Analysis of simulated trap counts arising from correlated and biased random walks

Abstract

Traps are routinely used in insect ecology, conservation, and pest control, but the understanding of trap counts remains limited. A well developed theory only exists for non-baited traps (e.g. pitfall traps) and the simplest animal movement modes, such as Brownian motion, but not for more complex or realistic situations. In particular, important questions as to how the trap counts may differ in case of a baited trap and what its effect can be on the population distribution in the domain where the trap is installed are largely open. In order to bridge this gap in our knowledge, here we use straightforward yet powerful simulation framework of individual-based modelling. A baited trap has a strong effect on the animal movement pattern changing it from the Correlated Random Walk to the Biased Random Walk. This, in turn, is shown to have a dramatic effect on the trap counts. We show that a baited trap can introduce strong heterogeneity into the spatial population distribution, hence resulting in spatiotemporal pattern formation. We also consider a system of multiple traps and show that the trap efficiency can decrease if the traps are installed close to each other.