West, C.J. Consideration of rat impacts on weeds priorIsland to rat and invasives: cat eradication on eradication Raoul Island, Kermadecs and management
Consideration of rat impacts on weeds prior to rat and cat eradication on Raoul Island, Kermadecs
C. J. West Department of Conservation, PO Box 10-420, Wellington 6143, New Zealand.
INTRODUCTION Plants and animals have been introduced to many islands comm.). Raoul Island has been intermittently occupied by around the world (Abbott et al. 2000; Sax and Gaines people since C. 960 A.D. and permanently occupied since 2008; Towns et al. 2006). Raoul Island is no exception, 1937 (Sykes et al. 2000). having been colonised through human agency since the th Raoul Island, a Nature Reserve managed by the first Polynesian voyagers arrived in the 10 century AD Department of Conservation, is the northernmost (29°15’ S, (Anderson 1980). Rattus exulans Peale (Pacific rat, kiore) 177° 55’ W) and largest (2934 ha) of the Kermadec Group, were well established in the mid 1800s and were assumed situated in the South Pacific Ocean north of New Zealand to be native (MacGillivray 1854). However, they are most (Fig. 1). It is an active volcano with rugged topography, likely to have been introduced by Polynesian voyagers subject to cyclones in the summer months and occasional many centuries before (Harper and Veitch 2006). Felis heavy rain that triggers landslips. The Raoul volcano catus L.(cat) accompanied the earliest human settlers and last erupted on 17 March 2006. Forest dominated by were definitely present in 1854 (MacGillivray 1854).Rattus pohutukawa (Metrosideros kermadecensis W.R.B.Oliv.) is norvegicus Berkenhout (Norway rat) colonised Raoul the main vegetation cover. Grasses, strand vines and ferns Island after the schooner “Columbia River” was wrecked dominate coastal cliffs and the beach ridges at Denham there in September 1921 (Harper and Veitch 2006). Bay and Low Flat. Exotic plants deliberately introduced were food and Approximately 20% of the vascular plant flora is utility species such as Cordyline fruticosa (L.) Goepp. (ti endemic (Sykes et al. 2000). This comparatively low pore, ki) and Aleurites moluccana (L.) Willd. (candlenut), level of endemism is principally due to the young age of introduced by Polynesians, and a range of fruiting trees, the Raoul volcano coupled with its remoteness. The high pasture grasses, vegetables and flowering garden plants degree of disturbance associated with active volcanism introduced by European settlers (Sykes et al. 2000). Many may also be a contributing factor. Many of the endemic plant species were accidentally introduced including plant species closely resemble species endemic to the New Bidens pilosa L. (beggar’s ticks) with Polynesians, Conyza Zealand mainland and offshore islands e.g., Kermadec bonariensis (L.) Cronquist (fleabane) with early European pohutukawa, Kermadec ngaio (Myoporum kermadecense settlers (Sykes et al. 2000) and more recently, Selaginella Sykes) and Kermadec fivefinger (Pseudopanax kraussiana (Kunze) A.Braun (selaginella), first recorded kermadecensis (W.R.B.Oliv.) Philipson). Little is known in 1999 (West 2002) and Soliva sessilis Ruiz and Pav. about elements of the flora that might be now extinct as (Onehunga weed) first recorded in 2008 (David Havell pers. a result of human occupation or eruption history: pollen
Fig. 1 Location of Raoul Island and location of places mentioned in the text.
Pages 244-247 In: Veitch, C. R.; Clout, M. N. and Towns, D. R. (eds.). 2011. Island invasives: eradication and management. IUCN,244 Gland, Switzerland. West: Rat impacts on weeds analyses may shed some light on species turnover in this highly disturbed island ecosystem. The 2006 eruption has caused the extinction of at least one native species – Ophioglossum petiolatum Hook. (stalked adder’s tongue) – that was known from a small area in Green Lake crater that is now both under water and buried in a thick mud deposit. More than half the vascular plant flora of Raoul Island is introduced and a number of vines, trees and shrubs are transformer weeds (Pyšek et al. 2004). The majority of the introduced species are herbaceous and associated with disturbed ground along roads and open tracks, the accommodation for staff, meteorological station and abandoned rough pasture along the northern terraces of the island. There is an active surveillance programme, and this is how the selaginella and Onehunga weed were detected. Selaginella, in particular, could become a problem; however, prompt action has prevented this. Weed eradication has been a focus of management Fig. 2 Raoul Island food web from 1984–2002. Relationships on Raoul Island since 1972, with 29 species targeted for derived from published sources (Fitzgerald et al. 1991; Harper and Veitch 2006) and from observations (C.R. eradication in 1996 (West 2002). Goats were eradicated Veitch pers. comm., and pers. obs.). from Raoul Island in 1984 (Sykes and West 1996). The impact of goats on the native and exotic flora was evident both before and after they were eradicated: no exotic plant Island was constructed, based on published information species are known to have increased significantly after and observations (Fig. 2). Particular attention was paid to goats were eradicated (Parkes 1984; Sykes and West 1996). known prey and competitors of rodents and cats and those Rats and cats were eradicated from Raoul Island in 2002 organisms known to eat or disperse seeds. It was considered and 2004 respectively (Broome 2009), thus removing all that the plant eradication regime on Raoul was resourced introduced mammals. The need to identify the effect of rat sufficiently to be able to cope with any increase inthe eradication on the exotic flora was considered well before abundance or range of the target species if they responded the rat eradication operation (West 2002; Sykes and West positively to rat eradication, since a key action in achieving 1996) in order to avoid unintended outcomes (Zavaleta et eradication is to detect all individuals before they produce al. 2001; Caut et al. 2009). ripe fruit/seeds (West 2002). The effect of the Pacific rat on seeds and seedlings of The list of exotic vascular plant species present on Raoul native tree species of northern New Zealand islands has Island, excluding those currently targeted for eradication been documented (Campbell and Atkinson 1999; Campbell or control of seedlings (species of historic value), was and Atkinson 2002; Towns et al. 2006). Rats were shown to reviewed to identify those that might become invasive after eat seeds and seedlings, thus depressing the populations of rats were eradicated. The eradication of goats showed that at least 11 tree species. Norway rats also have been shown the primary vegetation response was a significant increase to suppress regeneration of native tree species by eating in density and abundance of native plant species, indicating seeds and seedlings (Allen et al. 1994; Towns et al. 2006). that forest regeneration was not impaired by the majority of Rats eat many plant parts including flowers, seeds, fruits the exotic species which are herbaceous (a full list of exotic and seedlings (Atkinson and Towns 2005; Innes 2005). species can be seen in Sykes et al. 2000). Woody trees and There is no published information on impacts of rats on shrubs, vines and clonal semi-woody perennials pose the exotic plant species in New Zealand. However, the factors greatest threat to the forest communities on Raoul Island. that predispose native species to predation by rats were The primary question, therefore, was would a species, if assumed to apply to exotic species as well: seedlings of all not checked through seed or seedling predation, expand its species, and plants with fleshy fruit and/or fruit with large population to the detriment of native plant communities edible seeds were considered to be vulnerable (Towns et on the island? Many species were eliminated at this point, al. 2006). particularly light-demanding, herbaceous species. Non- Cats on Raoul Island primarily preyed upon rats, forest communities, where light-demanding weeds might particularly the Pacific rat, and secondarily upon birds, thrive, are in comparatively harsh environments (e.g., some of which are effective seed dispersers (Fitzgerald coastal cliffs, rocky ridges, heated ground in the crater, et al. 1991). Rats also prey upon some of the same seed back dunes) and have not indicated susceptibility to weed dispersing bird species, mainly introduced passerines invasion to date, with the exception of the dune slack at (Towns et al. 2006). On Raoul, the principal fruit-eating Denham Bay where airplant has spread vegetatively. and seed-dispersing bird species are tui (Prosthemadera The reproductive status and dispersal potential of the novaeseelandiae Gmelin), blackbird (Turdus merula remaining species were then considered. Gravity or wind- L.), thrush (T. philomelos Brehm) and starling (Sturnus dispersed species of short stature were considered to pose vulgaris L.). lesser risk as any progeny would be readily located near Ahead of the mammal eradication, therefore, direct and parent plants (Clark et al. 2005). Bird dispersed, woody indirect effects of rats and cats on population dynamics species were identified as the group posing the greatest of weeds were evaluated. Direct effects of rats that were risk as their seeds could be dispersed over longer distances considered were predation of flowers, seeds, fruit and and their seed shadow would be poorly defined (Gosper seedlings. Indirect effects were predation by rats and cats et al. 2005). It would be difficult to find any progeny, and on seed dispersing bird species. populations of these species would be likely to expand. In examining the exotic flora, it was clear that five METHODS species were not regenerating from seed: all expansion was by vegetative means alone. In addition, fruit had been A food web describing the interactions observed rarely (Hibiscus tiliaceus L.) or never observed on these between the predominant species and guilds on Raoul species in the decade leading up to the rat eradication.
245 Island invasives: eradication and management
Therefore, the potential for each of these species to bear fruit and regenerate via seedlings was investigated.
RESULTS Two introduced tree species that flowered and fruited freely were targeted for eradication prior to the rat and cat eradication. Vitex lucens Kirk (puriri), a fleshy-fruited tree, native to New Zealand, was represented on Raoul by three trees planted by European settlers. On Tiritiri Matangi Island in the Hauraki Gulf near Auckland, Pacific rats had suppressed all regeneration of the two puriri trees native to the island (pers. obs.). Puriri has fleshy fruits, the smallest of which can be dispersed by tui. The three trees on Raoul were felled in 1997 and did not regenerate. New, mature, trees have since been reported in a new location and are being investigated. A small group of Macadamia tetraphylla L. Johns. (macadamia) trees was felled in the Fig. 3 Number of individuals of grape in different size same year and three seedlings were pulled out from the classes removed from Raoul Island from 1997 to 2008: same site in 2003 (West 2002). Although the macadamia small (< 30 cm tall: black bars); medium (30-100 cm tall: trees produced heavy nut crops in some years, the rats on grey bars) and large (>100 cm tall: white bars). Raoul Island efficiently devoured all seeds as evidenced by the many rat-gnawed shells beneath the trees. and rosy periwinkle control in 2008. Control of both species The five species that were not known to fruit in the is aimed at containing the populations so they don’t spread presence of rats were Vitis vinifera L. (grape), Hibiscus beyond their former extent. Both species are confined to tiliaceus L. (shore hibiscus, fou), Catharanthus roseus Denham Bay with airplant intermittently spread along 1–2 (L.) G. Don (rosy periwinkle), Bryophyllum pinnatum km of the dune slack. Rosy periwinkle occupies a smaller (Lam.) Oken (airplant), and Phoenix dactylifera L. (date). area at the forest edge. Shore hibiscus is currently not Firstly, the reproduction of these species was investigated controlled. Further investigation of its status is required as to determine what factors might limit them (Table 1). Dates the plants in Denham Bay and at Low Flat are thought to are dioecious and the one clump of dates at Denham Bay be introduced whereas a plant at Coral Bay is likely to be is likely to be either male or female. Although the date native (West 1996). flowers periodically, flowers have never been examined to Two unexpected outcomes have been observed. Firstly, determine the gender of the trees. Thus dioecy, with just when a large grape vine in the canopy of pohutukawa one gender present, probably limits sexual reproduction of and other trees was eradicated from an old (19th century) the date. For the remaining four species no reproductive garden site at the south end of Denham Bay, a dense stand limitations were detected other than the possibility of rat of the introduced canna lily (Canna indica L.) arose. predation on developing fruit or, in the case of grapes, Canna lily was first recorded on Raoul Island only after possibly the flowers. eradication of another weed (Mysore thorn – Caesalpinia The risk posed by the four species potentially limited by decapetala (Roth) Alston) began (Sykes and West 1996). rat predation was evaluated (Table 2). Grapes were found to Establishment of the canna lily from seed that had lain be the only high risk species: there were nine separate plants dormant in the soil for perhaps a century was facilitated growing in a range of locations and they are bird dispersed. by increased light levels after grape eradication and the Therefore they could arise at locations some distance from absence of rat predation. parent plants. One of the grape vines occupied an area of Secondly, aroid lily (Alocasia brisbanensis (F.M. C. 0.5 ha at Low Flat and another in Denham Bay was Bailey) Domin), a widespread weed on Raoul which was estimated to be 4 ha in extent (C.R. Veitch pers. comm.). suppressed by canopy closure after the goat eradication, is Grapes were formally targeted for eradication in 1998 as now patchily heavily browsed by the tropical army worm a result of this analysis. Initially Grazon (triclopyr-based (Spodoptera litura Fabricius). The type of browsing damage herbicide) and Roundup (glyphosate) were used to kill now seen on the aroid lily was never observed prior to the grape vines. Later Vigilant (a gel formulation of picloram) rat eradication. The aroid lily appears to be the only plant was used. Grapes took considerable effort to kill but by obviously being browsed by this noctuid caterpillar. 2008–09 just two grape sprouts were found (Fig. 3) and none were ever observed flowering or fruiting (they were DISCUSSION killed before getting to this stage). Some studies have indicated that exotic plants can The other three species not targeted for eradication – become invasive after herbivore or seed predator removal shore hibiscus, rosy periwinkle and airplant all began to (Caut et al. 2009; Zavaleta et al. 2001; Abbott et al. 2000). fruit after rats were eradicated. Seedlings are commonly These effects were anticipated on Raoul Island and the observed near parent plants. Airplant control began in 2007 Table 2 Risk posed by each of the species that could have Table 1 Reasons for non-fruiting of exotic plant species fruited but did not when rats were present on Raoul Island. on Raoul Island. Fruit Invasion Species Pollinator Climate Dioecious Rat browse Species Distribution type Disperser risk Grape wind suitable no ? Grape 9 sites fleshy tui, blackbird high Shore insects suitable no ? Shore 2 sites, hibiscus hibiscus inland capsule gravity, sea low Rosy insects suitable no ? Rosy periwinkle periwinkle 1 site capsule gravity low Airplant insects suitable no ? Airplant 1 site capsule wind low Date wind suitable yes ? (no plantlets)
246 West: Rat impacts on weeds exotic flora was scrutinised for species that could change Anderson, A. J. 1980. The archaeology of Raoul Island (Kermadecs) and its place in the settlement history of Polynesia. Archaeology and their status following rat and cat eradication. Just three Physical Anthropology in Oceania 15: 325-327. species were identified that warranted eradication prior Atkinson, I. A. E. and Towns, D. R. 2005. Kiore. In: King, C. M. (ed.). to the mammal eradication operation. Despite grapes not The Handbook of New Zealand Mammals. 2nd edition Melbourne, being recorded as an invasive species outside of North Oxford University Press. Bourgeois, K.; Suehs, C. M.; Vidal, E. and Medail, F. 2005. Invasional America (Reichard 1994), this species possessed sufficient meltdown potential: facilitation between introduced plants and mammals characteristics (Buckley et al. 2006) to indicate it could on French Mediterranean islands. Ecoscience 12: 248-256. become a widespread transformer once it was released from Broome, K. G. 2009. Beyond Kapiti - A decade of invasive rodent suppression by rat predation and available for dispersal eradications from New Zealand islands. Biodiversity 10: 7-17. Buckley, Y. M.; Anderson, S.; Catterall, C. P.; Corlett, R. T.; Engel, T.; by native and exotic birds which were also released from Gosper, C. R.; Nathan, R.; Richardson, D. M.; Setter, M.; Spiegel, O.; predation by rats and cats. Vivian-Smith, G.; Vogt, F. A.; Weir, J. E. S. and Westcott, D. A. 2006. Although the evidence of direct effects of rats on Management of plant invasions mediated by frugivore interactions. Journal of Applied Ecology 43: 948-857. fruiting and seedling production of the five weed species Campbell, D. J. and Atkinson, I. A. E. 1999. Effects of kiore (Rattus investigated is circumstantial, the response of airplant, rosy exulans Peale) on recruitment of indigenous coastal trees on northern periwinkle and shore hibiscus to release from rat predation offshore islands of New Zealand. Journal of the Royal Society of New is sufficiently striking to suggest that grapes also were Zealand 29: 265-290. Campbell, D. J. and Atkinson, I. A. E. 2002. Depression of tree impacted by rats and their eradication was justified. recruitment by the Pacific rat Rattus( exulans Peale) on New Zealand’s The eradication of Pacific rats, Norway rats and cats northern offshore islands. Biological Conservation 107: 19-35. from Raoul Island has changed the ecosystem from one Caut, S.; Angulo, E. and Courchamp, F. 2009. Avoiding surprise effects on Surprise Island: alien species control in a multitrophic level perspective. dominated by predators to one where competition for Biological Invasions 11: 1689-1703. resources is the primary driver (Broome 2009). Seabirds Clark, C. J.; Poulsen, J. R.; Bolker, B. M.; Connor, E. F. and Parker, V. are returning, and so have red-crowned parakeets T. 2005. Comparative seed shadows of bird-, monkey-, and wind- (Cyanoramphus novaezelandiae cyanurus) (Ortiz-Catedral dispersed trees. Ecology 86: 2684-2694. Fitzgerald, B.M.; Kark, B.J. and Veitch, C.R. 1991. The diet of feral cats et al. 2009) and spotless crakes (Porzana tabuensis) (pers. (Felis catus) on Raoul Island, Kermadec Group. New Zealand Journal of obs). Invertebrates are prominent whereas before they Ecology 15(2): 123-129. were not. Gosper, C. R.; Stansbury, C. D. and Vivian-Smith, G. 2005. Seed dispersal of fleshy-fruited invasive plants by birds: contributing factors Rats are seed dispersers as well as seed destroyers and management options. Diversity and Distributions 11: 549. (Bourgeois et al. 2005) and cats can also disperse seeds Harper, G. and Veitch, C. R. 2006. Population ecology of Norway rats (Bourgeois et al. 2005; Nogales et al. 1996). However, (Rattus norvegicus) and interference competition with Pacific rats R.( exulans) on Raoul Island, New Zealand. Wildlife Research 33: 539- they are functionally replaced, to a degree, on Raoul Island 548. by the return of red-crowned parakeets and an increase in Innes, J. G. 2005. Norway rat. In: King C. M. (ed.). The Handbook of New other resident bird and insect species. Birds and insects Zealand Mammals. 2nd edition Melbourne, Oxford University Press. will now be the species that disperse and destroy seeds Macgillivray, J. 1854. MacGillivray’s journals. Adm 7/851 and 852, Public Record Office, Kew. on plants, on the ground and in the seed bank, as well as Mack, R. N. and Lonsdale, W. M. 2002. Eradicating invasive plants: seedlings. Reintroduction of fruit pigeons (Hemiphaga hard-won lessons for islands. In: Veitch, C. R. and Clout, M. N. (eds.). novaeseelandiae) which previously occurred on Raoul, Turning the tide: the eradication of invasive species, pp. 164-172. IUCN will add to the suite of dispersers of larger fruited native SSC Invasive Species Specialist Group. IUCN: Gland, Switzerland and Cambridge, U.K. plants, such as the palm Rhopalostylis baueri. Mulder, C. P. H.; Grant-Hoffman, M. N.; Towns, D. R.; Bellingham, P. Unanticipated or surprise effects arising from the rat J.; Wardle, D. A.; Durrett, M. S.; Fukami, T. and Bonner, K. I. 2009. and cat eradication detected so far are relatively minor Direct and indirect effects of rats: does rat eradication restore ecosystem functioning of New Zealand seabird islands? Biological Invasions 11: and in the case of tropical army worm defoliating aroid 1671-1688. lily somewhat beneficial. Monitoring of changes post- Nogales, M.; Medina, F. M. and Valido, A. 1996. Indirect seed dispersal eradication is essential (Zavaleta et al. 2001) in order to by the feral cats Felis catus in island ecosystems (Canary Islands). be able to respond to any detrimental changes (Mack and Ecography 19: 3-6. Ortiz-Catedral, L.; Ismar, S. M. H.; Baird, K.; Brunton, D. H. and Hauber, Lonsdale 2002). M. E. 2009. Recolonization of Raoul Island by Kermadec red-crowned Over time, and with the continuation of the weed parakeets Cyanoramphus novaezelandiae cyanurus after eradication of invasive predators, Kermadec Islands archipelago, New Zealand. eradication programme, the ecosystem may be restored to Conservation Evidence 6: 26-30. an approximation of its state before Polynesians arrived. Parkes, J. P. 1984. Feral goats on Raoul Island II: Diet and notes on the The remoteness of Raoul Island (1100 km NNE of New flora. New Zealand Journal of Ecology 7: 95-101. Zealand) may be sufficient to reduce the likelihood of weed Pyšek, P.; Richardson, D. M.; Rejmánek, M.; Webster, G. L.; Williamson, M. and Kirschner, J. 2004. Alien plants in checklists and floras: towards invasion increasing as seabirds return (Mulder et al. 2009), better communication between taxonomists and ecologists. Taxon 53: so long as weeds continue to be managed, surveillance 131-143. is programmed, and island quarantine procedures are Reichard, S. 1994. Assessing the potential of invasiveness in woody enforced. plants introduced in North America. University of Washington. Sax, D. F. and Gaines, S. D. 2008. Species invasions and extinction: the future of native biodiversity on islands. Proceedings of the National ACKNOWLEDGEMENTS Academy of Sciences 105: 11490. Sykes, W. R. and West, C. J. 1996. New records and other information Thanks to Raoul Island managers for their wise on the vascular flora of the Kermadec Islands. New Zealand Journal of management of the restoration of Raoul Island and to Botany 34: 447-462. Island field staff and volunteers for their hard work. Ellen Sykes, W. R.; West, C. J.; Beever, J. E. and Fife, A. J. 2000. Kermadec Islands Flora - Special Edition: A Compilation of Modern Material Cieraad provided the first evidence of fruiting and seedling about the Flora of the Kermadec Islands, Manaaki Whenua Press. establishment of some exotic species. Dick Veitch provided Towns, D. R.; Atkinson, I. A. E. and Daugherty, C. H. 2006. Have Fig. 1. Thanks to the two referees who made helpful the harmful effects of introduced rats on islands been exaggerated? comments on the manuscript. Biological Invasions 8: 863-891. West, C. J. 1996. Assessment of the weed control programme on Raoul Island, Kermadec Group. Science and Research Series no. 98. REFERENCES Wellington, Department of Conservation. West, C. J. 2002. Eradication of alien plants on Raoul Island, Kermadec Abbott, I.; Marchant, N. and Cranfield, R. 2000. Long-term change in the Islands. In: Veitch, C. R. and Clout, M. N. (eds.) Turning the Tide: the floristic composition and vegetation structure of Carnac Island, Western Eradication of Invasive Species. IUCN SSC Invasive Species Specialist Australia. Journal of Biogeography 27: 333-346. Group. IUCN: Gland, Switzerland and Cambridge, UK. Allen, R. B.; Lee, W. G. and Rance, B. D. 1994. Regeneration in Zavaleta, E. S.; Hobbs, R. J. and Mooney, H. A. 2001. Viewing invasive indigenous forest after eradication of Norway rats, Breaksea Island, species removal in a whole-ecosystem context. Trends in Ecology and New Zealand. New Zealand Journal of Botany 32: 429-439. Evolution 16: 454-459.
247