Optimising Control Device Luring Strategies for Invasive Predator Control: A Modelling Approach

Invasive predators pose a serious threat to native biodiversity, with trapping being one of several methods developed to manage and monitor their populations. Many individuals in these predator populations have been found to display trap-shyness, which hinders eradication and results in inaccurate estimates of population size. Lures are used to help overcome trap-shyness by increasing the probability of interaction with the device, but the extent of trap-shyness in wild populations, and the best timing for the introduction of a new lure or combination of lures, are uncertain. A key challenge for wildlife managers is maximising the efficacy of invasive predator control, particularly in relation to baiting and trapping, so that pests are extirpated, or survivors are reduced to a minimum. We first use a Bayesian estimation method to quantify trap-shyness in a population of brushtail possum (<i>Trichosurus vulpecula</i>) in a New Zealand forest; the resulting estimated parameters are then used to calibrate a stochastic, individual-based model simulating the outcomes of different luring scenarios. We show that the brushtail possum population analysed was likely split into a smaller, very trappable group and a larger trap-shy group, with a low mean nightly probability of interaction with traps of 0.28 [0.14-0.56]. Furthermore, our results show that under the assumption of independent attraction levels towards different lures, using a combination of lures simultaneously can result in a greater and faster population knock-down than using a single lure, or than to switch from one lure to another. The model presented can be used to infer wildlife population trappability from capture data, and our simulation results highlight the potential of improved luring strategies to capture individuals in post-control residual populations.