How do owners perceive dominance in dogs?

Dominance is a well-established phenomenon in ethology, however, in connection with dogs, the public often misuses the term. A questionnaire study was launched to investigate the validity of owner-derived estimates of dominance in dog dyads sharing the same household (N=1151). According to the owners, dogs rated as dominant (87%) have priority access to resources (resting place, food, and rewards), undertake certain tasks (defend and lead the group, bark more), display dominance (win fights, lick the other’s mouth less, and mark over the other’s urine), have a certain personality (smarter, more aggressive and impulsive), and are older than their partner dog (all p<0.0001). An age related hypothesis has been suggested to explain formal dominance in dogs; however, we found that the perceived dominance status was a better predictor than age status for 11 of the items examined. Results suggest that dog owners’ estimates of dominance rank correspond to previously established behavioural markers of dominance. Size and physical condition were unrelated to the perceived dominance. Surprisingly, in mixed-sex dyads, females were more frequently rated as dominant than males. For future studies that wish to allocate dominance status using owner report we offer a six-item survey.

also select "Similar" if both dogs fitted the description or "N/A" if the question did not apply

Statistical Analysis
Next, for the dogs that were allocated a "dominant" or a "subordinate" status, based on the 1 5 8 response of the owner to item 1 ("Which of your dogs is the boss/dominant?"), binomial tests 1 5 9 were used to compare the distribution of observations between the dogs for each of the replies 1 6 0 to items 2 to 21. Please note we did not consider dyads where owners indicated that their dogs 1 6 1 were "Similar" in dominance status, or where they marked "N/A" (N=148). Then, for items 1 6 2 (2-21), dyads were also excluded from the analysis pairwise, if the owner marked them as 1 6 3 "Similar" or "N/A" in that particular behaviour or characteristic (sample sizes are indicated in  We examined whether each behaviour/physical attribute was equally likely to occur in 1 6 6 dominants and subordinates (derived from item 1) using a two-tailed test. We lowered the p 1 6 7 level to 0.0023 from 0.05 as suggested by a Bonferroni correction for the 22 comparisons. Binomial tests using Age Status on the full sample 1 7 0 We then repeated the binomial analyses but instead of dominance status, we used the response 1 7 1 of the owner to Age (item 19, "Which of your dogs is older?"), to assess differences between 1 7 2 dogs allocated an "older" or "younger" status (dogs which were "Similar" in age, or that were 1 7 3 marked "N/A", N=72, were excluded). Next, we used two-sample tests for equality of proportions with continuity correction in order to determine which factor (Dominance status 1 7 5 or Age status) best explained the behavioural and demographic differences between the dogs. Binomial tests on the mixed-sex and same-sex dyads In order to examine any effect of the dyad composition on dominance status allocation, we 1 7 9 created subsets of data including mixed sex dyads (N=491), and same-sex dyads (N=512), 1 8 0 and ran additional binomial tests to inspect possible associations for items 2 -21. We again 1 8 1 adjusted for multiple comparison using Bonferroni correction, and lowered the significance 1 8 2 level to 0.0025. We also aimed to examine how large the difference was in ranks between same-sexed, mixed-  Eighty-seven percent of owners indicated that their dogs differed in their social status, 10% 1 9 2 perceived them as similar, and 3% marked the question as "N/A" (Fig. 1).

9 3
Approximately third (30.1-35.1%) of the dyads were reported to be similar in greeting the 1 9 4 owner, smartness, and physical condition, but only 7.1-7.3% of owners claimed that their 1 9 5 dogs were similar in size and age (percentages are higher and lower than the 1 9 6 mean(16.1)±SD(8.6), Fig. 1).

9 7
Every 4 th -5 th owner did not link a particular dog to the items lick mouth, fight, overmark, and 1 9 8 aggressive (16.2-24.3%), but most respondents (98.6-99.3%) assessed differences between 1 9 9 their dogs regarding size, age and obedience (percentages are higher and lower than the dogs were "Similar" more often than 1 SD above mean (>24.7%) are indicated with #. Items 2 0 5 where owners indicated "N/A" more often than 1 SD above mean (>16.1%) are marked with 2 0 6 *. Item numbers are in brackets. Sample sizes are indicated in the table below the graph. We tested which items (from items 2-21) were associated with the perceived dominance rank 2 1 0 (item 1). The binomial tests revealed that dogs the owners considered as dominant (i.e. the 2 1 1 "boss" at home, item 1) bark sooner/more, lick the other's mouth less, eat food and obtain 2 1 2 rewards first, win most fights, and walk in the front during walks. They more often obtained 2 1 3 better resting places, marked over the other's urination, and defended the group in case of 2 1 4 perceived danger. "Dominant" dogs were also reported to be smarter, more aggressive, and 2 1 5 more impulsive, than their partner dog, and they were more often the older dog in the dyad  Table S1 for an overview of the results). Twelve items predicted Age status. Six in the same direction as the "dominance" status (bark, 2 2 5 lick mouth, fight, resting place, defend group, and smart), one in the opposite direction, as 2 2 6 owners found older dogs to be less impulsive, but "dominant" dogs more impulsive.

7
Age but not dominance predicted five items. Older dogs bark more, play with the ball less, 2 2 8 greet the owner less, are in worse physical condition, are larger, and are less often intact than 2 2 9 their partner dog, according to the owners (p < 0.001). Dominance status was a stronger 2 3 0 predictor of 11 items in comparison to age status (For statistical details see Table S1). In mixed-sex pairs (N = 491), females were more often dominant over males (57% females, 2 3 4 binomial test z = 3.249, p < 0.001). There was also a higher proportion of neutered individuals 2 3 5 compared to intact (58.7% neutered). As in the full sample, dominant individuals were more 2 3 6 often older than the subordinates (N=296 dyads, 65% older, binomial test z = 6.38, p < 0.001).

3 7
All of the remaining items that were found to describe individuals rated as dominant in the 2 3 8 full sample were also significant after Bonferroni correction in the mixed pairs subsample, apart from overmark which were the same between "dominant" and "subordinates". However, when "dominants" were more closely examined regarding their sexes, we found 2 4 1 that when a male was dominant in a mixed-sex pair, he more often marked over his female 2 4 2 partner, defended the group in case of perceived danger, and he was often larger in size than  P  r  o  p  o  r  t  i  o  n  w  i  t  h  w  h  i  c  h  t  h  e  c  h  a  r  a  c  t  e  r  i  s  t  i  c  w  a  s  a  s  s  o  c  i  a  t  e  d  w  i  t  h  t  h  e  "  d  o  m  i  n  a  n  t  "  s  t  a  t  u  s the female subordinate. When a female was the dominant individual, she was more often 2 4 4 neutered than when the male was the dominant (female neutered 72%, male neutered 51%).

4 5
Please refer to Supplementary Table S2 and S3 for more information. In same-sex pairs (N = 512, 48.5% neutered) there was no significant difference between the reported to be more aggressive, impulsive and smarter than subordinates were. In a previous 3 5 0 study aggression towards people, controllability, and leadership were also associated with  Aggressivity increase the likelihood of exhibiting dominance via agonistic interactions.

5 3
Depending on whether dogs in the dyad affiliate or not, they can form formal or agonistic dominance (see also (de Waal, 1989)). Unfortunately in our study affiliation was only 3 5 5 measured using the item "lick mouth", which is also a signal of submission, therefore we 3 5 6 could not distinguish between these two dominance types (Trisko & Smuts, 2015).

5 7
Impulsivity, without utilising a multi-dimensional assessment, (Wright, Mills & Pollux, 2011) 3 5 8 is difficult to distinguish from aggression, although it can alternatively co-vary with anxiety  We also did not include items on affiliative behaviour in the questionnaire, so it was not  Owner estimates of dominance rank corresponded to previously established behavioural 4 1 2 markers of dominance displays, which supports that dominance relationships are robust and 1 3 reward, walks at the front, acquires the better resting place, defends the group, and is more 4 1 9 aggressive. Asking which dog wins fights or which dog licks the mouth of the other might 4 2 0 also be useful, as both were highly predictive of owner perceived social status if they did 4 2 1 occur, in approximately 70% of cases. We thank Borbála Turcsán for her contribution to the design of the questionnaires, and for her  The procedures applied complied with national and EU legislation and institutional    The authors declare that they have no conflict of interest.