GUARD DOGS - SLEEP, WORK AND THE BEHAVIORAL-RESPONSES TO PEOPLE AND OTHER STIMULI

Sleep-wake cycles and the responses to naturally occurring stimuli wer e studied in 17 guard dogs; detailed video recordings were made of ten of these dogs. The guard dogs came from two different backgrounds: ei ther they lived permanently on-site, or they were commercially owned a nd regularly brought onto premises, removed before the workers started , and rested off-site at kennels. These two groups were studied to dis cover any difference in sleep-wake patterns, territoriality and respon ses to intruders. During 8 h recordings the guard dogs were inactive f or 84 +/- 17% (mean +/- standard deviation, SD) of the time at night, and 70 +/- 23% of the time during the day. The dogs had 2.7 +/- 2.9 se ssions of activity h(-1) at night, which was significantly less than d uring the day when there were 4.2 +/- 2.7 sessions h(-1) (P < 0.001). The guard dogs barked five times more often during the day (1.06 +/- 1 .6 barks h(-1)) than at night (0.2 +/- 0.4 barks h(-1)), (P < 0.001). Of these barking sessions 70% were apparently stimulated by human acti vity, 29% by activity of other dogs and 1% were of indefinite origin. During the day guard dogs permanently on-site had far more sessions of activity h(-1) (5.8 +/- 2.2) than commercial guard dogs which were re sting off-site (2.1 +/- 0.6) (P < 0.025). However, at night, there was no significant difference between the sleep-wake cycles and activity levels of the two groups. When the dogs were guarding premises after b usiness hours, most (14/17) rested close (< 15 m) to their front fence s yet they were usually out of the obvious view of the observer drivin g past the guard sites. All but one dog showed aggression towards pass ers-by, and especially towards other dogs. However, when 14 dogs were directly challenged at their fences, only three held their ground, six barked and then either backed or ran away, and the other five remaine d hidden. For most of the time spent resting 16 of the 17 dogs lay upo n available soft materials such as car seats, cloth or sand. All the d ogs that were off-site at kennels during the day (4/4) used resting-si tes at night within 15 m of their pick-up and delivery points. Of the dogs remaining permanently on-site, four rested at sites provided by t heir owners and the other nine rested close to where their owners were active during business hours. In order to increase their deterrent va lue, guard dogs should be provided with suitable housing or materials for resting sites in strategic areas so that they may see and be seen by passers-by.