were actually quite curious about our kind and were actually fairly easy to habituate to living near us. Over time, these wolves became incorporated into our society.

Much has been made that dogs and wolves have different reproductive strategies, but the truth is wolves actually have two reproductive strategies. One of these is the pair bond, where a male and female become partners and their grown offspring help care for the puppies. This is the most successful strategy for wolf reproduction because all the resources and attention of the pack can be devoted to a single litter. Another strategy goes on in parallel. Young male wolves leave their natal packs, but they often cannot find a mate or suitable territory. So they often try to mate with the grown daughters that are part of an established pack. These daughters cannot mate with their father, who is pair-bonded to their mother, who will attack them if she catches them in the act. So these females do often mate with these roaming male wolves. They often become pregnant and even have puppies, but in the wild, they almost never get a chance to raise those pups.  In the early years of wolf reintroduction to Yellowstone, prey was so abundant that many packs raised multiple litters every year. One would be born to the mated pair, while others would be born to these unpaired daughters and the roaming males. The way dog genes get into wolf populations is almost always when one of these non-paired females in a pack hooks up with a roaming male dog, which is why dog hybridization in European wolves went unnoticed. Most studies on wolf DNA looked at mitochondrial DNA alone, and if a wolf had dog ancestry coming from a male dog, it simply never would have been noticed.

The pair bond strategy exists because it about the only way to raise wolf pups. It is very common in all other dog species. In fact, a study of Chicago coyotes revealed that they are almost 100 percent monogamous, and similar findings have been discovered in golden jackals. However, in a wolf pack, preference toward the bonded pair during the mating season means lots of stress. They mated pair has to spend lots of time making sure no one mates with the wrong wolf, and they also will try to kill any of the roaming males that come near the unpaired females, which is also why the roaming male strategy isn’t as successful with wolves.  Natural selection would favor pair-bonding over roaming male mating strategies.

Domestication changed this equation. When wolves began to hook up with people, people began to provide food.  These wolves may have pair-bonded and mated in that fashion, but the chances of these secondary females to raise litters to maturity were much greater. Over time, pups born to roamers and unpaired females would start to outnumber those that were born to the pair-bonded wolves, and thus, a relaxation of selection pressures for pair bonding would become ubiquitous in these wolves. When pair-bonding became broken, then it became easier to selectively breed them. One could have a stud that mates with many females, and this promiscuous behavior could have been heavily selected for. Dogs are able to reproduce at much faster rates than wolves do, and one of the weird effects of domestication has been that female dogs are no longer monestrus and become sexually mature at younger ages than wolves or virtually any other wild dog.

Dogs and wolves have continued to exchange genes since the initial split. Black wolves in North America derive from domestic dogs that mated with wolves, as do black Italian wolves.  Wolves in Italy also can have dewclaws on their hind legs, which also originated from dog and wolf matings. Historians ranging form Pliny the Elder to the Plott hound historian Bob Plott have documented cases of hunters breeding their dogs with wolves, a practice that still goes in parts of Russia. It was just recently revealed that the livestock guardian dogs of Georgia have a rather significant amount of gene flow with the wolves of the same region.

So yes, I do recognize there are differences between wolves and dogs, but dog is a modification on the original wolf template.

The final important reason why I classify dogs as being part of the wolf species is that evolution has within it a nested law. This is the law of common descent. One can never evolve out of one’s ancestry. Humans are always going to be great apes, and humans are always going to be primates, not matter how different we become from our ancestors. A whale will always be a mammal, even if it somehow evolved gills.

A dog is always going to be a wolf. We change them through selective breeding, as we have with all our other domestic animals, but we are never going to change their fundamental ancestry.

All that I’m doing when I use Canis lupus familiaris is that I’m putting dogs where they fit on the tree of life. I’m showing my respect to their evolutionary heritage. I am paying homage to their phylogeny.

I’m not making excuses for Cesar Millan or anyone else.

Many people who promote science in our understanding of dogs are actually engaging in what I call “phylogeny denial.”  Many people bend over backwards to show how dogs aren’t like wolves, which I supposed is harmless, but I think it gives people a false impression of what a dog actually is in terms of its evolutionary history. It’s not a domesticated golden jackal or coyote or African wild dog. It is a domesticated ancient wolf, but that wolf was just an older form of the modern Canis lupus.

When you classify an animal according to its phylogeny, you aren’t doing anything else but classifying it. If a whale is a mammal, it does not automatically follow that it is a land mammal, does it? And classifying a dog a subspecies of wolf doesn’t mean that it evolved to hunt moose in Alaska.

I really wish people were taught to think about natural history in this fashion more often. It clarifies a lot of misconceptions people have about evolution. If I had a nickel for every time I get asked about humans evolving from modern chimpanzees, then I’d be a pretty wealthy individual. The last common ancestor between humans and chimps was not a chimpanzee. It may have looked more like a chimpanzee, because chimpanzees retain a lot more of the original African ape’s features than humans do, but it was not a chimpanzee like we have today.

By contrast, the wolves that gave rise to domestic dogs were probably indistinguishable from Eurasian wolves living today. Further, dogs and wolves continue to affect each other’s evolution through a rather significant gene flow. Humans affect chimpanzee evolution only through hunting them for bushmeat, destroying their forest habitats, and spreading disease. There is no gene flow between the two species, and because we have a different chromosome number, any “humanzees” would likely be sterile.

Finally, Canis familiaris creates a stumbling block in understanding the natural history of dogs, which is why you still run into people who think dogs derive from any number of different species of wild dog. Canis lupus familiaris neatens up that confusion very nicely.

Classifying a dog as a wolf shouldn’t be any more controversial than classifying a pekin as a mallard, but dog people just have a much harder time thinking in this way. I have never seen an internet flame war erupt between pekin duck owners over the classification of their ducks. In fact, I don’t think many pekin duck owners actually know that their ducks are mallards and do not actually occur anywhere in the wild.

But with dogs, charlatans have used the dog as wolf idea to justify all sorts of bad human behavior towards dog, but scientific facts remain scientific facts, whether charlatans misuse them or not. In terms of their ancestry, as has been revealed through copious analyses of their DNA, dogs are in the Canis lupus lineage.

They simply aren’t anywhere else in the tree of life. This is where they belong. Accept it, and move on.






Read Full Post »

Eucyon and the wolf

The likely ancestor of all extant species of Canis and probably the dhole and African wild dog is featured in this print.

The smaller canid featured here is Eucyon, a type of canid that lived in North America during the Late Miocene (10-5 million years ago). Eucyon is  both the genus and common name for these canids, which weighed about 20 pounds.

Eucyon means “original dog,” and it believed that all the extant Canis and possible Canis species evolved from some sort Eucyon species around 6 million years ago.

The oldest extant Canis species is the black-backed jackal (C. meslmelas) of Southern and East Africa. Mitochondrial DNA studies suggest that it split off the Canis lineage between 2.5 and 4.5 million years ago.

I have seen depictions of Eucyons that show them looking a lot like black-backed jackals.

However, black-backed jackals are still quite derived from this ancestor, and from my reading of the literature, this depiction is more accurate.  The morphology of this animal is believed to have been very similar to a crab-eating fox, which isn’t actually a fox.

For most of the dog family’s history on this planet,North America was the main center for their evolution. Canis came from North America and then spread throughout Eurasia and Africa.  The dire wolf eventually colonized South America, and then through the domestic dog, the genus eventually spread virtually every else.

Yes. Dogs are much more American than any of us African great apes!


Read Full Post »

Courtesy of Mashka Petropolskaya.


Read Full Post »

Here’s a new way of looking at dog breeds:

This comes from a newly released study that appears in PLoS Genetics.

More than 170,000 SNP’s were genotyped in 46 breeds and wolves.  SNP’s are short for single-nucleotide polymorphisms, which is a DNA sequence variation that occurs when there is a difference in a single nucleotide between individuals in the same species. One can make comparisons about relationships through examining these variations.

The recent studies that found the red wolf to be mostly coyote and found that dogs were mostly derived from Middle Eastern wolves also examined SNP’s, but they examined only around 50,000 of them.

This one examined 170,000 SNP’s, so the resolution, so to speak, is much clearer.

Let’s look at the neighbor-joining tree. Here are some highlights that I think can be confirmed in the historical record.

  • Modern flat-coated retrievers fit within the golden retriever breed. As I’ve noted, golden retrievers reflect much of the diversity in phenotype that existed within the old wavy and flat-coated retriever breed.  Goldens and flat-coats share many foundational sires and dams, and one would expect that flat-coats would be closely related to goldens.  The flat-coated retriever nearly became extinct in the Interwar Period, while the golden became extremely popular following the Second World War. This change in fortunes happened after the two breeds were officially placed in their own separate registries. Because the flat-coated retriever became rare, the modern representatives of that breed reflect only those flat-coats that survived that population crash. Goldens continued to reflect much more of the diversity in type and in genes that once made up the entire population– even though they are of a recessive color.
  • Newfoundlands and retrievers are close relatives. I’ve always said that a Newfoundland is just a big retriever, but really, the Newfoundland is an offshoot of the St. John’s water dog that was imported to the United Kingdom from Newfoundland. It was selected to be a much larger dog than one normally found in Newfoundland, and it became a popular family pet for much of the nineteenth century. The golden, Labrador, and flat-coated retrievers derive St. John’s water dogs that were imported from Newfoundland and were selected to be gun dogs. Richard Wolters made the assessement that the large Newfoundland was derived from the St. John’s water dog, and this breed should be regarded as the common ancestor of the retriever-Newfoundland family.
  • Nova Scotia duck-tolling retrievers are more closely related to border collies and Australian shepherds than retrievers. I have said that there is no evidence that they are related to golden retrievers, even though they superficially resemble dogs of that breed. I also suggested that we should think of the toller as a small retrieving collie.  It also suggests that the recent cross-breeding of a toller to an Australian shepherd actually has greater merit than breeding it to another retriever.
  • Dalmatians are related to Weimaraners, which are related to other pointing breeds.  The postulate that the Dalmatian is derived from pointer crosses appears to have some merit.
  • Border terriers are just one of the fancy varieties derived from the fell or Patterdale terrier.  Patterdales are widely acknowledged to have Staffordshire bull terrier and bull terrier in them. But this map suggests that this infusion of bulldog or bull and terrier blood happened before the border terrier became fancy.
  • Gordon setters are widely said to have border collie in them. It now appears that some Gordon setters do derive from border collies or from the ancestral collie that gave us the border collie, but others are more closely related to English.
  • The theory that Large Münsterländers are derived from English or Gordon setters that were crossed with German long-hairs appears to have some merit. They are more closely related to one lineage of Gordon setter than to the Weimaraner.
  • The pug may be more closely related to the spitz breeds because much of its development occurred in the West, particularly in the Netherlands. Spitz breeds were very common as pets in the seventeenth century Dutch Republic, and they were probably used to develop the pug in its present form. There were not likely vast numbers of pugs coming from China at this time, and the Dutch just used a local dog to make the breed more genetically diverse.
  • The boxer is related to the bulldog, and we know that the bulldog was often bred into boxer lines. The boxer lines are longer because the boxer has had its full genome sequenced.
  • This study does not necessarily show that German shepherds are derived from wolves. Rather the authors write:

The most obvious clustering of breeds is exhibited by two wolf hybrids: Sarloos and Czechoslovakian wolf dog, which exhibit a closer relationship to the wolf than other breeds as predicted by their known origin. The German shepherd also clusters with this group, although this is likely to be a result of its close relationship with the Czechoslovakian wolf dog, rather than with wolf. The tree is consistent with previous studies and supports the accuracy and reliability of the array.

Each time more SNP’s are analyzed, greater resolution is provided, and the result will be different. So let’s not assume that these results are going to remain unchanged.

However, I am amazed at how much this tree appears to confirm certain aspects that appear in various breed histories.

Read Full Post »

A great lecture by Robert Wayne on the evolution of domestic dogs and how they came to be so diverse in appearance and genetics.

Read Full Post »


He takes down Ray Comfort’s nonsense.  Comfort has no clue.

He also argues that Lycaon pictus ought to be in the genus Canis, for exactly the same reason I feel it should be. It and the dhole are more closely related to the other species of the genus Canis than the black-backed jackal and side-striped jackal are. Xenocyon lycoanoides, the extinct ancestor of the African wild dog, should also be part of this very important genus in the dog family.  (See the dog phylogenetic tree).

He is a little off on the origins of the dog. The East Asian theory of dog origins seems to have been falsified through the genome-wide study that utilitized SNP chip technology found that dogs had greater genetic similarity with Middle Eastern wolves, which suggests that the Middle Eastern wolves, not the East Asian wolves, are the main ancestors of domestic dogs.

Epicyon and the Amphicyonids make an appearance in this video, as to the “dog-bears” (Hemicyonids).

The Caniformia suborder of Carnivora has the most diverse species. Not only does it have domestic dogs, which have greater diversity in head morphology than the whole order Carnivora combined, it includes the smallest member of the order (the least weasel) and the largest (the southern elephant seal). Yes, Carnivora includes the seals, walruses, and seal lions, which are now classified within the Caniformia suborder.

Red pandas are fascinating because of their status as a “living fossil.’

And giant pandas are bears with fused chromosomes.  The two animals evolved their bamboo diet and their very similar s specialized wrist that acts like a thumb in parallel with with each other. Their common ancestor in the basal Caniformia didn’t have that thumb wrist or the specialized bamboo diet. Because both of these animals are derived from meat-eating Carnivora ancestors, they have not developed the ability to digest cellulose, so they have to eat tons of bamboo to survive. If these animals had been designed, one would think the designer would have put in some digestive bacteria in them to help them digest cellulose.

Please note that hyenas are missing from the Caniform cladistics video. Simple reason:  Hyenas are not in Caniformia. They are in the other big suborder of Carnivora, Feliformia.  Yes. Hyenas have a closer common ancestor with cats than with dogs.

This is a very good video.

Read Full Post »

One of the most interesting pieces of evidence for speciation through evolution is the ring species concept.

Ring species are several interbreeding populations that can generally interbreed, but there are always two populations that make up the “end series “that are too genetically divergent to interbreed. The populations gradually range from one “end” to the other.

Such is the case with the Ensatina salamanders, which range through the coastal and forested regions of California. Those that came down the coast evolved to look like the poisonous newts of the Pacific Coast, while those that came down the forested regions of the Sierra Nevada evolved splotches for camouflage.

Their distribution forms a horseshoe around the Central Valley of California, which has acted as a geographic barrier. The two populations at the southern end of the horseshoe cannot interbreed. These two populations actually overlap in range, but they do not produce offspring. They have become too genetically distinct. And any hybrids that happen between the potentially interbreeding newt mimic populations and camouflage populations are poorly adapted to be either.


Something very similar is happening with dogs. The reproductive barrier isn’t a geographic one. It also isn’t the result of having closed registries, because even dogs within closed registry breeds still exchange genes with other dogs, whether the breeders like it or not. Closed registries could become a reproductive barrier that could lead to the development of a ring species, but another factor is probably much more likely to be the cause of the first domesticated ring species.

The factor is size.

It is likely that all the dog breeds are going to be able to interbreed with each other, closed registry or not. However, an exception exists at the two extremes in size. Very small dogs cannot breed with giant dogs unless someone uses AI. It is simply mechanically impossible for a male chihuahua or yorkie to mate  with a Great Dane or mastiff. It would be a near death sentence if a female toy breed tried to mate with a giant dog. These two extremes are genetically isolated from each other and probably will be forever.

Now, toy dogs can breed with large dogs. I know of several goldendoodle breeders who use toy poodle studs for their golden bitches without having to use AI. So large and very small dogs will not be reproductively isolated. Large dogs can cross with giant ones with very little trouble, so they will not be reproductively isolated. And there are always stories of relatively small dogs mating with big ones. For example, one of my childhood beagles knocked up a dobermann. He was a medium-small dog, and he was able to exchange genes with a very large dog with no human assistance. (The dobermann’s owners didn’t want that breeding to happen at all, so there was no assistance.)

But no one is going to find a Great Dane/Chihuahua mix that happened without human intervention.

Over time, those two extremes are going to become so genetically distinct from each other that they can’t interbreed at all.

And thus, we would have the first artificially created ring species.

It is not that far fetched.

It is what is likely to happen, but it is going to take a while.

Read Full Post »

Older Posts »

%d bloggers like this: