Behavioral Repertoire Differences in Breeds Deserves Further Study

Some dogs still look much like wolves, and some dog populations have continued to interbreed with wolves. This raises the possibility that dogs that are closely genetically to wolves might also be closer behaviorally. Many of the interactions between domestic dogs are similar to the dominance and submission interactions of wolves within their packs. One team considered the question of whether the variation of signaling repertoire between breeds can be correlated with the degree of dissimilarity in overall appearance from the wolf. They also asked whether those breeds with the most restricted repertoires were more commonly using the behavioral patterns emerging earliest in the development of the wolf cub. Deborah Goodwin, John W.S. Bradshaw, and Stephen M. Wickens, “Paedomorphosis Affects Agonistic Visual Signals of Domestic Dogs,” 52(2) Animal Behaviour 297-304 (February 1997).

The breeds this group studied were the Cavalier King Charles Spaniel, Norfolk Terrier, Shetland Sheepdog, French Bulldog, Cocker Spaniel, Large Munsterlander, Labrador Retriever, German Shepherd, Golden Retriever, and Siberian Husky. This list is from least wolf-like to the most wolf-like, as determined by 14 members of the Association of Pet Behaviour Counsellors, a UK organization (www.apbc.org.uk). The dogs were observed for “agonistic behavior”--behavior involving disputes--in situations where something was present that could bring about conflict—unfamiliar people, food, toys, shelter, and other dogs. Groups of each breed were studied interacting with each other so as to determine what behaviors would be used by members of each breed. The behaviors detected in each breed group are listed below. Growling and displacing, when they are found, generally occur in less than 20 days. Standing over and the inhibited bit occur within 20 to 30 days. The rest of the behaviors occur after 30 days. As can be seen, the breeds closest to the wolf (the list is again from furthest to closest) are the ones with the broadest behavior repertoire. Submissive behaviors are in italics.

• Cavalier King Charles Spaniel (CK) (growl, displace)
• Norfolk Terrier (NT) (growl, displace, stand erect)
• Shetland Sheepdog (SS) (growl, displace, bare teeth; muzzle lick)
• French Bulldog (FB) (growl, displace, stand erect; look away)
• Cocker Spaniel (CS) (growl, displace, stand over, stand erect, body wrestle; look away)
• Large Munsterlander (LM) (growl, displace, stand over, inhibited bite, stand erect, body wrestle; muzzle lick, crouch)
• Labrador Retriever (LR) (growl, displace, stand over, inhibited bite, stand erect, aggressive gape; muzzle lick, crouch, passive submit)
• German Shepherd (GS) (growl, displace, stand over, inhibited bite, stand erect, body wrestle, aggressive gape, bare teeth; muzzle lick, look away, crouch)
• Golden Retriever (GR) (growl, displace, stand over, inhibited bite, stand erect, body wrestle, aggressive gape, bare teeth, and stare; muzzle lick, look away, submissive grin, passive submit)
• Siberian Husky (SH) (growl, displace, stand over, inhibited bite, stand erect, body wrestle, aggressive gape, bare teeth, stare; muzzle lick, look away, crouch, submissive grin, passive submit, active submit)

It would have been better had the genetic similarity to the wolf been used rather than a group of pet behavior specialists, but perhaps that is a future research project. Nevertheless, the result was impressive: the dogs that were the least wolf-like from some perspectives also exhibited the fewest wolf-like patterns of agonistic behavior. Thus, dogs become morphologically more separated from wolves, and arguably showing a higher level of pedomorphosis (retention of juvenile traits into adulthood), they have more limited signaling repertoires, at least in situations involving aggression and submission.

The researchers noted that three of the four gundog breeds—-Cocker Spaniels, Labrador and Golden Retrievers—-retained more wolf-type behavior patterns than their appearance might suggest. They argue that the purpose for which the gundog breeds were developed may have required keeping a fuller range of ancestral behavior patterns than the two breeds derived from shepherding stock German Shepherds and Shetland Sheepdogs. It has been argued that the German Shepherd was developed from shepherding stock with the deliberate intention of producing a physically wolf-like animal, (M. Willis, The German Shepherd Dog: A Genetic History (Witherby 1991)), yet it displayed fewer wolf-type signals than the golden retriever. The researches suspect that once a behavior has been lost from the breed’s repertoire, “it cannot be reconstructed merely by altering the physical appearance of the breed.”

Another observation from the data was that dogs showed less submissive behavior than is found in wolves. They argue that the exhibition of escalated aggression towards other dogs is likely to be less costly than would be the case with wolves because humans intervene to stop conflicts between dogs and the physical cost of failing to display submissive behavior is reduced, allowing for such behaviors to be dropped from the repertoire.

A study published in 2009 looking at the social signaling of puppies in litters of some of the same breeds considered whether the puppies compensated for the morphological limitations on signaling (e.g., not being able to direct their ears because the breed has floppy ears) by more frequent use of those signals that they could make. They found no evidence for this. Keven J. Kerswell, Pauleen Bennett, Kym L. Butler, and Paul H. Hemsworth, “The Relationship of Adult Morphology and Early Social Signalling of the Domestic Dog (Canis familiaris),” 81 Behavioural Processes 376-382 (2009). Nor did they find any new types of signals replacing those that a breed had lost. The researchers acknowledged that “communication modalities, other than those that can be detected by human visual and auditory observation,” might yet exist. It is possible, perhaps probable, that dogs whose shape has been significantly altered from that of the ancestral wolf just have simpler behavior patterns.

This does not mean that King Charles Spaniels are dumb and Siberian Huskies are smart. A recent review of the literature on dogs’ ability to follow human pointing gestures concluded that no reviewed studies had established significant differences between breeds in their ability to understand pointing gestures. Nicole R. Dorey, Monique A.R. Udell, Clive D.L. Wynne, “Breed Differences in Dogs Sensitivity to Human Points: A Meta-analysis,” 81(3) Behaviour 409-415 (July 2009).