Probabilistic approach reveals the toxicity threshold values of free-living raptors in Great Britain, United Kingdom, for the lethal effect of second-generation anticoagulant rodenticides

•Second-generation anticoagulant rodenticides (SGARs) are recognised as an effective tool for rodent pest control. However, the use of SGARs enhances their exposure of non-target species and can cause lethal and sublethal effects on these species, including free-living raptors. Although certain liver SGAR residue concentrations from laboratory experimental studies have been considered as conventional thresholds, determining the threshold values for free-living predatory birds remains a challenge. In this study, we estimated the toxicity threshold values for liver SGAR residues associated with lethal SGAR poisoning (i.e., mortality by coagulopathy due to SGAR contamination), using a probabilistic modelling approach with data of red kites (Milvus milvus) and common buzzards (Buteo buteo) from Great Britain, United Kingdom. We also assessed factors influencing the relationship between lethal SGAR poisoning and liver SGAR concentrations. •From our results, we suggest 55 and 40 ng/g wet weight of the sum of liver SGAR residues as the toxicity threshold values for red kites and buzzards, respectively, assuming 1% of extra mortality as a negligible effect on the population size. The estimated threshold values did not significantly differ by the sex, age class, season, or British country. However, the threshold values for red kites, but not for buzzards, significantly increased over years, which suggests a decrease in their susceptibility to lethal toxicosis. The reason for this trend is unclear, although improved nutrition or physical conditions of red kites could be possible factors. Further studies are still needed to understand the impacts of SGAR exposure on free-living predatory birds.