Search the SPREP Catalogue

Considerable bene?ts can be achieved for indigenous biodiversity when invasive vertebrates are removed from islands. In New Zealand, two logistically challenging eradications were undertaken, one to remove cats (Felis catus) and the other Paci?c rats (Rattus exulans) from Te Hauturu-o-Toi/Little Barrier Island (Hauturu). Here we document the short- and long-term impacts of these interventions on the biodiversity of Hauturu. We also assess the extent to which predicted outcomes were re?ected in the measured responses for a wide range of species. Short-term impacts of the eradication program encompassed individual mortality for some native species but no measurable impact to populations. In contrast, at least 11 native vertebrates and one native invertebrate species increased in abundance after rat and cat removal. Fifteen of 34 plant species monitored had signi?cantly more seedlings on Hauturu after rat eradication compared with control islands, indicating future changes in forest composition. Several native species previously not recorded on the island were discovered, including the New Zealand storm petrel (Fregetta maoriana) (formerly considered extinct), the forest ringlet butter?y (Dodonidia helmsi) and eight species of aquatic invertebrate. The chevron skink (Oligosoma homalonotum) has been found in increasing numbers and tuatara (Sphenodon punctatus), raised in captivity on the island, are now re-established and breeding in the wild. These results illustrate an island gradually recovering after a long period of modi?cation. We conclude that more success stories such as Hauturu must be told if we are to allay the public’s concerns about such eradication campaigns. And more public support is required if the conservation community is to tackle invasive species at a scale commensurate with the threats they pose.

Invasive rodents are successful colonists of many ecosystems around the world, and can have very flexible foraging behaviours that lead to differences in spatial ranges and seasonal demography among individuals and islands. Understanding such spatial and temporal information is critical to plan rodent eradication operations, and a detailed examination of an island’s rat population can expand our knowledge about possible variation in behaviour and demography of invasive rats in general. Here we investigated the movements and survival of Pacific rats (Rattus exulans) over five months on sub-tropical Henderson Island in the South Pacific Ocean four years after a failed eradication operation. We estimated movement distances, home range sizes and monthly survival using a spatially-explicit Cormack-Jolly-Seber model and examined how movement and survival varied over time. We captured and marked 810 rats and found a median maximum distance between capture locations of 39 ± 25 m (0–107 m) in a coastal coconut grove and 61 ± 127 m (0–1,023 m) on the inland coral plateau. Estimated home range radii of Pacific rats on the coral plateau varied between ‘territorial’ (median: 134 m; 95% credible interval 106–165 m) and ‘roaming’ rats (median: 778 m; 290–1,633 m). The proportion of rats belonging to the ‘roaming’ movement type varied from 1% in early June to 23% in October. There was no evidence to suggest that rats on Henderson in 2015 had home ranges that would limit their ability to encounter bait, making it unlikely that limited movement contributed to the eradication failure if the pattern we found in 2015 is consistent across years. We found a temporal pattern in monthly survival probability, with monthly survival probabilities of 0.352 (0.081–0.737) in late July and 0.950 (0.846–0.987) in late August. If seasonal variation in survival probability is indicative of resource limitations and consistent across years, an eradication operation in late July would likely have the greatest probability of success.

Rodent eradications are a useful tool for the restoration of native biodiversity on islands, but occasionally these operations incur non-target mortality. Changes in cereal bait colour could potentially mitigate these impacts but must not compromise the eradication operation. Changing bait colour may reduce mortality of Henderson crakes (Zapornia atra), an endemic globally threatened flightless bird on Henderson Island, Pitcairn Islands, South Pacific Ocean. Crakes had high non-target mortality in a failed 2011 rat eradication operation and consumed fewer blue than green cereal pellets. We examined which cereal bait properties influenced its acceptance by captive Pacific rats (Rattus exulans) on Henderson Island. We held 82 Pacific rats from Henderson Island in captivity and provided them with non-toxic cereal bait pellets of varying properties (blue or green, moist or dry). We estimated the proportion of rats consuming bait using logistic generalised linear mixed models. We found no effect of sex, females’ reproductive status, bait colour or bait moisture on rats’ willingness to consume baits. Rats’ bait consumption was unaffected by cereal bait properties (colour or moisture). The use of blue bait is unlikely to affect future eradication operational success but may reduce non-target mortality of Henderson crakes. Timing cereal bait distribution in relation to precipitation may also reduce crake mortality without compromising palatability to rats.

Rodent eradications in tropical environments are often more challenging and less successful than those in temperate environments. Reduced seasonality and the lack of a defined annual resource pulse influence rodent population dynamics differently than the well-defined annual cycles on temperate islands, so an understanding of rodent ecology and population dynamics is important to maximise the chances of eradication success in the tropics. Here, we report on the recovery of a Pacific rat (Rattus exulans) population on Henderson Island, South Pacific Ocean, following a failed eradication operation in 2011. We assessed changes in the rat population using capture rates from snap-trapping and investigated seasonality by using capture rates from live-trapping. Following the failed eradication operation in 2011, rat populations increased rapidly with annual per capita growth rates, r, of 0.48–5.95, increasing from 60–80 individuals to two-thirds of the pre-eradication abundance within two years, before decreasing (r = -0.25 – -0.20), presumably as the population fluctuated around its carrying capacity. The long-term changes in rat abundance may, however, be confounded by short-term fluctuations: four years after the eradication attempt we observed significant variation in rat trapping rates among months on the plateau, ranging from 36.6 rats per 100 corrected trap-nights in mid-June to 12.6 in late August. Based on mark-recapture, we also estimated rat density fluctuations in the embayment forest between 20.4 and 42.9 rats ha-1 within one month in 2015, and a much lower rat density on the coral plateau fluctuating between 0.76 and 6.08 rats ha-1 in the span of two months. The causes for the short-term density fluctuations are poorly understood, but as eradication operations on tropical and subtropical islands become more frequent, it will be increasingly important to understand the behaviour and ecology of the invasive species targeted to identify times that maximise eradication success.

Attempts to eradicate invasive terrestrial arthropods are often regarded as gambles. They o?er the possibility of long term freedom from a pest but are usually confronted with substantial uncertainty and come with a range of technical, economic, environmental, social and political risks. Few guidelines are available for evaluating eradication attempts against terrestrial arthropods. Here, we build on scienti?c literature, including six criteria previously developed to evaluate the feasibility of vertebrate eradications, and our own experiences to de?ne nine criteria that are intended to both assist experts to evaluate proposed arthropod eradication attempts and guide attempts that are underway. The criteria are straight forward and easily interpreted, though evaluating them for a particular programme relies on expert group assessment that will often benefit from rigorous supporting statistical and/or modelling analyses.

Pieris brassicae, large white butter?y, was ?rst found in New Zealand in Nelson in May 2010. The Ministry for Primary Industries (MPI) responded with a monitoring programme until November 2012 when the Department of Conservation (DOC) commenced an eradication programme. DOC was highly motivated to eradicate P. brassicae by the risk it posed to New Zealand endemic cress species, some of which are already nearly extinct. DOC eliminated the butter?y from Nelson in less than four years at a cost of ca. NZ$5 million. This is the ?rst time globally that a butter?y has been purposefully eradicated. Variation in estimates of bene?ts, costs, the e?cacy of detection and control tools, and the probability of eradication success all contributed to uncertainty about the feasibility. Cost bene?t analyses can contribute to assessing feasibility but are prone to inaccurate assumptions when data are limited, and other feasibility questions are equally important in considering the best course of action. Uncertainty does not equate to risk and reducing uncertainty through data gathering can inform feasibility and decision making while increasing the probability of eradication success.

Tiritiri Matangi Island in the Hauraki Gulf, Auckland, New Zealand is a 220 ha restoration island managed by the Department of Conservation as an open sanctuary. Following eradication of the only mammalian predator, the Paci?c rat (Rattus exulans) in 1993, a variety of threatened birds, lizards and a giant invertebrate have been transferred to the island. In March 2000, Argentine ant (Linepithema humile) (Hymenoptera: Formicidae) was discovered and delimiting surveys revealed a 10 ha infestation. Managers were concerned that the ant could have signi?cant negative impacts on invertebrates, birds and lizards. Early surveys con?rmed a dramatic decline in all other ant species within the infested area. In February 2001, an eradication programme commenced with paste baits (a.i. 0.01% ? pronil) spread manually in a 2 m × 3 m grid over the entire area. The second year employed a 1 m × 3 m spacing. A second incursion part way through the programme extended the area to 11 ha. The same toxic bait was used throughout the programme to kill residual colonies and a non-toxic version was used as a lure to intensively monitor progress. Eradication was declared in 2016. Critical parts of the programme included detection of post treatment survivors and the level of e?ort required to con?rm successful eradication. New treatment techniques were developed to kill the last small nests by placing toxic baits inside vials on the ground to prolong bait life. Such nests exhibited non-invasive behaviour, short foraging distances, and were prone to disturbance leading to foraging cessation. Bait densities and ?eld placement were critical to success. Sites with residual nests were deemed free of Argentine ant once there had been no detections over three consecutive years of ongoing monitoring. With few successful Argentine ant eradications in the world the techniques used here can inform and improve success rates for other ant eradication attempts.