Genetic distance from wolves affects family dogs' reactions towards howls

Abstract

Domestication dramatically changes behaviour, including communication, as seen in the case of dogs (Canis familiaris) and wolves (Canis lupus). We tested the hypothesis that domestication may affect an ancient, shared communication form of canids, the howling which seems to have higher individual variation in dogs: the perception and usage of howls may be affected by the genetic relatedness of the breeds to their last common ancestor with wolves ('root distance') and by other individual features like age, sex, and reproductive status. We exposed 68 purebred dogs to wolf howl playbacks and recorded their responses. We identified an interaction between root distance and age on the dogs' vocal and behavioural responses: older dogs from more ancient breeds responded longer with howls and showed more stress behaviours. Our results suggest that domestication impacts vocal behaviour significantly: disintegrating howling, a central, species-specific communication form of canids and gradually eradicating it from dogs' repertoire.

Fig. 1. Effect of the interaction of Root distance/Breed group and age on the Reply scores (howling moaning, yelping, and bark-howling response).

In a, X-axis shows the values of breeds’ genetic distance from wolves; the Y-axis shows the scores of the PCA factors, and the dots represent partial residuals, controlling for the effects of other variables in the model besides the plotted ones. The different types of lines represent the age categories: −1SD = 1.89 years, mean = 4.56 years, +1 SD = 7.23 years of age. In b, X-axis shows the age as a scaled variable, the Y-axis shows the scores of the PCA factors, and the dots represent the partial residuals from the model. The different types of lines represent the breed categories (ancient, modern).

Fig. 2. Effect of the interaction of Root distance and Playback part/Age on the frequency of stress behaviours.

In a, X-axis shows the values of breeds’ genetic distance from wolves; the Y-axis shows the number of stress behaviours of the subjects, and the dots represent partial residuals. The different types of lines represent the playback parts (Solo1, Chorus, Solo2). In b, X-axis again shows the values of breeds’ genetic distance from wolves, the Y-axis shows the scores of the PCA factors, and the dots represent partial residuals, controlling for the effects of other variables in the model besides the plotted ones. The different types of lines represent the age categories: −1SD = 1.89 years, mean = 4.56 years, +1 SD = 7.23 years of age.

Fig. 3. The setup of the testing room.

The owner sat on a chair, listening to music  through headphones, and reading a book during the test. The dog moved freely in the room, and the stimulus was played from a speaker hidden behind the opaque, removable wall. Drawn markers served for coding the dogs’ proximity to the owner, sound source, and exit.

Fig. 4. Construction of the howling stimulus.

It consisted of three parts in a defined order: Solo1, Chorus and Solo2.