Towns, D.R. Eradications of vertebrate pests from islands around New Zealand: what have we delivered and what have we learned?Island invasives: eradication and management
Eradications of vertebrate pests from islands around New Zealand: what have we delivered and what have we learned?
D. R. Towns Research and Development Group, Department of Conservation, Private Bag 68 908 Newton, Auckland 1145, New Zealand. . Abstract
Eradications of invasive mammals have become increasingly complex and expensive. Increased public exposure and involvement in decisions about island eradications mean that conservation scientists must be prepared to justify the benefitsof proposed eradications and defend the science used to measure cause and effect of agents of decline. Here I assess the biological, scientificand political outcomes of eradications on those islands in New Zealand from which all introduced mammal populations have been removed. By 2010, 147 populations of 13 species of vertebrates had been removed from a least 95 islands with a total area of 32,000 ha. Identified benefits to biodiversity were through in situ
recovery, translocations or metapopulation management on the islands. These include improved prospects for 16 species of invertebrates, two species of frogs, three taxa of tuatara (Sphenodon
spp.), 23 species of lizards, 32 taxa of terrestrial birds and 16 taxa of seabirds. The eradications can also be used to test hypotheses about the impacts of invasive species on native ecosystems. Considerable effort has been applied to understanding the effects of Pacific rats ( Rattus exulans
). There are now published accounts of the effects of these rats on plants
, lizards, tuatara and seabirds, often using well designed fieldexperiments. However, the effects of most other invasive vertebrates are poorly documented. Furthermore, impressive accounts of biodiversity achievements obscure potential problems. These include the genetic effects of small relict populations or small founders from translocations. Nonetheless, there has been acceptance of the value of these eradications at the highest political levels, government support for assistance in developing countries, and global export of technologies developed. A deeper understanding of the effects of invasive species, good reporting systems, and frequent communication and defence of benefits will be needed to gain public acceptance of increasingly ambitious projects
Keywords: Biodiversity benefits
, birds, reptiles
, amphibians, invertebrates, plants, cause and efect, invasive mammal
sINTRODUCTION Invasive species are now recognised as major agents of global change (Mack et al. 2000; Simberloff 2003). The effects of invasive species are particularly severe on islands (Paulay 1994) where they are implicated in two thirds of recorded animal extinctions (Cole et al. 2005). On the other hand, there are increasing numbers of successful eradications, especially of introduced mammals. These include exotic foxes from 40 islands covering 210,000 ha in the Aleutian Islands of Alaska (Ebbert and Byrd 2002), 45 populations of introduced mammals from 29 islands in northwestern Mexico (Aguirre-Muñoz et al. 2011) and 21 species of introduced mammals from 17 islands in the Galapagos archipelago off Ecuador (Donlan et al. 2003). The upper limits of areas attempted have risen greatly since the 1990s (Donlan and Wilcox 2008). The eradication of mice (Mus musculus) and ship rats (Rattus rattus) is now being attempted on 3881 ha Rangitoto-Motutapu Island, New Zealand (Griffiths 2011), and the eradication of Pacificrats ( Rattus exulans) has been achieved on 3083 ha Hauturu (Little Barrier) Island, New Zealand (Towns et al. 2006); Norway rats (R. norvegicus) on 11,300 ha Campbell Island, New Zealand (McClelland and Tyree 2002); cats (Felis catus) on 12,800 ha Macquarie Island, Australia; rabbits (Oryctolagus cuniculus) on 3450 ha Norfolk Island, Australia; and in the Galapagos Islands, Ecuador, goats (Capra hircus) on 458,812 ha Isabela Island, and pigs (Sus scrofa) on 58,465 ha Santiago Island (Donlan and Wilcox 2008).
Eradications on large islands are expensive and are likely to include sites with a high public profile,or inhabited by people. For example, since 1996 the Department of Conservation has undertaken ten large and complex island eradication campaigns at a total cost of over NZ $8 million (updated from Broome 2009). Among these, there was intense debate within the scientific community and Maori tribal groups (iwi) over the removal of Pacific rats from Hauturu in 2004 (e.g., Kapa 2003; Towns et al. 2006), which incurred legal costs of at least NZ$ 200,000 (Broome 2009). Elsewhere, eradication attempts have been stiffly resisted on the grounds of unacceptable collateral damage or concerns from animal rights activists
364(Towns et al. 2006). In the UK, a US$1.6 million attempt to remove hedgehogs (Erinaceus europaeus) introduced by the inhabitants of the Uist Islands of Scotland proved ineffectual – at least in the initial years – largely because animal rights activists convinced Scottish Natural Heritage to use live capture and relocation rather than kill trapping (Carrel 2007; Webb and Raffaelli 2008). Such examples pose a dilemma. Because of the extent to which invasive species can disrupt ecological processes and human welfare (Mack et al. 2000), increasingly ambitious eradications of these species should be attempted (Simberloff 2002). But as the public profile of these attempts increases, so does resistance to them, despite likely benefits to biodiversity, native ecosystems, and ultimately human welfare.
Since 1996, eradications requiring toxins in New Zealand have often been publicly notified through the Resource Management Act 1991 (RMA). Proposers must compile an Assessment of Environmental Effects (AEE), which is available for public submissions. The AEE and submissions are then examined by independent commissioners who may reject the application or place conditions on the way the project is conducted. Two key biological questions often arise during this process (pers. obs.). Firstly: “Do the benefits to biodiversity outweigh financial and short term environmental costs?” Secondly: “How good is the evidence for cause and effect between losses of biodiversity and purported agents of decline?”
Neither question is exclusive to eradication attempts on islands. Any attempted eradications should include measures of the benefits to species and ecosystems. In addition, treating the eradications as large-scale experiments should over time illustrate the relationship between introduced organisms and those that they affect (Towns et al. 1997).
In this review I describe the outcomes of eradications of vertebrates from islands around New Zealand and ask how measured outcomes have informed our understanding of the effects of invasive species. I firstsummarise the biological benefits attributable to eradications on islands from which all vertebrate pests have been permanently removed. I then
Pages 364-371 In: Veitch, C. R.; Clout, M. N. and Towns, D. R. (eds.). 2011. Island invasives: eradication and management. IUCN, Gland, Switzerland.
Towns: Eradications of vertebrate pests from NZ islands
Table 1 Number of invasive vertebrate populations removed from 95 islands around New Zealand, their general effects on native biota (King 2005) and the type and quality of evidence of their effects on island ecosystems
Brushtail possum Trichosurus vulpecula
Rabbit Oryctolagus cuniculus
Pacific rat Rattus exulansNo. Opsseedlings of forest plants; wide range of invertebrates; lizards; eggs and chicks of some birdsHarper 2007Single pest
dwelling birds including seabirds Between island comparisons; stable isotopes (seabirds)
spp of native plants; extensive canopy defoliation; predation of invertebrates (e.g., large snails), eggs, nestlings and adult birds (including seabirds)Forest canopy recovery after eradicatio
nGeneral diet in New Zealand Evidence for effects References
Weka 3 1 Invertebrates
, reptiles, ground-
3 0 Foliage, flowers, fruit and bark of >90
12 2 Grasses and shrubs Recolonisation by broadleaved coastal shrubs after eradicationAtkinson 1992
Towns et al. 199
742 26 Foliage, flowers, fruit, seeds andBetween island comparisons of plants, reptiles and seabirds; exclosure experiments with plants; post eradication recovery of invertebrates, plants, lizards, tuatara and seabirds Whitaker 1978; Atkinson 1985; Towns 1991, 2002, 2009; Towns et al. 1997, 2007; Pierce 2002; Campbell 2009; Campbell and Atkinson 1999, 2002; Rayner et al. 2007
Norway rat R. norvegicus26 11 Foliage, fruit, seeds and rhizomes of plants; wide range of invertebrates, lizards; eggs and chicks of some birdsNewman 1986; Allen et al. 1994; Campbell 2002Observed post invasion declines of tuatara; posteradication responses of forest plants
Ship rat R. rattus 6 5 Fruits of native plants; wide range of invertebrates, lizards; eggs, chicks and adults of some terrestrial and arboreal birdsPost invasion declines of invertebrates and forest birds and bats; stable isotopes (forest bids)Atkinson and Bell 1973; Harper 2007; Towns 2009
House mouse Mus musculus 13 4 Seeds of native plants; wide range of invertebrates; some lizards and birdsBetween island comparisons of invertebrates, posteradication responses by invertebrates and lizardsNewman 1994; MacIntyre 2001; Roscoe and Murphy 2005
Stoat Mustela erminea7 5 Invertebrates, lizards and birds; introduced rodents and rabbits Post invasion declines of birdsKing and Murphy 2005
Cat Felis catus
Pig Sus scrofa8 1 Invertebrates, lizards, birds (esp. seabirds); introduced rodents and rabbits
10 1 Fruits and foliage of plants, wide range of invertebrates; frogs and lizards; ground-nesting birds and their eggs; introduced rodents and rabbitsPost invasion declines of birds; post eradication recolonisation by land and sea birds
Exclosures; recovery of seabirds post eradicationFitzgerald and Veitch 1985; Fitzgerald et al. 1991; Girardet et al. 2001, Veitch et al. 2004; K. Baird (pers comm.)
Harper 1983; Coleman et al. 200
1Cattle Bos taurus3 0 Wide range of herbs, grasses shrubs and trees None recorded
Goat Capra hircus
Sheep Ovis aries10 1 Fungi, ferns, grasses and broadleaved shrubs and trees Post invasion destruction of vegetation; diet analysis; post eradication recovery of plant communities
4 0 Grasses and some shrubs Post removal recovery of native herbs and grasses Dilks and Wilson 1979; Meurk 1982; Meurk et al. 1994Sykes 1969; Parkes 1984; Brook 2002; Bellingham et al. (2010b
Island invasives: eradication and management
ask whether the eradications provide less obvious benefits through scientifc knowledge, communication, political support and international uptake.
STUDY SITES Islands used here are those beyond the range of natural recolonisation by the eradicated vertebrates. Successful eradications are those with no recolonisation for two years or more after the original campaign. A few islands have occasional incursions of mammals through natural dispersal, but if these are consistently eliminated on arrival, the site is regarded as permanently clear and is included in the analysis. Guidance about motives for eradications was obtained from legal status of the land, statutory plans and interviews with project managers. Evidence of the effects of invasive species was regarded as available if accessible with search engines such as the Department of Conservation library catalogue, Google Scholar and BIOSIS.
Up to 2010, all invasive mammals and one species of bird had been removed from 95 islands; a total of 147 populations of 13 species of vertebrates within an area of 32,000 ha (updated from data In: Veitch and Bell 1990; Clout and Russell 2006). Eradications on an additional 20 islands (total 4700 ha) of eight species of vertebrates have yet to be confirmed.The most frequently eradicated species were Pacific and Norway rats (Fig. 1), but also included one species of out-of-range flightless predatory bird and one arboreal marsupial (Table 1). Most of the remaining species were farm animals that became feral, although domesticated livestock removed from islands retired as farms were not included in these totals. Assessments of the effects of feral species were complicated by the previous presence of stock on 20 (21%) of the islands, which in most cases were also cleared of forest for agriculture. Additionally, even the forested islands were burned during Maori or early European history (Bellingham et al. 2010a), although they have now had many decades to recover. Furthermore, on 25 (26%) islands, multiple species of terrestrial vertebrates coexisted, with potential for complex interactions between them (e.g., Courchamp et al. 1999, 2000). On the other hand, for most of the earlier eradications, multispecies removals were conducted over long time intervals, with the potential to measure responses between the eradications. Finally, all of the islands are inhabited by introduced birds such as European starlings (Sturnus vulgaris) and blackbirds (Turdus merula) whose effects are unknown. Many such species are now found through the entire archipelago and are assumed to have equal effects across the sample.
Fig. 1 Composition of 147 populations of invasive vertebrates removed from 95 islands around New Zealand.
366BIOLOGICAL OUTCOMES OF ERADICATIONS Species and communities
Given that eradications were designed to protect and enhance depleted biodiversity, what were the benefits? Based on assessments of eradications over the last 20 years, in situ recovery or subsequent translocations to islands now free of introduced mammals around New Zealand improved the long term prospects for at least 16 species of invertebrates and 76 species of vertebrates. The latter included two of the four species of frogs, all three taxa of tuatara, 23 of the 80 species of lizards, 32 of the 73 taxa of terrestrial birds and 16 of the 84 taxa of seabirds (Bellingham et al. 2010a). Furthermore, earlier eradications of goats from Great Island (Three Kings Group) may have enabled the recovery of more than 200 species of plants and up to 30 species of endemic snails (Brook 2002; P.J. de Lange pers comm.; Bellingham et al. 2010b). Similarly, the removal of pigs from Aorangi Island (Poor Knights Group) likely provided benefits for numerous rare species, including 18 species of plants, five species of snails, 13 species of insects, six species of reptiles and two species of birds (Towns et al. 2009b; Bellingham et al. 2010a).
For many species, range contractions have been reversed after eradications as species are either returned to sites they previously occupied or released into new ones as a conservation measure. Excluding planting for island reforestation, translocations alone have involved at least 139 populations of 63 taxa of animals (Fig. 2). The results of species translocated to or between islands must be treated with caution because determining the success of translocations can be difficul. If we use self - sustaining populations as the minimum criterion for success (e.g., Dodd and Seigel 1991), birds have the highest proportion of identified successful translocations to islands after pest eradication 44/72 (61%). The proportion is much lower for invertebrates 3/21 (14%) and reptiles 3/37 (8%). None of the populations of amphibians and seabirds translocated to new islands can yet claim to have met basic criteria for success. In part, lack of data on success relates to the ease of locating released animals. With the exception of terrestrial birds, which often have flexible and high reproductive output, many invertebrates and reptiles are cryptic and difficult to locate at low density. Some, such as tuatara, also have low reproductive output and late age at maturity (Cree 1994). For such species the outcome of translocations may not be measurable for years or even decades after release (e.g., Towns and Ferreira 2001).
Furthermore, aside from at least three known failures (4%), there are also populations (all birds) that are maintained in island environments where they are unlikely
Fig. 2 Composition of 139 translocations of 63 taxa of native vertebrates and invertebrates to islands cleared of all introduced mammals
Towns: Eradications of vertebrate pests from NZ islands
to ever form self-sustaining populations, but where their prospects can be improved away from introduced predators. Examples of these include kakapo (Strigops habroptilus), kiwi (Apteryx spp.), takahe (Porphyrio mantelli) and hihi (Notiomystis cincta). Here success is based on overall increases in metapopulations, even though contributing populations may be very small (see also Bellingham et al. 2010a).
Populations that are expanding after invasive species removals may carry a legacy of past problems. For example, when Pacific rats threatened populations of northern tuatara (Sphenodon punctatus) on Hauturu, the remaining eight adults were taken into captivity to breed until Pacificrats were eradicated in 2004. Since 2006, over 100 tuatara raised in captivity have gradually been released (MacAvoy et al. 2007). This appears to be an exemplary breeding programme but the adult tuatara on Hauturu have lost genetic variation, with potential attendant problems of low fitness (MacAvoy et al. 2007; Miller et al. 2008). Furthermore, around 78% the released progeny were sired by one male (Moore et al. 2008). Tuatara can take over 10 years to reach sexual maturity and each female has an annual reproductive output of about 2 offspring (Cree 1994). Consequently, even determining numerical success or failure of the Hauturu population may take many decades. Establishing the genetic effects of a predation bottleneck and restricted paternity on tuatara may take even longer.
Similar problems can arise in translocated populations. Miller (2009) assessed the genetic heterozygosity of three populations of translocated lizards, each of which had selfsustaining populations (sensu Dodd and Seigel 1991). She found that when the founder population is low (15), or in larger populations when there is relatively low founder survival, inbreeding depression can erode genetic diversity sufficiently to jeopardise the long term prospects for the populations.
Such problems aside, natural recovery in situ, recolonisations, and translocations can greatly change the structure of communities on islands once invasive species have been removed. Some of these changes are subtle. For example, on Korapuki Island, lizard assemblages in the presence of Pacific rats and rabbits were dominated by diurnal species of skinks. After the two mammals were removed, dominance within the assemblages shifted as previously rare nocturnal geckos become increasingly abundant (Towns 1991, 2002). Similar subtle effects of rats such as Pacific rats have been reported for plant communities. Comparisons of seedling composition on islands where Pacific rats are present, have been excluded using cages, and have been eradicated, indicate that the rats have measurable effects on at least 11 and perhaps over 30 species of coastal and forest plants. These effects are sufficiently severe to result in impaired recruitment, sex imbalances and declines to local extinction of canopy and subcanopy species (Campbell and Atkinson 1999, 2002; Campbell 2011). There may also be a feedback loop, where predation on the large seeds of some plants by Pacific rats reduces their incidence in the canopy, thereby reducing visits from fruit pigeons (kereru: Hemiphaga novaeseelandiae) and dispersal of large-fruited plants that remain (Campbell and Atkinson 2002). The extent to which changed seedling recruitment after release from the effects of Pacific rats might change forest composition is as yet unclear.
More extensive changes in community structure can follow the removal of grazing species such as sheep and goats. On subantarctic Campbell Island, removal of sheep from the island in 1990 was followed within four years by recovery of tall native grasslands, reinvasion of the old pasture by native megaherbs, and declines in coverage by native species resistant to grazing. Full recovery of native plant communities is likely within a few decades (Meurk et al. 1994). Likewise, after the removal of goats from Great King Island in 1946, grazing-induced turf was 40 years later replaced by early successional forest up to 2m tall, and reappearance in coastal forest of endemic tree species (Wright and Cameron 1990; Bellingham et al. 2010b). However, the spread of some endemic species has been slower than expected, largely due to the absence of birds able to disperse large seeds. The importance of dispersers was illustrated when the translocation of a small number of kereru to Great King Island was rapidly followed by the appearance of new populations of seedlings (Bellingham et al. 2010b).