BACKGROUND: House mice (Mus musculus) cause significant, ongoing losses to grain crops in Australia, particularly during mouse plagues. Zinc phosphide (ZnP) coated grain is used for control, but with variable success. In a laboratory setting, we tested if mice would (i) switch from consumption of one grain type to another when presented with an alternative and (ii) consume ZnP-treated grains when presented as a choice with a different grain.RESULTS: Mice readily switched from their background grain to an alternative grain, preferring cereals (wheat or barley) over lentils. Mice readily consumed ZnP-coated barley grains. Their mortality rate was significantly higher (86%, n = 30) in the presence of a less-favoured grain (lentils) compared to their mortality rate (47%, n = 29; 53%, n = 30) in the presence of a morefavoured grain (wheat and barley, respectively). Mice died between 4 and 112 h (median = 18 h) after consuming one or more toxic grains. Independent analysis of ZnP-coated grains showed variable toxin loading indicating that consumption of a single grain would not guarantee intake of a lethal dose. There was also a strong and rapid behavioural aversion if mice did not consume a lethal dose on the first night. CONCLUSIONS:The registered dose rate of 25 g of ZnP/kg wheat (∼1 mg of ZnP/grain) in Australia needs to be re-evaluated to determine what factors may be contributing to variation in efficacy. Further field research is also required to understand the complex association between ZnP dose, and quantity and quality of background food on efficacy of ZnP baits.

No abstract

Irregular plagues of house mice, Mus musculus, incur major economic impacts on agricultural production in Australia. The efficacy of zinc phosphide (ZnP), the only registered broadacre control agent for mice, is reported as increasingly variable. Have mice become less sensitive over time or are they taking a sub-lethal dose and developing aversion? In this laboratory study, the sensitivity of mice (wild caught; outbred laboratory strain) was assessed using oral gavage of a range of ZnP concentrations. The estimated LD 50 values (72-79 mg ZnP/kg body weight) were similar for each mouse group but are significantly higher than previously reported. The willingness of mice to consume ZnP-coated grains was determined. ZnP-coated grains (50 g ZnP/kg grain) presented in the absence of alternative food were consumed and 94% of wild mice died. Mice provided with alternative food and ZnPcoated wheat grains (either 25 or 50 g ZnP/kg grain) consumed toxic and non-toxic grains, and mortality was lower (33-55%). If a sublethal amount of ZnP-coated grain was consumed, aversion occurred, mostly when alternative food was present. The sensitivity of wild house mice to ZnP in Australia is significantly lower than previously assumed. Under laboratory conditions, ZnP-coated grains coated with a new higher dose (50 g ZnP/kg grain) were readily consumed. Consumption of toxic grain occurred when alternative food was available but was decreased. Our unambiguous findings for house mice indicate a re-assessment of the ZnP loading for baits used for control of many rodents around the world may be warranted.

Context. Recent studies have shown that the sensitivity of wild house mice to zinc phosphide (ZnP) in Australia is significantly lower than previously assumed, which may account for the reported variability in efficacy of ZnP baits used for broadacre control of house mice in grain-growing regions. Under laboratory conditions ZnP-coated grains with a new higher dose (50 g ZnP/kg grain) were readily consumed but the efficacy of using grains with this higher dose under natural field conditions has not been tested. Aims. To test whether the newly derived ZnP50 (50 g ZnP/kg grain) was more effective under field conditions than the currently registered ZnP25 (25 g ZnP/kg grain) in reducing populations of house mice during a mouse population irruption. Methods. We used a before-after-control-impact (BACI) design to assess changes in mouse population size under different baiting treatments in a replicated field trial. We assessed changes in mouse abundance in recently sown paddocks with either ZnP50 (n = 3) or ZnP25 (n = 3) compared with unbaited control sites (n = 3). Key results. Baiting with ZnP50 led to a median reduction in mouse numbers of >85%. Our modelling showed that under similar circumstances, using the ZnP50 formulation should deliver >80% reduction in population size most (>90%) of the time. In contrast, the current registered bait (ZnP25) achieved approximately 70% reduction in population size, but with more variable results. We would be confident of getting an 80% reduction in population size only 20% of the time by using the currently registered ZnP25 bait under similar field conditions. Conclusions. Consistent with laboratory studies, this study demonstrated the higher probability of achieving a consistently high kill rate under field conditions with the new ZnP50 bait compared with the currently registered formulation (ZnP25). Implications. By using the new ZnP50 bait, farmers are far more likely to get good kill rates, thereby reducing the need for repeated baiting (which is costly and generally ineffective at protecting newly sown crops). Using the new bait should result in lower control costs for farmers and fewer toxic grains being spread to control mice.