Habitat structure mediates the non-lethal effects of predation on enclosed populations of house mice

1. Prey behavioural changes in response to predation risk can result in significant effects on prey body growth rates and reduced reproductive output, with resultant impacts on prey population dynamics. This paper examines the influence of habitat structure on these non-lethal impacts of predation using a model, field-based experimental system, with house mice as prey. 2. Three treatments were employed in eight 50 x 50 m pens that contained mice, but allowed access to a suite of free-living vertebrate predators, which included feral foxes, feral cats and native raptors: a treatment where the natural grassland vegetation in the pens was maintained at a height < 10 cm; a treatment where small, felled cypress pine trees covered with wire netting were added to low grassland vegetation to create refuge areas covering 10-15% of the area in a pen; and a treatment where predators were excluded from a 25 x 25 m section of some pens with an underlying grassland structure. A 5 x 5 grid of felled trees was added to grassland and predator-exclusion pens to allow assessment of mouse behaviour. 3. Mice in grassland pens avoided open areas, had lower body growth rates, and began breeding later in spring than mice in both predator-exclusion areas, where they foraged more readily in the open, and in refuge pens, where mice avoided open areas but had safe access to supplementary food located within the refuge. These results occurred despite mouse population densities being much lower in grassland pens, and presumably competition for food being much less, compared with under the other treatments. 4. The results indicate predators can have significant non-lethal impacts on prey, and these effects can be mediated by habitat structure.