Context. In New Zealand, the aerial application of toxic baits containing sodium fluoroacetate (1080) can consistently achieve significant reductions in populations of multiple vertebrate pest species including brushtail possums (Trichosurus vulpecula), ship rats (Rattus rattus) and stoats (Mustela erminea). Reductions in house mouse (Mus musculus) populations by 1080 baiting appear less consistent, possibly due to low acceptance of 1080 bait by mice in field conditions.

Aims. We tested the effect of pre-feeding and 1080 concentration on the acceptance of pellet food by mice.

Methods. Wild-caught mice were individually housed and presented with a series of two-choice laboratory feeding tests, using estimates of the daily amount eaten to indicate relative acceptance of different types of pellet food.

Key results. Pre-feeding mice on non-toxic food did not increase their subsequent acceptance of the same food containing 0.15% 1080. Mice showed low acceptance of food containing 0.08 and 0.15% 1080 (by weight), with similar mortality (25%). Acceptance of food containing 1.5% 1080 was also very low in comparison with non-toxic food, although mortality in mice was higher (~66%). In comparison with other concentrations, mice ate comparatively more of food containing 0.001% 1080 with no mortality, although the non-toxic food was still significantly favoured. Presentation of a choice between non-toxic food and food containing 0.08, 0.15 or 1.5% 1080 to mice was followed by a significant decrease in average total daily food intake over the following 2 days. In surviving mice this ‘drop feed’ effect was followed by an increase in average daily intake of non-toxic food over the next 3 days until normal daily intake levels were again reached.

Conclusions. We suggest that wild mice can rapidly identify food containing 1080 and subsequently will avoid it.

Implications. This feeding response partly explains the variable success of 1080 baiting operations against wild mouse populations (M. musculus) in New Zealand.