p. 204

All five coyotes attacked the neck/throat area of collared lambs. Four collars containing 45 to 105 ml of 3% capsicum oleo resin were penetrated during the attacks. Coyotes that bit through collars reacted immediately by head shaking, mouth gaping, muzzle pawing and licking, muzzle rubbing in snow and grass, and snow eating. One coyote pulled the collar from a test lamb without puncturing the collar and then made a throat kill.

p. 205

Coyotes that reacted to collar contents immediately stopped their attacks on lambs for a mean of 21.0 min (n=4, ange= 1-60 min) and then resumed their attacks (Table 2), usually at the sides and rear of the lamb. The single coyote that did not resume attack for 60 min was subsequently retested twice, at two day intervals. During the second test, the test lamb was immediately attacked at various locations and the collar was punctured. Attack was interrupted for 4 min. Including this result with the times of the other coyotes provided a mean latency of 17.6 min (n=5, range= 1-60 min) for all tests with collar punctures and deterred attacks. Two days later, after collar punctures, the coyote attacked a third collared lamb at the sides and rear, but made no attempt to grasp the neck/throat area.

p. 206

Vichos collars briefly interrupted attacks on lambs, but coyote predation was not substantially deterred. This result is consistent with the general finding that sensory repellents do not stop predation by coyotes (Lehner 1987; Linhart 1984). It is worth noting that collars appeared to redirect attacks by coyotes away from the throat, resulting in less efficient killing than would have otherwise occurred.