Abstract
Invasive mammalian predators are a threat to native island biota because native biota evolved without mammalian predation defences. The invasive house mouse Mus musculus is one such invasive predator that thrives in a wide range of habitats, including alpine environments. In southern Aotearoa New Zealand, a mountainous archipelago with substantial alpine area, the invasive mouse is the main alpine rodent. As large, ground-dwelling invertebrates are a key food for alpine mice, I set out to understand mouse impacts on the invertebrate abundance, community composition, and mean body size in two alpine areas. I used pitfall traps to catch invertebrates, then identified the invertebrates to family level and measured their size, and used tracking tunnels plus existing data to understand mouse presence in a correlative study.
I caught invertebrates and tracked mice in alpine Homer Gertrude cirque, northern Fiordland, during the 2023-24 austral summer. This valley has a long history of mouse presence. I recorded a mouse population irruption in response to abundant food resources over the course of four field trips conducted monthly from December to March. I found that mouse presence did not significantly impact total abundance, community composition, or size of all invertebrates. However, a particular vulnerable group, the Anostostomatidae (wētā), were less abundant where mice were present, and there was some evidence to suggest mice were targeting larger-bodied invertebrates. I concluded the history of mouse presence in this area may have already altered the invertebrate community substantially, and as such effects of mice observed over a single season were minimal.
To support this, I conducted a comparison over a larger spatial and temporal scale. Along with my 2023-24 Homer Gertrude data, where invertebrates were suppressed by active mouse predation, I used invertebrate community data from Homer Gertrude in 2020-21, where I hypothesised invertebrates were recovering from intensive mouse predation the previous year, where tracking rates were over 60%. I also collected invertebrate community data with three sampling trips in 2023-24 to Kā-Tū-waewae-o-Tū/Secretary Island, which has alpine habitat but no mice present. Therefore, its invertebrate community was unaltered by mouse predation. I found that vulnerable soft-bodied Amphipoda were more abundant on the island, while Lycosid spiders were less abundant. All three invertebrate communities were significantly different, with the community that was recovering from extensive mouse predation the previous year the least diverse, and the community with current mouse predation similarly diverse, but different from the island community. The average body size of Anostostomatidae was larger on the island, and the recovering mainland group had lower mean body size of carabid beetles. Based on my findings, I concluded that past mouse predation alters an invertebrate community for more than one year, but many years of mouse predation meant that regardless of mouse levels, the invertebrate community on the mainland was less different to each other than to the island invertebrates unaltered by mouse predation. Overall, I showed that mouse predation affects the ground-dwelling invertebrate community in the understudied alpine environment. These clear mouse effects on invertebrates emphasise the need for research into, and implementation of, mouse control in alpine environments.