Coyotes, Sheep and Lithium Chloride

The use of LiCl-treated baits and carcasses has been advocated as a means of controlling coyote predation on sheep through a process known as "aversive conditioning or taste aversion." While some investigators have made well publicized claims of damage reduction through the use of LiCl on sheep ranges, other researchers have experienced difficulty establishing prey aversion in captive coyotes. The conflicting results suggest a need for extensive, carefully controlled research in both pen and field situations before valid conclusions can be reached regarding aversive conditioning as a depredations control method. Aversive conditioning through the use of LiCl-treated baits has recently been proposed as a method of reducing sheep losses to coyotes. Theoretically, if coyotes eat prey-like baits and/or prey carcasses treated with a physiological illness producing drug (LiCl), negative associations will be formed between the baits and subsequent gastrointestinal disorders. The dislike for the bait will then be transferred to live prey, inhibiting future attacks on that prey. Although the concept of druginduced bait aversion has received widespread attention (Riley and Baril, 1976; Riley and Clarke, 1977), the use of LiCl baits to suppress predatory behavior in coyotes (Gustavson et al., 1974) represents an extension of that concept. As used in this paper, the terms aversive conditioning, bait aversion, and prey aversion refer to the process of conditioning coyotes to reject or avoid normally desirable food items as a result of negative physiological experiences following consumption of LiCl treated baits and/or prey carcasses. A coyote is truly averted to a prey or food item only if he refuses to attack or eat the item when it is readily available to him. In order to apply aversive conditioning to the problem of coyote predation on sheep, coyotes must be induced to eat lamb-like baits that make them ill. However, it may be difficult to get all of the coyotes on a range to eat the baits. In most field trials to date, sheep carcasses and/or sheep meat baits containing LiCl placed on the range have been fed on by coyotes. Unfortunately, no study has shown that the baits were eaten by sheep-killing coyotes, and there is reason to suspect that some coyotes will not take carrion baits when live lambs are available. A study of predation on a Montana ranch showed that coyotes rarely returned to feed on old carcasses; instead, new kills were made almost every night (Henne, 1977). Under these circumstances, it is unlikely that the presence of LiCl in carcasses would reduce predation. On the contrary, the use of LiCl in carcasses could reinforce the avoidance of carrion in favor of live prey. In baiting studies on larger areas, investigators were able to mark only 28-34% of the coyotes with 16 baits per square mile (Linhart et al., 1968). Theoretically, illness subsequent to ingestion of a bait should cause coyotes to become averted to similar baits and transfer that aversion to the live sheep. This suggests that baits should resemble sheep as closely as possible, in odor, taste and texture, so that coyotes will become averted to the baits rather than the chemical they contain, since some researchers have shown chemical rather than bait aversion (Burns, 1977; Griffiths, 1978). The extent to which an aversion will transfer from dead baits to live prey is uncertain since prey-killing may involve behavior not directly related to consumption. The final requirement for effective use of the conditioned aversion concept is that the aversion must persist for a long time either with or without periodic reinforcement from additional treated baits. Assuming the above problems can be resolved, taste aversion would present an attractive alternative to traditional lethal methods of depredation control. Interest in aversive conditioning as a means of reducing coyote depredations on sheep was sparked by a report of successful pen tests by Gustavson (1974). Subsequently, Gustavson and his collaborators have claimed generally promising results in field tests conducted in Washington, California, and Saskatchewan (Stream, 1976; Ellins et al., 1977; Gustavson et al., 1974, 1975, 1976, 1977). However, other investigators working in laboratory situations have been less successful. Conover et al. (1977), Lehner and Horn (1977), Burns (1977), and Griffiths (1978) experienced difficulty in establishing an aversion to live prey in captive coyotes. Additionally, some of the methodology and interpretations from field tests of LiCl baits for reducing coyote predation have been questioned (Bekoff, 1975; Sterner and Shumake, 1978). Stream (1976a) originally reported that LiCl baiting was effective at the Honn Ranch, but later revised that conclusion (Stream, 1976b). As noted above, not all investigators have had favorable results. In this report we examine the research conducted to date and offer some possible reasons for the contradictory results. We have also outlined a series of pen and range studies that we feel are needed before sound recommendations can be made regarding the use of LiCl in controlling coyote depredations.