Evidence is mounting that the earth is entering a sixth mass extinction event. However, species conservation, essential to prevent further loss, is costly and some management actions may have unintended negative effects for the species they aim to protect. Mammalian predator exclusion is a key action implemented for many species. In New Zealand the scale of predator control is set to increase drastically as we progress towards Predator Free 2050 goals. However, the response of some taxa, such as invertebrates, to predator eradication even at a small scale is not well understood. To better use conservation resources and improve outcomes for threatened species, unintended detrimental effects of conservation must be identified and mitigated.

I examined the effect of a newly designed low predator exclusion fence on a population of the Nationally Endangered robust grasshopper (Brachaspis robustus). This flightless, large bodied insect is believed to be suppressed by mammalian predators, so I expected that exclusion of mammals would increase the grasshopper’s abundance. However, I also hypothesised that predator exclusion and the fence structure itself could have unintended effects on other organisms and the environment with possible negative consequences for B. robustus. As such, as well as confirming predators had been excluded, I also monitored skink populations, microclimate, and vegetation change.

Robust grasshopper abundance within the fence did not differ from open sites in the first five months after exclusion fencing was installed, possibly because of a lag in response related to grasshopper lifecycle, or suppression by an apparent increase in bird abundance related to the exclusion fence itself. Following that, B. robust abundance was lower within the fence compared with open sites one year after fence installation, possibly because of meso-predator release of native skinks, or potential fence mediated alterations to microclimate affecting B. robustus development. There was no evidence that lagomorph (rabbit, Oryctolagus cuniculus; and hare, Lepus europaeus) exclusion by the predator fence affected vegetation. However, this conclusion is limited by the timing of vegetation sampling so vegetation changes affecting B. robustus could not be ruled out. This negative response by B. robustus indicates that management in addition to mammal control may be required to conserve this species.

The observed response of B. robustus to increased skink abundance suggests that native meso-predator release has the potential to prevent or slow recovery of threatened invertebrates. These findings have implications for threatened invertebrates within areas subject to wide-scale mammal control, as expected with the approach of a Predator Free 2050. Overall, this result reinforces that a better understanding of the consequences of certain conservation methods on invertebrates is required so that effective and efficient management action can be undertaken to give the best outcome for threatened species.