Eradication or control to zero density on near-shore islands?
Resolution Island, Fiordland. Image - Peter McMurtrie.
Lessons from a stoat-removal operation on Resolution Island, Fiordland
The problem of reinvasion: animal movements on land and water
The number of successful pest eradication programmes on islands is growing internationally and resulting in important increases in native biodiversity. Most of these successes have been on isolated islands where reinvasion can only occur with human assistance. Attention is now turning to near-shore islands, which are at high risk of invasion or reinvasion. A common perception is that once an island is cleared of pests, the ‘job is done’ and management can shift from intensive control to periodic surveillance. Eradication on near-shore islands challenges this expectation. It is still possible to achieve ambitious biodiversity outcomes, such as the reintroduction of threatened fauna, but it requires a different mind-set because it involves long-term investment. In this article, Dean Anderson and colleagues use a case study to demonstrate the biological and management complexity associated with removal of invasive pests from nearshore islands. Because of the inherent reinvasion risks and the required long-term commitment for management, they use the concept of ‘control to zero density’ in place of ‘eradication’.
In July 2008 the Department of Conservation (DOC) began a project to eradicate stoats from Resolution Island, Fiordland (Fig.1), to create a sanctuary for endangered species such as kākāpō. Resolution Island is approximately 20,800 ha and 550 m from the mainland. Stoats first swam across in the early 1900s. Stoats continue to persist on the island despite more than 2500 traps having been checked and reset three times a year (January, July and November) since 2008. Given these results, the current management questions are (1) is control to zero density feasible and (2) what effort is required to achieve success? Trapping data collected by DOC were analysed in a two-stage modelling approach. First, data on the number and locations of stoats trapped on the island were used to estimate population size at each stage of the operation, and to estimate immigration rate, population growth rate, and the probability of capturing a stoat. Second, forward-prediction modelling was used to simulate different management scenarios to identify the effort required to achieve control to zero success. The probability of sustained zero density (i.e. no stoats on the island for the final five years of the simulation (2016–2020)) was quantified for each simulated scenario. Immigration was allowed during this time, but for control to zero to be successful new immigrants had to be captured in the subsequent trapping session.
How can control to zero density be achieved?
The predictive simulation of the current trapping regime showed that there was a 17% chance of successful sustained zero density from January 2016 through November 2020 (Fig. 2A). If the trapping programme was stopped, the stoat population would rebound to its starting population size within 2–3 years. If the July trapping session in each year was discontinued, because of funding cuts for example, the population would steadily increase to high levels unsuitable for reintroductions of endangered species (Fig. 2B).
The probability of sustained zero density could be increased by increasing the trapping effort (more traps or trapping sessions), reducing reproductive potential of the population, decreasing immigration rate, or increasing trap success. A scenario in which 264 additional traps were placed in large gaps between trapping lines resulted in a 30% chance of sustained zero density from 2016 to 2020. An alternative increase in trapping effort was simulated by adding an annual trapping session in March, which increased the chance of sustained zero density to 53%. These substantial increases in trapping effort failed to reach high probabilities of zero density (e.g. >90%) because of the persistent risk of immigration. However, the predictive simulations showed that addressing immigration alone would not provide the desired outcome. A 50% reduction in immigration would result in a 43% chance of sustained zero density (Fig. 2C), likely due to high reproductive rates and insufficient trappability. The reproductive potential of the population could be reduced by capturing a disproportionate number of females. A 50% reduction in the population growth rate increased the chance of sustained zero density to 32% (Fig. 2D). Increasing the trappability of both sexes by doubling trap attractiveness resulted in a 47% chance of sustained zero density (Fig. 2E).
Increasing the chance of sustained zero density to 90% or higher was only obtained when three additional management actions were combined; namely. a reduction in immigration, an increase in trappability, and either the addition of an extra trapping session or a reduction in the reproductive potential of the population. For example, when immigration and population growth were both reduced by 50%, and trap success was doubled, the chance of sustained zero density was 90% (Fig. 2F).
Conclusions
Near-shore pest management is complicated by reinvasion, but may benefit native fauna. For control to zero density to be feasible, the following three rules must be met: (1) all pest animals must be put at risk; (2) pests must be removed faster than they reproduce; and (3) immigration must be stopped or new invaders captured before they reproduce. Our predictive-simulation results for Resolution Island indicated that only the first condition has been met. Control to zero density of stoats on Resolution Island is feasible but will depend on a concerted effort to reduce immigration, increase trappability, and include either an additional trapping session or specific targeting of females to reduce the population growth rate. Until such time as threatened or endangered species are reintroduced onto Resolution Island, the programme offers the opportunity to act as a long-term research project into how to conduct and learn from near-shore control to zero operations.
This work was funded by the Ministry of Business, Innovation and Employment and the Australian Invasive Animals Cooperative Research Centre.
Dean Anderson & Andrea Byrom
Peter McMurtrie (DOC)
Kerri-Anne Edge (DOC; Edge-Effect Consulting)
Fig. 1 Location of Resolution and Secretary islands in Fiordland.
Fig. 2 Results of six simulation experiments in which the post-trapping stoat-population size is plotted over time, showing the eff ect of: (A) the current trapping programme; (B) stopping the July trapping session; (C) reducing the immigration rate by 50%; (D) reducing the mean population-growth rate by 50%; (E) increasing the trappability of both sexes by doubling trap attractiveness; and (F) reducing immigration and population growth by 50% and doubling trap attractiveness.
