Article (refereed)Dowding (Singleton et al., 1999;Stenseth et al., 2003).72 Consequently, a range of methods is employed to reduce rodent density and 73 associated damage. This is most commonly done in developed countries using 74 anticoagulant rodenticides, vitamin K antagonists that prevent the synthesis of 75 functional prothombrin and related blood-clotting factors. Extensive use of first-76 generation anticoagulant rodenticides (FGARs) during the 1950s, however, led to the 77 evolution of genetic resistance in brown rats (Rattus norvegicus), with widespread 78 cross-resistance to other compounds (Cowan et al., 1995;Thijssen, 1995). As a 79 result, more potent second-generation anticoagulant rodenticides (SGARs) were 80 developed which have a greater affinity to binding sites, resulting in greater 81 accumulation, persistence and toxicity (Parmar et al., 1987; Huckle and Warburton, 82 1986). 83Given their mode of action, both FGARs and SGARs are potentially harmful to all 84 vertebrates, and so users are expected to adopt measures that limit direct exposure 85 to non-target species. However, the degree to which these preventive measures are 86 adhered to, particularly by non-professionals, is unknown. For example, in Britain 87 some products are readily available to householders who may be less aware of the 88 risks of non-target poisoning and/or less likely to follow manufacturer's guidelines.89 Non-target species may also be deliberately poisoned (Barnett et al., 2006). 90Most studies investigating indirect exposure of non-target species to 91 anticoagulant rodenticides have focussed on the consumption of poisoned rodents by 92 predatory birds and mammals (Newton et al