Effects of altitude, seedfall and control operations on rat abundance in South Island Nothofagus forests 1998–2016
- Centre for Integrative Ecology, School of Biological Sciences, University of Canterbury, Private Bag 4800, Christchurch 8140, New Zealand
In New Zealand, ship rats (Rattus rattus) have been implicated in many extinctions, declines, and range contractions of native birds, so ship rats are an important target of predator control. The outcomes of ship rat control operations are difficult to predict due to other factors which affect rat populations including altitude, Nothofagus seedfall, and control of other mammalian pests, particularly brushtail possums (Trichosurus vulpecula) and stoats (Mustela erminea). Here we used long-term rat tracking-tunnel data 1998–2016 from seven Nothofagus forest sites in the northern South Island to address three questions: (1) How do rat tracking rates vary with altitude and Nothofagus seedfall? (2) Which forms of rat control are more effective at reducing rat tracking rates? (3) Is there evidence for mesopredator release of rats in Nothofagus forests when stoats are controlled by trapping? Analysis with binomial GLMMs found that rat tracking rates significantly declined with altitude and increased with Nothofagus seedfall, especially during high-seed years in 2000, 2006, and 2014. Diphacinone, and especially brodifacoum and aerially applied 1080, significantly reduced rat tracking rates, whereas intensive snap-trapping did not. Contrary to earlier studies from North Island forests, we found that rat tracking rates increased significantly with mustelid trapping, suggesting mesopredator release of rats following stoat control. Therefore, in Nothofagus forests where rats are present, land managers should consider the relative threats to native wildlife from stoats and ship rats when deciding whether to trap stoats in an area without effectively controlling ship rats. This study highlights the value of long-term data sets for identifying relationships that may otherwise go undetected.