Reactive and risky: The behavioural structuring of ‘dog reactive dogs’

Reactive behaviour is a commonly used term among dog trainers and owners to describe rapid, high-arousal responses by dogs directed towards specific stimuli. When directed towards other dogs, the term ‘dog reactive dog’ (DRD) is often used. There has been little attempt to examine these dogs’ responses within the broader biological framework of agonistic behaviour. We aimed to generate an owner-centred definition of DRDs, then investigate whether the commonly reported reactive behaviours could be used to define specific population subtypes. A dog reactive dog was defined as a dog exhibiting one or more of the following behaviours in response to another dog: barking, growling, snarling (raised lip), lunging, snapping, nipping, biting, whining, and a stiff posture with raised hackles accompanied by intense staring. A questionnaire (n = 1959; 1360 complete responses for reactive behaviour profiles) collected data on the frequency of observation of these behaviours in DRDs, alongside dog demographics and contextual factors. Three components were derived using principal component analysis, labelled: PC1 ’Oral Attack’ (snapping, nipping, biting), PC2 ‘Frustration ’ (barking, lunging, whining), and PC3 ‘Posturing’ (growling, snarling, stiff posture). Hierarchical agglomerative cluster analysis of dogs’ component scores indicated four clusters within the population. Cluster 1 dogs (40% of the sample) exhibited moderate ‘Posturing’ and ‘Frustration’ but very low ‘Oral Attack’ , suggesting lower escalation risk. Cluster 2 (18%) displayed very high ‘Frustration’ , high ‘Posturing’ , and low-moderate ‘Oral Attack’ . Cluster 3 (32%) displayed moderate ‘Frustration’ but high levels of ‘Posturing’ and moderate levels of ‘Oral Attack’ , suggesting a greater tendency for contact. Cluster 4 (10%) showed high scores across all components and represented the highest escalation risk. Demographic associations indicated that source of acquisition and general social reactivity were predictive of cluster membership, whereas sex, neuter status, age, and size were not. Overall, these results demonstrate that DRDs are a heterogeneous group whose behaviours cluster in psychobiologically meaningful ways. The findings provide partial empirical support for the structure of behaviours noted in the ‘Ladder of Aggression’, while highlighting an independent influence from ‘ Frustration ’. Clinical management should be tailored according to cluster type, with lower-risk dogs benefiting from avoidance measures and offering alternate communicative behaviours, and higher-risk groups requiring stricter risk management with a longer-term focus on strategies to reduce ‘ Frustration ’. As a result, an owner-friendly screening tool was developed to estimate a dog’s cluster type and associated management guide. • Working definition of dog reactive dogs (DRDs) developed from owner survey • Reactive behaviours: bark, growl, snarl, lunge, snap, nip, bite, whine, posture • PCA showed 3 components: ‘ Oral attack’ , ‘ Frustration ’, and ‘ Posturing’ • Clustering revealed 4 DRD behavioural subtypes differing in escalation risk • Clusters linked to source and reactivity, not sex, age, neuter status, or size