mammal invaders on islands: impact, control and control...

Courchamp, Chapuis & Pascal Introduced mammals and their control

23/01/02 1submitted to Biological Review

Mammal invaders on islands:

Impact, control and control impact

FRANCK COURCHAMP1, JEAN-LOUIS CHAPUIS2 and MICHEL PASCAL3

1 Corresponding author. Ecologie, Systématique & Evolution, Université Paris-Sud,Bat. 362, 91405 Orsay Cedex, France (email: [email protected]; Fax: 003316915 5696 Phone: 0033 16915 5685)

2 Laboratoire d'Evolution des Systèmes Naturels et Modifiés, (UMR 6553, Rennes 1)Muséum National d'Histoire Naturelle - 36, rue Geoffroy Saint-Hilaire - 75005 ParisFrance

3 INRA - Equipe Faune Sauvage et Biologie de la Conservation, Station SCRIBE,Campus de Beaulieu - F 35 042 Rennes Cedex France

ABSTRACTThe invasion of ecosystems by exotic species is currently viewed as one of the most

important sources of biodiversity loss. The largest part of this loss occurs on islands, whereindigenous species have often evolved in absence of strong competition, herbivory, parasitism orpredation. As a result, introduced species thrive in those optimal insular ecosystems affectingtheir plant food, competitors or animal prey. As islands are characterised by a high rate ofendemism, the impacted populations often correspond to local subspecies or even unique species.

One of most important taxa concerning biological invasions on island is mammals. Most ofthe time, the introduction of mammal has been voluntarily: most domestic mammals and many ofthose hunted for food, fur or recreation. Other mammals have been introduced inadvertently, likerodents, or for early (and futile) attempts of biological control of these rodents (cats, mongoosesand mustelids). A small number of mammal species is responsible for most of the damages toinvaded insular ecosystems: rats, cats, goats, rabbits, pigs and a few others.

The effect of the presence of alien invasive species may be simple or very complex,especially since a large array of invasive species, mammals and others, can be presentsimultaneously and interact among themselves as well as with the indigenous species. In mostcases, these effects are very important and often lead to the impoverishment of the local flora andfauna.

mailto:[email protected];



The best course of action is almost always to control the alien population, either byregularly reducing their numbers, or better, by eradicating the whole population from the island.Several types of methods are currently used: physical (trapping, shooting), chemical (poisoning)and biological (e.g., use of diseases). Each has its own set of advantages and disadvantages,depending on the mammal species targeted. The best strategy is almost always to combine severaltypes of methods. Whatever the strategy used, it is crucial to have solid support for the operationto be successful (financial support, staff commitment, public support, …).

In many cases, the elimination of the alien invasive species is followed by a rapid, andoften spectacular, recovery of the impacted plant communities and animal populations. However,in many other cases, the removal of the alien is not sufficient for the damaged ecosystem to getback to its former state, requiring complementary actions, such as species re-introduction. In toomany other cases, the sudden removal of the alien species may generate a further desequilibrium,which can lead to the explosion of another invasive species, previously held in check by the firstone, with resulting similar or greater damages to the ecosystem. Because actual interactions arenumerous and often complex, and potential interactions are even more so, it is, at best, difficult topredict the outcome of the removal of key species such as a top predator, be it introduced andcausing damages. This justifies careful pre-control study and preparation for an alien specieseradication not to turn into an ecological catastrophe. Similarly, long term monitoring of the post-eradication ecosystem is crucial both to assess success and prevent reinvasion.

Key words: alien species, biological invasions, pest management, introduced species,

island restoration



CONTENTS

I. Introduction.................................................................................................4(1) Biological invasions: increased concern for a cause of major change......................4(2) Islands: fragile and threatened ecosystems.....................................................5(3) Nature and purpose of this review..............................................................6

II. Ecological consequences of introductions.............................................................6(1) Impact assessment................................................................................6(2) Reasons for high impact.........................................................................7(3) Simple ecological effects....................................................................... 10

(a) Simple effects due to single species introductions.................................... 10(b) Notorious invaders....................................................................... 11(c) The case of competition and parasitism................................................. 13(d) Simple ecological effects due to multi-species introductions........................ 14

(4) Complex interactions........................................................................... 15III. Control: traditional vs. biological methods......................................................... 18

(1) Traditional methods............................................................................. 19(a) Large mammals............................................................................ 20(b) Medium and small sized mammals...................................................... 21(c) Advantages and disadvantages of traditional methods................................ 23

(2) Biological methods............................................................................. 25(a) Decreasing the survival of alien mammals............................................. 25(b) Decreasing the reproduction of alien mammals........................................ 26(c) Advantages and disadvantages of biological methods................................ 27

(3) Strategies......................................................................................... 29(a) Eradication of established populations.................................................. 29(b) Eradication of new incursions........................................................... 30

IV Consequences of alien mammal control.............................................................. 32(1) An increased number of successful control programs...................................... 32(2) On the definition of successful programs.................................................... 36(3) Positive and negative effects of mammal eradication....................................... 37

IV. Conclusion.............................................................................................. 41V. Acknowledgement...................................................................................... 41VI. References.............................................................................................. 42



I. Introduction

(1) Biological invasions:increased concern for a cause ofmajor change

The invasion of species into naiveecosystems, a process known as biologicalinvasion, has historically been overlooked bymost scientists, as well as politicians andeconomists. It is now, however, consideredas the second most important cause ofbiodiversity loss, after habitat destructionand fragmentation (Vitousek et al., 1997b).This recent change in awareness has severalreasons: the increased concern forbiodiversity, the increase of biologicalinvasions with accelerating global exchange(and the associated gain of knowledge abouttheir consequences) and the delay ofecological effects, such that theconsequences of some ancient invasionshave only become visible in recent decades(Williamson, 1996; Williamson, 1999).

In this context, islands that havemaintained their original ecosystems withouthuman-induced species introduction arevirtually non-existent: islands are therecipients of 80% of documented bird andmammal introductions (Ebenhard, 1988). Aswe will see in this paper, insular ecosystemsare particularly sensitive to such disturbances(Holdgate, 1967; Williamson, 1999), andbiological invasions frequently result inprofound modifications of initialcommunities.

It is thus justified to assess the actualimportance of biological invasions, eventhough their significance may not beappreciated by non-ecologists. The evolutionof ecosystems is indeed something natural,so why should we worry? What differencecan it make if a species is introduced into atrophic web by humans, rather than by a“natural process”? Why would it be aproblem if this species out-competes orexploits a native species? Isn’t it what naturalselection is about? Isn’t it a process

contributing to the evolution, as defined byDarwin himself?

The most immediate answer to thesequestions is that these artificial changes areoccurring at a much faster rate than naturalextinction/colonisation processes and that,compared to natural events, or even to earlyhuman interference, they concern a muchhigher number of species, occur with greaterfrequency, have far greater dispersaldistances and thus affect a much largergeographic range. Consequently, they affectvirtually all insular ecosystems. Thanks tohuman migration and commerce, the numberof species whose colonisation of new rangeshas been facilitated has increased by ordersof magnitude in the last two centuries (diCastri, 1989), in (Mack et al., 2000). Notonly do human activities promote biologicalinvasions by being very effective dispersalagents, but also they facilitate theestablishment of new immigrants, throughhusbandry activities that protect small andvulnerable immigrant populations until theyget larger and more resistant (Mack et al.,2000). In fact it has recently been estimatedthat the current rate of species extinction isbetween 50 and several hundreds timesgreater than was estimated on the basis ofgeological data (Wall et al., 2001). Thisecological homogenisation is mainly causedby the relatively recent increase in humanmovements on the globe, which led to theerosion of geographical barriers and thedecrease of insular ecosystems isolation. Fortens of millions of years, species not capableof long-range dispersion have divergedbehind sea barriers on separate land habitats(Holdgate, 1986). Accelerated fluxes ofspecies between land habitats by humandispersal activities are now acting toeliminate such isolation. Biological invasionscurrently constitute one of the main causes ofmajor global ecological change (Vitousek etal., 1997a), and could very well influenceother agents of global change (Mack et al.,2000). For the purpose of this review, wewill follow this distinction between naturaland man-induced additions of species intoecosystems, focussing on the latter.



(2) Islands: fragile andthreatened ecosystems

Among introductions of species intonew habitats, most are failures (deVos &Petrides, 1967). For example, of the 150 orso species of birds introduced onto Hawaii,and of the 145 introduced to New Zealand,“only” 30 and 36 have become established,respectively (Roots, 1976; Veitch & Clout,2001). This relates to what has been termedthe 10’s rule (about 10% of introductionssucceed and about 10% of these will besignificantly ecologically disruptive(Williamson, 1996)). However, the sheernumber of attempts still makes the lowproportion of successful introductions a verylarge number of cases, concerning virtuallyall major islands. In addition, some species,whose ecological equivalents are naturallylacking in the native ecosystem, may have adisproportionate impact, not following the10% rule. Such is the case for a number ofmammal species. Of the 55 mammal speciesintroduced to New Zealand, 32 have becomeestablished, far more than the 10% (Veitch &Clout, 2001). A little more than a decadeago, Ebenhard recorded 644 mammalintroductions on islands (Ebenhard, 1988). Itfollows that very few islands escaped theproblems caused by species introductions.For example, over 80% of all islands havebeen invaded by rats, Rattus sp. (Atkinson,1985), and at least 65 major island groupshave also been invaded by cats, Felis catus(Atkinson, 1989).

Most of contemporary world-wideextinction events have occurred or arecurrently occurring in insular ecosystems.For example, more than 90% of the 30species of reptiles and amphibians reportedas extinct since the 17th century were islandforms (Honnegger, 1981). Similarly, 93%of 176 species or subspecies of birdextinctions that have occurred during thisperiod, as well as 81% of 65 mammalspecies, have also occurred on islands(Ceballos & Brown, 1995; King, 1985).These of course include only the species thatare known to have disappeared: thevertebrate fauna of mainland ecosystems is inmany cases better known that the insular onepredating humans arrival, and the figure

ought to be accordingly much higher.Similarly, it is very likely that the lownumber of reptiles and amphibians here onlyreflects the low number of studies on thesetaxa. Few data are available concerningextinct insular species of plants and ofinvertebrates, but the figure is undoubtedlyproportional. The number of speciesextinctions is thus very likely to be fargreater than the currently accepted figures.

When research is being undertaken toconserve threatened species, resources andtime are limited and are usually focused onsingle species, but it is clear that there willnever be sufficient time and resources tomanage each taxon individually (Fitzgerald &Gibb, 2001; Saunders & Norton, 2001).From this established fact has emerged theconcept of biodiversity hotspots: areasfeaturing exceptional concentrations ofendemic species and experiencingexceptional loss of habitat (e.g., (Myers etal., 2000b; Reid, 1998)). In fact, islands orisland groups such as New Caledonia, NewZealand, the Caribbean,Polynesia/Micronesia, Madagascar and thePhilippines have been defined as biodiversityhotspots (Myers et al., 2000b). No less thannine of the 25 hotspots defined by Myers andcol. are entirely or mainly made up ofislands, and almost all tropical islands fallinto one or another hotspot (Myers et al.,2000b). Conservation of these insularecosystem is not only among the mosturgent, it might also be among the mostfeasible: much of the time, the biodiversitycrisis taking place on islands has a singlecause, biological invasions, and it is lessdifficult to fight it on a closed, and oftenremote area, than on the same surface on thecontinent.

The successful introduction of alienspecies is the major reason of ecosystemperturbation and biodiversity losses on theseislands (Atkinson, 1985; Lever, 1994;Moors & Atkinson, 1984; Williamson,1996). For example, predation by introducedanimals has been a major cause of 42 percentof island bird extinction in the past, and is amajor factor endangering 40% of currentlythreatened island bird species (King, 1985).In this context, it is said that introduced



mammals induced more problems than anyother vertebrate group (Ebenhard, 1988;Lever, 1994). Mammals also are responsiblefor the best documented and most spectacularof ecological disturbances owing tobiological invasions (Holdgate, 1967). Thisis certainly due in part to the lack of anynaturally occurring terrestrial mammal onmost remote islands (Atkinson, 2001).Plants and insects, for example, are alsonotoriously devastating invaders, butbecause mammals constitute a large enoughsubject by themselves, we will focus only onmammal species in this review.

(3) Nature and purpose of thisreview

In this paper, we will attempt to reviewthe effects that can follow the introduction ofone or several mammal species onto anisland. In addition, we will outline the mostutilised methods of control of introducedmammal populations. We will then discussthe main possible impacts, both positive andnegative, of such mammal control. Becauseexhaustiveness is not possible on thissubject, we will instead provide an overviewof the main effects of mammal introductions,of the main methods of their control and ofthe main consequences of such control,illustrating most cases with typical examples.By the nature of studies on biologicalinvasions (often triggered by conservationproblems, and thus by effects of invasions),and by our choice of giving strikingexamples of effects of such invasions, thisreview may be biased towards reportingnegative effects of introduced species.

Following the Invasive SpeciesSpecialist Group’s (ISSG) definition ofterms (I.U.C.N., 2001), alien, foreign,exogenous and exotic species will be usedsynonymously to qualify species occurringoutside of their natural range of potentialdispersal, by opposition to native,indigenous or autochthonous species. Alieninvasive species will be used for an alienspecies that becomes established and is anagent of change in the invaded ecosystem,threatening the native biological diversity.Introduced species refers to a movement(unintentional or intentional) by human

agency of a species outside its historicallyknown natural range.

II. Ecological consequences ofintroductions

(1) Impact assessment

On islands, the successfulestablishment of introduced species can leadto dramatic changes in the original trophicwebs. As we will detail below, the presenceof introduced species often causes significantchanges of the abundance and/or spatialstructure of autochthonous species it interactsdirectly or indirectly with, and thereby canaffect significantly the global functioning ofconcerned ecosystems.

It is however generally very difficult toassess the impact of an introduced species onthe ecosystem it has invaded. Most of thetime, data are not available to comparecommunities before and after the invasion.The recovery of native flora and faunafollowing exotic mammal eradication isgenerally a good illustration of their realimpact on invaded ecosystems (Newman,1994; Taylor, Kaiser & Drever, 2000;Towns, Daugherty & Cree, 2001). Analternative approach is possible in the case ofarchipelagos, where several combinations ofintroduced species can occur on nearbyislands. Such islands are similar in mostbiotic and abiotic aspects, which allows theestimation of single species effects (e.g.,comparing islands with and withoutintroduced goats, Capra hircus) or ofcombinations of introduced species (e.g.,islands with rats and cats vs. islands withcats only, islands with rats only and islandswith no introduced species). Unfortunately,studies on the effects of alien controlprograms based on comparison betweenareas with or without alien removal (orcomparisons before/after removal) cannotalways be considered as unconditionalevidence. Indeed, the communities may notrestore themselves to their former states, dueto the deep modifications caused by the alienprior to its removal, to the presence of otheraliens, or to other factors. Even evidence



based on diet of invaders and correlated withshifts of abundance of native species can berightly criticised, since a causative process isnext to impossible to demonstrate in thiscontext. In absence of such evidence, therehas even been doubt concerning the actualimpact of species considered by mostconservationists as the most notoriouslyharmful invading species (Norman, 1975).As stated by Ebenhard, “The decisionwhether an introduction has had anecological effect is often hard to make onthe basis of published information only,and hence is open to criticism by peoplewith first-hand experience of someparticular situation” (Ebenhard, 1988).There is however a very long list ofcircumstantial evidence of the effects ofintroduced species, which, taken together, isa body of “ circumstantial proof ”(Peabody, 1961) that can overpower eventhe slightest doubt about the dramatic impactof biological invasions on oceanic islands.

According to Ebenhard, introducedanimal species may affect native species inone or more of six different ways. Theymay: (i) affect plant populations and speciesliving in the habitat structured by theseplants; (ii) be predators of native prey; (iii)induce interference or resource exploitationcompetition; (iv) spread micro- andmacroparasistes into natives populations; (v)induce genetic changes to native speciesthrough hybridisation; (vi) act as prey tonative predators (Ebenhard, 1988). The twolast categories are probably relativelyunimportant as far as islands are concerned.Effects through competition or release ofpathogens are probably much moreimportant, but so far this is not reflected inthe literature, as it is extremely difficult tounambiguously demonstrate such effects,and thus very few studies have focused onthese points. Moreover, such examples maybe rare simply owing to the scarcity of nativeterrestrial mammals on oceanic islands.Although not entirely based on the aboveclassification, the examples given in thisreview will therefore be mostly restricted tothe effects due to exploitation of nativespecies by introduced species, as well as

resulting indirect effects (habitat destructionand competition with other species).

(2) Reasons for high impact

Not only are most alien invasivespecies highly adaptable (Apps, 1986;Ebenhard, 1988; Flux, 1993; van Aarde,1986), but they also encounter favourableconditions in the new environment,compared to their native environment.Because of their long evolution in isolationfrom other species, insular species have notevolved traits to provide an adapted responseto efficient interference and exploitation(Atkinson, 1985; Atkinson, 2001; Burger &Gochfeld, 1994; Holdgate & Wace, 1961;Moors & Atkinson, 1984; van Aarde &Skinner, 1981). For example, plants onoceanic islands have often less efficientdefensive mechanisms against intensiveherbivory, such as high fecundity,compensated growth, production of toxins orunpalatable substances, or physicaldeterrents, such as spines (Atkinson, 1989).Therefore the plant communities of oceanicislands are mainly composed of species thatare highly palatable and vulnerable(Holdgate, 1967). Similarly, smallvertebrates may not have learned to hide theirsounds, odours, or their shelters frompredators, and once discovered, they maynot flee or defend their young. Flightlessbirds are a good example of this: Dodos inMauritius, Moas and Kiwis in New Zealand,Kagus in New Caledonia. In addition tobeing behaviourally maladapted, these preyspecies lack the life-history traits associatedwith many continental species, such as anearlier age of first reproduction and highfecundity (some examples are given for NewZealand shorebirds in (Dowding & Murphy,2001)). These species constitute a major (andsometime unique) and easy trophic resourcefor the few founders of introduced predators.Introduced cats, rats and dogs, Canisfamiliaris, all have participated in theextermination the flightless birds of Tristanda Cunha, Amsterdam, Ascension and SaintPaul islands, among others (Lever, 1994).The main islands mentioned in the text arelocated on the map Figure 1.



In addition to a great abundance ofresources, successfully introduced speciesare also favoured by a lower pressure fromnatural enemies. Introduced species alsogenerally suffer from a smaller subset ofparasites (which are often less virulent thanon the mainland, at least for oceanic islands(Dobson, 1988)). Indeed, first, the limitednumber of hosts involved in introductionprocesses results in fewer pathogenicparasites being transmitted to the newenvironment. Second, many introducedparasitic species may be lost either becausethe new environment is unsuitable, becausesome key species are lacking that prevent thecompletion of its life cycle, or because thedensity of the introduced population isinitially too low to allow the maintenance ofthe parasite population (Keeling & Grenfell,1997). It is also notable that native predatorsare often relatively inefficient in preyingupon introduced species. After more than acentury, the skua, Catharacta skua, stilldoes not prey very efficiently on rabbits,Oryctolagus cuniculus, introduced to theKerguelen islands, preying mostly on youngand sick individuals (Moncorps et al., 1998).

With an abundance of resources and alack of pressure from natural enemies,invasive animals often build up numbers inan abnormally short period afterintroduction. Because these introducedspecies have not evolved to adapt theirpopulation dynamics in a closed (and oftensmall) geographic area, their populationsnewly freed from biotic pressures canincrease beyond the limit of sustainability ofthe new environment (in terms of foodresources as well as shelters). As a result,when these resources become lacking apopulation crash may follow a marked peak(Nugent, Fraser & Sweetapple, 2001). Insome cases, this crash is fatal to the alieninvasive population. After this crash,depending on the damage on the resources,the population may either recover to a lowerpoint of stability, with much impoverishedresources and often with changes incomposition, or it may simply die out. Forexample, the 29 reindeer, Rangifertarandus, introduced onto St Matthew Islandresponded so well to the high quality and

quantity of forage that they increased in lessthan 20 years up to 6000 individuals. Thehuge impact of such a large population on arestricted environment led to an importantimpoverishment of the flora, resulting in acrash of the dramatic population, down to 50individuals (Klein, 1968). Such a reductionof the population can lead to an increase ofthe vegetation, which may allow anotherbuild up of the herbivore population,sometimes leading to another crash. In 1957,a pair of mouflons (Ovis musimon) wasintroduced on a small island (650 ha) of theKerguelen Archipelago. From those twoindividuals, a population developed thatnumbered around 700 individuals 20 yearsafter. Between 1977 and 1995, at least 5population crashes occurred, mainly becauseof insufficient food availability at highdensity. Each crash, concerning 400-450individuals, was followed by a recovering ofthe flora, despite an increasing degradationof the environment by erosion, leading to arebuilding of the population. This circle ofevents was broken in 1995, the date of thelast population crash, with populationreduction by a yearly shooting campaign(Boussès, Réale & Chapuis, 1994; Chapuis,Boussès & Barnaud, 1994a; Réale, Boussès& Chapuis, 1996).

Alternatively, the damage to theenvironment may be such that the habitat orthe resources themselves can be totallydestroyed, also resulting in the extinction ofthe introduced species. In the long term,such a natural extinction of introducedspecies (more likely to occur on small islandswith simple ecosystems) may be moregeneral than currently witnessed. Forexample, it has been reported than for the607 islands where the fate of introducedrabbits is known, the population died outnaturally more than ten percent of the time(Flux, 1993). However, following exoticdisappearance or removal, recolonisation bylocal species is not always possible, owingto the very high endemism of oceanic islandsand the lack of neighbouring populations thatcan act as recolonisation sources, or becausethe alien invasive species caused irreversiblemodifications of the environment (such asextreme erosion).



(3) Simple ecological effects

(a) Simple effects due to singlespecies introductions

Ecological effects of introducedspecies may be better appreciated if one candisentangle the different effects of differentspecies, and the effects of speciescombinations. Because trophic webs onoceanic islands are often very simplified,with little ecological or taxonomicalredundancy (Holdgate, 1967), dramaticincreases of introduced species may lead toimportant unbalances in the wholeecosystem. The decline of endemic speciesthat are victims of interference or exploitationfrom introduced species is often the firstconsequence, and may then lead to furtherdamage at the ecosystem level. Three typesof ecological effects can be distinguished: (i)shift of relative species abundances, withoutaffecting the total number of species; (ii)disappearance of some species from parts ofthe habitat; (iii) extinction of some species.Some ecological effects, habitat change orshift of native abundance of plant or animalspecies caused by predation or competitionhave been reported in at least 40% of allmammal introductions (Ebenhard, 1988).These dramatic results are better appreciatedwhen compared to the effects of birdintroductions, which have resulted inobserved ecological effects in only 5% of allcases (Ebenhard, 1988). The figure formammals is much higher if onlyintroductions onto oceanic islands areconsidered.

Ebenhard also calculated that 93 out of95 introductions of strict herbivore speciescaused habitat change on oceanic islands,and the number of plant species affected byintroduced herbivores is probablyinnumerable (Ebenhard, 1988). Becausethere is little published information on theindirect effect of plant community changeson invertebrates, it is difficult to assess thedirect and indirect effects of mammalintroduction on local invertebrates.However, they are certainly considerable.For example, the extinction of a single plant

may cause the extinction of insects whoselife cycle relies on that plant, in addition tothe extinction of other invertebrates relyingon that insect. For example, in the Kerguelenislands, the introduction of rabbits led to thenear elimination of the Kerguelen cabbage,one of the main vascular plant of theecosystem. Such biodiversity loss led to adisplacement of the trophic niches of a stiltfly, Calycopteryx moseleyi, which is nowfound on sea algae (Tréhen et al., 1987). Theeradication of the rabbit from these islandstriggered the return of this phytophagousinsect back to its original trophic niche(Chapuis et al., 2000).

The list of infamous cases of trophicdisruptions is sadly becoming very large, sothat exhaustiveness becomes impossible.Therefore, only a very few typical exampleswill be given in this review. A morecomprehensive list of the effects of mammalintroductions is given by Ebenhard(Ebenhard, 1988). The bad start of BuckIsland (West Indies) as a newly classifiedU.S. National Monument in the early 1960sis a good illustration of the potentially severeimpact that a single species can have on anindigenous species. Ten years before thatdesignation, two pairs of mongooses,Herpestes auropunctatus, had beenintroduced to this small island, and despitean immediate but unsuccessful eradicationcampaign, a further decade was all that wasneeded for the descendants of these foundersto exterminate the Saint Croix ground lizard,Ameiva pelops (Lever, 1994).

A single introduced species can alsoaffect, directly or not, a whole range ofindigenous species, thereby affecting thewhole ecosystem it invades. It is the case ofthe Indian mongooses again. Nine foundersintroduced to Jamaica from Calcutta in 1872seem to have been the source of all thepopulations on 29 Caribbean and 4 Hawaiianislands before the end of the XIX century(Espeut, 1882; Lever, 1994; Nellis &Everard, 1983). These mongooses eradicatedmany invertebrates, reptiles, amphibians,birds and mammals, mostly in the WestIndies (Lever, 1994; Nellis & Everard, 1983;



Westermann, 1953). Another goodillustration of this is given by domesticgoats, generally seen as one of the mostdestructive feral mammals to nativevegetation (Coblentz, 1978; Daly & Goriup,1987; Keegan, Coblentz & Winchell, 1994;Parkes, 1990; Turbott, 1948). Introducedgoats have a very large spectrum diet: theyconsumed at least 48 species of vascularplants on Raoul Island, mostly indigenousspecies, and several of them rare andendangered (Parkes, 1984). Within 35 yearsof overgrazing, over-browsing andtrampling, introduced goats reduced thenumber of plant species from 143 to only 70on Great Island, New Zealand (Turbott,1948). Through direct competition, or byseverely reducing the vegetation cover, goatsintroduced in the Galapagos archipelago alsoindirectly threatened several species of birdsand reptiles (including the local race of thegiant tortoise (Eckhardt, 1972; Hamman,1975)). On Santa Catalina Island, they alsoalmost wiped out the native Artemisiacalifornica, and the resulting reducedcoverage seriously threatened the Californiaquail, as well as several native rodents;consequently, several species of reptilian andmammalian predators have decreased as well(Coblentz, 1978). The multiple impacts madeby single species are also well illustrated bythe brushtail possum, Trichosurusvulpecula, in New-Zealand (Clout, 1999).The primary conservation impact of thisinvading marsupial is through the damagethey cause to native forest through browsing.Possums also feed on flowers and fruits,reducing fruit crops of native plants andhence both forest regeneration and the foodsupplies of several native birds, which mayfail to reproduce and even survive. They alsocompete for hollows with endemic hole-nesting species, such as kiwi, Apteryx spp.Although mostly herbivorous, possums haverecently been shown preying on the eggs andchicks of birds, including those of threatenedspecies (Brown et al., 1998). Finally, theyhave an economic impact as well, as theirtransmit a virus, the bovine tuberculosisvirus, to cattle and deer (Clout & Sarre,1997). Because of their multiple ecologicaland economic effects as an invasive species,the introduction of possums to New Zealand

is now recognized as “an unmitigateddisaster for the natural ecosystems andnative biota of this archipelago” (Clout,1999).

(b) Notorious invaders

Most of man’s commensal species,and many domestic animals, have beenintroduced onto islands. Some have beenintroduced inadvertently; for example theywent ashore from ships by themselves orthey came from newly established humancolonies (rats, mice, Mus musculus, cats,dogs). However, in contrast to plants andinvertebrates, the majority of mammals weredeliberately introduced, sometimesrepeatedly and tenaciously. These weregenerally introduced either to control thespread of previously introduced rodents(e.g., cats, mongooses), for attempts of furor stock farming or hunting (e.g., cattle, Bostaurus, goats, pigs, Sus domestica, wildboar, S. scrofa, horses, Equus caballus,donkeys, E. asinus, foxes, possums,reindeer, mouflon, deer) or to provide adurable source of food for shipwreckedmariners (e.g., rabbits, goats, sheep, Ovisaries, and the brown and Arctic hares,Lepus timidus, L. europaeus). A large partof research on biological invasion isfocussed on characterising the biologicalattributes of successful invading species(Crawley, 1986; Kolar & Lodge, 2001;Mack et al., 2000; Williamson, 1996).Indeed, some species are notoriously moresuccessful in this regard, both because theyhave been more often introduced and becausesuch introductions have been more oftenfollowed by establishment. Among these,some are also well known to conservationbiologists because their establishment isalmost invariably followed by rapid andimportant changes in the originalcommunities. The most important introducedmammals, both in number of introductionsand in damages, are the common rabbit, thecat, three commensal species of rats, thehouse mouse, the goat, the pig, the dog andcattle. Apart from the two latter, all are in thelist built by the ISSG listing 100 of the worstalien invasive species(http://www.issg.org/database/welcome/).

http://www.issg.org/database/welcome/



Many of these invasive species have had amajor impact, as illustrated by the number ofspecies allegedly extirpated by theirintroduction. For example, (Jackson, 1977),in (Ebenhard, 1988) estimated that 61 birdtaxa have been exterminated by introducedpredators, with cats (33 extinctions), rats (14extinctions) and mongooses (9 extinctions)being the most important predators.

Specialists on several of these speciesrival in the use of strong terms to describethe impact of these mammals whenintroduced into naive ecosystems:

- “The domestic goat undoubtedlyhas been the most destructive feralmammals to native vegetation, particularlyas it travels over all types of terrain andconsumes all kinds of browse andherbaceous material.” (deVos, Manville &Van Gelder, 1956).

- “With the exception of man, the ratis the mammal that has undoubtedly hadthe greatest impact on the environmentand economy of most Pacific islands...”(Storer, 1962).

- “Of all the introduced pests, therabbit is by far the worst. Itsacclimatization in Australia was thegreatest single tragedy that the economyand the native animals have eversuffered.” (Frith, 1979).

- “Cats are now ubiquitous oninhabited islands, and frequently presenton those uninhabited. Probably no otheralien predator has had such an universallydamaging effect on seabirds.” (Moors &Atkinson, 1984).

- “The avifaunas of some islandshave suffered drastically from predationby rats. (…) On a few islands, theproportion of bird species that havedeclined or become extinct following theintroduction of R. rattus is so great that theterm catastrophe is appropriate.”(Atkinson, 1985).

- “Rats … have been implicated inthe greatest number of extinctions due topredators (54 percent… Of the threespecies which have shared the blame forthis record of predation, the Black Rat orRoof Rat (Rattus rattus) has been the mostserious problem…”(King, 1985).

- “Introduced predators haveaccounted for about half of island birdextinctions. Rats and cats are the mostnotorious killers and island birds the mostnotorious victims, in this regard.”(Diamond, 1989a).

- “… there can be little doubt thatthe mongoose in the West Indies … hashelped to endanger or exterminate morespecies of mammals, birds and reptileswithin a limited area than any otheranimal deliberately introduced by mananywhere in the world.” (Lever, 1994).

A short review of the effects of theintroduction of three of these species (anherbivore, an omnivore and a carnivore) isprovided in Boxes 1 (rabbits), 2 (rats) and 3(cats), in order to provide an idea of thephenomenal impact of mammal introductionsworld-wide. If these three species areamongst the most notorious mammalsnuisance on islands, many other species haveunfortunately caused conservation problemsall over the word. Some are close behind theabove calamities in terms of threat on insularspecies, especially goats (Coblentz, 1978;Daly & Goriup, 1987; Parkes, 1984; Parkes,1990; Parkes, 1993; Rudge, 1984; Turbott,1948), pigs (Hone, 1990; Hone, 1992;Hone, 1995; Sterner & Barrett, 1991), smallIndian mongooses (Seaman, 1952), mice[Rowe-Rowe, 1989 #397; Gleeson, 1982#396; Le Roux, 2001 #405; Newman, 1994#370; Crafford, 1990 #411], foxes (arcticAlopex lagopus and red Vulpes vulpes,(Bailey, 1992; Bailey, 1993; West & Rudd,1983)) and domestic dogs (Iverson, 1978).For example, dogs in the vicinity of humansare reputed to hunt principally for “sport”,and can thus exceed other alien predators inexterminating species. A noticeableillustration of this is the individual dog who



is said to have systematically hunted andkilled more than half of the remaining 900New-Zealander brown kiwis, Apteryxaustralis, in less than 3 weeks (Diamond,1989b; Taborsky, 1988).

In addition to the above, severalauthors list mammal species reported to havebeen introduced onto islandse (deVos et al.,1956; Ebenhard, 1988; Holdgate, 1967;Lever, 1994). There have been at least 80mammal species introduced since 1700(Atkinson, 1989), but some of them causemore damage than others. The other mostcurrently introduced mammals are mustelids(stoat, Mustela erminea, weasel, M.nivalis, ferret, M. furro, north mericanmink, M. vison, polecat, M. putorius,Japanese weasel, M. sibirica itatsi),wallabies (Dama, Macropus eugenii,Parma, M. parma and brush-tailed rock,Petrogale penicillata) and deer (Axis,Cervus axis, fallow, C. dama, red, C.elaphus, Rusa, C. timorensis, while-tailed,Odocoileus virginianus and reindeer). Someare less frequent but nonetheless important,for example the Virginia Opossum,Didelphis marsupialis, the house shrewCrocidura russula, the musk shrew, Suncusmurinus, the crab eating macaque, Macacafascicularis, the green monkey,Cercopithecus aethiops and the waterbuffalo, Bubalus arne.

Even species that are almostanonymous in the world of alien mammalcalamities can have a very important impact.For example, the house shrew, recentlyintroduced in some Brittany islands,generated the disappearance of the insularshrew form, Crocidura Suavelons, fromMolène and Sein in less than 30 years(Cosson, Pascal & Bioret, 1996); the Indianor musk shrew, introduced in Réunion islandbefore 1730, and to Guam islands in the1950s generated much damage to the nativefauna of these islands (Fritts & Rodda, 1998;Probst, 1997).

(c) The case of competition andparasitism

Apart from the classically studieddirect effects of herbivores and predators,which we have detailed above, introductionsmay be deleterious for native populationsthrough two other processes, bothunderstudied but potentially very important:competition and spread of parasites.

Although competition is rather rare,when it occurs it often affects the abundanceof the native species (Ebenhard, 1988).Competition can be of two types. It can becalled interference competition, as is the casewith grey-bellied squirrels (Callosciuruscaniceps) introduced on the island ofOshima, and which chase local birds awayfrom the flowers they feed on and pollinate(Lever, 1994). It can also be a resourceexploitation competition, a process bothmore likely to occur and more difficult todemonstrate. Typical examples of suchinteractions include the competition of rabbitsand seabirds for nesting burrows (Cooper &Brooke, 1982) or the competition ofintroduced ungulates which overgraze plantsthat constitute the basis of diet or shelter fornative species (e.g., (Coblentz, 1978;Honnegger, 1981; Roots, 1976; Turbott,1948) in Lever, 1994). These relationshipscan be complex, because some birds maybenefit from rabbit burrowing, or from thevegetation opening. Direct competition forfood with introduced goats is considered oneof the direct reasons for the decline or evenextinction of several populations of birds andreptiles, among which is the giant Galapagosgiant tortoise Geochelone elephantus (Daly& Goriup, 1987). Obviously, strictherbivores are more likely to get involved incompetitive relationships with native speciesthan are omnivores or predators, especiallysince terrestrial predators are scarce onoceanic islands. However, some examplesconcern the competition for food with nativepredators, notably in the Kerguelen islandsbetween cats and skuas: where cats arepresent, there are not enough petrels for theskua to reproduce, despite the presence ofalternative prey, such as mice and rabbits(Chapuis et al., 2001). Similarly, the



domestic mouse can be a competitor ofgranivorous and or insectivorous indigenousspecies. In subantarctic islands, caracterisedby simplified trophic webs and by a rarity ofpredators in terrestrial systems (Holdgate,1977; Smith, 1977; Vernon, Vannier &Tréhen, 1998), the introduced domesticmouse has an important impact oninvertebrate communities (Crafford, 1990;Crafford & Scholtz, 1987; Le Roux et al.,2001). This impact leads to an increasedrarity of invertebrates having a larval cycle ofseveral years, such as flightless moths (e.g.,Pringleophaga kerguelensis; P. marioni)whose caterpillars constitute an importantpart of the mice diet (Crafford, 1990;Gleeson & Van Rensburg, 1982; Le Roux etal., 2001; Rowe-Rowe, Green & Crafford,1989). Control of mice can lead to anincrease in the density of some arthropods,especially beetles and spiders (Fitzgerald &Gibb, 2001). Such impact on invertebratescan have consequences on the trophic web,for example through competition. As theyencountered favourable conditions, miceintroduced on Marion island increased tosuch a point that they could be at the originof the decline of the chionis, Chionis minor,a terrestrial bird whose survival over winterdepends on invertebrate communities(Huyser, Ryan & Cooper, 2000; Smith &Steenkamp, 1990).

The spread of new parasites withinvading animals is probably the least studiedof the major effects of species introductionson native ecosystems. This may besurprising if one considers the upsurge ofconcern for biological control of introducedmammals: the understanding of naturalexperiments that represent introductions ofmammals (and of their parasites) should beof prime interest for applied research. Thisarea also embodies a great potential forfundamental research, since the processes ofestablishment and spread of parasites are farfrom understood. For example, severalmechanisms are likely to act in oppositedirections, and it is not obvious what theresulting effect will be. On one hand fewfounders bring fewer parasites, most ofwhich may disappear during the introductionprocess. On the other hand, the nativespecies may represent new niches to be

exploited, or naive (non-immunised) hostpopulations, which will greatly favour thespread of the new parasite, and introducedspecies may constitute new hosts for nativeparasites. However, the scarcity of nativeterrestrial mammals on oceanic islands maybe one of the reasons for such a lack ofstudies.

A typical example of the impact ofintroduced parasites on the native fauna is theavian malaria introduced in Hawaii andpresumably responsible in part for theextinction of many local bird species. Fewerexamples exist concerning mammalpathogens. These include the introduction ofunidentified pathogens by black rats onChristmas Island, that seems to have led tothe extinction of the endemic bulldog rat(Rattus nativitatis; (Day, 1981)) and theintroduction to Bering Islands of the cestodEchinococcus multilocularis by theNorthern red-backed vole (Clethrionomysrutilus) which affected native foxes (deVoset al., 1956). All these are in fact goodillustrations of the apparent competitionprocess (Abrams, Holt & Roth, 1998; Holt& Lawton, 1994), whereby two speciescompete indirectly via differentialexploitation by a common natural enemy(predator or parasite).

(d) Simple ecological effects due tomulti-species introductions

Unfortunately, there are many well-known examples of coincident introductionof several species, leading to more complex,and often detrimental, effects on the originaltrophic webs. For example, the constructionof a tourist hotel on Caicos Islands led withinthree years to the near extirpation of the15000 endemic ring-tailed iguanas that werehunted for sport by introduced cats and dogs(Iverson, 1978). On Stewart Island, cats andrats of three species have exterminated fivespecies of birds, three of which wereendemic (Karl & Best, 1982). Similarly, inthe Marianne Island, Rota Island numberedno more than 121 adult trees of theendangered Micronesian Serianthesnelsonii, while the neighbour Guam had



only a single one, these trees being mostlythreatened by introduced red deer, pigs andgranivorous insects (Wiles et al., 1996).There are also many cases of biodiversitydecrease generated by the parallel andsimultaneous effects of herbivore andpredator alien species (see below).

Of course, in many cases, severalinvaders belonging to distinct trophic levelshave been introduced into naive ecosystems,leading to some of the most complete casesof community disruptions. The effects ofmammal introductions on Lord Howe Island,off Australia, are a good illustration of this.The endemic forest of Howea forsterananow contains very few small individuals,due to the combination of grazing by goatsand seed consuming by rats (Pickard, 1982).In addition, in less than 20 years, ratseradicated on this island at least five endemicbird species (more than 40% of theindigenous species of land birds) and threeendemic invertebrates. They also extirpatedall the seabird colonies and much reduced thenumbers of lizards, land snails, and otherinvertebrates of the island (Atkinson, 1985;Recher & Clark, 1974). Even consideringonly two mammals, goats and rats, there islittle doubt that many other effects have beencaused indirectly by the change of thestructure of the habitat and by removal ofnative species. One can easily imagine howdeep changes can be caused by theintroduction of more mammal species, as isthe case for many islands. In addition toincreased effects owing to increased numbersof actors, there are many cases ofinteractions between introduced speciesplaying a synergetic effect in the damagesuffered by the local species. This is whatwe call here complex interactions.

(4) Complex interactions

On most islands, several mammalspecies interact to deeply modify the trophicwebs. Without listing exhaustively all thedifferent reported combinations, it is clearthat islands that have been inhabited byhumans even for a short time shelter a greatnumber of alien species, among which manyare mammals. Even for the most remoteislands, as long as they had a useful location

for ships, or worse, a permanent colony, agreat number of alien mammals wereintroduced, most of which were on purpose.As an example, the Kerguelen Archipelago isone of the most remote islands groups, in thesubantarctic zone of the India Ocean, 4000km off the coasts of Africa and 3000 awayfrom Australia. Nevertheless, rats, mice, catsand dogs rapidly followed sealers andwhalers (the two latter predators disappearedspontaneously but cats were introduced againto control rodents). In addition, rabbits, pigsand sheep were introduced to constitute self-maintaining reserves of food, whileAmerican mink were introduced for their fur,reindeer to serve as draught animals, andCorsican mouflon for hunting purposes(Pascal, 1983). Apart from pigs, dogs andAmerican minks, all these species are stillcurrently present in the wild of these islands(Chapuis et al., 1994a). Islands that areinhabited, or that have been heavily used fortheir resources, may face even darkersituations in this regard. For example, theislands of Hawaii shelter now one of themost complete arrays of alien species, withmany mammals, including the three speciesof commensal rats, mice, cats, mongooses,dogs, pigs, goats, cattle, sheep, rabbits,Axis deer and even brush-tailed rockwallabies. The very complex interactionsbetween indigenous and one or severalintroduced species make it very difficult tocharacterise the impact of introduced specieson the indigenous flora and fauna.

The simplest example of interaction isthe facilitation of invasions by other speciesprovided by established alien species thatmodify the habitat. For example, some exoticbirds need the forest to be modified bybrowsing herbivores and are unable tocolonise native forest undisturbed in theabsence of alien mammals. After the removalof introduced browsers (and predators) fromCuvier Island, the indigenous forest was ableto regenerate, and the exotic bird speciesdisappeared (Diamond & Veitch, 1981). Acomparison of Pacific islands with andwithout introduced ungulates suggests thatmany island species of plant can “resist”alien plant invasions in the absence ofungulates (Merlin & Juvik, 1992 in (Cabin etal., 2000).



There are also many cases of directinteractions between several established alienspecies. For example, if a predator isintroduced into a trophic web where twoprey species are competing for a resource,the complexity of its effects may increase ifone of the prey is preferred over the other.An exacerbation of this is when both a preyspecies and its predator are introduced into anaive ecosystem. In this case, the introducedprey is well adapted to the predation and hasan indirect competitive advantage over thenative prey. Worse, the introduced prey’sadaptation to predation may be such that theyare less favoured (i.e., more difficult tocatch), but at the same time this allows anincrease of the predator population withoutsuffering from it (because only thepopulation surplus is killed), leading tohyperpredation (see below). In this context,it is interesting to compare the effect ofintroduced cats on seabird communities, withand without alternative prey.

Effects of the presence of alternativeprey are well illustrated by the example ofrabbits, whose effects on endemic vertebratespecies can be more complex than the onespresented above, especially when predatorshave also been introduced. In this case,rabbits are only a secondary prey item inmonths when seabirds are present, butappear to enable predators (mostly domesticcats) to subsist over winter, when seabirdsare absent or rare (Brothers, Skira &Copson, 1985; Chapuis et al., 1994a;Derenne & Mougin, 1976; Jones, 1977;Pascal, 1980). They also allow predatorpopulations to reach remote colonies orpopulations of indigenous prey on islandswith heterogeneous indigenous preydistribution (Brothers & Copson, 1988).They can also exacerbate the predationpressure on indigenous species through whathas been termed hyperpredation (Smith &Quin, 1996). This process is well describedby the example of Macquarie Island, wherepredation by introduced cats caused thedecline of burrow-nesting petrels (Brothers,1984) and the extinction of an endemicparakeet and a banded rail (Taylor, 1979a).Cats were introduced to this island 60 yearsbefore the introduction of rabbits. However,the dramatic impact of cat predation on bird

populations dates back to just ten yearsfollowing the introduction of rabbits (Taylor,1979a). It is believed that the rabbitpopulation allowed not only the maintenanceof the cat population in winter (whenseabirds are absent from the island), but alsoa significant increase, therefore resulting inincreased predation pressure on the land birdspecies. Such increase was well supportedby the rabbit population, more adapted topredation by cats. It was however fatal to thebird population, which on its own couldsupport a small cat population, but wasextirpated by an over-sized cat populationthat no longer depended on the presence ofbirds to survive (Courchamp, Langlais &Sugihara, 1999b; Courchamp, Langlais &Sugihara, 2000). Since a great many islandsharbour both populations of introducedpredators and their natural prey, thishyperpredation process could be quite awidespread cause of further endangerment ofsmall indigenous vertebrate populations.

The complexity of the situationincreases again if an indigenous predator ispresent in the system, as is the case insubantarctic islands where the brown skua isoften the only natural terrestrial predator(Furness, 1987). It becomes then moredifficult to predict the global impact ofintroduced mammals on native predators,when they possibly both suffer fromcompetition with introduced predators andgain from arrival of introduced prey thatcompete both directly and indirectly (throughhyperpredation) with the native prey. On theKerguelen islands, the main prey of skuasare petrels and prions, especially the bluepetrel, Halobaena caerulea (Moncorps etal., 1998). The 2000 to 4000 breeding pairsof skuas on the Kerguelen (Weimerskirch,Zotier & Jouventin, 1989) are located eitheron the peripheral islands where cats areabsent, (Mougeot, Genevois & Bretagnolle,1998), or on the main island, near penguincolonies, and in parts where cat density islow. In the absence of seabird prey, thepresence of rabbits and mice does not seemto be sufficient for skuas to breed. Forexample, except for a few pairs of largepetrels (white-headed, Pterodromalessonii, and white-chinned, Procellaria



aequinoctialis, petrels), there are nobreeding pairs on Guillou Island, wherepetrels have been eliminated by cats, despitethe presence of introduced mammals(Chapuis et al., 2001).

A fascinating example of complexinteractions through single speciesintroduction comes from the California GulfIslands. There, an endemic fox species hasdangerously declined over the past decade, toa number less than a tenth of their previouspopulation size. In parallel, its onlycompetitor, an endemic skunk, has beenincreasing. The reason for this decline wasfar from obvious. Careful studies showedthat skunks were not responsible: they arelargely dominated by foxes in the competitiverelationship, and their increase was aconsequence of the fox decline. Pathogenswere not the cause either, and prey wereplentiful (they were in fact increasing,following the release of fox predation). Theonly remaining explanation was a toppredator which presence, or its effect, wasnew to the ecosystem (since foxes wereknown to be the top predators on theseislands, and their decline was recent).Researchers on these islands pointed out thatthe decline in foxes was linked to acoincident increase in golden eagle (Aquilachrysaetos) predation ((Roemer, 1999;Roemer et al.)). A combination of field dataand mathematical modelling suggested ahyperpredation process, engendered by theintroduced population of pigs (Roemer,Donlan & Courchamp). On these islands,golden eagles are not able to survive as apopulation on the local prey only, but pigletsnow constitute year round sources ofabundant food, which allows eagles tobreed. Pigs have thus facilitated thecolonisation of the islands by eagles: whilethe pig population does not suffer too muchfrom predation (only the surplus of piglets istaken by eagles), they attracted a threat tofoxes. The endangerment of this endemic foxis directly linked to an important increase ofan endemic skunk, its natural competitor,and is caused by a predator whose survival isthreatened in other places, thus complicatingconservation measures.

Some other interesting complex effectsof mammal introductions include thedisplacement of native species from preferredhabitat by introduced predators, or habitatdestruction by introduced herbivores, both ofwhich lead to increased exposure topredators. Both cases have beendocumented. On Bermuda islands,introduced rats have forced cahows(Pterodroma cahow) to nest in suboptimalhabitat where chicks are threatened by thewhite-tailed tropicbird (Phaethon lepturus,(Jackson, 1977), in (Ebenhard, 1988)). Inthe Galapagos islands, goat browsing hasdestroyed the cover of land iguanas andthereby increased their susceptibility topredation by hawks (Dowling, 1964;Gibbons, 1984).

There are also mutualist interactionsbetween introduced species, which can act insynergy to inflict damages to localcommunities. An interesting illustration ofthis is the facilitation of access of introducedbrush-tailed possums to forests, followingthe destruction of the understorey byintroduced red deer, a process that dries theforest, making it more suitable for possums.The possums in turn cause a partialelimination of the tree canopy, whichfacilitates understorey regeneration, andthereby benefits red deer (Wallis & James,1972).

All this illustrates the near-impossibility to accurately predict theecological consequences of relationshipsamong several introduced and native species.Nevertheless, conservationists generallyagree about the need to control mammalsintroduced on oceanic islands. We will seethat such complex necessitate a deeperunderstanding of the system in order toproperly predict the result of drasticmanagement actions such as the removal ofone species from the ecosystem.



III. Control: traditional vs.biological methods

It is nowadays trivial to affirm thatpreventing species introduction is the firstand best action to be taken [I.U.C.N., 1999#298]. Unfortunately, as the number ofintroductions increases, action often has tobe taken after the fact. There are threestrategies to alleviate problems caused byintroduced species, all three involving thereduction of the numbers of the animalscausing these problems: exclusion, controland eradication (see for example (Bomford &O'Brien, 1995). Exclusion is only a localsolution, as it concerns a spatially delimitedzone of effort, from where the alien specieswill be removed. Control can have twomeanings: it can be a general term of actionagainst an alien species ranging from simplereduction up to eradication, and it can morespecifically mean lowering of the introducedpopulation numbers, that is mitigation orreduction. In the review, we use control fora general term, and will rather use reductionor mitigation for the partial populationelimination. Control in the mitigation senseis the reduction of the size of the pestpopulation, down to acceptable levels, inecological or economic terms. As such it isopposed to eradication (see below). Since itis not complete, such removal strategyinvolves constant and/or repeated actions, tokeep the population at low density after theinitial decline. Although it is generally morefeasible than eradication, the gains achievedby control are temporary and it usually mustcontinue periodically for an indefinite period.Many pest mammals have a densitydependent reproductive rate with highreproductive rates at low density, leading tofaster recovery rates for controlledpopulations. For example, after an 80%reduction, a goat population returned toprevious levels within four years, thedoubling time of a controlled goat populationbeing around 20 months (Parkes, 1984).

For these reasons, populationreduction is more realistic for non-isolatedlarge areas, such as continental areas, or onislands with a constant source of immigrationof the pest species. Eradication is thecomplete removal of all the individuals of the

population, down to the last potentiallyreproducing individual, or the reduction oftheir population density below sustainablelevels (Myers et al., 2000a). This isgenerally, although not always, the beststrategy for islands, but it is often limited byits high cost, logistically as well aseconomically. There are a number of factorsthat make some islands more difficult to freefrom alien species, or some alien speciesmore difficult to be removed from islands.An example is given in Figure 2 concerningthe island characteristics, but thecharacteristics of the alien species are alsocrucial. Some species are said to beparticularly robust and difficult to eliminatefrom large remote surfaces. A somewhatseparate but nonetheless fascinating anecdoteis the nuclear testing that occurred three timesduring the 1950s on some Australian islets,which allegedly failed to eliminate either ratsor cats. In addition, for most islands andalien species, when pushed down to a verylow density, the last individuals of thepopulations are generally very difficult toremove. This difficulty has two causes.First, the few remaining individuals aresimply more difficult to find, and second,because of the density dependence thatcharacterises most populations of invaders,these last individuals will quickly reproduce,which is a force that the control program willhave to fight against. An amazing example ofthe difficulty to find the last remainingindividuals is that of the last remaining muledeer on San Clemente Island, USA. During apig eradication program in 1991, huntingdogs turned up an old male mule deer, whichwas shot. This was the first sighting of adeer on the island since the Navy's NaturalResources Office "completed" deereradication in 1977. Of course, being asingle male, eradication had been essentiallyachieved, but this illustrate well how evenlarge animals can remain undetected for quitea long time (14 years in this case; R. BrandPhillips, pers. comm.).

However, unless the island has a highlikelihood of repeated invasions, theeradication should have to be implementedonly once, as opposed to regulative control.It may be initially more expensive, but willbe more cost effective than repeated (or



continuous) population reductions because itis permanent (Rice, 1991; Veitch & Bell,1990). For example, the cost of ratmitigation for the Mediterranean Lavezziislands was estimated around 3000 €, a yearwhile its eradication cost five times thanamount (Pascal, unpublished data). Althoughthe larger amount is more difficult to obtain,it is more cost-efficient in the long term: thesame funds will be spent for than 5 years ofpopulation reduction, after which further

control will be needed, while eradicationresult is permanent. In addition, if poisoningis one of the chosen action methods,eradication will require lesser amounts to beused than would be necessary for sustainedcontrol efforts. Similarly, fewer animals willhave to be killed during an eradicationcampaign than would be for a long-termmitigation, which may be more acceptable toanimal rights people.

Fig. 2. Eradication probability, according to island area, accessibility and distance tocontinent. This 3D graph shows that, for the same control effort, the eradication is less likely inlarger, less accessible and more isolated islands. A similar graph could be drawn focussing on thealien species characteristics (such as dispersal abilities, diet width, ecological plasticity, …)

(1) Traditional methods

We will here compare two main typesof control: traditional and biological. Thecontrol of introduced mammals is obviouslyadapted to the characteristics of thesespecies. Therefore, traditional control ismostly mechanical, by hunting, trapping andpoisoning (insects are mostly controlled bychemicals) and biological control mostly uses

pathogens (the natural enemies used forinsect control will rather be their predatorsand parasitoids). Each method comes with itsown set of advantages and disadvantages andoften these depend on the species to becontrolled, resulting in some control methodsbeing very well adapted to a given species.We provide in Table I a summary of methodtypes, with the associated main advantages,disadvantages and typical target species.



Consequently, rather than describingall existing traditional control methods forintroduced mammals, we will now provide afew typical examples of the techniquesfavoured to control the main mammalinvaders. Table I summarises the mainadvantages and disadvantages for each typecontrol method (physical, chemical andbiological).

(a) Large mammals

Because of their large size, ungulatesare among the most easy mammals toexclude from areas using fences. Thissolution is most appropriate when the area tobe controlled is too large for an eradicationprogramme, or if the herd is not to beentirely destroyed. An example of this is theferal cattle of Amsterdam Island (55 km2) inthe Indian Ocean (Micol & Jouventin, 1995).In 1871, a farmer gave up his livestockattempt on this island because of the harshconditions, abandoning five cattle. Theseindividuals adapted well enough, and bred torapidly form a population reaching 2000individuals in 1988. Owing to the grazingand trampling of this large population, aswell as several major fires from theeighteenth century (1792, 1853, 1899,1974), the native vegetation regressed infavour of introduced species. From the1980s, introduced cattle started to colonisehigh elevation zones, which are especiallysensitive to trampling and which shelter arelic population of Amsterdam albatross(Diomedea amsterdamensis), discovered inthe early 1980s and numbering only 12breeding pairs (Jouventin & Roux, 1983).However, despite their impact, eradication ofthe cattle was not initiated because theyrepresented one of the very few feral herdsof this species anywhere in the world, a largeherd that had evolved in total isolation formore than a century. The restoration strategyadopted was to divide the island into twosectors with a fence, and the cattle wereeliminated from the main and most sensitivepart of the island while the herd wasregulated on a smaller sector with no birds.Fences are also used to exclude smalleranimals, such as foxes, cats, rabbits, rats oreven mice, in conjunction with other controlmethods, to clear the fenced area from the

unwanted species. An impressive example isthe Karori Wildlife Sanctuary, nearWellington, New Zealand (Saunders &Norton, 2001). It is a mammal-proofenclosure of 252 ha from which allintroduced animals have been eliminated:black rats, mice, possums, stoats and ferrets,in order to restore the local hardwood forestand to reintroduce extirpated species.

Large mammal will be evidently moreeasily shot, and therefore hunting hashistorically been privileged to control them.This is also linked to the associated gain inmeat, and sometimes to the recreationalproperty of hunting. The main disadvantageof this method is the poor accessibility ofmany oceanic islands, which makes it costlyand logistically difficult to maintain enoughmanpower for a sufficient task force. Thisproblem maybe at one or both of two scales:accessibility to the island itself, andaccessibility to some parts of the island. Forexample, it takes more than twenty daysround trip by ship to access the Kerguelenislands, with only 5 to 6 trips a year. Inaddition, the main area of this archipelago isextremely ragged: it has a surface ofapproximately 6500 km2 and a coastal lengthof more than 1000 km (that is, between threeand four times more than a perfect circle ofthe same area). Most parts of the island arereachable only by foot, requiring severaldays of trekking. This precludes the use ofmany control methods, and it seriouslyhinders others such as hunting. Yet, whenaccessibility can be dealt with, hunting is oneof the most efficient way of controlling oreven eradicating a herd of ungulates,especially if large incentives are madeavailable for hunters (Gosling & Baker,1989). Using helicopters, it would bepossible to eliminate the reindeer populationfrom Kerguelen, but this would incur a highcost.

The most difficult part of largemammal control by hunting is the localisationof the last few individuals, since the verylow density may seriously hinder theelimination of these few survivors, therebythreatening the whole control programme.Two main methods are used to overcome thisproblem. In many cases, dogs can be trained



to assist hunters with rifles, and their help indetecting isolated individuals may make thedifference (Cowan, 1992; Wood et al.,2001). Another, very clever method,especially designed to eradicate goats, iswhat has been called the Judas goat method(Keegan et al., 1994; Parkes, 1984; Parkes,1990). It consists in fitting one individual,preferably a female in oestrus, with a radiocollar. Goats are highly gregarious animals,and this female will be very efficient atfinding isolated groups throughout theisland. At regular times, the radio signal willbe used to localise the female, and thehunters will have little difficulty to reach thegroup, either by foot or by helicopter. Oncein view, the whole group will be shot, exceptfor the “Judas”, which will be spared so thatit can find another group, thus involuntarilyleading the hunters to it. When the “Judas” isregularly found alone, the population isconsidered eliminated and that last goat isremoved.

(b) Medium and small sizedmammals

Next to the shooting, trapping hashistorically been used for medium sizedmammals, such as small carnivores and largerodents, because they are less easy to shoot,but still valuable to collect either for fur orfor meat. Trapping is mostly used forrodents, cats, mustelids, mongooses, rabbitsand hares, but only for population reduction,since their eradication is not feasible bytrapping alone (except for the carnivores).Trapping has also been very efficient forpossums (Cowan, 1992). Leg-hold traps aremore rarely used now for ethical reasons,rather live door-traps are routinely used.However, leg-hold trapping may be the onlyefficient method in some cases, especially forcats (Veitch, 2001). The control methodsbased on trapping are very similar to thosebased on poisons: both will generally need tocover a substantial area and to developattractive baits. However, traps may be oflimited use if the area to control is too large,with limited accessibility, if the population istoo large, or if the animals are trap shy. Theeradication of a small population of cats canbe considered by trapping on a small island ifthe terrain is not too irregular, but

populations on large and/or rugged islandscannot be so, since it is just impossible tocarry and disseminate enough traps on verylarge or poorly accessible areas. On the otherhand, the traps have the advantage ofselectivity, because they can be designed toexclude or reduce the access to nativeanimals.

Poisoning is the other major controlmethod for medium and small mammals, andis certainly the best current option forrodents. The main difficulty of poisoning,which often requires very large amount ofpoisoned baits to be more or less evenlydistributed over large areas, is the logisticdifficulty of such delivery (and the associatedtransport). Recent advances in automaticdelivery from containers transported byhelicopters, with a Global PositioningSystem assisted timing, have lowered thisdisadvantage, making it much lessproblematic to consider rodent eradicationfrom larger islands. Poisoning programsoften attempt to avoid non-target species killswith baits targeted for the pest species.However, discrimination can be difficult toachieve when native and alien species aretaxonomically or ecologically close, and theimpact of poisoned bait on native species cansometimes be important. For example, theuse of poison to eliminate feral cats from theislands from the California Gulf, off Mexico,had to be abandoned because of the presenceof endemic carnivores (e.g. ringtail cat,Bassariscus astutus) and avian scavengers(e.g. turkey vulture, Cathartes aura) onsome of them (R. Rodriguez-Estrella comm.pers.). Losses of non-target individuals topoisoning must thus be assessed with respectto losses caused by the uncontrolledintroduced species [I.U.C.N., 1999 #298].In fact, research consistently suggests thatthe harmful effects of introduced mammalsare greater (and longer) than those of thetoxin used to remove these mammals. Evenif many non-target individuals may die frompoisoning, populations of these individualsoften recover rapidly once the pressure fromthe introduced species is removed (Merton,1987). As an example, Pukekos (Porphyrioporphyrio) have re-populated Tiritiri Matangiand Motuihe Islands, New Zealand, after an



initial reduction of around 90% due to apoisoning campaign with brodifacoum(Dowding, Murphy & Veitch, 1999). Insome cases, non-target elimination throughpoisoning has been deemed unavoidable, andthe eradication program has been designed sothat the native species could be reintroduced,once the alien species eliminated. This wasthe case for the burrowing bettong Bettongialesueur, which was eliminated during thepoisoning of the black rats on Boodie Island,off Australia. However, this had beenforeseen by the programme leaders, who haddesigned their programme so that once the ratwas eradicated, bettongs could bereintroduced from a nearby island used as areservoir (Morris, 2001).

Another possibility is to capture at-risk non-target endemics in order to protectthem away from poison during theeradication program and then to releasethem once the invasive population iseradicated and the poison removed ordegraded. This has been done for severalprograms already in New Zealand. In 1996Pacific and Norway rats, Rattusnorvegicus were eradicated from KapitiIsland (2000ha), New Zealand (Empson &Miskelly, 1999). During the operationthree pairs of takahe, Porphyrio mantelli,a giant flightless rail, were held in captivityon the island for just over two months. Inaddition approximately 200 weka,Gallirallus australis, a large flightless rail,were removed from the island, 50 of whichbeing held in pens in a mainland reserveand the remainder released in a remoteformer part of the species mainland range.The captive birds were held for aroundthree months then returned to theisland. During the period 1996 – 2000, theeradication of five introduced mammalspecies (feral cat, rabbit, black rat, Norwayrat and house mouse), was attempted onfour inhabited islands in the Seychelles.Since no rat-free island was available towhich native animals at risk from poisoningmight be transferred, it was necessary tomaintain c. 560 individuals of threethreatened animal species in captivity forthe three month duration of the eradicationprogramme. Very low mortality was

detected during captivity, and all captiveanimals were released in their former, nowsafe habitat, once the rats eradicated(Merton et al., 2001).

Paralleling the increased interest inmammal pest control and island restoration,there has been some technical progress in thedevelopment of mammal attractants and lures(e.g., (Clapperton et al., 1994; Saunders &Harris, 2000) and further research iscurrently under way. There has also beenmuch progress to prevent aversiondevelopment in animals exposed topreviously used poison (Cook, 1999;Hickling, Henderson & Thomas, 1999).Much effort has also been devoted for thedevelopment of baits that do not attract non-target species, which has been done bydiminishing both hazard and exposure (e.g.,(McDonald et al., 1999; Morgan, 1999).

In addition, there has been importantprogresses in the methods of baits delivery,from fixed and evenly distributed baitingstations, to the more recent broad scale useof aerial baiting (usually helicopters) withGPS assisted distribution, both of whichbeing major factors in the increasing numberof eradication successes.

Several toxins are routinely used forthe control of small introduced mammals,mostly in Australia and New Zealand):sodium monofluoroacetate (or 1080),pindone, cholecalciferal, brodifacoum,cyanide, strychnine. However, 1080 andbrodifacoum are currently the most used,with brodifacoum, an anticoagulant, beingnowadays more used than 1080 (Innes &Barker, 1999). For example, brodifacoumwas used in 28 of 33 mammal eradicationprogrammes undertaken by the New ZealandDepartment of Conservation (DOC) in thelast decade (Innes & Barker, 1999).Although 1080 is usually rapidly eliminatedfrom live animals and rapidly broken downby microbial activity in baits, water and soil,it is still highly toxic to a wide range ofanimals and may pose problems for non-target species (Innes & Barker, 1999). Bycontrast, brodifacoum shows a greaterpersistence: it accumulates in vertebratetissues, it is insoluble in water and only



slowly broken down by microbial action(Dowding et al., 1999; Eason et al., 1999).It is however less toxic for non-target species(Godfrey, Reid & McAllum, 1981).Although it is often known for the control ofbrushtail possums in New-Zealand,brodifacoum has also been used insuccessful mammal eradication programmeson islands for both rats (Brown, 1997;Buckle & Fenn, 1992; Empson & Miskelly,1999; Taylor & Thomas, 1993) and rabbits(Godfrey et al., 1981; Merton, 1987).

During control programmes, poisonedanimals (alive or dead) may be eaten bycarnivores or scavengers, potentiallyresulting in secondary poisoning. This mayprovide conservation managers with anefficient multi-species tool for controllingmammalian pests on off-shore islands wheretheir re-invasion is unlikely (Gillies &Pierce, 1999). Indeed, it is now known thatmammal predators numbers can beeffectively reduced by secondary poisoningwith both brodifacoum and 1080 followingrat control (Gillies & Pierce, 1999; Heyward& Norbury, 1998; McIlroy & J., 1992;Murphy et al., 1999), to such a point thatdeliberate secondary poisoning has beensuggested as a potentially cost-effectivepredator eradication tool (Alterio, 1996). Inthe context of secondary poisoning, alltargets are not as efficient. For example,rodents appear to be a better vector forsecondary poisoning of cats (Dowding et al.,1999). Biodiversity managers therefore needto take into consideration the relativeabundance of prey in each case. Notably, it

must be taken into account that other prey,such as rabbits may provide target predatorswith a sufficient non-toxic prey “buffer” toprevent large scale secondary poisoning.

There is also a risk that pest speciessurviving poison operation aimed at otherpest species in the same area maysubsequently develop an aversion to thepoisoned baits and become exceedinglyresistant to further toxin-based control(Hickling et al., 1999). This not only caninduce a learned aversion to the chemicalused in the poisoned paste but also can causeanimals to exhibit “enhanced neophobia”, asseen in possums following rabbit control(Hickling et al., 1999). All these pointshighlight the need of developing medium andlong-term strategies for the use of poisoningfor mammal pest control.

(c) Advantages and disadvantages oftraditional methods

Despite an improved efficiency,especially concerning the recent progressesfor the eradication of rodents from largerislands (e.g., with aerial poisoning),traditional methods remain logisticallydifficult, are costly in material, manpowerand time, and have in general littlespecificity. These methods can have a lowcost-efficiency on large and/or poorlyaccessible islands. This has led severalbiodiversity managers to look for alternativestrategies, among which biological methodshave an interesting potential (see Table I).



Method typeTypicaltarget

speciesAdvantages Disadvantages

Physical methods

Trappingsmall

Carnivores,possums

- efficient for smallpopulations ofsmall andmedium-sizedspecies inaccessible andsmall areas

- can be selective- environmentally

clean

- certain traps may pose ethical problems- requires a limited infested area- requires a limited population- requires good accessibility to carry traps

on the field- may be expensive to buy sufficient traps- may require good trapping experience to

be efficient- requires attractive baits for target species

that remains non attractive (or repulsive)for non-target

- may be selective (age or sex)- rarely achieves eradication unless used

with another method

Shooting Ungulates

- very efficient forlarge animals

- very selective- environmentally

clean- more ethical

method (butusual targetshave a highpublic profile)

- requires good accessibility for hunters inthe field

- requires experienced and dedicated staff- may be costly and logistically difficult to

maintain enough manpower for asufficient amount of time

- requires public information andsensibilisation

- may require special authorisation forguns

- may imply additional techniques: dogs,Judas technique, helicopters, etc…

Chemical methods

Poisoning

Rodents,Carnivores,

smallHerbivores

- very efficient forsmall Rodents

- requires special authorisation- may affect non-target species (protectionmeasures may be expensive and not fullyefficient)- implies complete coverage of the whole

infested area- targets can develop bait aversion- implies the use of adequate baits and

often of bait stations- may cause secondary poisoning

(accumulation in animal tissues)- some toxins are only slowly degraded- can be costly, especially for large islands- requires public information and

sensibilisationBiological methods

Predatorintroduction

Rodents,small

Herbivores

- environmentallyclean

- very low cost

- great risk of specificity loss, causingfurther ecological disequilibria

- historical reason for the existence of aliencat and mongoose populations

- add a species and its parasites to theecosystem



ecosystem

Competitorintroduction Carnivores

- can be veryefficient if wellplanned

- released individuals must to be removedor die off without reproducing

- very little information available- add a species and its parasites to the

ecosystem

Pathogenintroduction

Carnivores,small

Herbivores


- very low cost- can be very

specific

- poor efficiency at low density.- ethical problems (may induce suffering)- sanitary risks- requires good knowledge of host-

parasite system- requires absence of risk for non-target

species- potential for host adaptation (evolution of

immunity)

Immuno-contraception

Rodents,Carnivores,

smallHerbivores

- most ethicalmethod

- low cost (whenselfdissemination)

- can be veryspecific


- not yet operational- concerns for loss of control- very slow- irreversible process- low public acceptance of release of

genetically engineered organisms- potential for development of host

resistance

Table I . Advantages and disadvantages of different control methods

(2) Biological methods

Biological control is the control of pestspecies by an enhancement of a differentspecies that decrease reproduction or survivalof the target pest species. Several categoriesof natural enemies are used in this context,but predators and pathogens have historicallybeen the most used for mammals.

(a) Decreasing the survival of alienmammals

The main biological control strategy isthe introduction of natural enemies of thetarget species, aiming to a decrease of itssurvival. Parasitism, predation andcompetition are the three possible processesinvolved, although the former is by far themost common: the natural enemies most usedfor mammal pest being microparasites(viruses and bacteria). Concerning pathogenintroductions, the best known example is themyxoma virus, introduced in Australia and

then on several islands to control rabbitpopulations (Brothers et al., 1982; Chapuis,Chantal & Bijlenga, 1994b; Chekchak et al.,2000; Fenner & Ross, 1994; Skira, Brothers& Copson, 1983). Despite problems linkedwith virulence and immunity evolutions inlarge populations, this virus was successfulat removing entire populations in many cases(Flux, 1993). More recently, the introductionof the rabbit haemorrhagic disease (RHD) inNew Zealand and Australia has led to massmortality in rabbit populations, but thesewere not scientifically planned eradicationsfor conservation biology purposes and areextremely contentious. Regarding RHD inAustralia, the disease has now beenestablished in Australia for over 5 years andhas reduced the abundance of rabbits acrossAustralia, most spectacularly in arid, inlandAustralia (Brian Cooke, pers. com.).Another example is the eradication of feralcats from Marion island, using the FelinePanleucopenia Virus (van Rensburg, Skinner& van Aarde, 1987). In this case, acombination of biological control (the virus)



and traditional control (hunting) was requiredto complete the eradication of cats, neithertechnique alone would have been successful.This has been termed integrated pestmanagement.

Generally, biological control bypredator introduction has been disastrous.The presence of mongoose and or domesticcats on most islands is the result of failedattempts to control rodents, rabbits orsnakes. Most of the time, these mammalianpredators ignored the target prey and insteadturned to native prey, which were ofteneasier to locate and/or kill, having evolved inabsence of these terrestrial predators(Atkinson, 2001). The consequences ofpredator introductions for biological controlof rodents are sometimes more subtle orunpredicted. For example, in the WestIndies, rat control has often been attemptedby mongoose introduction. But mongoosesare typically terrestrial, and their presenceoften caused the rats to shift habitats infavour of arboreal niches (Seaman, 1952).As a result, not only have rats become lessavailable to mongooses, which startedpreying on many ground species of birds,but also the introduction of mongooses led toan increased predation on tree nesting birdsby the now more arboreal rats.

One of the rare attempts at controllingmammal pest through competitive processesconcerns a successful program to eradicatearctic foxes from the Aleutian islands. Thisprogram used the competitive superiority ofred foxes: some sterilised individuals wereintroduced, and then removed once the arcticfoxes extirpated (Bailey, 1992; Bailey, 1993;West & Rudd, 1983). This original methodwas highly efficient and, although verycostly certainly deserves more interest.

(b) Decreasing the reproduction ofalien mammals

The common goal of control methods,either traditional or biological, is to decreasethe survival of individuals of the species. Insome species with a short generation timeand high reproduction rate (typically rodentsand most other pest species), decreasingsurvival may not be as efficient as decreasing

reproduction. During the last decade, asignificant impetus of research effort hasbeen devoted to the development oftechniques lowering the fertility of mammalspecies. Notably, as an alternative toincreasing death rate, immunocontraceptionaims to reduce birth rates (Alexander &Bialy, 1994; Tyndale-Biscoe, 1994).Immunocontraception is a process by whichthe immune system of an individual isinduced to attack its own reproductive cells,leading to sterility (Tyndale-Biscoe, 1994).This is achieved by infecting individuals witha protein derived from the follicular layers,which activates the production of antibodiesagainst its own gametes, thereby blockingfertilisation (Bradley, Hinds & Bird, 1997).Depending on the size of the animal and thelevel of control desired for its population,infection can be completed by injection, asfor large mammals (Kirkpatrick et al., 1997)or by bait, typically for small carnivores(Bradley et al., 1997). More recently, newmethods have received considerableattention, which would allow infectionthrough living vectors, typically geneticallyengineered viruses, bacteria ormacroparasites. Virus vectoredimmunocontraception (VVIC), for example,utilises a species-specific virus todisseminate this vaccine through a pestpopulation by placing the gene encoding thereproductive protein into the genome of thevirus (Tyndale-Biscoe, 1994). Thispotentially powerful new technique would beused mainly for rodents and smallherbivores, such as rabbits and possums(Cowan, 1996; Rodger, 1997; Smith,Walmsley & Polkinghorne, 1997), but couldbe very efficient for small carnivores as well(Bradley et al., 1997; Courchamp & Cornell,2000; Pech et al., 1997; Verdier et al.,1999).

Control by VVIC presents numerousadvantages over traditional control methodsbased on increases in mortality. Because itdoes not result in animal suffering, thismethod is better received by the generalpublic, ethics committees, and animal rightorganisations (Cowan, 1996; Loague,1993). Moreover, unlike chemical control forexample, this method is not harmful to theenvironment. It also shares the advantages of



biological control over traditional methods:being self-disseminating, VVIC can be usedto control large areas (even whenaccessibility is limited), for a minimal cost(Chambers, Singleton & Hood, 1997).Finally, it seems to be the most host-specificof current methods (Tyndale-Biscoe, 1994).Although genetic engineering of a pathogenis a prerequisite, preliminary studies are veryencouraging (Moodie, 1995), and suggest apossibility of a durable sterilisation of thehost without modifying its social behaviour.The main disadvantages of this emergingmethod include the irreversibility of theprocess once the vectors released (Nettles,1997), the potential for development of hostresistance, the need for the engineering of agenetically modified vector (Bradley et al.,1997), a slow response time for definitiveresults as well as for monitoring progress(McCallum, 1996) and a low publicacceptance of the release of geneticallyengineered organisms.

We cannot, however, overemphasisethat biodiversity conservation planners mustbe cautious and responsible with respect tothe use of control through release ofpathogens in general, and of geneticallymodified pathogens in particular. VVICmethods should only be used once they aredeemed truly reliable, and should be so onlyin conditions of maximum security for non-target species in the ecosystem, and for non-target populations of the target speciesoutside this ecosystem. In this context, theapplication of VVIC methods, if the localbiotic and abiotic conditions are favourable,should be first (and perhaps only) used inremote uninhabited islands, where potentialunforeseen effects could be naturally andeasily circumvented (Courchamp & Cornell,2000). In addition, given its specificcombination of advantages (e.g., selfdissemination) and disadvantages (e.g., slowprocess), this type of method is better suitedfor large islands. However, this is oftenwhere there are more than one interactingalien mammal species, and thus, ifappropriate vectors are available, VVICshould be used in parallel on all interactingmammals. For example, controlling catswithout controlling rats, or controlling

rabbits without controlling cats, wouldalmost certainly lead to program failure.

In addition, one of the largest concernsrelated to control by VVIC is the possible useof a genetically modified virus by an ill-intentioned person. As has been shown withthe introduction of the myxoma virus inFrance to reduce rabbits numbers, and morerecently with the RHD introduced in NewZealand, for the same goal, it may bedifficult to fully control the intentionalintroduction of micro-organisms by personswho are not fully aware of (and/or interestedby) the potential ecological effects of suchactions. Apart from the potential sterilisationof whole populations over very vast areas,there could be more profound and significantconsequences at the ecosystem level.

(c) Advantages and disadvantages ofbiological methods

The main advantages of biologicalcontrol are a potentially very goodspecificity, and a higher cost efficiency sinceit corresponds to a self-disseminating controlmethod, what could be characterised as“release and watch” effort. This can becrucial when large areas, low accessibility orlow density prevent from using traditionalmethods, as is often the case on islands(Courchamp & Sugihara, 1999).

However, there are also a number ofdisadvantages with this type of method,which have precluded the generalisation ofits use. One drawback is the poor specificityof some of the species that have been used tocontrol pest species. Although it could beargued that most failed attempts atintroducing natural enemies to control exoticmammals could have been foreseen with aminimal scientific pre-release survey, thehistory of island restoration is virtuallydominated by catastrophic failures.

It is important to stress that biologicalcontrol meets with growing success asprograms are scientifically established andimplemented. The main criticisms ofbiological control concern programs thathave been conducted without adequatepreliminary study of the ecosystem (see



below). One must keep in mind thatbiological control is, at least in the case ofremote oceanic islands, nothing less than theintroduction of an alien species in order tosolve a problem of a previous alienintroduction. History has shown us thatoften more harm is generated from suchinterventions. An illustration of this lack ofevaluation concerns the repeated attempts atcontrolling the rodents in sugarcane fields inJamaica, during which cane growersintroduced ants (Formica omnivora), whichdid not reduce rat numbers but soon becamea problem themselves. To remove rats andants together, it was then decided tointroduce toads (Bufo marinus). But toadsstill did not control rats, and became a pestthemselves. Finally, small Indian mongooseswere introduced to control rats and toads.Mongooses failed to control either, andbegan preying on native birds, posing newthreats to wildlife (Silverstein & Silverstein,1974).

The other main type of disadvantage ofbiological control is the lack of control oncethe natural enemy population is released. Theadvantage of a self-disseminating systemmay become a problem if the control methodis not working properly, which it rarelydoes. Having no control over the releasedpopulation also implies difficulties for themonitoring of the efficiency of the operation.Introduced species and their newenvironment interact in ways which are hardto predict even if we know all the facts,which is seldom the case. Since we knowlittle in most cases, accurate predictions areextremely difficult and entail unknown risksto every single introduction, be it forrestoration purposes. This has led someauthors to describe scientifically controlledintroductions as “a serious fallacy”(Ebenhard, 1988).

The third important type of problem inusing biological methods to reduce mammalpest species is of an ethical nature. Becausepathogens are more often used for thebiocontrol of mammals, there are moreconcerns about safety and ethics than inclassical biocontrol programs in agriculturalecosystems. Most pathogens, for example,

induce diseases that cause suffering beforekilling the individual host, which can bedifficult to justify. Even though most animalright association representatives recognisethat the death of a few individuals by controlis better than that of hundreds of individualsfrom other species by extinction if nothing isdone, or of these individuals by starvation insome cases, they rarely stand for methodsthat induce suffering. The use of pathogenalso generates concerns for the security ofother populations and other species,including human beings.

The risk of the organism eitherunintentionally (or illegally) making its wayback to the ecosystem from which the pest isnative, which could be disastrous, is alsoworth mentioning. It is particularly relevantto the case of the possums invading NewZealand, and native to Australia. Biologicalcontrol of alien possums has been considered(Barlow, 1994; Cowan, 1996), but the risksuch a program would represent for nativeAustralian population is a major drawback.

It is also often stated that biologicalcontrol is more efficient at controlling thaneradicating populations. One argument is thatbecause of a long co-evolution, and theassociated arms race of both the targetspecies and its natural enemy (generally thehost and its pathogen), the natural enemy haslost the ability to completely eliminate thehost population (as species that do so woulddisappear too). One point that is not takeninto account in this reasoning is that theconditions may be very different from thosewhere the two species have co-evolved. Inparticular, the pathogen may rarely, if ever,have had to cope with a single population incomplete isolation. It is conceivable thatmany pathogens have maintainedcharacteristics that cannot eliminate a hostpopulation in contact with others, but whichwould eliminate a hypothetical populationwith no immigrants. Similarly, because ofthe founder effect, the genotype diversity ofalien host populations may be more reducedthan that of mainland populations, and maythus not encompass enough variability toallow immunity to a fraction of the hostpopulation. A mammal population that wouldnormally survive in presence of a virus may



not do so if no immune individuals arepresent. Another argument is that somepathogens can be used in conditions that arenot encountered naturally, for example if thevirus is genetically modified.

More generally, it can be questionedwhether the advantages of biological controloutweigh its risks (Simberloff & Stiling,1996a; Simberloff & Stiling, 1996b; Thomas& Willis, 1998). Without entering a longdebate that lies outside the purpose of thisreview, it is worth emphasising two points.Firstly, even though it is true that there havebeen more failure or semi-success than fullsuccess with biological control, one shouldbear in mind that most of the programsinvolving biological control have beencarried out in a period where carefulfeasibility and impact assessment were notthe rule. Thus past failure mostly shows thatthese programs were not optimised, but itdoes not rule out the possibility of using thesame tools to get better results. It isconceivable that the small number of studiesconcerning biological control of vertebratesis more the result of a lack of confidence andof effort than of a lack of potential (Wood,1985). Secondly, even in the days whenscientists and conservation managers did notbenefit as we do from the hard learnedlessons of initial failures or half success,there were nevertheless interesting cases ofimpressive achievements through biologicalcontrol. The elimination of the cats fromMarion Island, which would not havesucceeded without the elimination of some80% of the population by the introduction ofFeline Panleucopenia Virus, is one suchexample. Biological control thus remains apotentially powerful method that meritsfuture investigations.

But these counterexamples should notbe taken as evidence that neither biologicalcontrol, nor traditional methods are auniversal panacea. Each has its specificdisadvantages and each particular case ofinvasion should be managed with an adaptedand specific plan of action. The appropriatestrategy will be less often the use of a singlemethod, but rather the simultaneous use ofbiological, chemical and mechanical controlmethods. It is commonly accepted that a

combination of strategies, that is, integratedpest management, is the best response tomammal introductions.

(3) Strategies

(a) Eradication of establishedpopulations

Many authors have insisted oninvesting time in optimising the strategyused to reduce or eradicate an alien speciesfrom an island. In particular, to ensure thebest chances of success, an investigation ofthe historical, ecological, social andeconomic situation of the island must becompleted. Moreover, the aim of theprogram must be clearly defined in order toidentify the available options and theirrespective likelihood of success.

Applying methods that weresuccessful elsewhere, even if the same alienspecies is concerned, or some other aspectsseem identical, is not necessarily a guaranteeof success. Each case must be thoroughlystudied, with careful planning of theprotocol and how to institute it.

Many authors have given ratherprecise guidelines based on their experiencewith certain species, and much relevantinformation can be found either in publishedstudies or from the national agencies wholed or financed the programs (e.g., theDOC). There are however a number of basicpoints that can be cited as important for thesuccess of a programme. To be successful acontrol program requires (but is notguaranteed by):

- Good planning (on all aspects,from logistics to ecology)

- A good pre-control study of thesituation (see chapter below)

- A careful choice of the selectedmethod or suite of methods(including good timing)



- Sufficient and lasting financial andpolitical support to completeeradication (even if it takes time)

- Public support (e.g., (Veitch &Clout, 2001)

It is outside the scope of this review toexpand on each of these points. We provideinstead a figure summarizing the mostimportant recommendations for a controlprogram (Figure 3). But each aspectcertainly deserves more detailed exposition.As an example, a few words can be saidconcerning the timing of the program. Thispoint is crucial for the success of theprogram on several grounds, depending onthe type of alien population: established ornew incursion. For the former, the timing ofthe operation is of major importance becausethe biotic and abiotic conditions play a rolein the population dynamics of alien invasiveand indigenous species (Innes & Barker,1999). These conditions (climate, fluctuatingdynamics of other species, etc…) canpromote or hinder the growth of the targetpopulation, and it is better to have theseforces acting with the protocol rather thanagainst. For example, it may make adifference to launch an eradication programof a species when the food resource of thatspecies (either plant or prey) are lacking,e.g., in winter or during migratory birdabsence (Chapuis et al., 2001). It may bealso judicious to act when the target speciesis more vulnerable because it has less cover(e.g., vegetation cover is lacking in winteron many islands) or because it has arestricted home range (e.g., arctic foxesavoid the inner parts of the Aleutian islandsin winter, restricting their movement to acoastal strip, and they are therefore morereadily removed from those islands duringthis period, Steve Ebbert, pers. comm.). Ifseveral introduced mammals are present,careful though must be dedicated to thestrategy to be implemented. There is nocurrently admitted prioritisation scheme, butany such system would have to consider,among others, the known effect(s) of thealien invasives, the probability and rate ofspread of the alien invasives and probabilityand cost of achievable eradication, and the

interactions of the alien invasive speciesamong themselves as well as with nativecommunities. It is however our deepconviction that often all should be done toremove each introduced species, and thatprioritisation schemes should equate toeradication timing not to a choice of whichspecies to remove and which to leave. Wedescribe in the fourth part of this reviewwhy and how the timing of suchprogrammes is crucial for the restoration ofthe ecosystem.

(b) Eradication of new incursions

Eradication of new incursions alsoneeds good planning, good selection ofmethods, financial support, and publicsupport, but these need to be generateddifferently. Because there is no time to dothem in a case-by-case way, the bestapproach is to have a strategy/policy in placewhich allows for rapid action. In this regard,it is crucial that a law should exist thatallows action to be taken and the authorityneeds to be established so that appropriateagents or Institution know they have to (andare allowed to) make a decision about theaction to take. Similarly, the public supportmust exist beforehand for the fact that rapidaction may be needed at some stage.Accordingly, public consultation would beon a general strategy for emergencies, ratherthan on a particular case. As for establishedpopulation, timing is very important,althought in a different way. Indeed, oneneeds to react rapidly when an invasion isdocumented, because it almost invariablystarts with few individuals in a localisedplace and this is easier to deal with than alarge and expanding (or even established)population. So launching the operation asclose as possible to the beginning of theinvasion, and if possible before the speciesis established, is by far the best option forsuccess. The 1999 exemplary case of themussel invaded marina near Darwin (Kaiser,1999), immediately dealt with by Australia’sNorthern Territory government and theCSIRO has been so successful, it is to hopeit will help quick decision and quick actionfor future cases.



IV Consequences of alienmammal control

(1) An increased number ofsuccessful control programs

Control or eradication programs havebeen increasing in numbers in the lastdecades. This may be because of a shift inecologists' interests, with an increasedinterest in conservation biology andrestoration ecology (Young, 2000) ingeneral, and by biological invasions andalien species in particular (see Figure 4),This may also be because new protocolshave allowed successful eradication ofmammals from islands in conditions thatwere previously judged impossible toovercome (Dingwall, Atkinson & Hay,1978). For example, it was admitted as lateas 1976 that islands of more than one hacould not be freed of rats (Atkinson, 2001).However, improved protocols led to theireradication from 3253 ha Langara Island,(Bertram & Nagorsen, 1995; Drever &Harestad, 1998; Taylor et al., 2000). Thiswas made possible by both the availability of'one shot' anticoagulants in baits that areattractive to rodents and the development of asuccessful technique of dispensing poisonbaits from fixed stations spaced such that atleast one station lays within the home rangeof each rat (Clout, 1999; Taylor & Thomas,1989; Taylor & Thomas, 1993). This hasbeen recently replaced by aerial poisoningusing Global Positioning System toaccurately and evenly distribute baits, amethod less time-consuming than fixed baitstations. This innovative method has createda new impetus both to research on controlprotocols and to eradication campaigns onnew (large) threatened insular ecosystems.

Fig. 4. Evolution of the citationfrequency of the “biological invasions” and“alien species” keywords in the titles andabstracts of scientific articles referred in theISI database during the last two decades.

It not easy at the time to quantify themany successful eradication programmes,and the many more control programmes andfailure situations: these data are at bestdifficult to trace regarding most part of theworld. Such programmes are notsystematically reported in the classicalscientific literature. It is especially soconcerning failures, which are more rarelyreported. One good counterexample of this isNew Zealand, for which such data iscarefully recorded. The total number ofcompleted mammal eradication was 153 atthe end of 2001, with a further 24 ongoingoperations, to be compared with threestopped and nine failed programmes (C.R.Veitch, pers. comm.). In total, failures ordiscontinuation of programmes account for amere 7.3% of programmes not currently inprogress. In total, these eradicationprogrammes concern on 144 islands, threequarters of which are now successfullycompleted and 2.8% being stopped and 7%being failures (C.R. Veitch, pers. comm.).

Table II shows the largest islands fromwhich 14 of the most notorious mammalisland invaders have been successfullyeradicated. Because relative areas may not be



easy to grasp, we reported in Figure 5 therelative areas of the island mentioned inTable II so that they can be visuallycompared. By definition, this table isdoomed to be obsolete soon aftercompletion, although this will indicate thatother, larger, successful programs will havebeen completed. This table may also lacksome already completed programs that arelarger that those mentioned: not all programsare presented in classical scientific literature,and, mostly, some programs need time inorder for the managers to assuredly claim asuccess. For example, rabbits have probablybeen eradicated from the 8 km2 St PaulIsland, France, but the end of the program isvery recent, and several additional years ofchecking for recovering populations will berequired before it is classified as a definitivesuccess (T. Micol, pers. com.). Similarly,the Norway rat may have been eradicatedfrom the 113 km2 Campbell Island, NZ,which would make it the world’s largestsuccessful rat eradication project. As is thecase for St Paul Island, time is needed toassess the real state of the population and thepopulation cannot be officially declared

eradicated until the island is still rat-free aftertwo years. It seems however that this majoroperation, including an aerial poisoningcampaign by the DOC, involving the airdropof 120 tonnes of anticoagulant baits, isalready a great success. A coming projectalso aims at eradicating goats, pigs and catsfrom the 4600 km2 Isabela Island, Galapagos(Kaiser, 2001), which would make it thelargest eradication for all three species, aswell as an enormous task. There is also theproblem of eradicating of mammals living onsmall portions of very large islands. Forexample, the feral goats eradicated from the460 km2 Auckland Island, NZ actually livedonly in about 40 km2 (Chimera, Coleman &Parkes, 1995). This poses problem for thedefinition of islands. In this review, we donot consider Great Britain, Australia or NewZealand as islands, although they aregeographically defined as such. Therefore,we did not include cases such as the coypu,Myocastor coypus, eradicated fromEngland, although it is one of the mostimpressive achievements to date in thiscontext (Gosling & Baker, 1989).

Mammalspecies

Island andcountry

Size(km2) Method Supervising

InstitutionEstimated #eradicated

Date ofcompletion

Operationduration

Arctic fox Attu, Alaska,USA 905.8 Shooting,

TrappingUS Fish &

Wildl. Service 373 1999 ?

Black rat St Paul, F 8 Poison TAAF 8-12000 1999 3 years

Brushtailpossum

Rangitoto-Motutapu, NZ 38.5

Poison,Trapping,

DogsNZ Dpt Cons. 21000 1997 8 years

Brush-tailedrock

wallaby

Rangitoto-Motutapu, NZ 38.5

Poison,Trapping,

DogsNZ Dpt Cons. 12500 1997 8 years

Cat Marion, SA 190BiocontrolTrapping,Shooting

S. Afr. DptEnvir. Affairs 2790 + 1124 1990 4 years

CattleAmsterdam,

Indian Ocean,F

55* Shooting TAAF 1059 1989 2 years



Goat San Clemente,USA 148 Shooting US Navy 30000 1972 19 years

Mouse Enderby, NZ 7.1 Poison NZ Dpt Cons. ? 1995 3 years

Norway rat Langara, Can 32.5 Poison Can. Wildl.Service 3000 1995 <1 year

Pig San Clemente,USA 148 Shooting,

Trapping US Navy 2200 1989 16 years

Pacific rat Kapiti, NZ 19.7 Poison NZ Dpt Cons. ? 1996 <1 year

Rabbit Enderby, NZ 7.1 Poison NZ Dpt Cons. ? 1995 3 years

Red fox Dolphin, Aust 32.8 PoisonAust. Dpt

Cons. & LandManagt.

30 1980 10 years

Sheep Campbell, NZ112.2 Shooting NZ Dpt Cons. ? 1991 21 years

Table II . Largest islands where eradication was achieved for 14 introduced mammals (* :eradication on part of the island, managed population on the rest)

Some countries have been leaders bothin research and implementation, most notablyNew Zealand and Australia. This is probablybecause they are countries with manyislands, with a fauna that evolved in theabsence of many mammals (no indigenousterrestrial mammal at all for NZ, (Atkinson,2001)) and thus suffered more from speciesinvasions than more "central" mainlandcountries. Also undeniably important is thepolitical will, the social maturity onenvironment-related matters and theeconomic ability to act to preserve the nativefauna and flora against a major threat, threepoints unfortunately not present together inmost other countries. There is also a need ofpublic support and of conviction by seniormanagers and politicians that the proposedaction is achievable. The lead of NewZealand and Australia in the restoration andmanagement of invaded ecosystems iseloquently reflected in hard data showing thenumber of islands subject to mammaleradication program (see Table III). Forexample, in terms of mammals only, NewZealand alone has conducted more than 216

successful eradication programs in insularecosystems. The table also shows that evenvery large surfaces can be freed from alienmammals (see also Table II). However, asemphasised by the leaders of these programsthemselves, success requires the availabilityof solid financial support, careful protocoland planning, and experienced staffcommitment.



Fig. 5. Landscape of Sarigan Island,before (1996) and after goat control (1999).In the presence of the goats, the destructionof the vegetation is very important (the scaleis given by the goat in the centre, shown bythe black arrow). Two years after the goateradication, the introduced vine has spread,eventually covering and asphyxiating the

native vegetation that had already begun torecover. Photo courtesy of Curt Kessler.

According to Myers and colleagues,six factors contribute to successfuleradication (Myers et al., 2000a). First,financial support must be sufficient both inquantity and in duration. Second, all thenecessary authority must be invested in thegroup in charge of the program in order toensure the protocol can be enacted. Third,the life history, reproductive biology,behaviour and dispersal ability of theinvading species must make it susceptible tocontrol procedures (implying that theseparameters must be known before theimplementation of the program). Fourth, theprogram should take into account thenecessity of stopping the influx ofindividuals into the ecosystem. Eliminationof mammals from islands might in thiscontext be more successful because theprobability of reintroduction can be easilyreduced. Fifth, detection of the target speciesmust be possible even at low densities, sothat eradication can be best achieved and newintroductions can be recognised before thespecies become widespread. Last,complementary restoration actions mighthave to be taken to accompany theelimination of the pest species if it had animportant role in the ecosystem and if itsremoval can trigger further problems(Towns, 1997).

# ofislands

# oferadications species concerned methods

averagesize (ha)

[min,max]

NZ Is. 107 153

mouse, cat, rabbit, brushtailpossum, wallaby, ship rat,

Norway rat, Pacific rat,stoat, ferret, pig, goat, cattle,

sheep, red deer

Poison, Shooting,Dogs, Traps,

307 [1-11216]

W. Aus. Is. 45 57 mouse, Norway rat, fox, cat,rabbit, goat

Poison, Shooting,Traps,

528 [1-4267]



Aleutian Is. 34 36 fox, cattleShooting, Poison,Traps, Biological

control

2505 [25-9174]

Hawaiian Is. 9 13 cattle, sheep, goat, rabbit,Polynesian rat, ship rat


505.5 [4-1885]

French Is.* 39 46 rabbit, goat, cattle, ship rat,Norway rat, mongoose


273 [1-5500]

Mexican Is. 24 32 cat, rabbit, goat, sheep,mouse, ship rat, Norway rat

Poison, Shooting,Dogs, Traps,

330 [2-2000]

Table III . Effort of mammal eradication for a few island groups. Data were obtained fromIan McFadden and John Parkes (NZ Is.), Keith Morris (Western Australian Is.), Steve Ebbertand Vernon Byrd (Aleutian Is.), Fred Kraus (Hawaiian Is), Michel Pascal (French Is), BernieTershy and Josh Donlan (Mexican Is.). *French islands regroup islands in different parts of theworld: Britany, Mediterrannean, Carribean, subantarctic and Indian Ocean.

(2) On the definition ofsuccessful programs

The increased number of successfulcontrol programs must not mask the fact that,in too many cases, the so-called success wasbased on achieving elimination of the targetspecies, which might not always be rightlycalled a success. We must consider themeaning of "successful" in the context ofisland restoration. Initially, restorationprograms were solely aimed at removing analien species, without having either the willor the financial possibility to manage theresultant ecological effects of removal of thealien invasive species from the targetecosystem.

The removal of the alien invasivespecies might be sufficient to restore theecosystem if the invader did not havesufficient time to inflict damages that arenaturally irreversible in the normalfunctioning of the ecosystem. Often, theintroduced mammal species will only havehad a quantitative impact on species itinteracts with, and after its elimination,natural ecological processes will drive theinteracting indigenous species back to theirformer equilibrium. However, it may not

always be the case, as unexpected changesmay arise from the sudden removal of thealien species. Introduced species may havehad an irreversible impact, such that theinitial equilibrium state cannot be reachedwithout further restoration actions. Thesimplest case is when an entire population orspecies has been eliminated from theconsidered islands and potential sourcepopulations are too far away to allow a rapidnatural recolonisation. Another case is whenthe habitat has been degraded to a pointwhere regeneration cannot be made rapidlyenough to avoid the collapse of wholecommunities. In addition, the introducedspecies may have changed the local bioticand abiotic conditions in such a way that itssudden removal not only will be insufficientto naturally reach previous states, butsometimes may even lead to further threatsfor the native species. Following theeradication of introduced species dramaticchanges to islands have been reported buthave more rarely been quantified. Thenumber of cases where an introduced speciesremoval has led to even greater risks for theprotected species is unfortunately high, andcases are still occurring with currentrestoration programs. These situations, andactions which help to reduce theiroccurrence, are discussed in the next section.



(3) Positive and negativeeffects of mammal eradication

There are a number of basicrecommendations that can be madeconcerning the restoration of invaded island(see Figure 3, but also (Atkinson, 2001;Parkes, 1990; Saunders & Norton, 2001;Towns, 1997; Towns & Ballantine, 1993;Veitch & Bell, 1990) for example). Becauseremoving a species from an ecosystem, evenan alien species, can have diverseconsequences, both desired and undesired, itis crucial to be able to quantify and to predictthese effects. Indeed, quantification of thedesired effects can lead to improvement ofcontrol methods as well as a betterjustification of control program forbiodiversity conservation. Adequateknowledge can also help predict and thusprevent the undesired, and generallyunexpected, effects. Towards, this end, aswe will see below, one crucial point is thusto fully and precisely assess the current stateof the ecosystem, a type of zero statequalification and quantification. From someof the pre-removal studies that have beenconducted for long enough for subsequentstudies to quantify change, it has beenconcluded that 10 years of study prior to aneradication may be required for post-eradication observations to be adequatelycomparer (C.R. Veitch, pers. comm.). Inmany cases, it is however difficult to waitthat long before acting on the invadingpopulation.

The other major aspect, alsounfortunately overlooked in many controlprograms, is post-eradication monitoring.This long-term surveillance is crucial formany aspects. First, it allows an assessmentof the success of the proximate goal of theprogram: durable elimination of the targetspecies, with no further invasion. It also canquantify the ultimate goal of this elimination:the reversal of a decline of affected nativespecies, their increase, or their recolonisationof the ecosystem. This part is not to beneglected: it is surprising to see thatassessment of whether expected positiveeffects (restoration of initial ecologicalconditions) have been reached is still notsystematically included in restoration

programs. Historically, the lack ofassessment as part of control programmesmay have led to a slower realisation that alieninvasive elimination does not systematicallylead to ecosystem restoration. Third, post-eradication monitoring allows the earlydetection of the increasingly described"unexpected changes" that can be disastrousin closed and non-redundant ecosystems.

The before-after-control-impact studydesign, also called BACI, is often used toassess the impact of some event on variablesthat measure the state of an ecosystem. Thedesign involves repeated measures over time,made at one or more control sites and one ormore impacted sites, both before and afterthe time of the event that may cause animpact (Manly, 2000).

The need for long term monitoring iswell illustrated by the famous case of RoundIsland, Mauritius, a unique ecosystemendangered by introduced species. This 151ha island is an important refuge for severalendemic plants and reptiles that havedisappeared from the mainland of Mauritius(Bullock, 1977). The effects of goats andrabbits introduced in the early 19th centurywere so large that the main habitats weregreatly modified and several native specieswere threatened with extinction (Bullock,1986; North & Bullock, 1986). By 1986,rabbits and goats were eradicated from theisland (Merton, 1987), and a program wasimplemented to monitor the changes in theextent and composition of key elements ofthe biota, such as vegetation, invertebratesand reptiles (North, Bullock & Dulloo,1994). Long term monitoring now showsthat a satisfactory regeneration andrecolonisation of plant species formerlythreatened by alien grazers is somewhatimpeded by a new threat on native plants,linked to the colonisation and rapid spread ofseveral alien invasive plants (Bullock et al.,2001; North et al., 1994).

The unexpected and undesiredsecondary effects are in general more likelyto occur when ecosystems contain more thanone invading species (that is, the greatmajority of islands), when invading speciesare in the late stages of invasion and when



they have eliminated native species andreplaced them functionally (Zavaleta, Hobbs& Mooney, 2001). In general, secondaryeffects following the sudden removal of alienspecies are very damaging to the ecosystem,and careful monitoring of the communitiesprior to any control action has the potential toprevent such catastrophic chain reactions. Aswe have seen in detail, exotics interact withnative species as well as among themselves,creating a complex pattern of direct andindirect effects that can be extremely difficultto comprehend, let alone to predict. Forexample, the removal of one exotic speciescan favour the expansion of other exotics thatwere held in check by the removed species.This has been called the Sisyphus effect(Mack & Lonsdale, 2002). The removal ofherbivorous aliens such as rabbits and goatscan lead to a release of exotic plants that, inabsence of browsing, are more competitivethan native plants, leading to an explosion ofsuch weeds. There are very few examples ofislands on which the presence of feral goatsor other browser is more positive thannegative (Daly & Goriup, 1987). But "theimmediate removal of introduced browsingmammals … may lead to unexpected andeven undesired results" (Taylor, 1968).

This does not mean that such browsingmammals should be allowed to remain, butthat careful planning must be designed priorto removal, in order to avoid such undesired“surprises”. For example, a recent projectwas undertaken to remove goats and pigsfrom the Sarigan Island, one of the MarianaIslands, in order to stop and reverse the lossof forest and accompanying erosion, andthereby to protect the endangered nativefauna (Kessler, 2001). In only two months,nearly 1000 animals were shot from this 500ha tropical islands, allowing rapid recoveryof both plants and monitored vertebrates.The project has been successful in reversingthe trend of forest loss, and improving thehabitat for local birds and bats. However, theremoval of alien mammals has allowed theintroduced vine Operculina ventricosa tothrive and spread, to such a point that part ofthe island is now covered by an almostuninterrupted carpet of vine, withunpredicted consequences for the future ofthe whole ecosystem (see Figure 6). Theintroduced mammals had previously held thevine at a low density such that minimal pre-operation monitoring had not identified thisthreat.



Fig. 6. Relative size of the largest island from which 14 introduced mammal species havebeen successfully removed (see Table II for details).

A similar example, concerningpredatory relationships, is what has beencalled the mesopredator release effect(Courchamp, Langlais & Sugihara, 1999a;Soulé et al., 1988). This process predictsthat once top predators are suppressed, aburst in growth of intermediate predators, ormesopredators, may follow that leads to preyextinction. A classic illustration of this is theremoval of cats from islands were rats arealso present: the elimination of feral catpopulations from such ecosystems could leadto a severe negative impact on the endemicspecies, through a rapid growth of rodentpopulations following the removal of theirpredators. Attempted reduction of the catpopulation of Amsterdam Island is alleged tohave caused a compensating increase in thenumber of rats and mice, and so has beenabandoned (Holdgate & Wace, 1961). It isaxiomatic that removal of cats (as well assimilar predators) from islands must beaccompanied by a control of their introduced

prey (herbivores as well as omnivores orcarnivores). However, outcomes of changesof these already perturbed trophic webs arenot intuitive and intervention as dramatic asspecies eradication should, when possible,be preceded by careful empirical andtheoretical studies of the whole ecosystem.Sometimes, the presence of a few individualsof a species that may appear of minorimportance can mask powerful interspecificinteractions. Thus, avoiding unexpectedchanges may not be as simple as removingthe most obvious introduced species: in thiscase, removing both cats and rats might notbe sufficient. On Bird Island, in theSeychelles, eradication of the introduced ratpopulation led to an explosion of the exoticcrazy ant Anoplolepis longipes, nowthreatening the bird colonies whoseprotection was sought through rat removal(Feare, 1998).

These catastrophic chain reactionsemphasise how crucial the pre-control



studies are for successful ecosystemrestoration. As mentioned in Figure 3, it isnecessary to assess the current state ofinvaded ecosystems prior to interventions asdrastic as species removal, be it anintroduced species (Thomas & Willis, 1998).This step is essential to allow a rigorousestimation of the effects of the invadingspecies and of the expected effects followingits removal. It is also fundamental toimplement the best control strategiesqualitatively as well as quantitatively (e.g.,see (Choquenot & Parkes, 2001)), accordingto local conditions. Finally, it isindispensable to predict, and thereby avoid,the potential side effects of alien invasiveremoval. Researchers and managers areincreasingly aware of community levelinteractions because pest control can haveunforeseen repercussions such asmesopredator (or competitor) release andprey switching (Dowding & Murphy, 2001;Murphy & Bradfield, 1992).

Invertebrates are often overlooked ininsular restoration programs, perhapsbecause they are not as conspicuous asmammals, and may not be considered asharmful. The already mentioned case of theirruption of crazy ants is not a minor sideeffect: on Christmas Island in the IndianOcean crazy ants killed three million red landcrabs in 18 months. Taking invertebrates intoaccount when removing introduced mammalsis important, as both are often directly orindirectly related. The operation of rabbitspoisoning affected the density of introducedmice who were sensitive to the sameanticoagulant, which affected the mice prey,the introduced carab Oopterus soledadinus,a major predator of invertebrates in theKerguelen (Chevrier, Vernon & Frenot,1997). Controlling rabbits may thus lead tothe development of alien plants but also tomajor changes in invertebrate communities ofthe islands, that is, on species anchoring thebase of trophic pyramids. This implies that insome cases, anything but the most finelytuned intervention may lead to more damagethan benefit to biodiversity, begging thequestion as to whether intervention should bedone at all. Nevertheless, we can still gainfrom unwitting mistakes made in the past,since all results contribute to an

understanding of island ecology and can beused in future conservation actions on otherislands.

It is noteworthy at this point tomention that there are cases where theurgency of intervention, in terms of speciespreservation, does not allow years of carefulpreparation and study of the wholeecosystem in order to establish a controlprotocol. In these cases, action must be takenimmediate lest an important component of thenative communities be lost, or the alieninvasive may reach a point where thefeasibility of removal may be compromised.In general, these are cases where thesituation is critical because of one of tworeasons. One is that the invasive species hasbeen present for a long time in the invadedecosystem and remedial action has taken toolong to start, so that further delay may meanthe difference between saving or losingpopulations or species. Obviously, weshould not let these situations happen, butwhen they do it is clear that losing moremonths or years to pre-study may onlyamount to witnessing the extinction of nativeflora and fauna. In these cases, rapidintervention may be the option, but, assuccess will undoubtedly be related to themeticulousness of planning, all precautionsshould be taken, based on availableknowledge and experience, such that theprotocol is not likely to lead to unexpectedside effects. The second situation is when anew invasion has just been reported, and isin its initial stage, one that is often easier tocounter than when the species is betterestablished. In this case, fast action is notonly an option, but should be the rule: earlydetection and rapid response afford thegreatest opportunity to control pestinvasions, allowing elimination on arelatively small scale and with a greaterprobability of success owing to the invasivesless widespread distribution. In most ofthese cases, the invading species has not hadtime to modify the ecosystem enough for itssudden removal to generate furthercomplications.



IV. Conclusion

(1) There are many other aspects of theimpacts of mammal invaders and theircontrol that are also important to discuss, andwe hope this review will trigger follow-upson complementary elements of this theme.As a conclusion, two main points need to beespecially stressed. First, despite the ratherpessimistic image of the current situationdepicted here, there are now more and moresuccessful control programs, both becausethere are more programs, indicating thatresearchers and conservationists are gettingmore aware, interested and involved in theproblems of biological invasions, andbecause programs are better planned, withlessons being learned from past experiences.It would thus not be unjustified to beoptimistic and foresee a sustained increase ofsuccessful restorations of insularecosystems. The best strategy will alwaysdepend on the specific conditions and willoften involve integrated pest management,combining the advantages of mechanical,chemical and biological methods, whileavoiding their respective disadvantagesthrough careful planning. As scientific andmanaging tools improve and become moreavailable, we hope that other countries willfollow the example of New Zealand andAustralia, with respect to their cogentpolitical, financial and scientific resolutenessin controlling biological invasions and theireffects.

(2) Second, if one lesson only were tobe learned from past failures or semi-successes, it is that a restoration programcannot be limited merely to eradication. Athorough pre-eradication assessment and along-term post-eradication monitoring (notlimited to those communities directly linkedto the eradicated species, as there can beunexpected indirect consequences) both arenecessary. Only in this way will we ensurethe best chances of success of full anddurable ecosystem restoration. Moreover,only in this way will we fully benefit frompast experiences and enrich the knowledge ofconservation biology as well as more generalareas of modern ecology.

(3) Indeed, we wish to conclude on thepotential of insular ecosystem restorationprograms, as sources of valuable knowledgefor fundamental research. Experimentation inthe epidemiology of biological invasions hasbeen advocated as a way to obtain definitivesynthesis, generalisation and prediction(Mack et al., 2000). Future and ongoingeradication programs also provide excellentresearch opportunities for ecologists to studythe roles of species in communities, theimpact of the non-indigenous species, andthe behaviour and population dynamics ofexotics for which eradication is beingconsidered (Myers et al., 2000a). In general,we advocate that eradication projects beviewed as ecological experiments in whichthe modification of trophic webs (by additionand subtraction of species) can exposecommunity processes. This contributes tonot only our understanding of biologicalinvasion, but also more generally of manyareas of population biology and generalecology. As such, each alien control programshould be designed and implemented as ascientific experiment, with hypotheses,rigorous protocols, and control of expected(and unexpected) results (Pascal & Chapuis,2000). This will allow researchers inconservation biology to establish a wider andstronger theoretical framework, which ismuch needed in this discipline (Caughley,1994), and will also allow conservationmanagers to benefit from ensuinggeneralisations.

V. Acknowledgements

The authors would like to acknowledge manycolleagues who helped provide data, facts orcomments on the manuscript, and without which thisreview would have been much poorer. We areespecially grateful to Chris Buddenhagen, VernonByrd, Jenny Daltry, Josh Doolan, Steve Ebbert, JulieGeritzlehner, Fred Kraus, Jan Larson, Ian McFadden,Don Merton, Keith Morris, John Parkes, Gad Perry,Brand Phillips, Dan Simberloff, Philip Thomas,David Towns, and Alvin Yoshinaga. Thank you toMatthew Godfrey who minimized the heavy Frenchaccent in the text. An earlier version of the reviewwas improved by comments from Maj De Poorter,Mike Clout and Dick Veitch. This work is part ofthe two programs IFB n° 209/01 and IFRTP n° 276.



Box 1: A notorious herbivore: the rabbit

Because their impact is focussed at thebase of trophic webs, introduced herbivorescan have an important impact on invadedecosystems. One of the most documentedintroduced mammal species, often referred tofor its dramatic impact on endemic plantspecies, is the rabbit, which has beenintroduced (most of the time voluntarily) tomore than 800 islands so far (Flux &Fullagar, 1992). Rabbits have a highecological adaptability and introductionseasily succeed (Flux, 1993). They can adaptto harsh conditions, eat a wide range of plant

species (Chapuis, 1981; Chapuis, 1990;Homolka, 1988; Thompson, 1953), andhave an exceptional growth rate for theirbody weight (Smith & Quin, 1996). Thevery rapid increase of their populations inecosystems where indigenous grazers are ingeneral much less numerous andcompetitive, leads to a dramaticimpoverishment of the vegetation, bothquantitatively and qualitatively (Chapuis etal., 1994a; Costin & Moore, 1960; Gillham,1955; Holdgate & Wace, 1961; Norman,1967; Selkirk et al., 1983), and results insevere impacts on associated indigenousfauna.

Introduced rabbit grazing on native plants in the Kerguelen. Photo J.-L. Chapuis.

Rabbit overgrazing constitutes areduction of the plant cover for terrestrialnesting birds, which often affects theirreproductive success, as was seen on MeeuwIsland (Gillham, 1963) or on the Kerguelenarchipelago (Weimerskirch et al., 1989). Italso increases direct competition for foodamong birds depending directly (e.g.granivorous) or indirectly (e.g.insectivorous) on the terrestrial vegetation(Gillham, 1963). Overgrazing may lead tothe exposure of larger proportions of soil

surface to frost heaving, rain and winderosion (Scott, 1988), resulting in dramaticdenudation of the soil. Rabbits also competedirectly for nesting burrows (or places to digthem) with many burrowing seabird species(e.g., (Young, 1981). Furthermore, thisburrowing can accelerate erosion (Chapuis,1995; Norman, 1967). It is also believed thatas much as 10% of Hooker's sea lion pupsdie annually by being trapped and suffocatedin rabbit burrows on Enderby and RoseIslands (Sanson & Dingwall, 1995). Rabbitovergrazing also prevents any large-scale



regeneration in case of fire (Norman, 1967)or cyclones (Kirk & Racey, 1992), not tomention accounts of potentially significantdirect effects such as direct disturbance(Gillham, 1963), which could lead to colonydesertion by seabirds, or even direct eggdestruction (Brown, 1974). Lastly, and aswe will see later on, rabbits often constitutean alternative prey for indigenous (e.g.,skuas) or introduced predators (e.g., cats).By increasing the number of predators, or byallowing them to persist when/where localprey is scarce, they contribute to the declineof indigenous prey (Brothers & Copson,1988). By one or several of these factors,rabbits are believed to have been ultimatelyresponsible for the decline or extinction ofseveral reptile and bird species (King, 1985;North et al., 1994; Smith & Quin, 1996).One of the most classical examples concernsthe extreme damages caused to RoundIsland, in Mauritius. Helped in this by feralgoats, introduced rabbits reduced thecommunity of several endemic palm trees,which protected the island from hurricanesdamages, down to 5-15 individualsaccording to species, several others havingcompletely disappeared (Bullock, 1977). Of

course, not only the plant species weredirectly affected, but also the animaldepending on them disappeared as well, suchas four species of endemic reptiles. As aresult, this 151 ha island possessed thehighest density of threatened plant andanimal species in the world, beforeconservation efforts concentrated oneradicating both goats and rabbits (North &Bullock, 1986; North et al., 1994).

Another example concerns LaysanIsland in the Hawaiian chain, where rabbitsalone were responsible for eliminating 26species of plants between 1903 and 1923, arate of loss exceeding one species per year(Christophersen & Caum, 1931), in(Atkinson, 1989). On that island, they havealso led to the elimination of three birdspecies. A fourth, the Laysan duck, has gonefrom 7 individuals to a one hundred foldincrease since rabbits removal (Warner,1963). Rabbits are also believed to have beenresponsible for the decline or extinction ofseveral other vertebrate species by the impactthey have on their habitat.



Box 2: A notorious omnivore: the rat

The other main category of mammalpest invaders concerns the omnivores. Thesehave often been more present in the minds ofconservation biologists because of their moredirect and visible effect on indigenouswildlife. Three species of rats have becomeunequalled worldwide threats for smallinsular prey: the Kiore, Polynesian or Pacificrat Rattus exulans, the Black, Roof or Shiprat, R. rattus and the Brown or Norwegianrat R. norvegicus. These three commensalsare known to have colonised at least 82% ofthe 123 major island groups (Atkinson,1985). Because they show importantecological differences, the three species havehad a dissimilar impact, both quantitativelyand qualitatively on insular wildlife(Atkinson, 1985; Lever, 1994).

The ecological impact of the Pacific ratis difficult to assess, but is it said to be theleast harmful of all three species on

vertebrates (and more destructive on insectand plant communities). Nevertheless, inaddition to its impact on the native plantcommunities, this partly arboreal and partlyterrestrial rodent (Atkinson, 1977) isbelieved to be responsible for the extinctionof many invertebrate, amphibian, reptile andbird species (Atkinson, 1978; Crook, 1973).For example, on Kure Atoll, these rats havebeen observed killing the translocated Laysanalbatrosses (see photo below), as well asgreat frigate birds, red-tailed tropic birds,sooty terns, common noddies and BoninIsland petrels (Kepler, 1967). Similarly, theimpact of this rat on Murphy’s petrel seemsso great (100% of chicks killed within 4 daysof hatching) on Henderson Islands, that thecolony is believed to be subsisting onlythrough immigration from the species’largest breeding site world-wide, located ontwo nearby islands, Oeno and Ducie (Lever,1994).

Introduced rat attacking a Laysan Albatross, Phoebastria immutabilis on Kure Atoll, Hawaii. The rat on theback of the bird is approaching a wound it made earlier, and will eventually kill the bird (Kepler, 1967). PhotoCameron Kepler



The two other rat species have a muchlarger impact on the fauna of the islands theyhave invaded. The Norwegian rat has largelyterrestrial habits, putting at highest risk birdswith similar habits, in particular seabirds:eggs, chicks or even brooding adults areregularly kill ed, often resulting in localextinction. Because it is arboreal and affectsforest birds as well, the Black rat probablyhas had the largest impact of all three species(Lever, 1994). A well-known example of theimportant impact of black rats on nativevertebrate communities is their establishmentaround 1964 on Big South Cape Island,New Zealand. There, they caused the localloss of three endemic birds, and the completeextinction of two other bird species and ofone species of bat, in less than 2 years (Bell,1978). In both cases, the populations ofmany other native species were depleted,although not to complete extirpation. Anotherexample, illustrating well the dramatic impactof introduced mammals on biodiversity,concerns the emblematic tuatara, Sphenodonpunctatus, last of the Sphenodontida, anorder of reptile present 225-80 million yearsago in different parts of the world. Thisendangered species is now endemic to NewZealand, but has been extirpated from themainland and now occurs only on a fewsmall islands. They are however not safe inthose few restricted areas: the invasion ofWhenuakura Island by Norway rats in 1982led to the elimination of the entire localpopulation of more than 130 tuatara(Newman, 1988). Rats have had an impacton most remaining tuatara populations aswell as on other rare reptile species (Cree,Daugherty & Hay, 1995; Newman, 1988;Newman & McFadden, 1990; Towns et al.,2001).

The impact of rats on vegetation of theislands they have invaded is also extremelyimportant. Rats eat leaves, seeds and fruits,flowers, bark and stems of many endangeredspecies. By hindering regeneration andkilling sapling and small adult trees, theymodify entire plant communities, and therebyaffect the associated fauna (Clark, 1981). Byinterference and/or exploitation, black ratswhich now occur on all vegetated habitat ofthe seven main islands of the Galapagos aresaid to be responsible for the extinction of

four species of endemic rice rats (Brosset,1963). For example, black rats have an effecton the regeneration of maritime forest(Campbell, 1978) because they are majorconsumers of inland podocarp seeds, butalso because they compete with and killmany species of birds that disperse seeds(Clark, 1981). Inversely, although examplesare rare, introduced rodents may favour theseed dispersion: black and Pacific rats arebetter seed dispersers than the native bats onTonga (Drake & McConkey, 2001).

If the Pacific rat had been documentedkilling at least 15 different bird species morethan 15 years ago, predation by the black rathas been described in at least 39 bird species,while the brown rat had been seen killing atleast 53 bird species (Atkinson, 1985).Needless to say that evidence of predation onsmall vertebrates by rats on remote islands isdifficult, and that therefore these numberssurely represent an underestimation of theactual number of bird species killed by each ratspecies. Arthropods, snails, amphibians andreptiles are less studied than birds, and fewdata are available concerning the number ofspecies under threat by introduced rats, buthere again, there is little doubt that the numberis very large (e.g., Clark, 1981; McKenzie,1993).

As mentioned before, the impact of anintroduced species on a remote oceanic islandsis very difficult to prove and even moredifficult to quantify. Norman even suggeststhat the impact of introduced rats may havebeen overstated in some cases (Norman,1975). This author argues that evidence of birdpopulation decreases is often circumstantial,and that few data are available to conclude thatrats are solely responsible of some birdextinction events. Norman suggests that themajority of the impact of introduced predatorshave been imputed to rats while the impact ofcats on the same island may have been muchhigher. Although most specialists concur onthe nearly invariable significant impact ofintroduced rats, it is true that untangling theeffects of several introduced predators, such asrats and cats, is difficult at best. Cats areindeed a major problem on oceanic islands andare often present on islands where rats havebeen introduced.



Box 3: a notorious carnivore: the cat

Because they were the main means ofcontrolling rodents on ships, domestic catstravelled around the world with explorers,sealers, whalers and other seafarers, andthus gained easy access to most remoteoceanic islands (Todd, 1977). On thoseislands, in addition to all the cats that escapedashore during a stopover, and those thatstayed with colonies or lighthouse keepers,many cats were in fact (as were mongooses)introduced in the hopes to control growingpopulations of rodents near newlyestablished human populations. Domesticcats are very adaptable (Apps, 1986;Konecny, 1983; van Aarde, 1986) and havesurvived in the most inhospitable conditionson many remote oceanic islands, bothinhabited and uninhabited (Fitzgerald, 1988).They are found on most major island groups,from subantarctic islands to arid islands withno fresh water (Atkinson, 1989; Derenne,1976; Pascal, 1980; Tabor, 1983).

Despite sometimes harsh climaticconditions, these ecosystems were veryfavourable for domestic cats, being virtuallyexempt of natural enemies and often full ofdefenceless prey (as well as introduced preysuch as rodents and rabbits, which mayallow them to pass the season whenmigratory birds are absent). The cat is anopportunistic predator, and its diet onoceanic islands may include a largeproportion of reptiles (e.g., (Bamford, 1995;Konecny, 1987)), birds (e.g., (Fitzgerald,1988; Rodriguez-Estrella et al., 1996)), ormammals ((Derenne & Mougin, 1976;Jones, 1977; Pascal, 1980)), according tothe prey’s relative abundance in differentseasons and regions.

As a consequence of the wide diet ofthis very efficient predator and the lack ofadaptation of the local prey, introduced catsare known to be the direct cause of severereductions or extinctions of numerouspopulations of insular vertebrate species(Iverson, 1978; King, 1985; Moors &Atkinson, 1984; Taylor, 1979b). As is the

case for rats, to enumerate the exhaustive listof (suspected) victims of domestic catswould be too long and depressing: as Leverpoints out for birds, “the li st of speciesthey have helped to exterminate orendanger reads like a roll-call of aviandisaster” (Lever, 1994). Instead a fewfamous examples can help to grasp the hugeimpact that a few individual cats introducedonto an islands can have on the local faunasome decades later.

Cats have often been introduced insmall numbers, but their ease in proliferationin such favourable conditions makes them afearsome threat for local animalcommunities. For example, within 70 yearsof their introduction to the Fijian islands, catsincreased so much in numbers on Viti Levuthat they were hunted for food and sport(Gibbons, 1984). It has been estimated that25 years after the introduction of five cats onMacquarie Island, the established populationof 2000 cats killed nearly half a millionburrowing petrels per year (van Aarde,1980). Similarly, in the KerguelenArchipelago, 30 years after the introductionof a cat and her three kitten, the population of3500 cats destroyed 1.2 million of birds peryear (Pascal, 1980). Some predators killmore than their immediate need: foxes storesurplus birds and eggs for use in winterwhen few birds are present, and singlelarders have been found with more than 100seabirds of several species (Bailey, 1993).This may obviously unbalance the predator-prey dynamics if the prey population is notadapted to such behaviours. On smallislands, the impact of such a population mayresult in the extinction of the prey species,especially if alternative prey are present tosustain the predator population.

In a few decades, cats introduced ontoHerekopare islands destroyed all but a fewthousands of the 400 000 birds leaving there,extirpating no less than 6 entire species ofland birds, and large colonies of seabirds(Fitzgerald & Veitch, 1985). In addition tomany local extinctions, introduced domesticcats are said to have caused more than half of



the avian species extinctions due tointroduced predators ((Jackson, 1977), in(Ebenhard, 1988)). Here again, it is notunreasonable to suggest that since birds arefar more studied than other animal taxa on

oceanic islands, the lack of published data onelimination by cats of indigenousinvertebrates, amphibians and reptiles doesnot accurately reflect the assuredly very highnumber of such events.

Stomach content of a feral cat caught in New Zealand, with at least 34 native skinks. Photo by JohnDowding.



VII. References

ABRAMS, P. A., HOLT, R. D. & ROTH, J. D. (1998). Apparentcompetition or apparent mutualism? Shared predation whenpopulations cycle. Ecology 79, 201-212.

ALEXANDER , N. J. & BIALY , G. (1994). Contraceptive vaccinedevelopment. Reproduction, Fertility & Development 6,273-280.

ALTERIO, N. (1996). Secondary poisoning of stoats (Mustelaerminea), feral ferrets (Mustela furo), and feral house cats(Felis catus) by the anticoagulant poison, brodifacoum.New Zealand Journal of Zoology 23, 331-338.

APPS, P. J. (1986). A case study of an alien predator (Feliscatus) introduced on Dassen Island: selective advantages.South African Antarctic Research 16, 118-122.

ATKINSON, I. A. E. (1977). A reassessment of factors,particularly Rattus rattus L., that influenced the decline ofendemic forest birds in the Hawaiian Islands. PacificScience 31, 109-133.

ATKINSON, I. A. E. (1978). Evidence for effects of rodents onthe vertebrate wildlife of New Zealand islands. In Theecology and control of rodents in New Zealand naturereserves, vol. Information Series 4 (ed. P. R. Dingwall, I.A. E. Atkinson and C. Hay), pp. 7-31. Department of Landand Survey, Wellington.

ATKINSON, I. A. E. (1985). The spread of commensal speciesof Rattus to oceanic islands and their effect on islandavifaunas. In Conservation of island birds, vol. 3 (ed. P. J.Moors), pp. 35-81. ICBP Technical Publication.

ATKINSON, I. A. E. (1989). Introduced animals and extinctions.In Conservation for the twenty-first century (ed. D.Western and M. C. Pearl), pp. 54-75. Oxford UniversityPress, Oxford, England.

ATKINSON, I. A. E. (2001). Introduced mammals and modelsfor restoration. Biological Conservation 99, 81-96.

BAILEY , E. P. (1992). Red foxes, Vulpes vulpes, as biologicalcontrol agents for introduced Arctic foxes, Alopex Lagopus,on Alaskan Islands. Canadian Field-Naturalist 106, 200-205.

BAILEY , E. P. (1993). Introduction of foxes to Alaskan islands- history, effects on avifauna, and eradication, pp. 1-53.United States Department of thje Interior - fish and WildlifeService, Washington, D.C.

BAMFORD, M. J. (1995). Predation by feral cats upon lizards.Western Australian Naturalist 20, 191-196.

BARLOW, N. D. (1994). Predicting the Effect of a NovelVertebrate Biocontrol Agent - a Model for Viral-VectoredImmunocontraception of New-Zealand Possums. Journal ofApplied Ecology 31, 454-462.

BELL, B. D. (1978). The Big South Cape Island rat irruption.In The ecology and control of rodents in New Zealand

nature reserves, vol. Information Series 4 (ed. P. R.Dingwall, I. A. E. Atkinson and C. Hay), pp. 33-46.Department of Land and Survey, Wellington.

BERTRAM, D. F. & NAGORSEN, D. W. (1995). Introducedrats, Rattus Spp, on the Queen-Charlotte-Islands -Implications for seabird conservation. Canadian Field-Naturalist 109, 6-10.

BOMFORD, M. & O'BRIEN, P. (1995). Eradication or controlfor vertebrate pests. Wildlife Society Bulletin 23.

BOUSSÈS, P., RÉALE, D. & CHAPUIS, J. L. (1994). Highwintertime mortality-rate in a population of Corsicanmouflons (Ovis musimon) in the sub-Arctic Archipelagoof Kerguelen. Mammalia 58, 211-223.

BRADLEY , M. P., HINDS, L. A. & BIRD, P. H. (1997). Abait-delivered immunocontraceptive vaccine for theEuropean red fox (Vulpes vulpes) by the year 2002?Reproduction Fertility and Development 9, 111-116.

BROSSET, A. (1963). Statut actuel des mammifères des îlesGalapagos. Mammalia 27, 323-338.

BROTHERS, N. P. (1984). Breeding, distribution and status ofburrow-nesting petrels at Macquarie Island. AustrialianWildlife Research 11, 113-131.

BROTHERS, N. P. & COPSON, G. R. (1988). MacquarieIsland flora and fauna management - interpreting progressand predictions for the future. Papers and Proceedings ofthe Royal Society of Tasmania 122, 129-135.

BROTHERS, N. P., EBERHARD I. E., COPSON, G. R. &SKIRA , I. J. (1982). Control of rabbits on MacquarieIsland by myxomatosis. Australian Wildlife Research9, 477-485.

BROTHERS, N. P., SKIRA , I. J. & COPSON, G. R. (1985).Biology of the feral cat, Felis catus (L.), on MacquarieIsland. Austrialian Wildlife Research 12, 425-436.

BROWN, D. (1997). Chetwode island kiore and wekaeradication project. Ecological Management 5, 11-20.

BROWN, K. P., MOLLER, H., INNES, J. & JANSEN, P.(1998). Identifying predators at nests of small birds in aNew Zealand forest. Ibis 140, 274-279.

BROWN, W. Y. (1974). Rabbit destruction of tern eggs. Auk91, 840-841.

BUCKLE, A. P. & FENN, M. G. P. (1992). Rodent controlin the conservation of endangered species. In 15thVertebrate Pest Conference (ed. J. E. Borreco and R. E.Marsh), pp. 36-41, University of California Davis, USA.

BULLOCK , D. (1977). Round Island - a tale of destruction.Oryx 15, 51-58.

BULLOCK , D. (1986). The ecology and conservation ofreptiles on Round Island and Gunner's Quoin, Mauritius.Biological Conservation 37, 135-156.



BULLOCK , D. J., NORTH, S. G., DULLOO , M. E. & THORSEN,M. (2001). The impact of rabbit and goat eradication on theecology of Round Island, Mauritius. In Turning the Tide:the Eradication of Invasive Species (ed. C. R. Veitch andM. Clout), pp. 10. Invasive Species Specialist Group of theWorld Conservation Union (IUCN). University of Auckland,New-Zealand.

BURGER, J. & GOCHFELD, M. (1994). Predation and effects ofhumans on islands-nesting seabirds. In Seabirds on islands,threats, case studies and action plans, vol. 1 (ed. D. N.Nettleship, J. Burger and M. Gochfeld), pp. 39-67. BirdlifeConservation Series.

CABIN , R. J., WELLER, S. G., LORENCE, D. H., FLYNN , T.W., SAKAI , A. K., SANDQUIST, D. & HADWAY , L. J.(2000). Effects of long-term ungulate exclusion and recentalien species control on the preservation and restoration of aHawaiian tropical dry forest. Conservation Biology 14,439-453.

CAMPBELL, D. J. (1978). the effect of rats on vegetation. InThe ecology and control of rodents in New Zealand naturereserves, vol. Information Series 4 (ed. P. R. Dingwall, I.A. E. Atkinson and C. Hay), pp. 99-100. Department ofLand and Survey, Wellington.

CAUGHLEY, G. (1994). Directions in conservation biology.Journal of Animal Ecology 63, 215-244.

CEBALLOS, G. & BROWN, J. H. (1995). Global patterns ofmammalian diversity, endemism and endangerment.Conservation Biology 9, 559-568.

CHAMBERS, L. K., SINGLETON, G. R. & HOOD, G. M.(1997). Immunocontraception as a potential control methodof wild rodent populations. Belgian Journal of Zoology127, 145-156.

CHAPUIS, J.-L. (1981). Evolution saisonnière du régimealimentaire d'Oryctolagus cuniculus (L.) dans différentstypes d'habitats en France. In World Lagomorph Conference(ed. K. Myers and C. D. Mac Innes), pp. 743-760,University of Guelph, Canada.

CHAPUIS, J.-L. (1990). Comparison of the diets of twosympatric Lagomorphs, Lepus europaeus (Pallas) andOryctolagus cuniculus (L.) in an agroecosystem of the Ile-de-France. Zeitschrift für Saügetierkunde-InternationalJournal of Mammalian Biology 55, 176-185.

CHAPUIS, J.-L. (1995). Alien mammals in the FrenchSubantarctic Islands. In Progress in conservation of theSubantarctic Islands, vol. 2 (ed. P. R. Dingwall), pp. 127-132. IUCN - The World Conservation Union - Conservationof the Southern Polar Region.

CHAPUIS, J.-L., BOUSSÈS, P. & BARNAUD , G. (1994a). Alienmammals, impact and management in the French sub-Antarctic islands. Biological Conservation 67, 97-104.

CHAPUIS, J.-L., CHANTAL , J. & BIJLENGA, G. (1994b).Myxomatosis without vectors on the sub-Antarctic islands

of Kerguelen 3 decades after its introduction. ComptesRendus de l'Academie Des Sciences Série III-Sciencesde La Vie-Life Sciences 317, 174-182.

CHAPUIS, J.-L., FRENOT, Y., LEBOUVIER, M., BARNAUD ,G. & BRETAGNOLLE, V. (2000). Restauration d'îlessubantarctiques françaises par manipulation d'espècesintroduites, pp. 1-112. M.A.T.E., M.N.H.N. (eds), Paris.

CHAPUIS, J.-L., LE ROUX, V., ASSELINE, J., LEFÈVRE, L.& KERLEAU, F. (2001). Eradication of the rabbit(Oryctolagus cuniculus) by poisoning, on three islandsof the subantarctic archipelago of Kerguelen. WildlifeResearch 28, 323-331.

CHEKCHAK , T., CHAPUIS, J.-L., PISANU, B. & BOUSSÈS,P. (2000). Introduction of the rabbit flea, Spilopsylluscuniculi (Dale), to a subantarctic island (KerguelenArchipelago) and its assessment as a vector ofmyxomatosis. Wildlife Research 27, 91-101.

CHEVRIER, M., VERNON, P. & FRENOT, Y. (1997).Potential effects of two alien insects on a sub-Antarcticwingless fly in the Kerguelen Islands. In AntarcticCommunity, Species, Structure and Survival (ed. B.Battaglia, J. Valencia and D. W. H. Walton), pp. 424-431. Cambridge University Press, Cambridge.

CHIMERA , C., COLEMAN , M. C. & PARKES, J. P. (1995).Diet of feral goats and feral pigs on Auckland Island, NewZealand. New Zealand Journal of Ecology 19, 203-207.

CHOQUENOT, D. & PARKES, J. (2001). Setting thresholdsfor pest control: how does pest density affect resource

viability? Biological Conservation 99, 29-46.CHRISTOPHERSEN, E. & CAUM , E. L. (1931). Vasular

plants of the Leeward Islands, Hawaii. B. P. BishopMuseum Bulletin 81, 1-41.

CLAPPERTON, B. K., EASON, C. T., WESTON, R. J.,WOOLHOUSE, A. D. & MORGAN, D. R. (1994).Development and testing of attractants for feral cats, Feliscatus L. Wildlife Research 21, 389-399.

CLARK , D. A. (1981). Foraging patterns of black rats acrossa desert-montagne forest gradient in the GualapagosIslands. Biotropica 13, 182-194.

CLOUT, M. (1999). Biodiversity conservation and themanagement of invasive animals in New Zealand. InInvasive species and biodiversity managment (ed. O. T.Sandlund, P. Schei and A. Viken), pp. 349*361. KluwerAcademic Publishers.

CLOUT, M. N. & SARRE, S. D. (1997). Model marsupial ormenace? A review of research on brushtail possums inAustralia and New Zealand. Wildlife Society Bulletin25, 168-172.

COBLENTZ, B. E. (1978). The effect of feral goats (Caprahircus) on island ecosystems. Biological Conservation13, 279-286.



COOK, C. J. (1999). Smarter baits: The effects of stress on baitaversion and options to avoid the development of baitaversions. New Zealand Journal of Ecology 23, 275-279.

COOPER, J. & BROOKE, R. K. (1982). Past and presentdistribution of the feral European rabbit Oryctolaguscuniculus on Southern African offshore islands. SouthAfrican Journal of Wildlife Research 12, 71-75.

COSSON, J.-F., PASCAL, M. & BIORET, F. (1996). Origine etrépartition des musaraignes du genre Crocidura dans les îlesbretonnes. Vie et Mili eu - Life & Environment 46, 233-244.

COSTIN, A. B. & MOORE, D. M. (1960). The effect of rabbitgrazing on the grasslands of Macquarie Island. Journal ofEcology 48, 729-732.

COURCHAMP, F. & CORNELL, S. J. (2000). Virus VectoredImmunocontraception to control feral cats on islands: amathematical model. Journal of Applied Ecology 37.

COURCHAMP, F., LANGLAIS , M. & SUGIHARA, G. (1999a).Cats protecting birds: modelling the mesopredator releaseeffect. Journal of Animal Ecology 68, 282-292.

COURCHAMP, F., LANGLAIS , M. & SUGIHARA, G. (1999b).Control of rabbits to protect island birds from cat predation.Biological Conservation 89, 219-225.

COURCHAMP, F., LANGLAIS , M. & SUGIHARA, G. (2000).Rabbits killing birds: modelling the hyperpredation process.Journal of Animal Ecology 69, 154-164.

COURCHAMP, F. & SUGIHARA, G. (1999). Modeling thebiological control of an alien predator to protect islandspecies from extinction. Ecological Applications 9, 112-123.

COWAN, P. E. (1992). The eradication of introduced Australianbrushtail possums, Trichosurus vulpecula, from KapitiIsland, a New-Zealand nature reserve. BiologicalConservation 61, 217-226.

COWAN, P. E. (1996). Possum biocontrol: Prospects forfertility regulation. Reproduction Fertility andDevelopment 8, 655-660.

CRAFFORD, J. E. (1990). The role of feral house mice inecosystem functioning on Marion Island. In AntarcticEcosystems. Ecological Change and Conservation (ed. K.R. Kerry and G. Hempel), pp. 359-364. Springer-Verlag,Berlin.

CRAFFORD, J. E. & SCHOLTZ, C. H. (1987). Quantitativedifferences between the insect faunas of sub-antarctic Marionand Prince Edward Islands: a result of human intervention?Biological Conservation 40, 255-262.

CRAWLEY , M. J. (1986). The population biology of invaders.Philosophical Transactions of the Royal Society of London,B 314, 711-731.

CREE, A., DAUGHERTY, C. H. & HAY , J. M. (1995).Reproduction of a rare New-Zealand reptile, the Tuatara

Sphenodon punctatus, on rat-free and rat-inhabitedislands. Conservation Biology 9, 373-383.

CROOK, I. G. (1973). The tuatara, Sphenodon punctatusGray, on islands with and without populations of thePolynesian rat, Rattus exulans (Peale). Proceedings ofthe New Zealand Ecological Society 20, 115-120.

DALY , K. & GORIUP, P. (1987). Eradication of feral goatsfrom small islands, pp. 46. International Council for BirdPreservation, Cambridge.

DAY , D. (1981). The doomsday book of animals. EburyPress, London.

DERENNE, P. (1976). Notes sur la biologie du chat haret deKerguelen. Mammalia 40, 531-595.

DERENNE, P. & MOUGIN, J. L. (1976). Donnéesécologiques sur les mammifères introduits de l'île auxCochons, archipel Crozet (40°06'S, 50°14'E). Mammalia40, 21-35.

DEVOS, A., MANVILLE , R. H. & VAN GELDER, R. G.(1956). Introduced mammals and their influence on nativebiota. Zoologica: New York Zoological Society 41, 163-194.

DEVOS, A. & PETRIDES, G. A. (1967). Biological effectscaused by terrestrial vertebrates introduced into non-nativeenvironments. In Proceedings and Papers of the 10thTechnical Meeting, vol. 9, pp. 113-119. InternationalUnion for Conservation of Nature and Natural ResourcesPublication.

DI CASTRI, F. (1989). History of biological invasions withemphasis on the Old World. In Biological invasions: aglobal perspective (ed. J. Drake, F. di Castri, R. Groves,F. Kruger, H. A. Mooney, M. Rejmanek and M.Williamson), pp. 1-30. Wiley, New-York.

DIAMOND , J. (1989a). Overview of recent extinctions.Conservation for the twenty-first century (ed. D. Westernand M. C. Pearl), pp. 37-41. Oxford University Press,New York.

DIAMOND , J. & VEITCH, C. R. (1981). Extinctions andintroductions in the New Zealand avifauna: cause andeffects? Science 211, 499-501.

DIAMOND , J. M. (1989b). Nine hundred kiwis and a dog.Nature 338, 544.

DINGWALL , P. R., ATKINSON, I. A. E. & HAY , C. (1978).The ecology and control of rodents in New Zealand naturereserves. In Department of Lands and Survey InformationSeries, vol. 4, Wellington, New Zealand.

DOBSON, A. P. (1988). Restoring island ecosystems: thepotential of parasites to control introduced mammals.Conservation Biology 2, 31-39.

DOWDING, J. E. & MURPHY, E. C. (2001). The impact ofpredation by introduced mammals on endemic shorebirdsin New Zealand: a conservation perspective. BiologicalConservation 99, 47-64.



DOWDING, J. E., MURPHY, E. C. & VEITCH, C. R. (1999).Brodifacoum residues in target and non-target speciesfollowing an aerial poisoning operation on Motuihe Island,Hauraki Gulf, New Zealand. New Zealand Journal ofEcology 23, 207-214.

DOWLING, H. G. (1964). Goats and hawks - a theory ofpredation on the land iguana. Animal Kingdom 67, 51-56.

DRAKE, D. R. & MCCONKEY, K. R. (2001). Does seedpredation by introduced rats threaten or enhance therecruitment of Pandanus tectorius trees in Tonga? In Theecology of Insular Biotas (ed. C. H. Daugherty, D. R. Drakeand C. P. H. Mulder), pp. 19, University of Wellington,New-Zealand.

DREVER, M. C. & HARESTAD, A. S. (1998). Diets of Norwayrats, Rattus norvegicus, on Langara Island, Queen CharlotteIslands, British Columbia: implications for conservation ofbreeding seabirds. Canadian Field-Naturalist 112, 676-683.

EASON, C. T., WICKSTROM, M., TURCK, P. & WRIGHT, G. R.G. (1999). A review of recent regulatory and environmentaltoxicology studies on 1080: results and implications. NewZealand Journal of Ecology 23, 129-137.

EBENHARD, T. (1988). Introduced birds and mammals and theirecological effects. Swedish Wildlife Research 13, 1-107.

ECKHARDT, R. C. (1972). Introduced plants and animal in theGalapagos Islands. Bioscience 22, 585-590.

EMPSON, R. A. & M ISKELLY , C. M. (1999). The risks, costsand benefits of using brodifacoum to eradicate rats fromKapiti Island, New Zealand. New Zealand Journal ofEcology 23, 241-254.

ESPEUT, W. B. (1882). On the acclimatization of the Indianmungoos in Jamaica. Proceedings of the Zoological Societyof London, 712-714.

FEARE, C. (1998). The sustainable exploitation of sooty terneggs in the Seychelles. 6th annual unpublished report to theDivision of Environment, Government of the Republic ofSeychelles.

FENNER, F. & ROSS, J. (1994). Myxomatosis. In TheEuropean rabbit. The history and biology of a successfulcolonizer (ed. H. V. Thompson and C. M. King), pp. 205-239. Oxford University Press Inc, New York.

FITZGERALD, B. M. (1988). Diet of domestic cats and theirimpact on prey populations. In The domestic cat, thebiology of its behaviour (ed. D. C. Turner and P. Bateson),pp. 123-147. Cambridge University Press, Cambridge.

FITZGERALD, B. M. & GIBB, J. A. (2001). Introducedmammals in a New Zealand forest: long-term research in theOrongorongo Valley. Biological Conservation 99, 97-108.

FITZGERALD, B. M. & VEITCH, C. R. (1985). The cats ofHerekopare Island, New Zealand: their history, ecology andeffects on birdlife. New Zealand Journal of Zoology 12,319-330.

FLUX , J. E. C. (1993). Relative effect of cats myxomatosistraditional control or competitors in removing rabbitsfrom islands. New Zealand Journal of Zoology 20, 13-18.

FLUX , J. E. C. & FULLAGAR , P. J. (1992). Worlddistribution of the rabbit Oryctolagus cuniculus onislands. Mammal Review 22, 151-205.

FRITH, H. J. (1979). Wildlife Conservation, 1st edition.Angus & Robertson, Sydney.

FRITTS, T. H. & RODDA , G. H. (1998). The role ofintroduced species in the degradation of islandsecosystems: a case history of Guam. Annual Review ofEcology and Systematics 29, 113-140.

FURNESS, R. W. (1987). The skuas. T & AD Poyser Ltd,Calton, UK.

GIBBONS, J. (1984). Iguanas of the South Pacific. Oryx 18,82-91.

GILLHAM , M. E. (1955). Ecology of the PembrokeshireIslands. III. The effect of grazing on the vegetation.Journal of Ecology 43, 172-206.

GILLHAM , M. E. (1963). Some interactions of plants,rabbits and seabirds on South African islands. Journal ofEcology 51, 275-294.

GILLIES , C. A. & PIERCE, R. J. (1999). Secondarypoisoning of mammalian predators during possum androdent control operations at Trounson Kauri Park,Northland, New Zealand. New Zealand Journal ofEcology 23, 183-192.

GLEESON, J. P. & VAN RENSBURG, P. J. J. (1982). Feedingecology of the house mouse Mus musculus on MarionIsland. South African Journal of Antarctic Research 12,34-39.

GODFREY, M. E. R., REID, T. C. & MCALLUM , H. J. F.(1981). The oral toxicity of brodifacoum to rabbits. NewZealand Journal of Experimental Agriculture 9, 23-25.

GOSLING, L. M. & BAKER, S. J. (1989). The eradication ofmuskrats and coypus from Britain. Biological Journal ofthe Linnean Society 38, 39- 51.

HAMMAN , O. (1975). Vegetational changes in the GalàpagosIslands during the period 1966-1973. BiologicalConservation 7, 37-59.

HEYWARD , R. P. & NORBURY, G. L. (1998). Secondarypoisoning of ferrets and cats after 1080 rabbit poisoning.Wildlife Research 25, 75-80.

HICKLING , G. J., HENDERSON, R. J. & THOMAS, M. C. C.(1999). Poisoning mammalian pests can have unintendedconsequences for future control: Two case studies. NewZealand Journal of Ecology 23, 267-273.

HOLDGATE, M. W. (1967). The influence of introducedspecies on the ecosystems of temperate oceanic islands. InProceedings and Papers of the 10th Technical Meeting,vol. 9, pp. 151-176. International Union for



Conservation of Nature and Natural Resources Publication.HOLDGATE, M. W. (1977). Terrestrial ecosystems in the

Antarctic. Philosophical Transactions of the Royal Societyof London, B 279, 5-25.

HOLDGATE, M. W. (1986). Summary and conclusions:characteristics and consequences of biological invasions.Philosophical Transactions of the Royal Society of London,B 314, 733-742.

HOLDGATE, M. W. & WACE, N. M. (1961). The influence ofman on the floras and faunas of Southern Islands. PolarRecord 10, 475-493.

HOLT, R. D. & LAWTON, J. H. (1994). The ecologicalconsequences of shared natural enemies. Annual Review ofEcology and Systematics 25, 495-520.

HOMOLKA , M. (1988). Diet of the wild rabbit (Oryctolaguscuniculus) in an agrocoenosis. Folia Zool. Brno 37, 121-128.

HONE, J. (1990). Predator prey theory and feral pig control,with emphasis on evaluation of shooting from a helicopter.Australian Wildlife Research 17, 123-130.

HONE, J. (1992). Modeling of poisoning for vertebrate pest-control, with emphasis on poisoning Feral pigs. EcologicalModelling 62, 311-327.

HONE, J. (1995). Spatial and temporal aspects of vertebrate pestdamage with emphasis on feral pigs. Journal of AppliedEcology 32, 311-319.

HONNEGGER, R. E. (1981). List of amphibians and reptileseither known or thought to have become extinct since 1600.Biological Conservation 19, 141-158.

HUYSER, O., RYAN , P. G. & COOPER, J. (2000). Changes inpopulation size, habitat use and breeding biology of lessersheathbills (Chionis minor) at Marion Island: impacts ofcats, mice and climate change? Biological Conservation 92,299-310.

I.U.C.N. (2001). Guidelines for the Prevention of BiodiversityLoss caused by Alien Invasive Species.http://www.iucn.org/themes/ssc/pubs/policy/invasivesEn g.htm .

INNES, J. & BARKER, G. (1999). Ecological consequences oftoxin use for mammalian pest control in New Zealand - Anoverview. New Zealand Journal of Ecology 23, 111-127.

IVERSON, J. B. (1978). The impact of feral cats and dogs onpopulations of the west Indian rock iguana, Cycluracarinata. Biological Conservation 14, 63-73.

JACKSON, J. A. (1977). Alleviating the problems ofcompetition, predation, parasitism, and disease in endangeredbirds. In Endangered birds (ed. S. A. Temple), pp. 75-84.University of Wisconsin Press and Croom Helm, Madisonand London.

JONES, E. (1977). Ecology of the feral cat, Felis catus (L.),(Carnivora: Felidae) on Macquarie Island. AustrialianWildlife Research 4, 249-262.

JOUVENTIN, P. & ROUX, J.-P. (1983). The discovery of anew albatross. Nature 305, 181.

KAISER, J. (1999). Stemming the tide of invading species -News Focus. Science 285, 1836-1841.

KAISER, J. (2001). Galapagos takes aim at alien invaders -New focus. Science 293, 590-592.

KARL, B. J. & BEST, H. A. (1982). Feral cats on StewartIsland: their foods and their effects on kakapo. NewZealand Journal of Zoology 9, 287-294.

KEEGAN, D. R., COBLENTZ, B. E. & WINCHELL , C. S.(1994). Feral goat eradication on San Clemente Island,California. Wildlife Society Bulletin 22, 56-61.

KEELING, M. J. & GRENFELL, B. (1997). Diseasesextinction and community size: modeling the persistenceof measles. Science 275, 65-67.

KEPLER, C. B. (1967). Polynesian rat predation on nestingLaysan Albatrosses and other Pacific seabirds. Auk 84,426-430.

KESSLER, C. C. (2001). Eradication of feral goats and pigsfrom Sarigan Island, Commonwealth of the NorthernMariana Islands; methods and results. In Eradication ofislands Invasives: practical actions and results achieved(ed. M. Clout), pp. 18, University of Auckland, New-Zealand.

KING, W. B. (1985). Island birds: will the future repeat thepast? In Conservation of island birds, vol. 3 (ed. M. P.J.), pp. 3-15. ICBP Technical Publication.

KIRK, D. A. & RACEY, P. A. (1992). Effects of theintroduced black-naped hare Lepus nigricollis nigricollison the vegetation of Cousin Island, Seychelles andpossible implications for avifauna. BiologicalConservation 61, 171-179.

KIRKPATRICK, J. F., TURNER, J. W., LIU , I. K. M.,FAYRERHOSKEN, R. & RUTBERG, A. T. (1997). Casestudies in wildlife immunocontraception: wild and feralequids and white-tailed deer. Reproduction Fertility andDevelopment 9, 105-110.

KLEIN , D. R. (1968). The introduction, increase, and crashof reindeer on St. Matthew Island. Journal of WildlifeManagement 32, 350-367.

KOLAR, C. S. & LODGE, D. M. (2001). Progress ininvasion biology: predicting invaders. Trends in Ecology& Evolution 16, 199-204.

KONECNY, M. J. (1983). Behavioural ecology of feral housecats in the Galapagos Islands, Ecuador, University ofFlorida.

KONECNY, M. J. (1987). Food habits and energetics of feralhouse in the Galàpagos Islands. Oikos 50, 24-32.

LE ROUX, V., CHAPUIS, J.-L., FRENOT, Y. & VERNON, P.(2001). Diet of the house mouse (Mus musculus) atGuillou Island, Kerguelen archipelago, Subantarctic.Polar Biology in press.

http://www.iucn.org/themes/ssc/pubs/policy/invasivesEn



LEVER, C. (1994). Naturalized animals: the ecology ofsuccessfully introduced species. Poyser Natural History,London.

LOAGUE, P. (1993). Pest control and animal welfare. NewZealand Journal of Zoology 20, 253-256.

MACK, R. N. & LONSDALE, W. M. (2002). Eradicatinginvasive plants: hard-won lessons for islands. In Turning theTide: the Eradication of Invasive Species (ed. C. R. Veitchand M. Clout). Invasive Species Specialty Group of theWorld Conservation Union (IUCN), Auckland, New-Zealand.

MACK, R. N., SIMBERLOFF, D., LONSDALE, M., EVANS, H.,CLOUT, M. & BAZZAZ , F. A. (2000). Biotic invasions:causes, epidemiology, global consequences and control.Ecological Applications 10, 689-710.

MANLY , B. F. J. (2000). Statistics for environmental scienceand management. Chapman & Hall / CRC, Boca Raton.

MCCALLUM , H. (1996). Immunocontraception for wildlifepopulation-control. Trends in Ecology & Evolution 11,491-493.

MCDONALD, R., SHORT, J. A., CATE, L., LYALL , K. &ORCHARD, R. (1999). New technology for poison delivery.New Zealand Journal of Ecology 23, 289-292.

MCILROY, J. C. & J., G. (1992). Secondary poisoning hazardsassociated with 1080-treatezd carrot-baiting campaignsagainst rabbits, Oryctolagus cuniculus. Wildlife Research19, 629-641.

MCKENZIE, D. (1993). Eradication of kiore from MotuopaoIsland. Ecological Management 1, 16-18.

MERTON, D. (1987). Eradication of rabbits from Round IslandMauritius a conservation success story. Dodo 24, 19-43.

MERTON, D., CLIMO , G., LABOUDALLON , V., MANDER, C.& ROBERT, S. (2001). Alien animal eradication andquarantine on inhabited islands in the Seychelles. In Turningthe Tide: the Eradication of Invasive Species (ed. C. R.Veitch and M. Clout), pp. 18. Invasive Species SpecialistGroup of the World Conservation Union (IUCN). Universityof Auckland, New-Zealand.

M ICOL, T. & JOUVENTIN, P. (1995). Restoration ofAmsterdam Island, South Indian ocean, following control offeral cattle. Biological Conservation 73, 199-206.

MONCORPS, S., CHAPUIS, J.-L., HAUBREUX, D. &BRETAGNOLLE, V. (1998). Diet of the brown skuaCatharacta skua lonnbergi on the Kerguelen archipelago:comparisons between techniques and between islands. PolarBiology 19, 9-16.

MOODIE, E. (1995). The potential for biological control offeral cats in Australia. Australian Nature ConservationAgency, Canberra.

MOORS, P. J. & ATKINSON, I. A. E. (1984). Predation onseabirds by introduced animals, and factors affecting itsseverity. In Status and conservation of the world's seabirds,

vol. 2 (ed. J. P. Croxall, P. G. H. Evans and R. W.Schreiber), pp. 667-690. ICBP Technical Publication.

MORGAN, D. R. (1999). Risks to non-target species fromuse of a gel bait for possum control. New ZealandJournal of Ecology 23, 281-287.

MORRIS, K. D. (2001). The control of the introduced blackrat on Barrow and adjacent islands off the north west coastof Western Australia. In Turning the Tide: theEradication of Invasive Species (ed. C. R. Veitch andM. Clout), pp. 26. Invasive Species Specialist Group ofthe World Conservation Union (IUCN). University ofAuckland, New-Zealand.

MOUGEOT, F., GENEVOIS, F. & BRETAGNOLLE, V. (1998).Predation on burrowing petrels by the brown skua atMayes islands, Kerguelen. Journal of Zoology 244, 429-438.

MURPHY, E. & BRADFIELD, P. (1992). Change in diet ofstoats following poisoning of rats in a New-Zealandforest. New Zealand Journal of Ecology 16, 137-140.

MURPHY, E. C., ROBBINS, L., YOUNG, J. B. & DOWDING,J. E. (1999). Secondary poisoning of stoats after an aerial1080 poison operation in Pureora Forest, New Zealand.New Zealand Journal of Ecology 23, 175-182.

MYERS, J. H., SIMBERLOFF, D., KURIS, A. M. & CAREY,J. R. (2000a). Eradication revisited: dealing with exoticspecies. Trends in Ecology & Evolution 15, 316-320.

MYERS, N., M ITTERMEIER, R. A., M ITTERMEIER, C. G.,DA FONSECA, G. A. B. & KENT, J. (2000b). Biodiversityhotspots for conservation priorities. Nature 403, 853-858.

NELLIS, D. W. & EVERARD, C. O. R. (1983). The biologyof the mongoose in the Caribbean. Studies on the faunaof Curaçao and other Caribbean islands. Utrecht 64, 1-162.

NETTLES, V. F. (1997). Potential consequences andproblems with wildlife contraceptives. ReproductionFertility and Development 9, 137-143.

NEWMAN , D. G. (1988). Evidence of predation on a youngtuatara, Sphenodon punctatus, by kiore, Rattus exulans,on lady Alice Island. New Zealand Journal of Zoology15, 443-446.

NEWMAN , D. G. (1994). Effects of a mouse, Musmusculus, eradication program and habitat change onlizard populations of Mana Island, New- Zealand, withspecial reference to Mcgregor Skink, Cyclodinamacgregori. New Zealand Journal of Zoology 21, 443-456.

NEWMAN , D. G. & MCFADDEN, I. (1990). Seasonalfluctuations of numbers, breeding, and food of Kiore(Rattus exulans) on Lady-Alice-Island (Hen and ChickensGroup), with a consideration of kiore - tuatara



(Sphenodon punctatus) relationships in New-Zealand. NewZealand Journal of Zoology 17, 55-63.

NORMAN, F. I. (1967). The interactions of plants and animalson Rabbit Island, Wilson's Promontory, Victoria.Proceedings of the Royal Society of Victoria 280, 193-200.

NORMAN, F. I. (1975). The murine rodents Rattus rattus,exulans, and norvegicus as avian predators. Atoll ResearchBulletin 182, 1-13.

NORTH, S. G. & BULLOCK , D. (1986). Changes in thevegetation and populations of introduced mammals of RoundIsland and Gunner's Quoin, Mauritius. BiologicalConservation 37, 99-117.

NORTH, S. G., BULLOCK , D. J. & DULLOO , M. E. (1994).Changes in the vegetation and reptile populations on RoundIsland, Mauritius, following eradication of rabbits.Biological Conservation 67, 21-28.

NUGENT, G., FRASER, W. & SWEETAPPLE, P. (2001). Topdown or bottom up? Comparing the impacts of introducedarboreal possums and 'terrestrial' ruminant on native forest inNew Zealand. Biological Conservation 99, 65-79.

PARKES, J. P. (1984). Feral goats on Raoul Island : effect ofcontrol methods on their density, distribution, andproductivity. New Zealand Journal of Ecology 7, 85-94.

PARKES, J. P. (1990). Feral goat control in New Zealand.Biological Conservation 54, 335-348.

PARKES, J. P. (1993). Feral goats - Designing solutions for adesigner pest. New Zealand Journal of Ecology 17, 71-83.

PASCAL, M. (1980). Structure et dynamique de la population dechats harets de l'Archipel des Kerguelen. Mammalia 44,161-182.

PASCAL, M. (1983). L'introduction des espèces mammaliennesdans l'archipel des Kerguelen (océan Indien sud). Impact deces espèces exogènes sur le milieu insulaire. Comptes-Rendus de la Société Biogéographique 59, 257-267.

PASCAL, M. & CHAPUIS, J.-L. (2000). Eradication demammifères introduits en milieux insulaires : questionspréalables et mise en application. Revue d'Ecologie (LaTerre et la Vie) S7, 85-104.

PEABODY, F. E. (1961). Annual growth zones in Vertebrates(living and fossil). Journal of Morphology 108, 11-62.

PECH, R., HOOD, G. M., MCILROY, J. & SAUNDERS, G.(1997). Can foxes be controlled by reducing their fertility?Reproduction Fertility and Development 9, 41-50.

PICKARD, J. (1982). Catastrophic disturbance and vegetation onLittle slope, Lord Howe Island. Australian Journal ofEcology 7, 161-170.

PROBST, J.-M. (1997). Animaux de la Réunion. AzaléesEditions, La Réunion.

RÉALE, D., BOUSSÈS, P. & CHAPUIS, J. L. (1996). Female-biased mortality induced by male sexual harassment in a feral

sheep population. Canadian Journal of Zoology 74,1812-1818.

RECHER, H. F. & CLARK , S. S. (1974). A biologicalsurvey of Lord Howe Island with recommendations for theconservation of the island's wildlife. BiologicalConservation 6, 263-273.

REID, W. V. (1998). Biodiversity hotspots. Trends inEcology & Evolution 13, 275-280.

RICE, C. G. (1991). Goat removal from Aguijan Island:lessons for future efforts. Transactions of the WesternSection of the Wildlife Society 27, 42-46.

RODGER, J. C. (1997). Likely targets forimmunocontraception in marsupials. ReproductionFertility and Development 9, 131-136.

RODRIGUEZ-ESTRELLA, R., DE LA LUZ, J. L. L., BRECEDA,A., CASTELLANOS, A., CANCINO , J. & LLINAS , J.(1996). Status, density and habitat relationships of theendemic terrestrial birds of Socorro Island, RevillagigedoIslands, Mexico. Biological Conservation 76, 195-202.

ROEMER, G. W. (1999). The ecology and conservation ofthe Island Fox. Ph.D. thesis thesis, University ofCalifornia.

ROEMER, G. W., COONAN, T. J., GARCELON, D. K.,BASCOMPTE, J. & LAUGHRIN, L. (2001). Feral pigsfacilitate hyperpredation by golden eagles and indirectlycause the decline of the island fox. Animal Conservation4, 307-318.

ROEMER, G. W., DONLAN, C. J. & COURCHAMP, F.(2002). Golden eagles, feral pigs and island foxes: howexotic species turn native predators into prey.Proceedings of the National Academy of Sciences, USA.

ROOTS, C. (1976). Animal invaders. David & Charles,Nweton Abbot.

ROWE-ROWE, D. T., GREEN, B. & CRAFFORD, J. E.(1989). Estimated impact of feral house mice on sub-antarctic invertebrates at Marion Island. Polar Biology 9,457-460.

RUDGE, M. R. (1984). The occurrence and status ofpopulations of feral goats and sheep throughout theworld. In: Feral mammals, problems and potential. In 3rdInternational Theriological Conference, pp. 55-84,Helsinki.

SANSON, L. V. & DINGWALL , P. R. (1995). Conservationstatus of New Zealand's Subantarctic Islands. In Progressin conservation of the Subantarctic Islands, vol. 2 (ed.P. R. Dingwall), pp. 85-105. IUCN - The WorldConservation Union - Conservation of the Southern PolarRegion.

SAUNDERS, A. & NORTON, D. A. (2001). Ecologicalrestoration at Mainland Islands in New Zealand.Biological Conservation 99, 109-119.



SAUNDERS, G. & HARRIS, S. (2000). Evaluation of attractantsand bait preferences of captive red foxes (Vulpes vulpes).Wildlife Research 27, 237-243.

SCOTT, J. J. (1988). Rabbit distribution history and related landdisturbance Macquarie Island Australia. Papers andProceedings of the Royal Society of Tasmania 122, 255-266.

SEAMAN , G. A. (1952). The mongoose and Caribbean wildlife.Transactions of the North American Wildlife Conference17, 188-197.

SELKIRK , P. M., COSTIN, A. B., SEPPELT, R. D. & SCOTT, J.J. (1983). Rabbits, vegetation and erosion on MacquarieIsland. Proceedings of the Linnean Society N.S.W. 106,337-346.

SILVERSTEIN, A. & SILVERSTEIN, V. (1974). Animalinvaders. The story of imported wildlife. Athenum, NewYork.

SIMBERLOFF, D. & STILING , P. (1996a). How risky isbiological control? Ecology 77, 1965-1974.

SIMBERLOFF, D. & STILING , P. (1996b). Risks of speciesintroduced for biological control. Biological Conservation78, 185-192.

SKIRA , I. J., BROTHERS, N. P. & COPSON, G. R. (1983).Establishment of the European rabbit flea on MacquarieIsland. Australian Wildlife Research 10, 121-127.

SMITH , A. P. & QUIN , D. G. (1996). Patterns and causes ofextinction and decline in Australian conilurine rodents.Biological Conservation 77, 243-267.

SMITH , G., WALMSLEY , A. & POLKINGHORNE, I. (1997).Plant-derived immunocontraceptive vaccines. Reproduction,Fertility & Development 9, 85-89.

SMITH , V. R. (1977). A qualitative description of energy flowand nutrient cycling in the Marion Island terrestrialecosystem. Polar Record 18, 361-370.

SMITH , V. R. & STEENKAMP, M. (1990). Climatic change andits ecological implications at a subantarctic island.Œcologia 85, 14-24.

SOULÉ, M. E., BOLGER, D. T., ALBERTS, A. C., WRIGHT, J.,SORICE, M. & HILL , S. (1988). Reconstructed dynamics ofrapid extinctions of chaparral-requiring birds in urban habitatislands. Conservation Biology 2, 75-92.

STERNER, J. D. & BARRETT, R. H. (1991). Removing feralpigs from Santa Cruz Island, California. Transactions of theWestern Section of the Wildlife Society 27, 47-53.

STORER, T. I. (1962). Pacific island rat ecology. Bulletin of theB.P. Bishop Museum, Hawaii 225, 1-274.

TABOR, R. (1983). The wildlife of the domestic cat. ArrowBooks, London.

TABORSKY, M. (1988). Kiwis and dog predation: observationsin Waitangi State forest. Notornis 35, 197-202.

TAYLOR , R. H. (1968). Introduced mammals and islands:priorities for conservation and research. Proceedings ofthe New Zealand Ecological Society 16, 13-16.

TAYLOR , R. H. (1979a). How the Macquarie island parakeetbecame extinct. New Zealand Journal of Ecology 2, 42-45.

TAYLOR , R. H. (1979b). Predation on Sooty terns at RaoulIsland by rats and cats. Notornis 26, 199-202.

TAYLOR , R. H., KAISER, G. W. & DREVER, M. C. (2000).Eradication of Norway rats for recovery of seabird habitaton Langara Island, British Columbia. RestorationEcology 8, 151-160.

TAYLOR , R. H. & THOMAS, B. W. (1989). Eradication ofNorway rats (Rattus norvegicus) from Hawea Island,Fjordland, using Brodifacoum. New Zealand Journal ofEcology 12, 23-32.

TAYLOR , R. H. & THOMAS, B. W. (1993). Rats eradicatedfrom Rugged Breaksea island (170 ha), Fjordland, New-Zealand. Biological Conservation 65, 191-198.

THOMAS, M. B. & WILLIS , A. J. (1998). Biocontrol - riskybut necessary? Trends in Ecology & Evolution 13, 325-329.

THOMPSON, H. V. (1953). The grazing behaviour of the wildrabbit (Oryctolagus cuniculus). British Journal ofAnimal Behaviour 1, 16-19.

TODD, L. (1977). Cats and commerce. Scientific American237, 100-107.

TOWNS, D. E. A. (1997). Restoration of New Zealandislands: redressing the effects of introduced species.Pacific Conservation Biology 3, 99-124.

TOWNS, D. R. & BALLANTINE , W. J. (1993). Conservationand restoration of New-Zealand island ecosystems. Trendsin Ecology and Evolution 8, 452-457.

TOWNS, D. R., DAUGHERTY, C. H. & CREE, A. (2001).Raising the prospects for a forgotten fauna: a review of10 years of conservation effort for New Zealand reptiles.Biological Conservation 99, 3-16.

TRÉHEN, P., VERNON, P., DELETTRE, Y. & FRENOT, Y.(1987). Organisation et dynamique des peuplementsdiptérologiques à Kerguelen. Mise en évidence demodifications liées à l'insularité. C.N.F.R.A. 58, 241-253.

TURBOTT, E. G. (1948). Effects of goats on Great Island,Three Kings, with description of vegetation quadrats.Records of the Auckland Institute Museum 3, 253-272.

TYNDALE-BISCOE, C. H. (1994). Virus-vectoredimmunocontraception of feral mammals. Reproduction,Fertility & Development 6, 281-287.

VAN AARDE, R. J. (1980). The diet and feeding behaviour ofthe feral cat, Felis catus, at Marion Island. South AfricanJournal of Wildlife Research 10, 123-128.



VAN AARDE, R. J. (1986). A case study of an alien predator(Felis catus) introduced on Marion Island: selectiveadvantages. South African Antarctic Research 16, 113-114.

VAN AARDE, R. J. & SKINNER, J. D. (1981). The feral catpopulation at Marion Island: characteristics, colonization andcontrol. In Colloque sur les Ecosystèmes Subantarctiques,vol. 51, pp. 281-288. C.N.F.R.A., Paimpont.

VAN RENSBURG, P. J. J., SKINNER, J. D. & VAN AARDE, R.J. (1987). Effect of feline panleucopenia on the populationcharacteristics of feral cats on Marion Island. Journal ofApplied Ecology 24, 63-73.

VEITCH, C. R. (2001). The eradication of feral cats (Feliscatus) from Little Barrier Island, New Zealand. NewZealand Journal of Zoology 28, 1-12.

VEITCH, C. R. & BELL, B. D. (1990). Eradication of introducedanimals from the islands of New Zealand. In Ecologicalrestoration of New Zealand islands, vol. 2. ConservationScience Publication (ed. D. R. Towns, C. H. Daughertyand I. A. E. Atkinson). Department of Conservation,Wellington.

VEITCH, C. R. & CLOUT, M. (2001). Human dimensions inthe management of invasive species in New Zealand. In TheGreat Reshuffling (ed. J. McNeely). I.U.C.N.

VERDIER, Y., CHAFFAUX, S., ARTOIS, M. & BOUE, F. (1999).Characterization of sperm antigens for use as a target forimmunocontraception in the red fox (Vulpes vulpes). HumanReproduction 14, R112.

VERNON, P., VANNIER, G. & TRÉHEN, P. (1998). Acomparative approach to the entomological diversity of polarregions. Acta Oecologica 19, 303-308.

VITOUSEK, P. M., DANTONIO, C. M., LOOPE, L. L.,REJMANEK, M. & WESTBROOKS, R. (1997a). Introducedspecies: a significant component of human-caused globalchange. New Zealand Journal of Ecology 21, 1-16.

VITOUSEK, P. M., MOONEY, H. A., LUBCHENCO, J. &MELILLO , J. M. (1997b). Human domination of Earth'secosystems. Science 277, 494-499.

WALL , D., MOONEY, H., ADAMS, G., BOXSHALL , G.,DOBSON, A., NAKASHIZUKA , T., SEYANI , J., SAMPER, C.& SARUKHÀN , J. (2001). An international BiodiversityObservation Year. Trends in Ecology & Evolution 16, 52-54.

WALLIS , F. P. & JAMES, I. L. (1972). Introduced animal effectsand erosion phenomena in the Northern Urewera Forests.New Zealand Journal of Forestry 17, 21-36.

WARNER, R. E. (1963). Recent history and ecology of theLaysan duck. Condor 65, 2-23.

WEIMERSKIRCH, H., ZOTIER, R. & JOUVENTIN, P. (1989).The avifauna of the Kerguelen Islands. Emu 89, 15-29.

WEST, E. W. & RUDD, R. L. (1983). Biological control ofAleutian Island Arctic fox - a preliminary strategy.

International Journal for the Study of Animal Problems4, 305-311.

WESTERMANN, J. H. (1953). Nature preservation in theCaribbean. Uitgaven van der NatuurwetschappellijkeStudiekring voor Suriname in der Nederlandse Antillen9, 1-106.

WILES, G. J., SCHREINER, I. H., NAFUS, D., JURGENSEN,L. K. & MANGLONA, J. C. (1996). The status, biology,and conservation of Serianthes nelsonii (Fabaceae), anendangered Micronesian tree. Biological Conservation76, 229-239.

WILLIAMSON , M. (1996). Biological invasion. Chapmanand Hall.

WILLIAMSON , M. (1999). Invasions. Ecography 22, 5-12.WOOD, B., DONLAN, C. J., TERSHY, B. R., HERMOSILLO,

M. A., SANCHEZ, J. A., KEITT, B. S. & CROLL, D. A.(2001). Non-native mammal removals from islands inNorthwest Mexico. In Eradication of islands Invasives:practical actions and results achieved (ed. M. Clout),pp. 43, University of Auckland, New-Zealand.

WOOD, B. J. (1985). Biological control of vertebrates - areview, and an assessment of prospects for Malaysia.Journal of Plant Protection Tropics 2, 67-79.

YOUNG, C. (1981). Rabbit eradication on islands of theWestern Australian coast. Swans 11, 13-16.

YOUNG, T. P. (2000). Restoration Ecology andConservation Biology. Biological Conservation 92, 73-83.

ZAVALETA , E. S., HOBBS, R. & MOONEY, H. A. (2001).Maximizing the benefits of eradication: Why invasivespecies removal should be viewed in a whole-ecosystemcontext. Trends in Ecology & Evolution.

mammal invaders on islands: impact, control and control...

Documents