Livestock Guardian Coyotes and the Limitations of Research

This coyote showed up in a game camera in July - approximately 15 feet from our sheep! No sheep were harmed in the
making of this photograph!

Recently, several folks sent me a link to a blog post by Randy Comeleo from Oregon State University cooperative extension entitled, "Using Coyotes to Protect Livestock. Wait, What?" Comeleo cites research - some from California - that suggests that the more coyotes we kill indiscriminately in ranching environments, the more livestock losses we'll experience. The post seems to fit with my own perspective - I've often said that my use of nonlethal tools (livestock guardian dogs and electric fence, primarily) is a contractual arrangement with the predators in my environment. I promise not to kill predators indiscriminately; the predators, in turn, must promise not to eat lamb or mutton. While my statement is somewhat tongue-in-cheek, there does seem to be some merit to the approach. Comeleo's article indicates that research supports the idea that a non-sheep-eating coyote fills the niche that might otherwise be filled by an ovinivorous animal.

Like much scientific research, however, there are studies that seem to contradict this perspective. For example, research in Utah in the late 1990s suggests that preventive hunting of coyotes can have benefits for sheep producers in the following grazing season (see "Effect of Preventive Coyote Hunting on Sheep Losses to Coyote Predation," by Kimberly Wagner and Michael Conover (1999)). To me, this conflicting research emphasizes the limitations of pure research to inform the "art" of raising livestock in a rangeland environment. The scientific process tries to control all variables except the one in question; those of us who manage livestock on rangeland in an ever-changing climate are (by necessity) comfortable with variables we can't control.

As someone with a foot in both worlds (ranching and research), I sometimes struggle with this dichotomy. I'm intrigued by the question of whether my livestock guardian dogs displace the coyotes that live near my sheep, or if they simply disrupt their predatory behaviors. From a purely scientific perspective, I suppose, I would need to know how many coyotes live in close proximity. I'd need to know whether these coyotes would prey on sheep that weren't protected by dogs. I'd need to know whether my electric fence provided similar deterrence, and I'd need to know how many lambs were saved by both tools. From my more simplistic (realistic?) rancher perspective, I only care about how much my dogs cost and whether I suffer any death loss due to predators. As a rancher, I'm interested in knowing what my dogs are doing when a coyote (or any other predator) approaches my sheep - I really don't care how many coyotes are in the neighborhood. I'm simply happy that my sheep aren't dead and that I can afford to keep the dogs.

All of this, I suppose, is related to our own personal paradigms. I believe my dogs and my fences will work. When a dog causes a problem, or when sheep break through the electric fence, the filter of my own paradigm influences my response. What can I do to help the dog understand his (her) job? Was the fence working properly? If I felt like dogs and portable electric fence were not effective, I'd likely abandon these tools at the first hint of a problem.

Fortunately, there is a new generation of scientists studying these issues - scientists who combine bio-physical research with a social science foundation. These researchers understand that the real world is messy - we can't answer important questions if we insist on controlling all of the variables. This type of work - which combines scientific rigor with the everyday concerns of on-the-ground practitioners - will be critical answering questions like the one raised by my OSU colleague. Can the right kind of coyotes actually protect my sheep?!