A wolf in dog’s clothing?

7 February 2014 by Malcolm Campbell, posted in Biology, Science

Hence, we may conclude, that under domestication instincts have been acquired, and natural instincts have been lost, partly by habit, and partly by man selecting and accumulating, during successive generations, peculiar mental habits and actions, which at first appeared from what we must in our ignorance call an accident.” From On The Origin of Species by Charles Darwin  (1809-1882)

There’s nothing like a frigid northern winter to remind you that dogs are simply not wolves.

Well, to be more specific, whippets are not wolves. dogs_wolves_1

Whippet coats, sleek and smooth as they are, may be sufficient for British midland winters, but they are no match for -20oC temperatures, deep snow, and the biting winds of the polar vortex. Such conditions can reduce a whippet to a shivering mass, who would sooner curl up under a comforter than brave the weather to empty their bladder.

Not so with wolves. In the winter, wolves are insulated from the cold by their double-layered coat. The wolf’s winter coat comprises a dense and duffy underfur, penetrated by the longer, coarser hairs of their outer coat. Together this double-layered coat is an effective barrier to inclement winter weather – it repels snow, and insulates against temperatures down to -40oC.

Of course, the winter coat is one of many physical differences that one could document between wolves and domesticated dogs more generally, let alone whippets. The differences are not only many, but highly varied depending on the domesticated dog breed against which the comparison is made. But, of course, one could draw this same conclusion on the basis of comparing two different dog breeds.

Whippets are not German Shepherd Dogs.

Pugs are not Irish Wolfhounds.

Chihuahuas are not Great Danes.

And so on.

In fact, on the face of things, it would appear that the difference between dog breeds is every bit as great as the difference between any given breed of dog and wolves. Indeed, some dog breeds bear more than a superficial resemblance with wolves. Many sled dogs, including the Alaskan Husky and Malamute, might give cause for a second look in areas where wolves also reside. On the basis of such a comparison, a wolf would appear merely a breed of dog. A wild dog to be sure, but superficially, a dog. dogs_wolves_2

And yet we know that wolves are not merely a breed of dog, but a distinct lineage onto themselves.

Indeed, recent evidence suggests that dogs and wolves may have embarked on their distinct genealogical paths earlier than previously thought.

Until recently, dog domestication was thought to have started with the emergence of agriculture. The birth of agriculture created a new niche for dogs – one suitable for scavengers on a source of dependable, stationary food – human refuse. Wild canines that were bold enough to capitalise on this new niche – to interact with the humans there – became “proto-dogs”.  These proto-dogs were domesticated over time as humans capitalised on their presence for other purposes – protection, hunting, and even as another source of protein.

The “scavenging proto-dog” hypothesis was supported by a number of lines of evidence, including wolf and domesticated dog genome sequences. Amongst the most striking differences between the wolf genome and the dog genome are those related to the digestion of starch.

Starch is a polysaccharide – a long chain of sugar molecules, in this case glucose, joined end to end. Starch is abundant in many of the foods we eat, notably grains, like wheat and oats, and tubers, like potatoes and yams. In the parts of the world where dog domestication took place, early farmers grew grains, whose leftovers would have been found in their refuse, and, over time, might have served as part of a dog’s diet.

While they are functional omnivores, wild canines, like wolves, have a diet dominated by protein- and fat-rich meat relative to starch-rich plant matter, like grains. Even still, evolution has equipped wolves with the capacity to digest starch, but not as a prominent component of their diet. Therefore, the wolf genome encodes proteins that enable the digestion of starch, and the uptake of the sugars that are liberated when starch is broken down.

The digestion of starch is a three step-process in canines. First, starch is broken down to shorter chains of sugars, known as oligosaccharides, in the intestine. The starch chains are cleaved into oligosaccharides by a protein catalyst, an enzyme, known as alpha-amylase. Next, the oligosaccharides made by alpha-amylase are broken down to the simple sugar, glucose. Three enzymes break down the oligosaccharides, maltase-glucoamylase, sucrase and isomaltase. Finally, the glucose is transported from within the intestine to cells that then pass it to the rest of the body. The uptake of the glucose from the intestine into these cells is accomplished by a transport protein known as SGLT1. dogs_wolves_3

The wolf genome contains genes that provide instructions to make all components of the starch digestion pathway – alpha-amylase, maltase-glucoamylase, sucrase, isomaltase and SGLT1. These genes work in concert to enable wolves to have some starch in their diet.

At first glance, relative to the wolf genome, the dog genome has striking alterations in the genes involved in making the starch digestion machinery. First, the dog genome seems to have many more genes that encode the enzyme alpha-amylase. Depending on the dog breed in question, the dog genome has anywhere between two and fifteen times the number of genes encoding alpha-amylase relative to the wolves’ genomes. The net result is that the dog intestine can have as much as 5 times the activity of starch-degrading alpha-amylase as the wolf intestine. This makes the dog intestine much better at that initial step of starch breakdown, relative to wolves.

In comparison to the wolf genome, the dog genome also contains altered versions of the genes that encode maltase-glucoamylase and SGLT1. These altered versions result in an increased amount of maltase-glucoamylase, and enhanced efficiency of SGLT1.

All told, the changes to the genes involved in starch digestion in the dog genome relative to the wolf genome would appear to make dogs more effective starch consumers. The variants support the notion that domesticated dogs originated in an agricultural setting, where such changes would confer an advantage to canines that can now acquire dietary starch more readily. The thinking is that starch digestion is evidence of the domesticated dog’s origin in farming communities.

Recent evidence has challenged this agricultural-origin hypothesis.

Freedman and colleagues compared the genomes of wolves from across their range against those from dogs that reflect the geographical distribution of ancient domesticated dogs – including the Australian Dingo and the African Basenji.

By looking at the occurrence of mutations within these genomes, Freedman and colleagues challenged the agricultural-origin hypothesis with two strong pieces of evidence. dogs_wolves_4

First, the genome comparisons allowed Freedman and colleagues to infer a range of dates encompassing the time when dogs and wolves last shared a common ancestor – that is, a time of origin for domesticated dogs. Changes within the genome, mutations, arise at random over time. The rate at which mutations emerge is contingent on the species, due to factors such as the frequency at which they bear young. While the mutation rate of the domesticated dog isn’t know precisely, reasonable estimates have been established. By tracking the occurrence of differences, of mutations, between dogs’ and wolves’ genomes, and using the calibrated mutation rate, one can make an estimate of the time since dogs and wolves last shared an ancestor. On the basis of the extensive analyses conducted in this study, it appears that dog divergence from wolves – that is, the start of dog domestication – began minimally 11-13 thousand years go. This time predates the adoption of extensive agriculture by humans.

But what of the changes to starch digestion evident in the genomes of domesticated dogs?

Crucially, it turns out that not all dog breeds have high levels of genes encoding alpha-amylase. The Dingo and the Siberian Husky, breeds that have accompanied hunter-gatherers, have the same number of alpha-amylase-encoding genes as wolves; whereas, the Saluki, a breed associated with the fertile crescent where agriculture emerged, have almost 15 times a many alpha-amylase-encoding genes.  It would appear that starch digestion is not so much related to the origin of dogs, as it is to where the breed developed after domestication.

Of course, the question could arise, might it just be that some breeds are more closely related to wolves than to other dogs? Perhaps this is why the genomes of these breeds don’t have the hallmarks of an agricultural origin? Freedman and colleagues were able to confirm that dog breeds, including the Dingo, Husky and Basenji, are more related to each other than they are to wolves. That is, wolf genomes contain a suite of mutations that identify them as being related, and distinct from the mutations found in dog genomes. What’s more, the genome data were unable to identify any wolf from across their natural range that was more closely related to the dog than any other wolf. That is, Freedman and colleagues were unable to pinpoint to geographical origin of the dog. dogs_wolves_5

This suggests one of two possibilities.

The first is that the lineage of wolves that gave rise to dogs has not yet been identified. The other, more intriguing possibility, is that the wolf lineage that gave rise to dogs may not be in existence anymore. It may be extinct. That would suggest that dogs and wolves shared a common ancestor that was neither one nor the other, but a true common ancestor that was something of both. Over time, the two lineages went their separate ways – one lineage, now defunct, that gave rise to dogs, and another that gave rise to modern wolves. These are fascinating hypotheses that, in the absence of time travel, are likely only to be resolved by more extensive analyses of dog and wolf genomes.

But this leaves us with a conundrum – if some dog breeds look like wolves but are not – what are those things that truly distinguish wolves from dogs?

As Charles Darwin presaged, the answer almost certainly lies in behaviour – “instincts have been acquired, and natural instincts have been lost.

In keeping with this, Friederike Range and Zsófia Virányi recently reported research that defines a rather remarkable difference in dog and wolf behaviour.

Range and Virányi raised a group of 16 wolves from the time they were 10 days old. They were initially raised indoors – first bottle fed, and then hand-fed by humans, until they were one month old. Between one month and five months of age, they had access to a large enclosure, where they had continuous contact with humans. During their first five months, they were also periodically exposed to domesticated dogs – five in total – with whom they played and established social relationships. After five months, the wolves had access to a much larger enclosure, at which point exposure to humans was no longer continuous – but still daily, for training and socialising. Range and Virányi raised a group of mixed-breed dogs in the same manner.  In the end, Range and Virányi had a group of wolves and a group of dogs that were very comfortable with humans and visiting dogs alike. dogs_wolves_6

Range and Virányi then investigated the extent to which the wolves versus the dogs could learn to problem solve based on the lead provided by a visiting dog. The visiting dogs in question had already established “dominance” relationships with the wolves and the dogs from the groups.

The visiting dogs had been trained to manipulate a device so as to liberate a hidden food reward. Manipulation involved moving a lever with their mouth or with their paw so as to obtain the food reward. Some of the visiting dogs were trained to use their mouth, while others had been trained to use their paw to manipulate the lever. None of the wolves, nor any of the dogs from the enclosures, knew how to manipulate the device at the outset of the experiment. For them, the box was a de novo problem.

Each wolf, in turn, was given an opportunity to see how the visiting dog liberated the food reward from the device. The same was true for each of the resident dogs. Following this “demonstration”, Range and Virányi documented the capacity of the wolves and dogs to “solve” the device. Where only 20% of the resident dogs tested were able to follow the lead provided by the trained visitor, 100% of the wolves tested were successful in solving the device on their first attempt.  So as to account for any developmental delay in dogs relative to wolves, dogs were tested not only when they were the equivalent age of the wolves (6 months), but also when they were 2 months older. The results remained the same – 20% of the dogs only could make use of the information provided by the visiting dog to solve the device in their first attempt.

The results obtained by Range and Virányi suggest that wolves pay close attention to other canines to gather useful information.

Dogs, on the other hand, not so much.

What do these findings tell us about dog domestication? dogs_wolves_7

On the simplest level, they tell us that dogs are not merely wolves in dog’s clothing. Dogs and wolves have very different ways of interacting with even relatively closely-related canines. This is consistent with other research that shows significant differences in how and when dogs versus wolves respond to human-derived cues. As might be expected, throughout their lifetimes, dogs are generally more responsive to human-derived cues, such as directional pointing.

On a deeper level, Range and Virányi’s findings suggest that dog domestication selected canines that place a different weight on where their cues are coming from. As Range and Virányi suggest, it may be that this reflects a relaxation of dogs’ need to pay attention to cues provided by other dogs. After all, after domestication, the most vital cues were likely to come from an entirely different species – humans. An alternative view is that this capacity to prioritise human cues, perhaps even at the expense of interpreting other dogs’ cues, was crucial for dogs being domesticated in the first place. Gregarious, human-cue-driven proto-dogs were more likely to find themselves favoured by the humans doling out the cues, and the food. Of course, it may be that both forces have been in action over the past 13 thousand years – gregarious cue-readers were selected and became less reliant on being able to respond to other dogs’ cues over time. If you are no longer a member of a pack, the need to pay attention to what a potential pack member does is a lot less important.

Irrespective of the precise nature, or even the timing, of their evolutionary journey from wolf to dog, the fact remains that dogs and wolves are very different today.  Sometimes they look different, sometimes they don’t. Regardless, what is going on inside their heads – their respective instincts – have diverged markedly since they last shared a common ancestor. To treat one as though they are the same as the other ignores their past, and the wonderful things that make them so distinct today.

Images: All photographs by Malcolm M. Campbell.


Axelsson E, Ratnakumar A, Arendt ML, Maqbool K, Webster MT, Perloski M,  & Lindblad-Toh K (2013) The genomic signature of dog domestication reveals adaptation to a starch-rich diet. Nature 495: 360-364

Freedman AH, Gronau I, Schweizer RM, Ortega-Del Vecchyo D, Han E, Silva PM,  & Novembre J (2014) Genome sequencing highlights the dynamic early history of dogs. PLOS Genetics, 10: e1004016

Gácsi M, Gyoöri B, Virányi Z, Kubinyi E, Range F, Belényi B, & Miklósi Á (2009) Explaining dog wolf differences in utilizing human pointing gestures: selection for synergistic shifts in the development of some social skills PLOS ONE 4: e6584

Range F & Virányi Z (2014) Wolves are better imitators of conspecifics than dogs PLOS ONE 9: e86559



Leave a Reply

5 − = one