Unexpected consequences of vertebrate pest control: predictions from a four-species community model

Although indirect effects are important structuring forces in ecological communities, they are seldom considered in the design of pest control operations. However, such effects may cause unpredicted and deleterious changes in other populations that could reduce or even negate the benefit to endangered species for which control is undertaken. Furthermore, the complexity and nonlinearities inherent in interacting ecological communities may cause thresholds in the strength of pest control, on either side of which indirect effects could vary greatly in their magnitude and desirability. We constructed a four-species simulation model for a common pest community in New Zealand beech (Nothofagus spp.) forests: house mice, ship rats, stoats, and brushtail possums. When the model was perturbed to simulate common control techniques, marked increases in the abundance of nontarget pest species were observed at the next forest mast. Higher mouse numbers were observed following both toxin (1080) application and rat kill-trapping, and higher rat numbers were observed following stoat kill-trapping, due to a release from predation in all cases. In comparison, a marked decrease in stoat abundance at the next forest mast was observed following simultaneous control of rats and mice, due to the effects of decreased prey abundance on the stoat population. For rat control, the size of the indirect effect on mouse numbers increased monotonically with control strength. Because the curvature of the relationship is slight, the relationship between the direct benefits of control and the indirect costs incurred would remain relatively unchanged regardless of the strength of control employed. For simultaneous mouse and rat control, however, high levels of control (as initially simulated) were predicted to cause decreased peak stoat abundance at the next mast event, whereas intermediate and low levels of control were predicted to cause increased stoat abundance. Hence, this study demonstrates two points of concern for pest managers. First, indirect effects of control operations do have the potential to reduce the planned-for benefit. Second, thresholds in the strength of control employed can potentially occur, across which indirect effects switch from being of conservation benefit to being of conservation concern.