Behavioral differences among breeds of dogs -- current status of science

Recently, a new research paper was published in the Applied Animal Behavior Science entitled "Behavioral differences among breeds of domestic dogs: Current status of the science."

The study does not present new data, per se, but instead looks at the depth and breadth of canine behavior science over the past 50 years and provides a very detailed look at what we know about behavior differences among breeds, what we think we know about these differences, and opportunities for more research.

The study includes references more than 120 other scientific studies about canine behavior (several I've referenced on this blog before) to create a thorough, and well-rounded report. While I doubt the conclusions will really surprise too many readers of this blog, I think the information is very interesting throughout (and I now have a lot more reading material ahead of me).  

Among the key findings based on the American Veterinary Society of Animal Behavior:

1)  Most studies have found convincing evidence of differences between breeds and breed groups with respect to behavior

2) Mant of these studies are based on survey results  -- and such studies are vulnerable to inaccurate reporting on a dog's behavior because the survey responder may be infuenced by cultural and media biases. Studies based on survey results reported larger differences in behavior among different breeds and breed groups than controlled studies.

3) Many of the studies also show very large within-breed behavior variations. This variation is due to  genetic differences within breed, and individual learning history as it relates to environmental differences and specific contingencies that the individual dog has contacted throughout its lifetime.

4) Breeds are not equally represented in the research - -and specifically, some breeds are vastly overrepresented, making conclusions about the influence of breed on behavior difficult to generalize as a whole. 

5)  When breed differences have been identified in scientific literature, they do not typically match behavioral descriptions in breed standards which are largely unsubstantiated.

In my opinion, some other key findings worth noting:

1) The study indicates that modern breeds are not closely related to ancient breeds because of a loss of genetic diversity due to historical, cultural and geographic factors.

2) Breeds that were commonly targeted in the media, and politics for human-directed aggression (pit bulls, German Shepherds and Rotweilers), in scientific studies actually consistently scored low on stranger-directed aggression as breed groups.

3) There were significant differences in behavior within breeds.  This is partciularly true in differences between individual dogs within a breed group were bred for show vs field work. I think this is particularly highlighted in this German Study in which German Shepherds from show lines and those from working lines showed to be genetically identifiably different even though they were the same breed of dog. 

4) There is a very interesting section on dog-directed aggression that notes that different dog breeds vary somewhat in their social signaling. In particular, some physical characteristics (snout length, ear structure, reduced skull shape) may influence social signaling and how dogs react to each other. 

5) Many breeds were differently ranked for dog-directed aggression, owner directed aggressiona and stranger-directed aggression -- which suggests that environmental stimuli, rather than breed alone, play a major role in the propensity to exhibit aggression. 

6) Physical characteristics often influence problem-solving tasks. For instance, short-snouted (brachycephalic) breeds seem more likely to use their paws in manipulation tasks where as long-snouted (dolchocephalic) breeds rely less on their paws better allow them to manipulate objects in small spaces with their muzzles.  Physical attributes like muzzle length may also impact a dog's sense of smell which may cause some breeds to be more or less likely to rely on olfactory sense for seeking. 

7) A dog's age, and experience working with humans (often as a result of the function of a dog in its use by the human) have a significant impact on human-responsive tasks and ability to follow human cues.

------------------

I highly recommend anyone interested in this topic to read the entire study.  I think it's an interesting study, and further confirms what experts have said for decades about the influences of canine temperament and behavior. Termperament and behavior is complex system based on a variety of factors including:Breed, Breeding (within breed), upbringing, how they are rewarded/punished, envirornment, history of human interactions (social cues) and that ancient history doesn't play a close role in the behavior we see today.

I think this information largely helps support the notion that the idea of "aggressive dogs" should be based on an individual dog's behavior, and not its breed or appearance. Due to all of the factors that go into shaping behavior, there is a significant differential in behavior and temperament even within specific breeds of dogs.

I think the importance of a dog's ability to read its human cues is an important part of why the National Canine Research Council's differential between a "resident dog" and "family dog" is so important and relevant when assessing major dog attacks.

For more reading:

American Veterinary Society of Animal Behavior blog on the study.

The entire 13-page study: Behavioral Differences among breeds of domestic dogs: Current Status of the Science

Journal of the AVMA -- Co-Occurence of potentially preventable factors in 256 dog bite-related fatalities

Breed Differences in Canine Aggression -- KC Dog Blog

Factors Associated with Aggressive Responses in Pet Dogs -- KC Dog Blog

Academic Paper on the behavioral challenges when dogs are removed from litters too young -- KC Dog Blog

Comments

[Anthony]

Hey Brent, I'm kind of busy today and I was wondering if you could save me some time by just printing the list of peer reviewed, published articles by Merrit Clifton/Clifton Merrit...whatever that guy's name is...that this article cites to. Thanks in advance - ab

[Brent]

LOL. In what I'm sure was merely an oversight none of Clifton's writing was a part of this scientific paper :) Or Colleen Lynn. Or Kory Nelson for that matter.

[EmilyS]

There are so many holes in the research in this paper... studies like this make my head hurt. Why study breed traits and then use generic dog traits (like "aggression".. whatever that means?) Why study breed traits and use shelter dogs, of unknown pedigree, as participants?

Meanwhile, in the real world, police and military organizations tend to choose Malinois, sheepherders tend to choose border collies, duck hunters tend to choose Labs, ranchers with diverse flocks in predator country tend to use Great Pyrenees and their relatives. Etc. It's not just a fashion statement...

and p.s. are we not at all even a little upset at the use of the original CDC report??

[Cheryl Huerta]

Excellent. Thanks for the info. I'm always telling people that science doesn't 'prove' anything but a scientific study/reasearch is only the latest theory/belief. I think this is great evidence to that as it mentions that their findings differ based on where the information came from (survey's or controlled studies).

Thanks!

[Brent]

Emily,

The paper is not primary research, but is solely a secondary research paper. Thus, as part of the study, it reports many of the holes in the existing research -- including some of the holes in breed selection, and the fact that there is no universally agreed upon definition to the temperament traits.

And yes, the reference the CDC report, but also note that drawing definitive conclusions from bite statistics alone is acknowledged to be difficult. As a secondary research paper, I think they would have called into account their credibility by choosing to completely ignore it. As such, they acknowledged it, debunked it, and moved on.

As for your second paragraph, the study seems pretty clear that breed does play a role in behavior (both due to physical and emotional roles) - but that breed alone is not a true indicator as breeding (within breeds), and learned factors through environment and human interaction are also extremely important. Which is why the labs used by duck hunters are VERY different in behavior than the ones used in the show ring or the ones bred for household pets. This finding should in no way be surprising or controversial.

[Anthony]

I think the most glaring flaw is that any paper that leaves out the 3 foremost experts on dog behavior is obviously funded by "Big Pit Bull" and can't be trusted.

[Mona Lindauj]

Some of the studies, like the one by Svartberg's papers in the early 2000's, are not based on questionnaires, but on a corpus of thousands of dogs' scores from a temperament test in Sweden.

Svartberg's results showed some significant differences between breeds, but also large individual differences within breeds.

These are similar results to everybody else's, mostly from questionnaires. I guess that is good, but it does not seem a lot of progress. And I have not seen anything much new since Svartberg's study,

In addition, very very few of these studies include the mastiff type of guard breeds, where you may expect a lot of aggressive behaviors. That is a big hole in this type of science. I know, I am working on it....

[Mona Lindauj]

The study on muzzle length is interesting. True, dogs with almost no muzzle appear to have less odor faculties. But these breeds have so many other physical problems that these results should not count for much. The broadmouthed mastiff types were not included. They are large dogs with a shortened muzzle, but still a muzzle.

There are several bullmastiffs doing extremely well in the sport of Nose-Work, so there is no evidence that these breeds with shorter muzzles have any problems sniffing and finding stuff with their noses.

Where they DO have problems is in endurance. These breeds would mostly have problems working with their noses all day long, like the professional detection dogs. The shorter muzzles impede their heat exchange, so they overheat easier, and thus get tired.

[Brent]

Mona -- yes, not all of the studies were survey-based and Svartberg's papers are very interesting. And yes, difference between, and within, breeds.

And agreed on the muzzle-length study. I had always read that they were sometimes less valuable for nose-work, but also know that many "pit bull" type breeds also do amazingly well as scent work and that greyhounds are generally regarded as sight-hounds - so that would run counter to that conclusion. The endurance issues, particularly in in hot weather, makes sense for sure. I just found the use of paws to be interesting from the study as I'd never read that before, but it does follow what I've unknowingly observed over the years.

[Michelle]

"I think the most glaring flaw is that any paper that leaves out the 3 foremost experts on dog behavior is obviously funded by "Big Pit Bull" and can't be trusted." Posted by Anthony.

This statement does not make sense to me. Who are these three so-called experts that you speak of and what do you mean "Big Pit Bull"? Are you trying to say that Pit Bull advocates pushed this paper in a biased manner, with misleading information. Pleas elaborate, because I am very curious as to your meaning and your evidence.

[Brent]

Michelle -- I'm pretty sure Anthony is being completely sarcastic in that response...

[EmilyS]

Success in nosework is at least as much a matter of work ethic as sheer smelling ability (which all dogs have)... that's the reason why so many of the APBT/AST/SBTs do well at the game. And would be really great detection dogs and SAR dogs if people understood this.

"work ethic" can be a breed characteristic ..............

[Leif Persson]

Genetic studies of dog behaviour - Sweden
http://www.slu.se/en/departments/animalgenetics/research/hund/genetic-studies-of-dog-behaviour/

Measuring temperament in dogs for breeding purposes - Sweden

http://www.slu.se/en/departments/animalgenetics/research/species-dog1/measuring-temperament-in-dogs-for-breeding-purposes/

Svartberg, K. 2005. A comparison of behaviour in test and in everyday life: evidence of three consistent boldness-related personality traits in dogs

http://www.svartbergs.se/lasvart.html

[Lisa Lawstudent]

Responding to Anthony:

Please kindly share with us, who are these, "3 foremost experts on dog behavior?" I would argue that your opinion of who deserves that distinction is as worthy of debate as the study that Brent has shared with us.

Also, isn't it wonderful that Brent doesn't delete your comments as mine would be censured if I attempted to disagree with any opinions expressed on a place that talks about how dogs bite?

Yes, yes. I'm off-topic. B-b-but, he started it! Gigglegigglegiggle!

[Selma]

As a card-carrying member of Big "Pit Bull", I wish Anthony would not let the cat out of the bag, so to speak, regarding the three foremost experts on this subject being left out in the cold.

Harumph!