Inadequate socialisation, inactivity, and urban living environment are associated with social fearfulness in pet dogs
Abstract
Problematic behaviours are severe welfare issues for one of the world’s most popular pets, the domestic dog. One of the most prevalent behavioural problem that causes distress to dogs is social fearfulness, meaning fear of conspecifics or unfamiliar people. To identify demographic and environmental factors associated with fear of dogs and strangers, logistic regression was utilised with a large dataset of 6,000 pet dogs collected through an owner-filled behavioural survey. Social fearfulness was associated with several factors, including urban environment, poor socialisation during puppyhood, infrequent participation in training and other activities, small body size, female sex, and neutering. In addition, we identified several breed differences, suggesting a genetic contribution to social fearfulness. These findings highlight the role of inadequate socialisation, inactivity, and urban living environmental in fear-related behavioural problems in dogs. Improvements in the management and breeding practices of dogs could, therefore, enhance the welfare of man’s best friend.
Introduction
Fear is a major welfare problem in pet dogs1. As a transient feeling aroused by specific stimuli, fear is a normal, fundamental emotion conserved among species which aids an individual to survive from threatening situation2,3,4. Fearfulness, on the other hand, is a personality trait5. If fear is excessive, prolonged, or generalised in nature, fearfulness becomes a behavioural problem which can interfere with the normal performance of the dog, causing high levels of distress or anxiety and increasing the risk of diseases or even decrease lifespan1,6. Moreover, problematic behaviours might have a negative valence not only on the wellbeing of the dog but also on the wellbeing of its owner6,7,8,9. Undesirable behaviours, such as excessive fearfulness, are the leading cause for relinquishment or even euthanasia of pet dogs worldwide10,11,12,13,14. In the worst case, fearful dogs may resort to aggression and cause public health threats15,16,17, highlighting the gravity of canine fearfulness.
Based on the stimulus eliciting fear, fearfulness can be divided into two separate categories in dogs: social and non-social fearfulness5. Social fearfulness encompasses fear of conspecifics or unfamiliar people, whereas non-social fearfulness includes fear of different stimuli, such as loud noises, novel situations, or heights and surfaces. Fearfulness is one of the most frequent canine behavioural problems as the prevalence of fearfulness ranges from 26.2% even to 44%16,17,18,19, and around 10–19% of dogs show fear of strangers or dogs17,19,20.
Many behavioural traits are complex. Multiple genes with only small effects21, a plethora of environmental factors22, and the complex interplay between these genetic and non-genetic factors23all contribute to the development behaviour. Fearfulness in dogs is moderately heritable and associated with some candidate loci and genes24,25,26,27,28in addition to environmental effects22,29,30,31,32,33. However, more research is needed to reveal the genetic risk variants and environmental factors associated with social fearfulness in dogs. Enhanced understanding of these genetic and environmental factors interacting together, leading to the onset of a behavioural problem, would give us tools to better recognise, manage, and prevent these conditions.
As a part of a larger population-based canine behavioural survey with over 13,700 participants, the aim of this study was to investigate the demographic and environmental factors associated with social fearfulness in Finnish pet dogs. By identifying the factors increasing the risk for social fear-related problems, we could improve the wellbeing of pet dogs.
Results
Study cohort and demographics
We studied the demographic and environmental factors associated with fear of dogs and fear of strangers in datasets of 5,973 and 5,932 dogs, respectively. In the ‘fear of dogs’ data, the numbers of non-fearful and fearful dogs were 4,806 and 1,167, respectively. The age of the dogs in this dataset varied from 2 months to 17 years (mean age 4.6 years). In the ‘fear of strangers’ data, the numbers of non-fearful and fearful dogs were 5,036 and 896, respectively. The age of the dogs in this dataset varied from 2 months to 17 years (mean age 4.7 years). 51% of the dogs were females in both datasets. More detailed demographics and the lists of included breeds and the number of individuals per breed are presented in the Supplementary Table S1.
Demographic and environmental factors associated with fear of dogs
Logistic regression analysis identified several demographic and environmental factors associated with fear of dogs, including socialisation score, breed, body size, age, urban environment score, activities/training, daily exercise, and the interaction of sex and sterilisation (Table 1, Fig. 1).
Table 1 Associations between the demographic and environmental variables with fear of dogs and fear of strangers in the logistic regression analyses.
The effects of breed, socialisation, interaction between sex and sterilisation, body size, and living environment on fear of dogs in the logistic regression analysis. (a) Breeds differed in the likelihood of showing fear of dogs, with Chihuahua being the most and Pembroke Welsh Corgi the least fearful breed. (b) Dogs that had less socialisation experiences at the age of 7–16 weeks were more afraid of other dogs. (c) Intact males were less likely afraid of dogs than intact females, but no difference was observed between neutered males and females. In addition, intact individuals were less fearful in both sexes when compared to neutered individuals. (d) Small sized dogs were more likely fearful than medium and large sized dogs. There was also a difference between medium and large sized dogs. (e) Dogs living in a more urban environment had a higher likelihood of showing fear of dogs. Grey lines (b,e) and error bars (a,c,d) indicate 95% confidence limits. N = 5,973.
In the logistic regression model, socialisation score, body size, and breed had the strongest associations with fear of dogs. Dogs with less socialisation during puppyhood were more likely to show fear of dogs (χ2= 82.91, DF = 1, p < 0.001) (Table 1, Fig. 1). Small dogs were more likely fearful when compared to both large (OR = 3.29, p < 0.001) and medium sized (OR = 2.04, p < 0.001) dogs, and there was also a significant difference between medium and large dogs (OR = 1.61, p < 0.001) (Supplementary Table S2, Fig. 1). We discovered behavioural differences between breeds. Chihuahua, Shetland Sheepdog, and Spanish Water Dog were the most fearful breeds whereas Pembroke Welsh Corgi, Cairn Terrier, and Wheaten Terrier were the least fearful breeds (Fig. 1). The largest pairwise differences were seen between Chihuahua and Pembroke Welsh Corgi (OR = 8.64, p < 0.001), Shetland Sheepdog and Pembroke Welsh Corgi (OR = 8.13, p < 0.001), and Spanish Water Dog and Pembroke Welsh Corgi (OR = 7.32, p < 0.001). Cairn Terrier also had a significantly lower likelihood of being fearful when compared to Chihuahua (OR = 0.13, p < 0.001), Shetland Sheepdog (OR = 0.14, p < 0.001), and Spanish Water Dog (OR = 0.15, p < 0.001). Significant pairwise breed differences are summarised in the Supplementary Table S3and all pairwise breed differences are presented in the Supplementary Dataset. Based on previous research on breed differences of social fearfulness, we made an a priorihypothesis that Chihuahua, Jack Russell Terrier, Lagotto Romagnolo, and Shetland Sheepdog would be more fearful than German Shepherd Dog, Golden Retriever, Labrador Retriever, and Staff. Bull Terrier. Indeed, the first breed group was more likely afraid of dogs than the latter breed group (OR = 1.57, p = 0.024) (Supplementary Table S2).
There was an association between the age of the dog and fear of dogs, indicating that dogs from two to eight years old had the highest probability of being fearful, but this likelihood decreased after eight years of age (linear effect: χ2= 8.40, DF = 1, p = 0.004; quadratic effect: χ2= 12.67, DF = 1, p < 0.001) (Table 1, Supplementary Fig. S1). There was a significant interaction between sex and sterilisation of the dog, as intact males were less likely to show fear of dogs when compared to intact females (OR = 0.62, p < 0.001), but no difference was observed between neutered males and females (OR = 1.10, p = 0.807). Moreover, intact dogs were less likely fearful than neutered dogs in both sexes (intact male vs. neutered male: OR = 0.45, p < 0.001; intact female vs. neutered female: OR = 0.76, p = 0.036) (Supplementary Table S2, Fig. 1).
Dogs living in a more urban environment (χ2= 33.46, DF = 1, p < 0.001) (Table 1, Fig. 1), participating less frequently in activities and training, and getting less daily exercise had higher probabilities of being afraid of dogs (Table 1, Supplementary Table S2, Supplementary Fig. S1). Dogs participating in activities only seldom or never were more likely fearful than dogs training sometimes (OR = 1.54, p < 0.001) or weekly (OR = 1.56, p < 0.001). Dogs getting less than one hour of exercise per day were more likely fearful than dogs exercising more than three hours daily (OR = 1.58, p = 0.022). Dogs exercising 1–2 hours per day were also more likely fearful when compared with dogs exercising more than three hours in a daily basis (OR = 1.50, p = 0.005).
Demographic and environmental factors associated with fear of strangers
The best model explaining the difference between fearful and non-fearful dogs included several demographic and environmental factors, including socialisation score, breed, age, sex, sterilisation, body size, urban environment score, activities/training, weaning age, and family size (Table 1, Fig. 2).
The effects of breed, sex, sterilisation, socialisation, living environment, and body size on fear of strangers in the logistic regression analysis. (a) Breeds differed in their likelihood of showing fear of strangers, with Spanish Water Dog being the most and Wheaten Terrier the least fearful breed. (b) Female dogs were more afraid of strangers than male dogs. (c)Intact dogs showed less fear of strangers than neutered dogs. (d) Dogs that had less socialisation experiences in the age of 7–16 weeks showed more fear of strangers. (e) Dogs living in a more urban environment had a higher likelihood of showing fear of strangers. (f) Small sized dogs were more afraid of strangers than medium sized dogs. Grey lines (d,e) and error bars (a,b,c,f) indicate 95% confidence limits. N = 5,932.
Socialisation score and breed had the strongest associations with fear of strangers in the logistic regression analysis (Table 1, Fig. 2). Dogs with less socialisation experiences were more likely afraid of strangers (χ2= 129.71, DF = 1, p < 0.001) (Table 1, Fig. 2). We discovered behavioural differences between breeds. Spanish Water Dog, Shetland Sheepdog, and Chinese Crested Dog were the most fearful breeds whereas Wheaten Terrier, Finnish Lapponian Dog, and Labrador Retriever were the least fearful breeds (Fig. 2). The largest pairwise differences were seen between Spanish Water Dog and Wheaten Terrier (OR = 7.20, p = 0.001), Spanish Water Dog and Finnish Lapponian Dog (OR = 6.53, p = 0.001), Spanish Water Dog and Labrador Retriever (OR = 6.52, p = 0.001), Spanish Water Dog and Jack Russell Terrier (OR = 6.43, p = 0.001), and Shetland Sheepdog and Wheaten Terrier (OR = 6.40, p = 0.001). Significant pairwise breed differences are summarised in the Supplementary Table S5and all pairwise breed differences are presented in the Supplementary Dataset. Based on previous publications, we made an a priorihypothesis that Chihuahua, Jack Russell Terrier, Lagotto Romagnolo, and Shetland Sheepdog would be more fearful than German Shepherd Dog, Golden Retriever, and Labrador Retriever. We detected a significant difference between the groups, with the first breed group being more likely afraid of strangers than the latter one (OR = 1.65, p = 0.031) (Supplementary Table S4).
Dogs showing fear of strangers were more often females (OR = 0.85, p = 0.027) and neutered (OR = 0.62, p < 0.001) (Table 1, Supplementary Table S4, Fig. 2). Body size was also associated with fear of strangers, but the association was evident only between small and medium sized dogs (OR = 1.50, p = 0.01) (Supplementary Table S4, Fig. 2). Additionally, fearful dogs lived in more urban environments (χ2= 14.54, DF = 1, p = 0.001) (Table 1, Fig. 2) and participated less frequently in activities and training. Dogs participating in activities only seldom or never were more likely fearful than dogs participating sometimes (OR = 1.36, p = 0.013) or weekly (OR = 1.44, p = 0.001) (Supplementary Fig. S2).
Age was not associated with fear of strangers (linear effect: χ2= 0.24, DF = 1, p = 0.626; quadratic effect: χ2= 3.27, DF = 1, p = 0.070), and the overall effects of family size and weaning age of the dog were not significant after controlling for false discovery rate (Table 1, Supplementary Fig. S2). However, when examining pairwise comparisons between dogs living in different sized families, we found that dogs living with one (‘single’) or two adults (‘couple’) were less likely to show fear of strangers that dogs living in families with two children (single vs. two children family: OR = 0.67, p = 0.015, couple vs. two children family: OR = 0.74, p = 0.046) (Supplementary Table S4, Supplementary Fig. S2). In addition, dogs weaned later than eight weeks of age had a higher likelihood to be afraid of strangers than dogs weaned at a normal age (from seven to eight weeks of age; OR = 0.80, p = 0.014) (Supplementary Table S4, Supplementary Fig. S2).
Data: 26 febbraio 2020
Authors: Jenni Puurunen, Emma Hakanen, Milla K. Salonen, Salla Mikkola, Sini Sulkama, César Araujo & Hannes Lohi.
Fonte: www.nature.com