A strategy for Galapagos weeds

A. Tye, M. C. Soria, and M. R. Gardener Charles Darwin Research Station, Galapagos. Postal address: CDRS, AP 17-01-3891, Quito, . E-mail: [email protected]

Abstract Galapagos has a native vascular flora of some 500 species, 60 more species that are doubtfully native, and more than 600 introduced species. Introduced species are the most serious problem facing the native biota. The worst invasive are trees and other woody species, vines and grasses, and most of them were introduced deliberately. Many have invaded the Galapagos National Park and are also invasive in agricultural zones. A strategy for tackling the weeds problem includes prevention, control, eradication and restoration, the research required to develop and prioritise these management actions, and development of a legal framework for their implementation. Given limited resources for control, a risk assessment system for prioritising problem species and key sites is essential and is being developed. It will evaluate both species that are already present, and proposed introductions. A quarantine system for Galapagos has been designed and implementation commenced. Quarantine is essential if the balance between introduction and eradi- cation is to be tipped towards the latter. Research includes investigations of the ecology and distribution of introduced plants, to determine factors (such as reproduction and dispersal rates and longevity of plants and their soil seedbanks) essential for the design of successful management programmes. Research on control techniques is also essential, since many Galapagos invasives are useful species that have not been subject to control elsewhere. Restoration research is beginning, focussing on methods of control combined with active restoration, such as seeding with native species. Invasive plants have only recently been widely recognised as high priority in Galapagos, and the first projects investi- gating the ecology of serious weeds are now yielding results. Control trials are leading to the development and adoption of effective field methods. Attempts have begun to eradicate species with still-small populations, but which are known as invasive elsewhere. A pilot project is also beginning, to assess the feasibility of eradicating a well-established inva- sive tree species. These measures, aside from their scientific and conservation value, also act as confidence builders, demonstrating to the public and land managers both the dangers of introduced species and the possibilities for their control and eradication.

Resumen Galápagos cuenta con una flora vascular nativa de algunas 500 especies, más 60 especies que son dudosamente nativas y más de 600 especies introducidas adicionales. Los organismos introducidos constituyen el problema más grave que enfrenta la biota nativa. Las peores plantas invasoras son árboles y otras especies leñosas, trepadoras y pastos, y la mayoría de las especies que están causando o podrían causar problemas fueron introducidas a propósito. Muchas de estas especies han invadido el Parque Nacional Galápagos, y son igualmente agresivas en las zonas agrícolas. Una estrategia para enfrentar este problema incluye prevención, control, erradicación y restauración, las investigaciones necesarias para desarrollar y priorizar estas acciones de manejo, y el desarrollo de un marco legal para su implementación. El primer paso es desarrollar medidas para priorizar los problemas, tomando en cuenta los recursos limitados para el control. Un sistema de evaluación de riesgos se está desarrollando, para evaluar tanto especies que ya se dan en las islas y para introducciones propuestas. Este último forma parte de la prevención: un sistema de cuarentena para Galápagos ha sido diseñado, y su implementación iniciado. La cuarentena puede reducir pero jamás parar las introducciones, pero es necesario para cambiar el equilibrio entre la introducción y la erradicación. El control y erradicación tienen dos componentes: investigación y manejo. Las investigaciones de la ecología y distribución de las plantas introducidas nos permiten determinar los factores necesarios para diseñar programas de control y erradicación que sean exitosos, tales como tasas de reproducción y dispersión, longevidad de plantas y semillas etc. La investigación para desarrollar nuevos métodos de control también se necesita, por lo que muchas especies invasivas en Galápagos son especies útiles y no han sido sujetos del control en otras partes. La investigación para la restauración ya empieza, con su enfoque en combinar acciones de restauración positivos en combinación con el control, tales como siembra de especies nativas. El programa de plantas introducidas en Galápagos esta creciendo rápidamente, por lo que las plantas invasoras han sido solo recién ampliamente reconocidas como de alta prioridad. Los primeros proyectos para investigar la ecología de las malezas más graves ya son produciendo resultados. Los ensayos de control llevan al desarrollo y uso de métodos de campo eficientes. Se han iniciado intentos para erradicar especies aun representadas únicamente por pequeñas poblaciones, pero las cuales están conocidas como invasoras graves en otros lugares. También ha comenzado un proyecto piloto para evaluar la factibilidad de erradicar un árbol invasora bien establecido. Estas medidas, aparte de su valor científico y para la conservación, además pueden aumentar la confianza y cambiar la opinión general sobre el peligro de las especies introducidas y la factibilidad de su control.

Keywords Environmental weeds; strategic planning; Galapagos; prevention; control; eradication; restoration; research; islands.

Pages 336-341 In Veitch, C. R. and Clout, M. N. (eds.). Turning the tide: the eradication of . IUCN SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK. Tye et al: A strategy for Galapagos weeds

THE PROBLEM species are found in the more humid, higher altitudes of the four larger inhabited islands, and the settled areas are Galapagos is an isolated oceanic archipelago of volcanic the major source of invasion into the Galapagos National islands lying 1000 km west of Ecuador, straddling the equa- Park. tor (Fig. 1). The date of discovery of the archipelago is The Galapagos islands support a native vascular flora of usually recognised as 1535. The islands were uninhabited about 500 species, with an additional 60 doubtfully native at that time, and no evidence proving earlier human pres- species, principally pantropical ruderals, which may have ence has been found (Slevin 1959; Hickman 1985). The arrived naturally or may have been introduced by the ear- first visitors after discovery were mainly buccaneers, pass- liest human visitors to the islands. In this paper, “intro- ing sailors, whalers and sealers (Hickman 1985). Settle- duced” and “alien” are used interchangeably to mean in- ment began on Floreana Island in the early 1800s, but Santa troduced deliberately or accidentally due to the actions of Cruz Island was only settled as late as the 1920s (Slevin man. “Naturalised” means reproducing in natural or semi- 1959; Schofield 1989). natural habitat without the further assistance of man (be- The pirates and whalers deliberately or accidentally intro- yond habitat disturbance). “Invasive” means invading natu- duced some alien species, including goats, rats and, prob- ral (undisturbed) habitats. “Weed” means a naturalised ably, insects and plants. Even before permanent settlement, species. Floreana had large areas dominated by introduced plants Porter (1822) mentioned the first alien species (pumpkins such as Citrus spp. (Slevin 1959; Hamann 1984). The rap- Cucurbita sp. and potatoes Solanum tuberosum), which idly increasing settled population, growing at 8% per year were introduced about 1807. Numbers increased slowly in the 1990s through both immigration and births, has been until the 1960s, although true numbers are unknown, since accompanied by an enormous number of new introduc- earlier references (especially Wiggins and Porter 1971; tions of alien plants and animals (Mauchamp 1997). Al- Porter 1984) took into account only naturalised species. though agricultural development began at the time of set- The list continued to increase, reaching 438 in 1995 tlement, the process has been uneven, leading to different (Mauchamp 1997) and over 600 by November 2000 (Da- rates of introduction of alien species. Floreana has the long- tabase of the Galapagos Flora, Charles Darwin Research est history of the presence of a large introduced flora, while Station). The minimum detection rate has thus been more agriculture on Santa Cruz was minimal until about 1960 than 10 per year in the last 30 years (Fig. 2), and more (Moll 1990). than 120 during 2000. However, the recent apparent rate The Galapagos National Park forms 96.4% of the land of increase is obviously affected by increased interest in area of Galapagos; inhabited areas (urban and agricultural recent years in the introduction process, as well as increased zones, military bases and airports) make up the rest. Alien sampling effort and individual research projects, and the plants that have escaped from cultivation are mostly found inclusion of cultivated, non-naturalised species in more on the five inhabited islands, especially the four with agri- recent lists (Tye 2001b). cultural and urban zones (Floreana, Isabela, San Cristóbal, The principal threat to the terrestrial biota of Galapagos is Santa Cruz: Fig. 1); the fifth, Baltra Islet off the north coast introduced species (Loope et al. 1988). Most (c. 75%) of of Santa Cruz, is a military base and civil airport. There the alien species were introduced deliberately as use- are also a number of deliberately-introduced species on ful plants, for their ornamental, agricultural, medicinal or Santiago, which was formerly inhabited. Aliens that were timber value, although some were introduced accidentally. introduced accidentally often have a much wider distribu- An even higher percentage of the worst invaders was in- tion in the archipelago but are mostly less problematic than troduced deliberately (see Mauchamp 1997). Some 45% deliberately introduced species. Most introduced plant of introduced plant species have naturalised (A. Tye un- published data).

Fig. 2 Rate of increase in numbers of intro- duced plant species recorded in Galapagos Fig. 1 Galapagos archipelago. since their discovery in 1535 (Tye 2001b).

337 Turning the tide: the eradication of invasive species

Most introduced plant species have not significantly af- least 80 more species have naturalised and are still un- fected the ecological equilibrium of the islands. However, common but are known or suspected to have invasive ten- at least 37 species (Table 1) have invaded large areas and/ dencies. Another 140 or so have naturalised but do not or appear to be adversely affecting the natural ecosystem appear to be causing obvious ecological damage, while at to a degree more than simply occupying space within an least another 100 species present in the islands, which have existing community (e.g. altering community composition not yet escaped from cultivation, are known to be seri- or threatening a native species), or (in a few cases) are ously invasive in other parts of the world. The rest of the naturalised but with limited distributions and known to be introduced species are present in cultivation; some may extremely serious invasives in other parts of the world. At become invasive. Lawesson (1990) lists only eight seri- ous invasives, while Mauchamp (1997) lists 11, omitting Table 1. Invasive species in Galapagos that are one of those included by Lawesson. The 37 species in known or suspected to be causing significant Table 1 include these 12, the increase largely caused by a ecological change, including in natural areas. re-evaluation of the threat posed by certain species (four species listed by Mauchamp 1997 as potential invaders Family Species are included in Table 1), as well as a few recently-added species. Agavaceae Furcraea hexapetala (Jacq.) Urb. Aristolochiaceae Aristolochia odoratissima L. There have been few rigorous studies of the effects of the Asteraceae Pseudelephantopus spicatus invasions, but some species have caused drastic habitat (B. Juss. ex Aubl.) C.F. Baker changes, forming monospecific stands, shading out or oth- Bombacaceae Ochroma pyramidale erwise replacing native vegetation communities, or pre- (Cav. ex Lam.) Urb. venting seedling regeneration by forming impenetrable Boraginaceae Cordia alliodora (Ruiz & Pav.) Oken carpets. Where detailed studies have been made, dramatic Caesalpiniaceae Caesalpinia bonduc (L.) Roxb. community changes have been revealed (Jäger 1999). The Senna obtusifolia worst effects seem to be caused by woody species, espe- (L.) H.S. Irwin & Barneby cially trees such as Psidium guajava, Cedrela odorata and Capparidaceae Cleome viscosa L. pubescens, and bushes that form impenetrable Crassulaceae Bryophyllum pinnatum (Lam.) Oken thickets, such as Lantana camara and spp. Many Cucurbitaceae Cucumis dipsaceus Ehrenb. ex Spach vines and grasses are also causing serious problems. Euphorbiaceae Dalechampia scandens L. Ricinus communis L. Although most of the serious invaders in Galapagos were Lamiaceae Hyptis cf. atrorubens Poit. introduced deliberately as useful plants, most of the spe- Lauraceae Persea americana Mill. cies that are problematic in the National Park are also caus- Meliaceae Cedrela odorata L. ing problems for farmers, and this includes cultivated spe- Mimosaceae Leucaena leucocephala (Lam.) de Wit cies as well as accidental introductions. To a large extent, Myrtaceae Psidium guajava L. therefore, there is little conflict regarding priorities for Syzygium jambos (L.) Alston control or eradication of invasives. However, this is not Syzygium malaccense true in every case. For species such as Cinchona pubescens (L.) Merr. & L.M. Perry (which produces no economic yield of or wood), Passifloraceae Passiflora edulis Sims. all parties wish to see the pest eradicated, whereas Poaceae Digitaria decumbens Stent Pennisetum purpureum is a serious invader of the National Melinis minutiflora P. Beauv. Park but is a valued pasture grass and its removal from Panicum maximum Jacq. agricultural areas would be politically and economically Pennisetum clandestinum Hochst. difficult. In such cases, an eradication strategy would have ex Chiov. to include replacement by a non-invasive substitute. Pennisetum purpureum Schumach. Urochloa brizantha THE STRATEGY (Hochst. ex A.Rich) R.D.Webster Urochloa mutica The strategy for dealing with invasive plants comprises (Forssk.) T.Q. Nguyen five levels of action: (1) prevention, (2) control and eradi- Rubus niveus Thunb. cation, (3) restoration, (4) the research required to prioritise Cinchona pubescens Vahl and develop appropriate techniques to carry out these ac- Rutaceae Citrus aurantiifolia tions, and (5) development of a legal framework for their (Christm.) Swingle effective implementation. Citrus limetta Risso Citrus limon (L.) Bur.f. Research Solanaceae Cestrum auriculatum L’H¾r. Datura stramonium L. Prioritisation Solanum lycopersicum L. Ulmaceae Trema micrantha (L.) Blume Given limited resources for control, an essential first step Verbenaceae Lantana camara L. is prioritising the problems. A risk assessment system is

338 Tye et al: A strategy for Galapagos weeds used to evaluate both species that are already present in laboratory research is currently under way in preparation the islands and proposed introductions. Permitted lists for for what may become the first biocontrol field trial, for import of species and products have already been drafted, cottony cushion scale insect Icerya purchasii. Should this based on a preliminary risk assessment procedure, while first instance prove successful, the climate of opinion all introduced plant species have been subject to a subjec- should improve, and trials for the biocontrol of weeds, tive risk assessment. A formal risk assessment system is including Cinchona pubescens, Lantana camara, and Ru- currently being developed (Tye 2001a). Risk assessment bus spp.,may be considered. and prioritisation will also be applied to sites. This is al- ready done in a subjective manner, but a more formal sys- Given the pressure for control of the worst invaders, trial tem based on conservation value is being developed. results are often applied to control and eradication cam- paigns before completion of trial monitoring. In such cases Risk assessment is the major tool used to decide whether the control action is also monitored to enable better as- an introduced species should be merely controlled, or sessment of the efficacy of the techniques. Trial results whether an eradication attempt could be considered. The are also fed into the weed risk assessment system, since decision depends on the plant’s inherent biological char- availability of suitable techniques is an important factor in acteristics (including growth form, reproductive strategies, risk assessment, particularly in helping to decide whether dispersal mechanisms, reproductive rate, age at first re- a plant species should be controlled or whether an eradi- production, and longevity), its current distribution, abun- cation attempt could be considered. dance, rate of spread, ecological and other impacts, sus- ceptibility to treatment (visibility, availability of control Restoration research techniques cheap and effective enough to enable eradica- Restoration research is just beginning in Galapagos. The tion to be contemplated with available resources), and ur- initial focus is on methods of control combined with ac- gency of action required. tive restoration, such as seeding with native species. Only Ecology and distribution one project to investigate the possibility of assisted resto- ration has so far been undertaken, examining restoration The success of a risk assessment system depends on sound of native Scalesia forest following control of elephant grass scientific data about the distribution, biology, and ecology Pennisetum purpureum. In other cases, monitoring of se- of the introduced species. These factors are also essential lected control sites is being undertaken, to identify when to the design of control or eradication programmes. Re- positive intervention to restore native vegetation commu- search currently in progress or planned for the next five nities following control is needed, and to design such years includes complete surveys of introduced plants in projects. Two examples include monitoring of the agricultural and urban zones of the four major inhab- Zone vegetation following control of Rubus niveus and ited islands, ecological studies of the life cycle of the worst other introduced species at the unique highland crater-lake invaders and their effects on native vegetation, and gath- site of El Junco on San Cristóbal Island, and monitoring ering of information on the biology of introduced species of regeneration following control of Cedrela odorata in from sources outside Galapagos. So far, ecological and Transition Zone forest on Santa Cruz Island. Given the distributional studies have focussed on Cinchona limited resources available for introduced plant control in pubescens (Jäger 1999), Psidium guajava and Rubus spp. Galapagos in the foreseeable future, it is unlikely that major restoration projects will be able to be considered for some Control trials time. Research on control techniques is also essential, since many Management Galapagos invasives are useful species and have not been subject to control elsewhere. We need effective methods Prevention of killing the invaders that are both as cheap as possible and are appropriate for use within a national park, causing Prevention is the first stage of management action: a quar- minimal damage to the native vegetation. Limitations are antine system for Galapagos has been designed and its imposed by the unavailability of some herbicides in Ecua- implementation commenced (Zapata et al. 2000). Quar- dor and by the difficulties of obtaining, using, and servic- antine can slow but never stop introductions, although it is ing sophisticated application equipment. The trials take essential if the balance between introduction and eradica- these limitations into account, although where a herbicide tion is to be tipped towards the latter. The quarantine sys- is not locally available (but seems by far the most effec- tem will eventually implement primary control outside of tive), trials are undertaken along with negotiations with Galapagos, at the mainland port and airports from which potential suppliers and the Ecuadorian authorities in order traffic reaches Galapagos. Secondary quarantine control to try to ensure that the required materials become avail- is already in place on the islands, and an early warning able for use within the archipelago. Research includes tri- and rapid response strategy is being planned for species als of manual methods, herbicides, and herbicide applica- that evade these controls. The system deals also with trans- tion techniques. Trials so far have included 15 of the worst port between the islands, given the importance of inter- invasive plant species. Biological control has never been island differences in biodiversity, with many single-island applied in Galapagos and is controversial because of the endemic species. potential dangers of introducing yet another organism, but

339 Turning the tide: the eradication of invasive species

Control and eradication These cases are discussed in more detail by Soria et al. (2002). The remaining priority plants to make up the ini- Eradication is considered for plants at both ends of the tial target 30 species will include other timber trees and invasion scale in Galapagos. One priority group comprises ornamental plants. A preliminary list of some 60 potential plants that are known to be invasive in other parts of the targets has been drawn up based on the basic criteria of world but are present in Galapagos in very small limited distribution and known invasive tendencies, and populations and are not regarded as indispensable by the the final list of 30 will be selected following a risk assess- local community. At the other end of the scale are species ment based on a full range of criteria including feasibility that are seriously invasive already in Galapagos, but for of eradication. which an assessment indicates that eradication might be The second priority group of species for eradication com- feasible. We have two successful examples of eradication prises serious invaders that are widespread but whose char- of plants in the first group and projects in progress to eradi- acteristics suggest that eradication might still be feasible. cate more, but are only beginning to consider eradicating The pilot species for this work is Cinchona pubescens, a plants in the second group. small tree that has become the greatest single threat to the native highland vegetation of Santa Cruz. The principal Among the first group, plants known to be invasive but characteristics suggesting that it might be feasible to con- still present in small populations in Galapagos, two spe- trol it include that it is conspicuous, not regarded by the cies are considered eradicated from Galapagos. Echinopsis local community as a useful species but on the contrary pachanoi (Cactaceae) was known from a single garden widely recognised among conservation workers and local plant on Santa Cruz Island, which was cut down and burned people as a serious pest, and it is present on only one is- in the 1980s. The other species was tropical kudzu Pueraria land. In addition, whereas previously it had been difficult phaseoloides (Fabaceae), which is widely grown in main- or impossible to control Cinchona by acceptable herbi- land Ecuador for ground cover in orchards and as a forage cides, safe and effective chemical techniques for killing it, plant and soil improver. It was introduced by one farmer using picloram for cut stump treatment and picloram-me- in 1996 and planted in a single pasture field. The plant thyl metsulphuron mixtures for hack-and-squirt treatment, was spotted by a CDRS botanist, and after the potential have recently been identified by research in Galapagos. danger of the species was explained to the farmer (it is a These herbicide treatments for larger trees, combined with close relative of kudzu Pueraria lobata, one of the worst manual removal of seedlings and saplings, could eventu- invaders of the south-east U.S.A), he agreed to allow it to ally permit complete eradication. This project is still in be eliminated. The plot was treated with glyphosate and the planning stage, with larger-scale control trials under monitored regularly, with new growth spot treated. No way and an ecological study in progress that is designed plants have been seen since 1997 although monitoring to provide information on factors affecting the speed and continued until 2001 (further details in Soria et al. 2002). likelihood of eradication, such as density, distribution, Although considered eradicated, we still lack introduced seedbank longevity and dispersal ability. These studies will species inventory data on the agricultural and/or urban lead to better cost estimates, which are needed for the ul- zones of three of the four inhabited islands, so there may timate decision on whether to attempt eradication. yet be other plants of these species present there. Control is undertaken where eradication is not currently In the past, apart from these two isolated cases, eradica- considered feasible but where the plant is considered to tion has not been attempted. Instead, funds have been di- pose a significant conservation risk. Such plants include rected as a priority towards control of the most serious most of the most serious invaders, such as Rubus niveus, invaders in selected important sites. However, during the Lantana camara, vines such as Passiflora edulis, and very past year we have begun to attempt eradication on a larger widespread invasive trees such as Psidium guajava and scale. A project began in 2001, with sufficient funding to Cedrela odorata. Virtually every widespread introduced permit the selection and attempted eradication of 30 plant plant that is not a tree also falls into this category, as do species still present in small populations. Eradication in useful (indispensable to local people) but invasive species some cases will be from the entire archipelago and in other for which no effective substitute species has yet been iden- cases only from selected islands. Eradication attempts have tified. Some potentially invasive garden ornamentals are already begun with the following species: also considered not yet eradicable, since they are so popu- lar. v Citharexylum gentryi, a timber tree present in two parts Restoration of the agricultural zone of Santa Cruz. v Rubus megalococcus and R. glaucus, each present in a So far, no active restoration has been undertaken follow- single population on Santa Cruz. ing invasive plant control. Monitoring of selected control v Rubus niveus from the island of Isabela. This species is sites will provide data that will be used to identify needs already widespread and beyond current eradication ca- for intervention to restore native vegetation communities pability on Santa Cruz and San Cristobal islands, but is and to design such projects. present in only a small area on Isabela. v Rubus adenotrichos, present in single localities on Isabela and Santa Cruz.

340 Tye et al: A strategy for Galapagos weeds

Legal framework Jäger, H. 1999. Impact of the introduced tree Cinchona pubescens Vahl. on the native flora of the highlands of A comprehensive “Special Law for Galapagos Province” Santa Cruz Island (Galapagos Islands). Diplomarbeit. was passed by the Ecuadorian Government in 1998. It thesis, University of Oldenburg. covers all aspects of conservation and development in the Lawesson, J. E. 1990. Alien plants in the Galapagos is- province (which includes the entire archipelago), from lands, a summary. In Lawesson, J. E.; Hamann, O.; immigration and waste disposal to natural resource use, Rogers, G.; Reck, G. and Ochoa, H. (eds.). Botanical conservation of rare species, and control of pests. Regula- Research and Management in the Galapagos Islands. tions for the implementation of the new law are currently Monographs in Systematic Botany from the Missouri being drawn up, including sections on quarantine and in- Botanical Garden 32: 15–20. troduced species management. Weed risk assessment will Lawesson, J. E.; Adsersen, H. and Bentley, P. 1987. An be written into these regulations, as will requirements for updated and annotated checklist of the vascular plants planning for the control or eradication of introduced plants. of the Galápagos islands. Reports from the Botanical The regulations will place legal obligations on various Institute, University of Aarhus 16. public and private bodies, as well as on individual land- owners, to undertake specified actions in regard to declared Loope, L. L.; Hamann, O. and Stone, C. P. 1988. Com- weed species or new introduced plants, including moni- parative conservation biology of oceanic archipelagoes. toring, reporting, controlling, and eradicating. Such regu- Hawaii and the Galápagos. Bioscience 38: 272-282. lations and obligations are especially essential for the suc- Mauchamp, A. 1997. Threats from alien plant species in cess of eradication efforts, where all populations of the the Galápagos Islands. Conservation Biology 11: species, whether on public or private land, must be treated. 260-263. Moll, E. 1990. A report on the distribution of introduced THE FUTURE plants on Santa Cruz island, Galapagos. Cape Town, University of Cape Town. Recent advances in weed control in Galapagos have in- Porter, D. 1822. Journal of a cruise made to the Pacific cluded the development of an objective risk assessment Ocean, vol. 1. New York, Wiley and Halstead. system, the first properly designed, implemented and moni- Porter, D. M. 1984. Endemism and evolution in terrestrial tored control trials, the first comprehensive projects to in- plants. In Perry, R. (ed.). Key Environments. Galapagos, vestigate the ecology of serious weeds, the first strategic pp. 85–100. Oxford, Pergamon Press. eradication programme, the first attempt to eradicate a well- established invader, and the approval of the first compre- Schofield, E. K. 1989. Effects of introduced plants and hensive legislation to deal with the problem of environ- animals on island vegetation: examples from the mental weeds. All of these developments have taken place Galapagos archipelago. Conservation Biology 3: in the last five years, and serious attempts to eradicate in- 227-238. vasive species from the archipelago have begun to look Slevin, J. R. 1959. The Galapagos islands. A history of promising. These activities, aside from their scientific and their exploration. Occasional Papers of the California conservation value, should act as valuable confidence- Academy of Sciences 25: 1–150. builders, changing the climate of opinion about the dan- Soria, M. C.; Gardener, M. R. and Tye, A. 2002: Eradica- gers of introduced species and the feasibility of their con- tion of potentially invasive plants with limited distri- trol and eradication. If successful, we could be about to butions in the Galapagos Islands. In Veitch, C. R. and experience the first recorded decrease in the number of Clout, M. N. (eds.). Turning the tide: the eradication introduced plant species in Galapagos since their discov- of invasive species, pp. 287-292. IUCN SSC Invasive ery in 1535. Species Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK. ACKNOWLEDGMENTS Tye, A. 2001a. Invasive plant problems and requirements for weed risk assessment in the Galapagos islands. In: We thank Dick Veitch for encouraging this paper, Heinke Groves, R. H.; Panetta, F. D. and Virtue, J. G. (eds.). Jäger and Susan Timmins for comments on a draft, and Weed Risk Assessment, pp. 153–175. Melbourne, André Mauchamp for data used in his 1997 article. CSIRO Publishing. REFERENCES Tye, A. 2001b. Rising numbers of introduced plant spe- cies in Galapagos. Galapagos Report 2001. Quito, WWF – Fundación Natura. Hamann, O. 1984. Changes and threats to the vegetation. Wiggins, I. L. and Porter, D. M. 1971. Flora of the In Perry, R. (ed.). Key Environments. Galapagos, pp. Galapagos Islands. Stanford, Stanford University Press. 115-131. Oxford, Pergamon Press. Zapata Erazo, C.; Cruz, D. and Causton, C. 2000. El Hickman, J. 1985. The enchanted islands: the Galapagos sistema de inspección y cuarentena para Galápagos. discovered. Oswestry, Nelson. Informe Galápagos 1999–2000: 62–65. Quito, WWF– Fundación Natura.

341