Toxic Collar for Control of Sheep-killing Coyotes: a Progress Report

The toxic sheep collar is the most selective method known for killing coyotes that prey on domestic sheep. The concept dates back to the early 1900's, and has been studied at the Denver Wildlife Research Center (DWRC) since 1974. Field tests with sodium cyanide (NaCN) in 1975 were unsuccessful due to repellent properties of the toxicant and to the apparent reluctance of coyotes to attack tethered lambs wearing bulky collars. Coyotes attacked one or more tethered, collared lambs in 7 of the 19 test pastures. In all, 14 collared lambs were attacked. Eight of the collars were punctured but no dead coyotes were recovered. A smaller collar containing diphacinone was field tested in 1976. The diphacinone-filled collars were readily accepted by coyotes and lethal to them, but the slow action (5-16 days between dosing and death) of diphacinone made it difficult to assess the effectiveness of these collars under field conditions. Target flocks containing 1 to 12 collared lambs plus uncollared ewes were placed in 15 fenced pastures from which the larger ranch flocks had been removed after repeated coyote predation. One or more collared lambs were attacked in 11 of the 15 tests. An unknown number of coyotes was killed, and in most tests the subsequent incidence of predation was lower than that before the test. Captive coyotes continued to kill sheep for 4 or 5 days after they received a lethal dose of diphacinone; therefore a faster-acting toxicant is needed. This research has shown that problem coyotes can be killed with toxic collars, but further studies are needed to determine the feasibility of this approach compared with traditional means of control. In most tests to date the frequency of coyote predation has been too low and too irregular to permit effective use of the collar; target flocks were in the field for an average of 10 days before being attacked. The known disadvantages of the method include the need to sacrifice live lambs, the human hazards associated with the use of toxicants under field conditions, and the costs of managing target flocks and other sheep in the problem areas. For many years the Animal Damage Control (ADC) Program of the U.S. Fish and Wildlife Service has attempted to limit predator control as closely as possible to the individual predator or local predator population that is causing or about to cause damage. In keeping with that objective, the ADC Program, through its research arm at the Denver Wildlife Research Center (DWRC), has worked to improve the safety and selectivity of methods for reducing coyote depredations on livestock. Of all the methods available, the toxic sheep collar is the most selective one for killing coyotes that prey upon domestic sheep. Since coyotes typically kill sheep by biting them on or under the neck (Figure 1; Connolly et al., 1976; Nass, 1977; Henne, 1977), the coyotes that attack collared sheep usually rupture the collar with their teeth and thus receive a lethal, oral dose of the toxicant. The toxic collar concept dates back at least to Duncombe (1920) who patented a wire-ring device consisting of two toxicant-filled syringes attached to either side of a sheep's neck. The DWRC has worked with the toxic collar concept since 1974. Earlier, a collar containing 1080 solution had been developed in Texas by Roy McBride, a former employee of the ADC Program and the DWRC. Before the 1972 Presidential ban on the use of toxicants in predator control, McBride used the collars to stop depredations on sheep ranges where problem coyotes had eluded all other means of control. His procedure was to put collars on 15 to 20 lambs and place them with their ewes on the bedground where kills had been taking place. The remaining sheep were corralled or herded away from the trouble zone until one or more collared lambs were killed. On 13 of 14 ranches where these tactics were used, dead lambs with ruptured collars were found and the predation losses immediately stopped (R. McBride, personal communication). On the basis of these results, McBride (1974) patented the leather-sheathed collar containing a rubber bladder to be filled with either a toxicant or aversive solution. In February 1972, all uses of toxicants for predator control by Federal agents or on Federal lands were prohibited by Executive Order No. 11643. This prohibition applied to research as well, but when it appeared that the Order might be modified to permit the use of NaCN in the M-44 device, or sodium cyanide spring loaded ejector mechanism (Matheny, 1976), the DWRC began work on a toxic collar to use this toxicant. The collar research program of the DWRC proceeded through seven phases, as follows: (1) Development of a sheep collar that killed captive coyotes with NaCN. (2) Field tests of the NaCN collar. (3) Pen tests to develop alternate toxicants and to improve the collar configuration. (4) Field tests of the diphacinone collar. (5) More pen tests to develop alternate toxicants, including 1080, and to further improve the collar device itself. 197 (6) Negotiations with the Environmental Protection Agency (EPA) for an experimental use permit for 1080 in the toxic collar. (7) Field tests of the 1080 collar. Phase (1), the development of the NaCN collar, has been reported elsewhere (Savarie and Sterner, 1977), and the field tests with 1080 (phase 7) are now in progress under EPA Experimental Use permit No. 6704-EUP-14 issued in September 1977. Therefore this report is limited to the field tests with NaCN and diphacinone, together with the related chemical screening and collar modifications that led to the issuance of the current experimental use permit. FIELD TESTS OF THE SODIUM CYANIDE TOXIC COLLAR By early 1975, a polyvinylchloride (PVC) toxic collar had been developed at the DWRC. It consisted of 10 packets, each containing 50 ml of 33% NaCN in water, that were assembled and attached to the lamb with nylon cord (Figure 2). Of 12 captive coyotes that attacked sheep equipped with this collar or previous models, 9 were killed. Four of these coyotes attacked tethered, collared lambs, and all 4 were killed (Savarie and Sterner, 1977). Figure 1. Coyotes usually kill sheep by biting them on the underside of the throat. This characteristic attack pattern led to the idea of a sheep neck collar filled with toxicant, so that any coyote biting through the collar would be poisoned. University of California photo by G.E. Connolly. Figure 2. A sodium cyanide toxic collar at a test site in North Dakota in 1975. This collar contained 500 ml of 33% NaCN solution. The lamb is tethered by its right hind foot. USFWS photo by G.E. Connolly. Field tests of the NaCN collar were conducted in North Dakota (13 ranches), Montana (3 sites on 1 ranch), and Texas (3 ranches) from August through October 1975. Most of the work was done in North Dakota because, through previous experience, a key man on the project (R.E. Severson) had intimate knowledge of the coyote problem areas in that state. The intent of each test was to document this sequence of events: (1) Occurrence of frequent and regular coyote predation on sheep; (2) Removal of one or more coyotes with the toxic collar; (3) Reduced incidence of predation after removal of the problem coyotes. Because it would have been neither feasible nor safe to put a NaCn collar on every sheep in a large flock, collars were placed only on a few tethered lambs (Figure 2) in each test. Previous trials without collars had shown that coyotes could be induced to attack lambs that were tethered along the routes habitually traveled by sheep-killing coyotes. In each pasture where the NaCN collars were tested, we attempted to place the collared lambs where they would be encountered by approaching coyotes. The number of collars in each test varied from one to five. Under these test conditions, the NaCN collar was ineffective. Coyotes attacked collared lambs at only 7 of the 19 test sites. In 574 exposure nights (1 exposure night = 1 collared lamb tethered in the field for 1 night) only 14 collared sheep were attacked. Eight of the 14 collars were punctured. Because NaCN takes effect within minutes, we expected to find the carcasses of attacking coyotes within a few hundred yards of the punctured collars. However, no dead coyotes were recovered. Most of the attacks on collared sheep were followed by additional kills of uncollared sheep in the test areas; therefore we believe that the problem coyotes were neither killed nor repelled from the test areas. The failure of the NaCN collar was attributed to 3 causes: (1) reluctance of coyotes to attack tethered lambs; (2) frequent avoidance of collars by attacking coyotes, presumably because of repellent properties of the collar or toxicant, or both; and (3) repellent properties of the toxicant, which caused coyotes to break off the attack before they received a lethal dose.