Canid livestock predation research has become more robust, but gaps remain

Livestock have been and continue to be an important part of the agricultural landscape in North America. Research on how to mitigate livestock predation by North American carnivores developed quickly during the 20th century. We collected information on 75 field-based experiments published since 1970 that evaluated strategies to mitigate livestock predation by wolves and coyotes. Collected research came from journal articles and grey literature, and we identified 22 mitigation strategies. We developed a ‘robustness index’ to compare each experiment based on its empirical design, temporal/spatial coverage, and sample size. We found the robustness index values increased over time, particularly for lethal mitigation strategies. Overall robustness of research on lethal mitigation strategies was similar to the robustness of research on non-lethal mitigation strategies. Some strategies were not well evaluated as 12 of the 22 mitigation strategies were evaluated only once or twice; some common lethal mitigation strategies (e.g., shooting) were not formally evaluated until the 1990s. We identified some robust assessments of mitigation strategies that reported positive effects (e.g., predator sterilization, protection llamas). In some cases, these were the only evaluation or the only robust evaluation of a strategy. In the few cases where there were multiple robust assessments for a single strategy, the outcomes were inconsistent. No strategies evaluated more than once had consistently high robustness index values and positive outcomes. Importantly, older practices based on less robust research should be re-evaluated or discarded.