Validation of modified open field behaviour as a measure of trait anxiety in the dog

•Canine habituation to open field testing over multiple trials was characterised.•Behaviours with high test-retest repeatability identified and tested with diazepam.•Anxiolytic diazepam increases locomotor activity and alters noise stimuli response.•Potential measures of trait anxiety were identified and validated with diazepam.

Trait anxiety may be a predisposing factor for anxiety-related behavioural problems in dogs. This study aimed to measure and quantify trait anxiety using novel behavioural paradigms. Diazepam is an anxiolytic drug which can be used to validate behavioural paradigms that are designed to measure anxiety. Greyhound dogs from a Canine Blood Bank were recruited. Dogs were individually tested in two behavioural paradigms, named the modified open field and the unconditioned substrate preference. The modified open field test consisted of a standard open field test of 10 min duration, followed by six individual 10 kHz 110 dB 1-second tones played every 30 s, interspersed with continuous white noise played at increasing intensity from 10 dB up to 90 dB. For one group of dogs, the test was repeated three times, with two weeks between each test. The greyhounds were pre-treated orally with either an empty gel capsule (sham) or 1 mg/kg diazepam 90 min before the third trial (n = 39). A second group was treated 90 min before a single naïve exposure to the test (n = 24). Video recordings of the trials were analysed using software which tracked the position of the dog over time. Total distance travelled demonstrated a high test-retest repeatability within subjects (Spearman’s r = 0.813, 95% CI: 0.67–0.90, P < 0.001) and was increased on average in both test-naïve (167.6 m, 95% CI: 54.8–280.4, P = 0.005) and test-experienced (139.0 m, 95% CI: 89.1–189.0, P < 0.001) dogs that had been pre-treated with diazepam when compared with sham. A noise tone as a distance-increasing signal over time had a reduced effect on mean midline distance following diazepam pre-treatment when compared with sham in the test-naïve (−0.79, 95% CI: −1.6 to 0.0, P = 0.047), but not to the same extent in the test-experienced (−0.47, 95% CI: −1.5 to 0.6, P = 0.370) dogs. The unconditioned substrate preference test consisted of a room that had a concrete floor with half the flooring covered with plastic (a floor substrate unfamiliar to the greyhounds). There was little effect of diazepam pre-treatment on preference for the novel plastic floor substrate. This study suggests that the anxiolytic effect of diazepam on dog behaviour may be measured through an increase in exploratory behaviour and a reduction in noise aversion. Furthermore, exploratory behaviour and noise aversion may represent aspects of underlying trait anxiety as these behaviours have reliable test-retest repeatability.