4705360_monash_162308.pdf (1.05 MB)
Download file

The role of oxytocin in the ability of domestic dogs (Canis familiaris) to use human social cues and bond with humans.

Download (1.05 MB)
posted on 01.03.2017, 02:58 by Oliva, Jessica Lee
The domestic dog (Canis familiaris) demonstrates attachment/bonding behaviour towards humans, whilst wolves (Canis lupus) do not. Domestic dogs also use humans’ non-verbal social cues to solve problems better than wolves do, even wolves raised in the same manner as domestic dogs. The neuropeptide oxytocin has been implicated in mammalian bonding and non-verbal intelligence and therefore the oxytocinergic system may have evolved in the dog during domestication in such a manner as to enable the formation of human-dog bonds and facilitate human-dog communication. To test this hypothesis three related studies were conducted. The first investigated the influence of intranasally-administered oxytocin on the ability of domestic dogs to perform an object choice task involving a concealed food reward. It was hypothesized that food-finding would be enhanced after the administration of oxytocin. The second study investigated whether owner-perceived level of bonding with, and intelligence of, their dog could predict the dog’s performance on the object choice task. It was hypothesized that dogs highly-bonded to their owners and with owners who perceived their dog to have a high level of intelligence would perform better on the object choice task than the dogs of owners with weaker perceptions of bonding with, and the intelligence of, their dog. The third study investigated whether variation in tandem repeat length close to the oxytocin receptor gene could account for individual differences in performance on the object choice task and for the species difference in performance between dogs and wolves. Seventy-five pet dogs and their owners were recruited for the studies, which involved two testing sessions, 5-15 days apart. An intranasal spray of oxytocin or saline was administered to the dogs in a pseudo-random, counter-balanced order at the beginning of each session. A buccal swab was also taken from the dogs for subsequent genetic analysis and the owners were required to fill out several questionnaires. Forty-five minutes after the intranasal administration, dogs commenced the object choice task which required them to find a hidden food treat using pointing and gazing cues given by the experimenter. It was found that oxytocin improved dogs’ performance on the object choice task when pointing cues were available and that this enhanced performance was maintained for up to 15 days in the absence of further oxytocin administration. Oxytocin also decreased aversion to the gazing cue, whereby dogs actively avoided the gazed-at bowl after saline but performed at chance level after oxytocin. Anxious attachment to pets (measured with the Pet Attachment Questionnaire) negatively predicted performance on the object choice task with pointing cues, whilst perceived contagion of human emotions (measured with the Perceptions of Dog Intelligence and Cognition Survey) positively predicted performance using gazing cues. This suggests that human communication signals may be interpreted differently by dogs owned by anxiously-attached and non-anxiously attached humans. No differences in tandem repeat length close to the oxytocin receptor gene could implicate this gene in affecting performance on the object choice task. However, a species difference in tandem repeat lengths was observed, suggesting that mutations of the oxytocin receptor gene played an integral role in the domestic dog’s evolution from the wolf.


Campus location


Principal supervisor

Alan Lill

Year of Award


Department, School or Centre

Biological Sciences

Degree Type



Faculty of Science