ERADICATION OF BLACK RATS FROM ANACAPA ISLAND: BIOLOGICAL AND SOCIAL CONSIDERATIONS

GREGG R. HOWALD1, KATE R. FAULKNER2, BERNIE TERSHY1, BRADFORD KEITT1, HOLLY GELLERMAN1, EILEEN M. CREEL1, MATTHEW GRINNELL1, STEVEN T.

ORTEGA3 AND DONALD A. CROLL1

1Island Conservation, 100 Shaffer Road, Santa Cruz, CA 95060; ghowald@islandconservation.org 2Channel Islands National Park, 1901 Spinnaker Drive, Ventura, CA 93001

3Golden Gate National Recreation Area, Fort Mason, Bldg. 201, San Francisco, CA 94123

Abstract—The use of rodenticides to eradicate rats (Rattus spp.) from islands is a powerful tool forpreventing extinctions of other species. This tool has been underutilized in North America where therehave been less than 10 eradications. Furthermore, rat eradications in North America have deployedrodenticides by hand rather than aerial application from a helicopter as is common elsewhere. This limitseradication attempts to small islands with relatively flat topography. Aerial broadcast of a rodenticide wasthe only feasible method to eradicate the introduced Rattus rattus from the three islets which make up the296 hectare Anacapa Island. After two years of planning, testing and monitoring, a 25-ppm brodifacoumbait was aerially applied to East Anacapa in December 2001, and to Middle and West Anacapa Island inNovember 2002. No rats have been detected anywhere on Anacapa since the application. Extensivemitigation measures, including holding native deer mice in captivity, and capturing and relocating birds ofprey, were adopted to minimize impacts to non-target species. Nonetheless, there were short-term impactsto non-target individuals. Preliminary ecological monitoring is showing substantial recovery of Xantus’smurrelet (Synthliboramphus hypoleucus), Anacapa deer mouse (Peromyscus maniculatus anacapae), andother native species. This was the first aerial application of rodenticide to eradicate rats in North Americaand the first on an island with an endemic rodent. Individuals and organizations opposed to the rateradication attempted to halt the project through a legal challenge, media coverage and political pressure.Knowledge and support of the project by mainstream environmental groups and conservation biologistswas critical to maintaining support. Hopefully, the success of this project will facilitate future rateradications in North America.

Keywords: aerial broadcast, brodifacoum, deer mouse, rat eradication, Rattus rattus, restoration

INTRODUCTION

Island ecosystems are key areas forconservation because they are critical habitat forseabirds and pinnipeds that use thousands of squarekilometres of open ocean, but depend on islands forbreeding, raising young, and resting. In addition,islands tend to be rich in endemic species and arehome to 15–20% of all plant, reptile, and birdspecies (Whittaker 1998). Unfortunately, islandshave been disproportionately impacted by humans.Approximately 55–67% of recorded animalextinctions have occurred on islands, with most ofthese extinctions, including more than half of all

seabird extinctions, caused by invasive species(Island Conservation analysis of IUCN Global RedList Data and WCMC 1992).

Rats in the genus Rattus have been introducedonto about 82% of the world’s islands and/or islandchains (Atkinson 1985), where they frequently havea quantifiable negative impact on the distributionand abundance of native flora and fauna. This ismost pronounced on oceanic islands where nativespecies have evolved in the absence of mammalianpredators and thus have limited behavioural,morphological, and life-history defences againstrats (Brown 1997). Consequently, rats have beenimplicated in 40–60% of recorded bird and reptile

300 HOWALD ET AL.

extinctions since 1600 (Groombridge 1992). Thus,rats have been a focus of land managers as keyspecies for removal from islands.

Anacapa Island has been subjected tointroduced cats, sheep, rabbits and black rats(Rattus rattus). The black rat was introduced toAnacapa Island off the coast of southern Californiain the Channel Islands National Park, sometimeprior to 1939 (Sumner and Bond 1939), probably insupplies transported onto the island for sheepranching or lighthouse construction, or from a shipwreck (Collins 1979) such as the Winfield Scott in1853. The black rat was the last non-native mammalon the island and was thought to have a majornegative impact on seabirds, such as Xantus’smurrelet (Synthliboramphus hypoleucus; H. Carterpers. comm.), reptiles and amphibians (Collins1979, Towns 1991), intertidal invertebrates(Erickson and Halvorson 1990, Navarrete andCastilla 1993), terrestrial invertebrates (Ericksonand Halvorson 1990) and vegetation (Erickson andHalvorson 1990, G. Howald pers. observ.).

The Channel Islands National Park (NPS) hadlong been aware of the presence of rats on AnacapaIsland. Introduced rats were a nuisance to NPSstaff living on the island and to campers due todamage to food and equipment. The ecology of therats and impacts in the Anacapa island ecosystemand methods to control and/or eradicate rats werethe subject of a few studies (Collins 1979, Ericksonand Halvorson 1990, Howald et al. 1997). TheNPS initiated a rat control program on EastAnacapa in the late 1980s through the early 1990susing a combination of bait containing therodenticide warfarin and trapping. The controlprogram was stopped due to the long-term cost andtime investment of perpetual control.

In 2001, we implemented a program toeradicate rats from Anacapa Island followingtechniques developed in New Zealand andelsewhere. Compressed grain pellets containing thesecond generation anticoagulant rodenticidebrodifacoum (3- [3- (4’-bromobiphenyl- 4-yl)- 1, 2,3, 4-tetrahydro-1-naphthyl]- = 4- hydroxycoumarin)were distributed by helicopter and by hand in allpotential rat territories on Anacapa. The steepnessand ruggedness of the island dictated the need foraerial application of the bait. This project is uniquein that it was the first island rat eradication usingaerial broadcast of a rodenticide in North America,

and one of the world’s first with a native rodentliving sympatrically with introduced rats.

PLANNING AND ENVIRONMENTAL COMPLIANCE

By about 1985, conservationists refinedtechniques and a general approach to eradicatingrats from islands. As of 2002, rats have beensuccessfully removed from over 200 islandsworldwide by delivery of bait containing ananticoagulant rodenticide into every potential ratterritory on the island, a fundamental requirementfor successful eradication (see Taylor and Thomas1989, Kaiser et al. 1997, Tershy et al. 1997). In1996, the Channel Islands National Park contractedwith Island Conservation to assess the feasibilityof, and to provide options for, eradicating rats fromAnacapa Island. The resulting report formed thebasis of the Anacapa Island Restoration Project(Tershy et al. 1997).

Formal compliance with the NationalEnvironmental Policy Act (NEPA) began inNovember 1999. The Park prepared anEnvironmental Impact Statement (EIS) due to thepotential for the project to have short-term impactson the environment and the potential controversyover those impacts. The process of public commenton the Draft EIS indicated support for the purposeof the project (i.e., the eradication of rats to restoreseabirds and other native species) and concernswith the projected impacts to non-target species,particularly Anacapa deer mice (Peromyscusmaniculatus anacapae), birds, reptiles, andamphibians. There were requests for the NPS tochoose a different method to achieve eradication ofrats. However, after the review of the trial data andconsultation with other eradication experts, theNPS and Island Conservation felt strongly thatonly the chosen alternative offered a reasonableprobability of eradicating rats and that any negativeimpacts would be short-term and not significant tonative populations. The Record of Decision for theNEPA document was signed by the NPS on 17November 2000, outlining the preferred alternativeof aerial broadcast of the rodenticide brodifacoumto eradicate rats from Anacapa Island.

The planned eradication of rats from AnacapaIsland required compliance with a host of additionalenvironmental laws. One of the more significant

BLACK RAT ERADICATION 301

requirements was the need to obtain a registrationfor an aerial broadcast bait from the EnvironmentalProtection Agency (EPA) in compliance with theFederal Insecticide, Fungicide and Rodenticide Act.The EPA evaluated the benefits of the rat removalagainst the potential short- and long-term negativeimpact of the project and determined that the projecthad a high probability of eradicating rats with min-imal short-term negative impacts in the ecosystem.A three-year, Quarantine Exemption registration foran aerial broadcast bait containing 25-ppm brodifa-coum was granted on 12 December 2000.

METHODSTiming and Bait Application

The major logistical obstacles to eradicatingrats from Anacapa were steep and ruggedtopography, presence of breeding seabirds limitingaccess, and presence of native species, in particularlandbirds and the endemic Anacapa deer mouse, anative rodent which was attracted to and competedwith rats for the bait. These limitations requireddevelopment of novel techniques to limit thepotential environmental impact while ensuring thatenough bait would be delivered into every potentialrat territory.

The primary objective of the bait applicationwas to ensure that sufficient bait was available inevery potential rat territory on the island. If anyareas of the island were not treated, or treatedinsufficiently, there was the potential that somerats may not have been exposed to the bait and theeradication could fail. Therefore, it was critical thatthe entire island, especially the cliff sides, wasbaited sufficiently. Monitoring of the aerialapplication was necessary to identify any areas thatwere either not treated, or received too little bait.

Rat eradication is more likely to be successfulwhen the rat population is food stressed, andtherefore more likely to eat the toxic bait, and whenthe population is declining and not breeding (D.Veitch pers. comm.). The best time to apply bait onAnacapa was during the late dry season (typicallySeptember–November), as the rat population hadbeen shown to be declining during that period andrats were not likely to be breeding (Erickson andHalvorson 1990).

We implemented the eradication program inNovember 2000 with a 2.5-hectare trial in Sheep

Canyon on Middle Anacapa, followed by thebaiting of East Anacapa in December 2001 andMiddle/West Anacapa in November 2002. At eachstep, we monitored for efficacy and determinedthat the aerial bait application was successful inkilling all marked rats (see below). In addition, wemonitored for environmental impact to validatethat the mitigations were successful and theenvironmental effects were limited and within thescope predicted in the EIS.

The delivery of bait into all potential ratterritories, especially on the cliff sides, was criticalto the successful removal of rats from Anacapa.Rats were known to regularly use and live on thecliffs (Erickson and Halvorson 1990, Howald et al.1997) but the friable basalt that forms the steep 50–284-m cliffs precluded baiting by hand. Thus, onlyan aircraft could ensure delivery of bait onto thecliffs. Bait was broadcast at 15 kg/ha from ahopper suspended under a Bell 206 helicopterfitted with an onboard differential GPS (DGPS)and computer to ensure even and completeapplication of bait. The cliffsides were treated withthe hopper fitted with a deflector that spread bait toone side only, preventing significant bait spreadinto the marine ecosystem. Bait was applied byhand to the lower reaches of the cliffs, above thebeaches and the intertidal zone, ensuring adequatebait application in this prime rat habitat. Theremainder of the island was treated with thedeflector removed, so that bait was spread in a 360degree pattern. Bait stations were used in bufferzones around the landing cove, the buildings andthe campground on East Anacapa Island to ensurethat bait was always available to the rats in theseareas of many alternative food resources.

We used a large (approximately 1–2 g, 9-mmdiameter) green, unwaxed, compressed grain,pelleted bait containing 25-ppm brodifacoum thatwas highly palatable to rats (based on Bell Labslaboratory tests; field test in 2000; and field labtests in 2001). Brodifacoum blocks the vitamin K1oxidation-reduction cycle in the liver and impairsthe ability to produce active clotting proteins; deathfrom internal hemorrhaging occurs within 3–10days after ingestion of the bait (Smith and Greaves1987). The major benefit of brodifacoum is that itis potentially toxic to rats after a single feeding andthe delayed onset of symptoms prevents thedevelopment of bait shyness (Smith and Greaves

302 HOWALD ET AL.

1987), factors offering a high probability of killing100% of the rat population.

Effect on Rat PopulationWe used radio telemetry, live trapping, and

peanut-flavored wax indicator/ chew blocks toassess the efficacy of the baiting. In addition,studies on the ecological impact of rats on nestingseabirds and intertidal invertebrates are ongoingand function as additional indicators of ratpresence/absence. Findings from the ecologicalstudies will be reported elsewhere.

Eighteen rats (eight in 2001, 10 in 2002) wereradio-collared prior to bait application. The ratswere monitored at least every 2–3 nights prior toand after the bait application. We considered radio-collared rats to have died if the radio signalindicated no movement for three consecutivenights after the bait was applied. When possible,burrows were dug up and rats were collected forconfirmation of cause of death. Live trapping(Tomahawk Live Traps) to index the rat populationbefore and after the baiting was conducted in areasof known prime rat habitat. Traps were placedalong transect lines of various lengths, withspacing of 15–25 m as dictated by islandgeography. Traps were baited with peanut butter.We used a peanut flavored wax chew block as analternate indicator of rat presence/ absence. Rats ormice chewed on blocks, leaving behind animpression of their incisors. The difference inincisor widths was used to distinguish between thetwo species. Blocks were staked to the ground at10–25-m spacing, along transect lines of variouslengths as dictated by island geography.

Minimizing Risks to Native WildlifeWe realized that aerial broadcast of a grain-

based bait containing brodifacoum would placesome individuals of native birds and mice at risk ofpoisoning either directly through consumption ofbait pellets (primary poisoning), or indirectlythrough scavenging poisoned animals (secondarypoisoning). We therefore considered mitigationmeasures for those species for which baiting mayhave a population level or long-term impact.

Anacapa Deer MouseThe presence of the endemic Anacapa deer

mouse presented a unique challenge to rat

eradication; how to deliver the bait into every ratterritory without making it available to everymouse. Mice across the three islets were found to begenetically similar and functioning as ametapopulation with some movement of micebetween the islets (Pergams et al. 2000). PopulationViability Analysis (PVA) determined that apopulation of 1,000 mice across all three isletswould be necessary to maintain populationpersistence with minimal loss of genetic diversity(Pergams et al. 2000). The mitigation plan for theAnacapa deer mouse was based on the geneticconnectivity of the Anacapa islets and the PVA.

To insure that we always had two viablepopulations of mice, we chose a mitigation planthat included (1) staggering the eradication overtwo years so that a free-ranging population alwaysexisted on one or more islets and, (2) holding arepresentative sample of the mouse population incaptivity followed by a soft release (withsupplemental food) and subsequent monitoring toensure viability. To accomplish this, we conducteda limited captive holding experiment of EastAnacapa Island deer mice from October 2001–April 2002 to refine our methods and logistics forcaptive holding during the Middle and WestAnacapa rat removal planned for 2002. We thenreleased the mice to East Anacapa in April of 2002at different release densities, and monitored therecovery of the mouse population.

Prior to the eradication of rats from Middle andWest Anacapa Island we translocated >1000 micefrom Middle and West Anacapa Island to EastAnacapa Island. Furthermore, we supplementallyfed free-ranging mice on East Anacapa through thewinter, by hand broadcasting commercial graderodent chow within 20 m either side of the EastAnacapa ring trail. Finally, we held >700 micefrom Middle and West Anacapa in captivity fromSeptember 2002 through March 2003.

The early spring was the optimum time torelease mice because it was the start of thebreeding season and natural food was abundant.We released the captive mice back to Middle andWest Anacapa in late March 2003 and monitoredthe mouse population on all three of the Anacapaislets monthly through the 2003 breeding season.

LandbirdsAll bird species known to use Anacapa Island

BLACK RAT ERADICATION 303

were also found on the other northern ChannelIslands or on the adjacent mainland. Even thoughthe eradication of rats would not cause significantimpacts to bird species, we made a large effort toreduce the likelihood of poisoning individual birdsby limiting their exposure to the rodenticide by (1)timing the bait application to occur when some ofthe migratory birds had moved off the island forthe winter, and before the initiation of breeding bygulls, pelicans and other seabirds (limiting thepotential for exposure and disturbance), (2)maximizing the size of bait pellets to preventgranivorous birds from ingesting the bait, (3)coloring the pellets green to make them lessvisually attractive to gulls and other birds (Day andMatthews 1999, H. Gellerman unpubl. data fromAnacapa Island), (4) designing the bait to breakdown within days of application in the moistmaritime climate, (5) not applying an excessiveamount of bait, but ensuring enough was availableto all rats on the island, and, (6) live trapping andremoving, or holding captive, resident birds of preyprior to the baiting. Captive birds were re-releasedonto the island after the risk period had passed.

We evaluated the impact to non-target speciesby conducting searches for carcasses after thebroadcast. We actively searched for carcasses ofnative animals across all three Anacapa isletsbetween November 3rd (24 hours after the baitapplication) and December 17th (45 days post-primary application). We searched 46 circular 10-mdiameter plots on days 3, 7–8, 10, 14, 21, 28, and 35post-application. (We added random transects afternative bird carcasses were found opportunisticallyand few animals had been found in the plots.)Random transects were walked on MAI and WAIon days 19, 21, 24, 28, 35, and 82 post-application.All birds, rats and mice found (in plots) werecollected, placed into a pre-labeled bag and frozen.Rats and mice found dead outside the monitoringareas were buried or tucked under dense brush. Birdcarcasses were sent to the Illinois Department ofAgriculture Centralia Animal Disease Laboratoryfor necropsy and toxicological analysis. A total of168 person-search-hours were spent looking fornon-target animals, plus the time the field crewspent opportunistically searching for carcasses.

Environmental MonitoringTo evaluate the toxicological risks and impact

to non-target species, we studied the movement ofbrodifacoum in both the marine and terrestrialecosystems. In the marine environment, bait wasnot expected to enter the marine environment inany significant quantities because of a specially-designed deflector mounted under the hopper. Weassessed bait drift by sending SCUBA divers tolook for and count bait pellets on the ocean floor,to assess the degradation of the bait, makeobservations of fish or invertebrates in the area,and to determine if any organisms were consumingthe bait. The divers investigated these areas on twodives 1.5 and 5 hours post bait-drop. We collectedocean water samples (500 ml) 24 and again 48hours post bait-drop. The samples were held in arefrigerator and shipped to the CaliforniaDepartment of Fish and Game, PesticideInvestigation Unit, Rancho Cordova, California forbrodifacoum residue analysis. In addition, wesampled marine organisms as indicators for thepresence/absence of brodifacoum in the marineenvironment. We collected mussels (Mytiluscalifornianus) and crabs (Pachygrapsus crassipes)at days ~15 days and 30 post-application, andtidepool sculpins (Oligocottus maculosus) at days~15, 30 and 90 post-application. The samples werecollected individually or pooled into pre-labeledplastic sample vials, placed on wet ice and frozenas soon as possible after collection.

In the terrestrial environment, native specieswere at risk of exposure to the rodenticide byconsuming the bait directly, and via contaminatedprey items (secondarily) if brodifacoum moved intothe food chain. We qualified and quantified themovement of brodifacoum into the Anacapaecosystem and evaluated its impact to non-targetanimals. The Anacapa insect and microbialcommunities play a critical role in the nutrient cyclethrough the breakdown of organic matter. Similarly,the insects and microbes ultimately removed theorganic rodenticide from the environment bydigesting residual bait pellets and rodent carcasses,and ultimately breaking brodifacoum down to itsnon-toxic base components of water and CO2(Shirer 1992). Invertebrates are not affected by therodenticide because they lack a closed circulatorysystem and brodifacoum does not persist ininvertebrate tissue (e.g., Payne et al. 2000), thus, thepresence of brodifacoum in invertebrates isindicative of a recent exposure to the rodenticide.

304 HOWALD ET AL.

Therefore, the residues in invertebrate tissue can beused as an index of the biologically availablebrodifacoum in the Anacapa environment.Similarly, brodifacoum levels in the soil, sampledover time, can be used as an index of rate ofbrodifacoum breakdown from the Anacapaenvironment. We collected samples of theinvertebrate community found under rocks, shrubs,and under the grid of boards used for monitoringreptiles and amphibians. We conducted randomsampling of the invertebrate community versussampling specific groups. When a diverse samplewas unavailable (especially during the dry season),we focussed on the available insects, primarilyisopods and ants. We collected five samples (~5 geach) at approximately days 5, 30, 90, and 180 post-application. The samples were collected intochemically cleaned, sterile plastic sample jars andfrozen. Further, we collected three soil sampleseach at approximately 1, 3, 6 and 12 months post-broadcast. We sampled and analyzed the top 5 cm ofsoil collected from under a pellet or arbitrarilywithin a bait uptake monitoring plot.

Legal Challenge and Public ControversyDuring the summer and fall of 2001,

opposition to the project was voiced by severalindividuals and animal rights groups. Primaryissues raised were the killing of animals (rats aswell as non-target mice and birds) and the use of apoison. Shortly before the planned application ofbait to East Anacapa Island, the NPS was notifiedby the Fund for Animals, a nationwide animalrights organization, that it intended to file alawsuit, and would request a temporary restrainingorder prohibiting implementation of the project inthe United States District Court for the District ofColumbia. The Federal Government agreed not toimplement the bait application for one month, aftera hearing and until a ruling was made by the judge.

The plaintiffs in the lawsuit were The Fund forAnimals, Channel Islands Animal ProtectionAssociation, and two residents of Santa Barbara.The Plaintiffs alleged that “…defendants’ plan tospray rodenticide-laced pellets by helicopter ontoAnacapa Island in the Channel Islands NationalPark is arbitrary and capricious, in violation of theMigratory Bird Treaty Act,…the National ParkService Organic Act,…the Park Service’s ownManagement Policies, and the National

Environmental Policy Act…” (MemorandumOpinion, U.S. District Court for the District ofColumbia, Judge Ellen Segal Huvelle, 29November 2001).

The most substantive claim was the allegationof non-compliance with the Migratory Bird TreatyAct (MBTA). The MBTA prohibits the taking ofmigratory birds except under a permit issued by theU.S. Fish and Wildlife Service (FWS). The FWSregulations for management of the MBTA defined“take” of migratory birds to be pursuit, hunting,killing, or capture of covered species. FWS did notconsider the unintentional mortality of a migratorybird to fall within the permitting requirements ofMBTA. Nonetheless, subsequent to the lawsuit, theNPS decided to apply for an MBTA permit fromthe FWS. The FWS granted the NPS’ applicationfor the MBTA permit due to the benefits theAnacapa Island project would have for migratorybirds. The federal government believed that allother allegations were without substance.

On 29 November 2001, the District Courtissued a “Memorandum Opinion” denying theplaintiff’s request for a preliminary injunction. Thecourt concluded that “…plaintiffs’ showing on themerits is not sufficient to warrant the entry ofinjunctive relief…” (Memorandum Opinion, U.S.District Court for the District of Columbia, JudgeEllen Segal Huvelle, 29 November 2001).Interestingly, on the question of whether theAnacapa project required an MBTA permit, thejudge wrote “The only cases on this subject haveconcluded that unintentional or incidental take byfederal agencies do not violate the MBTA.” Thejudge’s ruling cleared the way for the initiation ofPhase I of the Anacapa Island restoration project.The lawsuit greatly heightened the interest of thepress in the project and resulted in substantialmedia coverage of the project throughout allremaining phases.

Project FundingThe primary financial support for the

eradication of rats from Anacapa Island and thepre- and post-eradication monitoring came fromthe American Trader Trustee Council (ATTC). TheATTC was formed as a result of an oil spill fromthe tanker American Trader near HuntingtonBeach in 1990 and consisted of representatives ofthe FWS, National Oceanic and Atmospheric

BLACK RAT ERADICATION 305

Administration, and the California Department ofFish and Game. The ATTC considered theeradication of rats from Anacapa to be a projectthat would significantly contribute to therestoration of seabird species that had been injuredby the oil spill.

RESULTS AND DISCUSSIONTiming and Bait Application

The DGPS data proved to be an effectivemonitoring technique especially for cliff sides thatcould not be validated visually (Fig. 1). The use ofthe DGPS had two major advantages: (1) it visuallyshowed what areas had been baited, and what areasneeded to be baited, and (2) it identified gaps in thebaiting that could support rats and needed to befilled in by hand application or by helicopter. Theaerial application thoroughly treated the islets withminimal gaps.

Effect on Rat PopulationRadio collars—All radio-collared rats stopped

moving between days 3 and 14 post-application(mean 6.3 days), coinciding with the known timelag between ingestion of brodifacoum and death(Taylor 1993, Howald et al. 1999). This suggeststhat the majority of rats began feeding on the baitwithin a few days after application. In 2002, nineradio-collared rats were recovered dead. Asexpected, the majority (6 of 9) died below groundwhere they were inaccessible to avian scavengers.Two were found dead above ground, one of whichhad been scavenged or preyed upon. These results

are consistent with other field studies evaluatingthe fate of anticoagulant poisoned rodents (Fenn etal. 1987, Taylor and Thomas 1989, 1993, Howaldet al. 1999).

Trapping and wax chew blocks—Trap successfluctuated between 10–55.8% prior to the baitapplication. After the bait application, no rats weretrapped on East Anacapa Island (>11,281 TrapNights; Figs. 2 and 3). As of November 2004trapping and chew block results from East, Middleand West Anacapa Island have failed to detect anyrats.

Taken together, the above indicators ofefficacy indicate that rats have been removed fromAnacapa Island. As of November 2004, after twocomplete rat breeding seasons (late winter throughearly summer), no rats had been detected, andAnacapa Island was declared rat free.

Minimizing Risks to Other SpeciesAnacapa deer mouse—The mitigation strategy

for the Anacapa deer mouse was successful. Notunexpectedly, it appeared that all Anacapa deermice present on an islet during bait applicationwere killed (Fig. 4). The staggering of theeradication over two years ensured a free-rangingpopulation existed on Middle and West AnacapaIslands prior to the baiting of East Anacapa Island.The reintroduction of mice to East Anacapa Islandin 2002 was successful in establishing a free-ranging population on the islet prior to theapplication of bait onto Middle and West AnacapaIsland in the fall of 2002 (Fig. 4). Conditions forthe release of mice and throughout the breedingseason in 2003 were excellent. Above averagerainfall through the spring and summer resulted inan abundance of food and protective cover. Mousedensities, on Middle Anacapa, measured onmultiple trapping grids approximately 6.5 monthsafter release, were comparable to densitiesmeasured prior to the eradication.

Monitoring is ongoing to ensure that thepopulation has reestablished, and if necessary wewill intervene to support the mouse population.The recovery of the Anacapa deer mouse hasdemonstrated that rats can be eradicated fromislands while concurrently occupied by nativemammals.

Birds of prey—The Predatory Bird ResearchGroup removed 37 individual birds of prey

Figure 1. Final Differential GPS/computer link printout of thebait spread by helicopter on East Anacapa Island, 2001. Therectangular shapes represent the flight lines of the helicopterwhile broadcasting bait. The width of the flight lines representshow far and wide bait was spread onto the island. Note that theflight lines follow along the shoreline.

306 HOWALD ET AL.

including 13 of 19 from East Anacapa in 2001 and24 of 40 raptors known to be present on Middleand West Anacapa in 2002. This represented 63%of the resident and over-wintering raptorpopulation prior to and just after the baitapplication (Table 1). Most birds were released onthe mainland in suitable habitat and the peregrinefalcons were held and released back onto AnacapaIsland after the risk period (estimated atapproximately 2–3 weeks). It was decided tocaptively hold the peregrine falcons because of thepast intensive efforts to restore their populations onthe Channel Islands. Thus, in an attempt to ensurethat peregrine falcons remained on Anacapa, theywere captively held and released back onto theisland.

The fate of the remaining birds of prey on theisland is unclear. There is evidence that some birdssurvived the bait application. A burrowing owl wasconsistently observed on West Anacapa for almosttwo months after the bait application, long after allrodents had died. However, three barn owls, sixburrowing owls and an American kestrel eitherdied while in captivity or were found dead on theisland. The American kestrel and a burrowing owlthat were captured in 2001, after the baitapplication, likely died from brodifacoumpoisoning. Our data indicate that there was animpact on individual raptors, but that the impactwill not be long term. All of the species originally

removed from the island have returned. Onebanded pair of peregrine falcons was observed tobreed successfully and raised at least one chick onMiddle Anacapa in the spring of 2003 (N. Toddpers. comm.), suggesting that the mitigation wassuccessful and with no lasting impact.

The impact to the population of raptor specieswas negligible. The birds of prey on the ChannelIslands are generally habitat limited (B. Waltonpers. comm.) and any territory left open was filledby other birds (B. Latta pers. observ.). Thediversity of predatory birds observed prior to thebaiting was observed on Anacapa Island in the fallof 2003, suggesting that project activities had nolasting impact on populations and raptorpopulation diversity has returned to pre-eradicationlevels.

Landbirds—Some incidental poisoning of non-target granivorous birds was expected. Themitigation efforts reduced the exposure of someindividuals, however it was impossible to reducethe risk to all individuals. The diversity andforaging strategies of birds found dead after thetreatment is consistent with primary exposure(granivorous birds) and secondary exposure ofpredatory birds (which prey or scavenge rodents),and is consistent with reports from other ratremoval projects (Empson and Miskelly 1999,Howald et al. 1999, McClelland 2001). A total of94 birds (49 in 2001; 45 in 2002) representing 16

Figure 2. Index of rat abundance (percent trap success) on East Anacapa Island, October 2001 through March 2003 (n = 11,281 TN).Rats captured were either euthanized or fitted with radio collars and released.

BLACK RAT ERADICATION 307

species were found dead in the treatment area afterthe application (Table 2). Six additional birds weretoo decomposed to be identified to species. Most ofthe birds were granivorous songbirds and likelypicked up bait fragments soon after bait application.

All birds that could be tested (i.e., enough tissueavailable) likely died from brodifacoum exposure.

The number of birds found dead after the baitdrop, even if increased by several orders ofmagnitude to allow for poor recovery rates by

Figure 3. Index of rat and mouse abundance (incisor marks per 100 block nights) on East Anacapa Island, October 2001 throughMarch 2003.

Figure 4. Anacapa deer mouse densities (# mice per/ha, calculated from a 3-night mark/recapture trap session) on two monitoringgrids (10 x 10; 7-m trap spacing), before and after bait application on East Anacapa Island, 2000–2003.

308 HOWALD ET AL.

searchers, was not significant at the populationlevel for any of the recovered species. Mainlanddensities of white-crowned sparrow (Zonotrichialeucophrys), golden-crowned sparrow (Zonotrichiaatricapilla), and song sparrow (Melospize melodia)are estimated to be 1.94–3.67 birds per hectare.House finch (Carpodacus mexicanus) globalpopulation estimate ranges between 250 million to1 billion individuals. The most abundant breedinglandbird species on the island are orange-crownedwarblers and Bewick’s wren. Both of these speciesas well as the house finch, song sparrow, rockwren, and peregrine falcon were breeding onAnacapa only six months after the bait drop. Basedon these observations the bait application had littlelasting impact on the landbird populations ofAnacapa.

Removing introduced black rats from AnacapaIsland will likely result in improved habitat for allbirds, including birds of prey, due to decreased nestpredation and harassment of nesting birds, andincreased densities of breeding seabirds and deermice. In other rodent eradications on islands, theimpacts of poisoning on populations of nativespecies have been short-term and outweighed bythe long-term benefits of rat removal (Towns 1991,Empson and Miskelly 1999) with native speciesrecovering quickly to pre-eradication levels orhigher (e.g., Davidson and Armstrong 2002).

Environmental MonitoringThe helicopter had a high degree of accuracy

of placing bait onto the island. The cliff sides were

treated with the hopper fitted with a customdesigned deflector to minimize bait being spreadinto the water. Boat and island-based observersreported no bait was directly spread into the ocean.These reports were consistent with observationsmade during calibration trials on the mainland,during the application trial in November 2000 andwith the East Anacapa Island bait application inDecember 2001.

Island-based observers reported a small amountof bait entering the water indirectly from bounce offthe cliff sides. The divers counted a mean of 72 baitpellets (range: 69–75) over 500 m, at a 1–4 m depthon the ocean floor. The pellets were starting todegrade at 1.5-hours post bait-drop, and at fivehours post bait-drop, there were just a few scatteredcrumbs. No fish or other animals were observedfeeding on the bait on either dive. Fish were notexpected to consume the bait (see AIRP EIS 2000)and the observations were consistent withobservations made during the East Anacapa Islandbait application in 2001. No brodifacoum residueswere detected in any of the water samples (n = 2 @24 hours and n = 1 @ 48 hours post-application).No brodifacoum residues were detected in any ofthe invertebrate or fish samples collected.

Terrestrial EcosystemTrace levels of brodifacoum were detected in

three of five invertebrate samples collected five daysafter the bait application on East Anacapa in 2001.Trace levels of brodifacoum were detected in two offive samples on Middle Anacapa Island, five days

Table 1. Summary of birds of prey removed from Anacapa Island prior to bait application, 2001/2002.

Species

East Anacapa 2001 Middle and West Anacapa 2002Number removed/resident

populationInitial population

estimateNumber removed/resident

populationInitial population

estimate

Peregrine falcon 1 of 2 3 7 of 8 8–10

Red-tailed hawk 4 of 6 6 5 of 8 8

Barn owl 2 of 3 3 2 of 6 6

Burrowing owl 3 of 4 2 3 of 6 6

American kestrel 1 of 2 2 2 of 5 4

Northern harrier 1 of 1 0 4 of 5 0

Short-eared owl - - 0 of 1 1

Common raven 1 of 1 2 - -

Merlin - - 1 of 1 0

TOTAL 13 of 19 18 24 of 40 33–35

BLACK RAT ERADICATION 309

after the application in 2002. Brodifacoum wasdetected in one subsequent sample at six monthsafter the application, suggesting that the biologicalavailability and subsequent movement ofbrodifacoum into the food chain was limited to anarrow time period after application. It was likelythat the invertebrates were degrading residual baitpellets that may have been on the island, however,brodifacoum concentration of caged pelletscollected six months after the application haddegraded by 92.2%. No brodifacoum was detected inany of the soil samples analyzed from East AnacapaIsland. Trace levels were detected in one of ninesamples collected at approximately six months postapplication on Middle Anacapa in 2003.

CONCLUSION

Rats have had a profound impact on globalbiodiversity, including within the Channel IslandsNational Park. The eradication of the black ratsremoves the last introduced mammal from AnacapaIsland and will have ecosystem wide benefits.Seabirds and terrestrial species such as deer mice,herpetofauna, landbirds, invertebrates and plantswill directly benefit from the removal. Theeradication of rats from Anacapa has increased the

availability of introduced predator-free seabirdbreeding habitat in southern California, and it isexpected that seabirds will recolonize naturally andideally local populations will increase. Somechanges in the Anacapa ecosystem have alreadybeen measured, including increased numbers ofbreeding Xantus’s murrelets and use of habitatwhere they were previously excluded by rats(Hamer et al. 2003, D. Whitworth pers. comm.). Inthe spring of 2003, within four months after the baitapplication, two Cassin’s auklet (Ptychoramphusaleuticus) nests were found, a seabird that is highlysusceptible to rat predation and never documentedon the island. The nests were found to contain twoyoung downy chicks (D. Whitworth pers. comm.),clearly demonstrating the benefit of eradicatingrats.

The lack of detection of rats after two yearsand the ecological changes on the island post baitapplication, together provide strong evidence thatrats have been successfully removed. However, theboat traffic to and around the island, and the NPSactivities of hauling equipment and supplies toAnacapa, present a risk that rats may bereintroduced to the island. The National ParkService together with Island Conservation haveimplemented procedures to reduce the chance thatrodents will reach the island. We plan to continue

Table 2. Species and numbers of birds found dead after bait application, Anacapa Island, 2001/2002.

Common Name (Scientific name) East Anacapa 2001 Middle and West Anacapa 2002White-crowned sparrow (Zonotrichia leucophrys) 26 7House finch (Carpodacus mexicanus) 2 1Orange crowned warbler (Vermivora celata) 0 1Rock wren (Salpinctes obsoletus) 1 0Western gull (Larus occidentalis) 2 0Spotted towhee (Pipilo maculates) 1 0Song sparrow (Melospiza melodia) 2 4Golden-crowned sparrow (Zonotrichia atricapilla) 7 3Western meadowlark (Sturnella neglecta) 2 11Barn owl (Tyto alba) 1 2American kestrel (Falco sparverius) 1 0Burrowing owl (Athene cunicularia) 4 2Dark-eyed junco (Junco hyemalis) 0 1Varied thrush (Ixoreus naevius) 0 2Hermit thrush (Catharus guttatus) 0 1Fox sparrow (Passerella iliaca) 0 4Unidentified sparrows NA 3Unidentified small birds NA 3

310 HOWALD ET AL.

to refine those techniques and to build capacity torespond to any reintroduction of rodents to theisland to keep the island rat-free. Components ofthe prevention plan include inspection of materialsand supplies before loading onto boats, armed baitstations on board NPS vessels and ongoingmonitoring on the island to detect any rats that mayreinvade the island.

The Anacapa Island Restoration Project wassuccessful because of the team approach andcooperative effort between the staff of the NationalPark Service, American Trader Trustee Counciland Island Conservation at the planning andcompliance phase, and during field work. Thisproject was the first of its kind off the coast of themainland United States. The project hasdemonstrated that rats can be eradicated fromislands while protecting sensitive native species,including endemic rodents. The aerial broadcastapproach to rat eradication is a powerfulconservation tool on topographically complex orvery large islands. We hope that the success of thisproject will facilitate future conservation efforts onother islands off the coast of North America.

ACKNOWLEDGMENTS

We thank the American Trader TrusteeCouncil (J. Boyce [NOAA], C. Gorbics [USFWS]and P. Kelly [CDFG-OSPR]), and the field andadministrative staff of Island Conservation and theNational Park Service without whose dedicationand enthusiasm, this project would have beenimpossible. The following field crews andvolunteers, A. Angel, S. Buckelew, S. Barocio, T.Comendant, J. Donlan, B. Fitz-Earle, M. Grinnell,K. Hulvey, H. Jones, A. Khadun, A. Meyer, S.Morrison (TNC), C. Nishida, B. O’Connor, D.Renoud, A. Samaniego, N. Sanchez, J. Sheppard,R. Wanless, and A. Wilder, spent many months onthe island with sustained enthusiasm in less thanideal conditions. We thank R. Bidwell (NPS) forhis immeasurable on-island project support. V.McDermott kept the whole project running andmade sure that there was always money to buy baitand pay the helicopter bills. The GIS team of N.Biavaschi (IC), C. Schwemm (NPS) and J. Vogel(USGS) made sure we knew where bait was beingspread in near real time. We also thank G. Alaks,

M. Ashley, B. Lacey, and O. Pergams for theconservation genetics and population modeling andguidance in the development of the mousemitigation. B. Badzik (GGNRA) and C. DiSalvo(WASO) provided NPS-IPM oversight. T. Coonan,D. Lerma, D. Kushner, P. Martin, Y. Morales andall the NPS Natural Resource division staff arethanked for their help and support. All the CINPstaff including R. Brooks, D. Brooks, T.Buffington, K. Bullard, D. Burgess, T. Dore, K.Duran, J. Fitzgerald, P. and J. Flynn, T. Glass, D.MacDonald, Y.Menard, M. Morales, R. Nelson, J.Provo, M. Short, D. Stoltz, B. Struble, D. Willey,and I. Williams provided planning, logistical,transportation and island support. In addition, wethank K. Miskel and C. McLaughlin of AspenHelicopters for ensuring that bait was deliveredwith remarkable precision and in so doing, made ahuge contribution to biodiversity conservation inCalifornia, the Predatory Bird Research Group,especially B. Latta, N. Todd and G. Doney forprotecting the birds of prey, and finally D. Veitch(NZDOC), P. McClelland (NZDOC) and B. Bell(WMI) for sharing their experience and advice sogenerously. We remember and thank FredRodriguez for his logistical support and sense ofhumor that will always be missed.

REFERENCES

Atkinson, I.A.E. 1985. The spread of commensalspecies of Rattus to oceanic islands and theireffects on island avifaunas. Pages 35–81. In:Moors, P.J. (ed.), Conservation of island birds.International Council of Bird PreservationTech. Pub. 3.

Brown, K.P. 1997. Predation at nests of two NewZealand endemic passerines; implications forbird community restoration. PacificConservation Biology 3(2):91–98.

Collins, P.W. 1979. Vertebrate zoology: thebiology of introduced black rats on Anacapaand San Miguel Islands. Pages 14.1–14.56. In:Power, D.M (ed.), Natural Resources Study ofthe Channel Islands National Monument,California. Santa Barbara Natural HistoryMuseum, Santa Barbara, CA.

Davidson, R.S. and D.P. Armstrong, 2002.Estimating impacts of poison operations on

BLACK RAT ERADICATION 311

non-target species using mark-recaptureanalysis and simulation modelling: an examplewith saddlebacks. Biological Conservation105:375–381.

Day, T.D. and L.R. Matthews. 1999. Do coloursthat deter birds affect cereal bait acceptance bypossums (Trichosurus vulpecula)?. NewZealand Journal of Ecology 23(2):261–266.

Empson, R.A. and C.A. Miskelly. 1999. The risks,costs and benefits of using brodifacoum toeradicate rats from Kapiti Island, NewZealand. New Zealand Journal of Ecology23(2):241–254.

Erickson, W.A. and W.L. Halvorson. 1990.Ecology and control of the roof rats (Rattusrattus) in Channel Islands National Park.University of California, Cooperative ParkStudies Unit, Technical Report No. 38.

Fenn, M.G.P., T.E. Tew and D.W. Macdonald.1987. Rat movements and control on anOxfordshire farm. Journal of Zoology(London) 213:745–749.

Groombridge, B. 1992. World ConservationMonitoring Centre, British Museum (NaturalHistory), International Union for Conservationof Nature and Natural Resources. Globalbiodiversity: status of the earth's livingresources: a report. Chapman and Hall,London, UK.

Hamer, T.E, S.M. Schuster and H.R. Carter. 2003.Radar monitoring of Xantus’s Murrelets(Synthliboramphus hypoleucus) at AnacapaIsland, California in 2003 after eradication ofblack rats (Rattus rattus). Unpublished report,Hamer Environmental L.P. Mount Vernon,WA, 36 pp.

Howald, G.R., B. Tershy and D.A. Croll. 1997.The ranging movements, distribution, andabundance of the ship rat (Rattus rattus) andthe Deer Mouse (Peromyscus maniculatus) onAnacapa Island, Channel Islands NationalPark, California. Island Conservation, 100Shaffer Road, University of California, SantaCruz, CA, 17 pp.

Howald, G.R., P.Mineau, J.E. Elliot and K.M.Cheng. 1999. Brodifacoum poisoning of avianscavengers during rat control on a seabirdcolony. Ecotoxicology 8:429–445.

Kaiser, G.W., R.H. Taylor, P.D. Buck, J.E. Elliott,G.R. Howald and M.C. Drever. 1997. The

Langara Island Seabird Habitat RecoveryProject: Eradication of Norway Rats (1993–1997). Technical Report Series No. 304.Canadian Wildlife Service. Pacific and YukonRegion, Delta, BC, 90 pp.

McClelland, P.J. 2001. Eradication of Pacific Rats(Rattus exulans) from Whenua Hou NatureReserve (Codfish Island), Putauhinu andRaratoka Islands, New Zealand. Pages 173–181. In: Veitch, C.R. and M.N. Clout (eds.),Turning the Tide: The eradication of invasivespecies, Proceedings of the InternationalConference on Eradication of Island Invasives.Occasional Paper of the IUCN SpeciesSurvival Commission No. 27, 414 pp.

Navarrete, S.A. and J.C. Castilla. 1993. Predationby Norway rats in the intertidal zone of centralChile. Marine Ecology Progress Series92:187–199.

Payne, D.J., D. de L. Brooke, J.K. Finnie and A.Jackson. 2000. Effects of brodifacoum on theland crab of Ascension Island. Journal ofWildlife Management 64(2):380–387.

Pergams, O.R.W., R.C. Lacy and M.V. Ashley.2000. Conservation and management ofAnacapa Island Deer Mice. ConservationBiology 14(3):819–832.

Shirer, M. 1992. In poison’s defence. Terra Nova17:3.

Smith, R.H. and J.H. Greaves. 1987. Resistance toanticoagulant rodenticides: The problem andits management. Pages 302–315. In:Donahaye, E. and S. Navarro (eds.),Proceedings 4th International WorkingConference on Stored Product Protection, TelAviv, Israel, September 1986.

Sumner, E.L. and R.M. Bond. 1939. Aninvestigation of Santa Barbara, Anacapa andSan Miguel Islands, California. 106 pp.Unpublished Report.

Taylor, R.H. 1993. The feasibility of rateradication on Langara Island, BritishColumbia, Canada. Canadian Wildlife Service,Delta, BC.

Taylor, R.H. and B.W. Thomas. 1989. Eradicationof Norway rats (Rattus norvegicus) fromHawea Island, Fjordland, using brodifacoum.New Zealand Journal of Ecology 12:23–32.

Taylor, R.H. and B.W. Thomas. 1993. Ratseradicated from rugged Breaksea Island (170

312 HOWALD ET AL.

ha), Fjordland, New Zealand. BiologicalConservation 65:191–198.

Tershy, B.R., D.A. Croll and G.R. Howald. 1997.Options for removing introduced black rats(Rattus rattus) from Anacapa Island, ChannelIslands National Park. Unpublished Report,Island Conservation and Ecology Group, SantaCruz, CA.

Towns, D.R. 1991. Response of lizard assemblagesin the Mercury Islands, New Zealand, to

removal of an introduced rodent: the kioreRattus exulans. Journal of the Royal Society ofNew Zealand 21:119–136.

Whittaker, R.J. 1998. Island Biogeography:ecology, evolution, and conservation. Oxford.

WCMC (World Conservation Monitoring Centre).1992. Groombridge, B. (ed.); GlobalBiodiversity: Status of the Earth's livingresources; Chapman and Hall, London, 585 pp.