Distribution of Parmarion cf. martensi (Pulmonata ... · The semi-slugParmarion cf. martensi Sim-roth, 1893, is a recent introduction to the is-land of Hawai‘i. The ﬁrst record

457

Distribution of Parmarion cf. martensi (Pulmonata: Helicarionidae),a New Semi-Slug Pest on Hawai‘i Island, and Its Potential as a Vector

for Human Angiostrongyliasis1

Robert G. Hollingsworth,2,3 Rachel Kaneta,3,4 James J. Sullivan,5,8 Henry S. Bishop,5,8

Yvonne Qvarnstrom,5,6,8 Alexandre J. da Silva,5,8 and David G. Robinson7

Abstract: The semi-slug Parmarion cf. martensi Simroth, 1893, was first discov-ered on O‘ahu, Hawai‘i, in 1996 and then on the island of Hawai‘i in 2004. Thisspecies, which is probably native to Southeast Asia, is abundant in eastern Ha-wai‘i Island, reportedly displacing the Cuban slug, Veronicella cubensis (Pfeiffer,1840), in some areas. A survey in July–August 2005 found P. cf. martensi primar-ily in the lower Puna area of Hawai‘i Island, with an isolated population inKailua-Kona (western Hawai‘i Island). It is now established in commercial pa-paya plantations, and survey participants reported it as a pest of lettuce andpapaya in home gardens. Survey respondents considered P. cf. martensi a pestalso because of its tendency to climb on structures where it deposits its fecesand because of its potential to transmit disease. Individuals of this species werefound to carry large numbers of infective third-stage larvae of the nematodeAngiostrongylus cantonensis (Chen, 1935), the causative agent of human angio-strongyliasis and the most common cause of human eosinophilic meningo-encephalitis. Using a newly developed polymerase chain reaction test, 77.5% ofP. cf. martensi collected at survey sites were found infected with A. cantonensis,compared with 24.3% of V. cubensis sampled from the same areas. The transmis-sion potential of this species may be higher than that for other slugs and snailsin Hawai‘i because of the high prevalence of infection, worm burdens, and itsgreater association with human habitations, increasing the possibility of hu-man-mollusk interactions.

The semi-slug Parmarion cf. martensi Sim-roth, 1893, is a recent introduction to the is-land of Hawai‘i. The first record was made inthe summer of 2004 in Paradise Park, a resi-dential area in the district of Puna (East

Hawai‘i Island) (Arnold Hara, University ofHawai‘i, pers. comm., 2005). The species wasrecognized as being similar to, or the same as,a semi-slug species collected for the first timeon the island of O‘ahu in 1996 and provision-

Pacific Science (2007), vol. 61, no. 4:457–467Work of the U.S. GovernmentNot under copyright

1 Manuscript accepted 9 January 2007.2 Corresponding author (phone: 808-959-4349;

e-mail: [email protected]).3 U.S. Department of Agriculture, Agricultural Re-

search Service, U.S. Pacific Basin Agricultural ResearchCenter, P.O. Box 4459, Hilo, Hawai‘i 96720.

4 Current address: 190 Southwest Brumback Street,Linfield College, McMinnville, Oregon 97128.

5 Centers for Disease Control and Prevention, 4770Buford Highway Northeast, Atlanta, Georgia 30341.

6 Atlanta Research and Education Foundation in con-junction with the Atlanta Veterans Administration Medi-cal Center, Decatur, Georgia.

7 U.S. Department of Agriculture, Animal and PlantHealth Inspection Service, Plant Protection and Quaran-tine, Academy of Natural Sciences, 1900 BenjaminFranklin Parkway, Philadelphia, Pennsylvania 19103.

8 The findings and conclusions in this report are thoseof the authors and do not necessarily represent the viewsof the Centers for Disease Control and Prevention/theAgency of Toxic Substances and Disease Registry.

ally identified as Parmarion martensi Simroth,1893 (Cowie 1997). The taxon Parmarionmartensi was originally described from Cam-bodia (Simroth 1893), but it has also beenreported from Vietnam, Malay Peninsula, Su-matra, Java, Borneo, Japan, Taiwan, Singa-pore, Samoa, and American Samoa (vanBenthem Jutting 1950, Minato 1975, Minatoand Okubo 1991, Ho 1995, Cowie 1998,Asato et al. 2004). However, due to the diffi-culty of identifying Parmarion to the specieslevel, the accuracy of the records listed hererequires further investigation. Plate I showsphotos of this semi-slug collected from a sitein Koa‘e, East Hawai‘i Island, in December2004. Voucher specimens (2) collected fromthe site at that time were deposited in theAcademy of Natural Sciences malacologicalcollection (Philadelphia, Pennsylvania) anddesignated as ansp A21014.

In December 2004, before learning that P.cf. martensi had been found on Hawai‘i Is-land, R. Hollingsworth was requested by alocal resident to investigate the presence of anew slug species on a property in Koa‘e, nearthe eastern tip of the island. The request wasprompted by the resident’s concern abouttransmission of rat lungworm disease causedby Angiostrongylus cantonensis (Chen, 1935), arodent nematode that develops to the infec-tive larval stage in a slug or snail host (Mac-kerras and Sandars 1955). The disease, whichmanifests itself in humans as eosinophilicmeningitis (Kliks and Palumbo 1992), can beacquired by the intentional or accidental con-sumption of raw or undercooked slugs orsnails or paratenic hosts (such as shrimps orflatworms) (i.e., animals capable of carryingthe infective stage of the parasite but not sup-porting further development [Alicata and Jin-drak 1970, Ash 1976, Kliks and Palumbo1992]). The resident requesting the visit andtwo of her dinner guests became ill withsymptoms consistent with angiostrongyliasisafter consuming home-grown lettuce report-edly contaminated with immature semi-slugs.Important intermediate hosts of A. cantonensisin Hawai‘i include veronicellid slugs [pri-marily the Cuban slug, Veronicella cubensis(Pfeiffer, 1840)]; the giant African snail, Acha-tina fulica (Bowdich, 1822); and the marsh

slug, Deroceras laeve (Muller, 1774) (Wallaceand Rosen 1969a, Alicata 1991).

Parmarion cf. martensi has the potential forbecoming an important vector of A. canton-ensis in Hawai‘i, as happened in Okinawa(Asato et al. 2004) after P. martensi becamemore prevalent there starting around theyear 2000. Our initial survey in Koa‘e indi-cated that P. cf. martensi was extremely com-mon; it was found in trash cans, in acomposting toilet, in an outdoor showerarea, in a planting of spider lilies (Crinum asi-aticum [Amaryllidaceae]), under plastic sheet-ing, and in a vegetable compost pile whereegg masses of P. cf. martensi were also found.Specimens of P. cf. martensi collected duringthe initial survey were sent to the Division ofParasitic Diseases, Centers for Disease Con-trol and Prevention (CDC), Atlanta, Georgia,to be examined for infection. The 26 semi-slugs examined were all positive for A. canton-ensis, as determined by pepsin digestion(Graeff-Teixeira and Morera 1995). The im-portance of P. cf. martensi as a vector of thisdisease may be exacerbated by its high popu-lation densities, climbing behavior, attractionto food items associated with human dwell-ings, and potentially high parasite load.

Our objectives for this study were to: (1)determine the geographical distribution ofP. cf. martensi on Hawai‘i Island; (2) surveyhomeowners to gain information about peststatus, feeding preferences, and foraging be-havior; (3) compare levels of infection of A.cantonensis in P. cf. martensi and V. cubensiscollected from the same sites; and (4) com-pare the feeding patterns of P. cf. martensiand V. cubensis in the laboratory on selectedtypes of food.

materials and methods

The Parmarion survey was publicized withadvertisements in two local newspapers on 7July 2005. The advertisement included ablack-and-white picture of an adult P. cf.martensi semi-slug, a caption detailing its dis-tinguishing characteristics, and a request forinformation from anyone who had seen thisspecies on his or her property. An articleabout this species and our survey that ap-

458 PACIFIC SCIENCE . October 2007

PLATE I. Parmarion cf. martensi Simroth from East Hawai‘i Island, photographed December 2004. The distinct keelalong the posterior dorsal midline helps distinguish this species from similar-looking species in Hawai‘i. A yellowishbrown, flattened, fingernail-shaped shell is present on the dorsum (A), but the shell is frequently covered by mantlefolds (B).

PS 61-4 color plate mech 4/13/07 2:14 PM Page 1

peared in both newspapers on 15 July gener-ated an even greater response.

Site visits were made to follow up allcredible reports of semi-slug sightings madewithin 7 weeks of the initial newspaper adver-tisement. Residents were asked where theyhad seen this species on their property, whattypes of food they had observed the semi-slugs to eat, and whether they consideredthis species a pest. At least 20 min per sitewas spent collecting semi-slugs and othermollusks. The locations where P. cf. martensiwere found were recorded.

Specimens collected from each site duringthe July survey were sorted by species anddivided into size groups (large, medium,small, or neonate). For V. cubensis and P. cf.martensi, large specimens were about 4.5–5.5cm and 3.5–4.5 cm long, respectively; me-dium and small specimens were about one-half and one-third as long, respectively, as‘‘large’’ specimens of the same species. Neo-nates were <0.5 cm in length. These speci-mens were shipped on dry ice to the CDC,where an experimental polymerase chain re-action (PCR) method was used to determinethe percentage infected with A. cantonensis.DNA from intact slug tissue pieces was ex-tracted using one of two methods: eitherusing selected reagents from the FastDNAkit (MP Biomedicals, Solon, Ohio) or theDNeasy tissue kit (QIAGEN Inc., Valencia,California). The FastDNA extraction wasperformed as described previously (da Silvaet al. 1999) with one modification: the sampledisruption was performed for 30 sec at speed5.5 in the FastPrep 120 Disruptor (Q-Biogene, Carlsbad, California). PCR inhibi-tors were removed from DNA extracted withFastDNA method by further purificationwith the QIAquick PCR purification kit(QIAGEN Inc., Valencia, California). DNAextracted by the DNeasy tissue kit did notneed further purification. The PCR methodamplified 1,134 base pairs from the small sub-unit ribosomal gene in Angiostrongylus species(Qvarnstrom et al. 2007). Angiostrongylus-specific primers AngioF1 (5 0-ATCA-TAAACCTTTTTTCGAGTATCCAG-3 0)and AngioR1 (5 0-TCTCGAGACAGCT-CAGTCCCGG-3 0) were designed based on

positions 456 to 482 and 1,569 to 1,590 ofAngiostrongylus cantonensis 18S rRNA gene;GenBank entry AY295804. PCR was per-formed with 0.4 mM of each primer, 2 ml ofDNA and AmpliTaq Gold PCR Master Mix(Applied Biosystems, Foster City, California)for a 50-ml total PCR reaction volume. PCRcycling parameters were 95�C 5 min, 45cycles of 95�C 15 sec, 65�C 15 sec, 72�C1 min, and 72�C 10 min. To achieve iden-tification at species level, PCR amplifiedproducts were subjected to DNA sequenceanalysis.

Data used to compare infection levels in P.cf. martensi and V. cubensis were derived fromcollections of mollusks from five sites, eachof which contained multiple individuals ofeach species. The statistical model consistedof logistic regression implementing the gen-eralized estimating equations (GEE) proce-dure to adjust for correlation among mollusksof the same species being collected from thesame site. Analyses were performed usingthe GENMOD procedure of SAS on-lineversion 9.1 (SAS Institute, Inc. 2000–2004).Alpha was set at 0.05.

Feeding preferences of P. cf. martensi andV. cubensis were compared in unreplicatedbioassays. The bioassay arena consisted of aventilated 2-liter plastic container holdingmoist soil (about 2.5 cm deep) and three ma-ture semi-slugs, held for 5 days in an environ-mental chamber (27�C, 80% RH, 12:12 L:D).Various types of plant foods were placed onthe soil surface after weighing. Data collectedincluded weights of plant material and obser-vations of feeding damage.

results

The majority of the 51 survey respondentswere from the Puna district, although callswere also received from residents in/nearKailua-Kona (West Hawai‘i), Waimea(Northwest Hawai‘i), Honoka‘a (NortheastHawai‘i), Hilo (East Hawai‘i), and OceanView Estates (South Hawai‘i). Based on tele-phone interviews, we determined that manyrespondents had actually seen other types ofslugs. We confirmed the presence of P. cf.martensi at 27 of 29 properties that in our

Distribution of Parmarion cf. martensi on Hawai‘i Island . Hollingsworth et al. 459

judgment were associated with credible sight-ings of this semi-slug. The greatest concen-tration of sightings was in the Paradise ParkSubdivision (Figures 1 and 2).

At the time of our survey, populations ofP. cf. martensi were very low throughout the

Puna district. At least four survey participantsindependently noted that populations hadcrashed within the previous 2–3 months. Wealso observed such a decline on an organicfarm near Kapoho, near the eastern tip ofHawai‘i Island, where we had regularly been

Figure 1. Survey locations for Parmarion cf. martensi in Hawai‘i Island (shaded circles). White circles indicate siteswhere P. cf. martensi was searched for but not found. A white square (in the Waimea area) indicates location of a pop-ulation detected during a 2006 survey carried out by University of Hawai‘i scientists. The large square shows Punadistrict. Towns are indicated by black triangles.


collecting P. cf. martensi adults since January2005. In early June, we found large numbersof P. cf. martensi egg masses on the under-sides of halved coconuts being used as mulchbut saw few individuals of any other life stage.Subsequent observations at various sites in

the Puna district indicated that populationsof P. cf. martensi returned to a higher levelduring the fall of 2005.

Thirty-seven percent of survey respon-dents indicated that they had first noticedthis semi-slug species within the few weeks

Figure 2. Map inset showing the lower Puna area where Parmarion cf. martensi was generally abundant. See Figure 1for explanation of symbols.


preceding the publicity generated by surveyannouncements. However, one respondent,an organic gardener, claimed to have noticedthis species in lower Paradise Park in 1999.This would be the earliest report of P. cf.martensi on Hawai‘i Island, and another resi-dent @3 km distant remembered seeing thesemi-slug on his property in 2000 or 2001.Sixty-seven percent of survey participantsconsidered this species a pest because of thedisease risk and/or fecal deposits on houses.However, 22% did not consider it a pest,and the remaining 11% were either indiffer-ent or did not provide a response.

The types of habitat where semi-slugswere found as determined by survey re-sponses and our observations are summarizedin Table 1 and Table 2, respectively. Manysurvey participants reported that P. cf. mar-tensi demonstrated a remarkable propensityto climb objects, including drainpipes, wooddecks, the walls of homes, water tanks, andbarbeque grills. Although this climbing be-havior occurred mainly at night, its extentwas evidenced by the large number of fecaldeposits we saw on the upper walls of housesand on water tanks.

Respondents observed semi-slugs mostcommonly on green plants, fallen fruits, andplastic surfaces (Table 1), including four re-ports on lettuce, two of them in home gar-dens and two involving lettuce purchased at

markets on the east side of Hawai‘i Island.There were five reports of P. cf. martensi be-ing attracted during daytime to food prepara-tion and sink areas. In this context, severalrespondents remarked that P. cf. martensimoved more quickly than other slug speciesthey had seen around their houses, appearingsoon after food sources first became available.Five people reported semi-slugs feeding ondog food, cat food, or parrot food that was ei-ther spilled or left in bowls (Table 1).

Six survey participants noted seeing thesemi-slugs feeding in their covered compost

TABLE 1

Types of Habitat Where Parmarion cf. martensi Was Observed by Survey Participants (Day and Night Observations)

Location No. of Reports

On green plants (lettuce, fennel, sweet potato, banana, passion fruit, lemongrass, Heliconia) 10On fallen fruit (avocado, guava, citrus, papaya, mango) 9On/under plastic or plasticlike materials, including black plastic sheeting, tarps, drainpipes,

plant pots8

In compost (especially covered compost) or inside trash cans 6In food preparation and sink areas (outdoor sinks, on dishes, grills, on toothbrush) 5In pet food bowls or eating spilled pet food 5On deck or stairs 5On concrete, tile, or asphalt 5On walls of home 4On palm fronds or trunks or under coconut shells 4On catchment tank (on outside, under black saran cover at top) 3On ripe papaya fruits (on tree) 2In plant debris other than palms 2

TABLE 2

Types of Habitat Where Parmarion cf. martensi WasCollected during Daytime Site Visits

LocationNo. of

Properties

On/under plastic or plasticlike materials,including black plastic sheeting,ground tarps, drainpipes, plant pots, tires

10

On palm fronds (fallen or within mulch pile) 4Under rocks or tiles 3In covered compost 2Under banana trees 2On drainpipe 1In citronella grass 1Under woven grass mat (‘‘goza’’) 1Palm leaves (living) 1Coconut shell halves (overturned) 1


bins or in trash cans, and we collected nu-merous P. cf. martensi adults and egg massesfrom several covered compost bins. Two sur-vey participants noted that P. cf. martensisometimes fed on ripe papaya still attachedto the tree.

In our own daylight searches, P. cf. mar-tensi was most often found on or under plasticor other objects, including tires, tarps, blackplastic sheeting, and drainpipes. Eggs foundwere generally in clutches of ca. 10–30 eggseach, laid inside overturned coconut shells,underneath plastic plant pots, or attached tothe underside of plastic sheeting that was incontact with the ground or compost. In onecase, eggs and neonate semi-slugs were foundsingly within rotting leaves of Alexanderpalms, Archontophoenix alexandrae (F. Muell.)H. Wendl. & Drude (Arecaceae). Oursearches frequently found V. cubensis under-neath plastic objects. However, unlike V. cu-bensis, P. cf. martensi life stages (includingegg masses) were seldom found in direct con-tact with soil. Egg masses of the two speciescan be easily distinguished; eggs in the eggmasses of the Cuban slug are larger, arechained together, and a black threadlike ma-terial (slug feces) runs through the masses. Inegg masses of P. cf. martensi, eggs are notchained together and no black threadlike ma-terial is present.

Survey participants were asked if they hadrecently noted rats or mice on their property.Rats are the usual definitive host for Angio-strongylus cantonensis, and mice are potentialhosts that can be infected in the laboratory.Seventy-four percent and 48% of respon-dents, respectively, indicated that they had re-cently noticed rats or mice on their property.

None of those surveyed considered P. cf.martensi to be a serious agricultural pest, andseveral people volunteered that P. cf. martensidid not eat plant leaves to the same extent asV. cubensis. These observations were sup-ported by results in laboratory bioassays com-paring the feeding preferences of these twospecies. Parmarion cf. martensi semi-slugscompletely consumed a large, tender hibiscusflower but left hibiscus leaves present withinthe same container untouched. In contrast,V. cubensis presented with the same two

choices consumed both flower and leaf mate-rial. A similar preference for tender plant ma-terial was noted in containers holding bothred-leaved and green-leaved varieties of ti(Cordyline sp.): P. cf. martensi fed only slightlyon the red-leaved variety, which was moretender than the green-leaved variety, and didnot feed on the green-leaved variety held inthe same container. Veronicella cubensis dem-onstrated a preference for the red leaves butalso fed on the green-leaved variety. In con-tainers holding a mixture of orchid flowers,orchid leaves, and pseudobulb material, bothmollusk species fed exclusively on flowersand avoided the other plant parts.

The average level of infection by A. canton-ensis in P. cf. martensi and V. cubensis was77.5 and 24.3%, respectively (Table 3). Thisdifference was significant (P ¼ 0:0007, basedon logistic regression using generalized esti-mating equations procedure). In both species,percentage infection was highest for large in-dividuals (Table 3).

discussion

Our survey indicated that P. cf. martensi hasessentially a continuous distribution in lowerelevations of the Puna district. Such a wide-spread distribution is surprising for a specieswhose presence was confirmed only in 2004.However, two residents of Paradise Park pro-vided credible reports of having first seen this

TABLE 3

Percentage Infectiona by Angiostrongylus cantonensis inSlugs Collected from Five Sites in Hawai‘i Island

Slug Species Size % Infecteda nb

Veronicella cubensis Small 0 9Medium 0 2Large 34.6 26All sizes 24.3 37

Parmarion cf. martensi Neonate 100 1Small 25 4Medium 76.9 13Large 86.4 22All sizes 77.5 40

a Infection determined by PCR and confirmed by sequencingof PCR amplicons.

b Number of slugs or semi-slugs analyzed.


species on their properties in 1999 and 2001.The rapid spread of the semi-slugs in thisarea has almost certainly been aided by theactivities of people. There has been a con-struction boom in the lower Puna area duringthe past 6 yr, and it is likely that semi-slugswere transported between construction siteson building materials, building machinery,and in potted plants. The semi-slugs foundin South Hilo were in trash that had beendumped in an isolated area near StainbackHighway (Figure 2). It is possible that thesemi-slugs were derived from trash that origi-nated in the Puna district. As for the presenceof P. cf. martensi in North Kona, the ownersof the residence have a second home in thePuna district (Paradise Park Subdivision),and it is possible that semi-slugs were acci-dentally transported from their Puna hometo their Kona residence. After our survey wascompleted, an additional population of P. cf.martensi was found in the town of Waimea(Northwest Hawai‘i Island) (Figure 1), de-tected during a statewide survey for molluskpests carried out by University of Hawai‘i re-searchers. In their 2006 survey, P. cf. martensiwas also found on the island of O‘ahu, in aplant nursery near Kahalu‘u and in a com-mercial farming area in Waimanalo (KennethHayes, University of Hawai‘i, pers. comm.,2006).

As previously mentioned, the taxon Par-marion martensi is apparently native to South-east Asia. It is not known how P. cf. martensiarrived in Hawai‘i or whether its presence isthe result of a single introduction. Intercep-tion data maintained by the U.S. Departmentof Agriculture’s Animal and Plant HealthInspection Service indicate that semi-slugsin the genus Parmarion were intercepted onDracaena plants shipped from Malaysia toKailua-Kona in July 2003. On two occasions(February 2001 and April 2004), Parmarionhas been intercepted in Honolulu during pre-clearance inspections on plant material beingexported to the U.S. mainland. Interceptionrecords from 1996 to 2006 document sevenother cases in which plants destined for loca-tions in the United States were infested with‘‘Parmarion.’’ Orchids were the infested com-modity in five of these seven cases, with ‘‘ori-

gin’’ listed as Vietnam, Thailand, Indonesia,or Malaysia [USDA APHIS PPQ Pest Inter-ception Database (PestID), Riverdale, Mary-land].

The ecological consequences of the inva-sion of P. cf. martensi into Hawai‘i are diffi-cult to predict. Anecdotal evidence suggeststhat P. cf. martensi has displaced V. cubensis asthe dominant large mollusk in certain resi-dential areas. However, during our survey wefrequently found both species in the sameproperty. Our survey was based on a conve-nience sample, and information provided byrespondents on pest status may constitute abiased sample of surveyed areas. The dietaryhabits of P. cf. martensi have not been studiedin detail, but V. cubensis is known as a seriouspest of ornamental and garden crops inHawai‘i (Furutani and Arita-Tsutsumi 1998).Many slug species are known to feed on awide variety of vegetable and animal matter,including fruits, foliage, decaying vegetablematter, dead arthropods and earthworms,and algae and fungi that grow on surfaces ofplants, rocks, or wood (Ebeling 2002). Her-bivorous slugs generally prefer tender planttissues, such as flowers, plant seedlings, ormature plants that have tender leaves orstems. Larger species (>2.5 cm) are morelikely than smaller species to feed on healthyplant tissue. In feeding tests, P. cf. martensipreferred flowers over leaves of the same spe-cies and readily fed on fruits such as papaya.Acceptable foliage included lettuce, cabbage,and hibiscus leaves. The acceptability of thesefood sources to P. cf. martensi (a relativelylarge species, with extended length sometimes>5 cm) is not surprising. What is unusual,relative to other slug species in Hawai‘i, isthe propensity of P. cf. martensi to climb andlocate rich food sources, including bird food,dog food, cat food, fish entrails, and papayas(including fruit on the tree and fruit set outon the railing of a deck high off the ground).This climbing behavior, in combination withan apparent attraction to rich food sourcesand a naturally high rate of infection by A.cantonensis, increases the likelihood that peo-ple will come into contact with semi-slugsand the parasitic nematodes they carry.Increased contact can occur when people


handle items contaminated by semi-slugsor accidentally ingest a semi-slug or part ofone. Current evidence suggests that the slimeof mollusks (if accidentally ingested on fruitsand vegetables) may not contain sufficientnumbers of A. cantonensis to cause diseasesymptoms (Hollingsworth and Cowie 2006).Accidental ingestion of the slugs themselveson poorly washed fruits and vegetables (con-sumed raw) probably represents a muchgreater risk and would presumably be morelikely to occur when semi-slugs are verysmall. Our data show that even neonate P. cf.martensi can be hosts for A. cantonensis. Onesurvey respondent reported finding neonatesemi-slugs on her garden-grown lettuce; theneonates were very difficult to see because oftheir small size (about 2 mm in length). Evenlarger specimens could be accidentally con-sumed on lettuce or other fresh greens ifthese are not thoroughly washed and checkedfor semi-slugs before chopping. Further re-search is necessary to determine whetherneonates and very small semi-slugs carry asufficient number of nematodes to potentiallycause disease symptoms.

The climbing behavior of P. cf. martension water tanks is a potential health concernbecause of the chance that semi-slugs mighttransmit various types of disease organismsinto drinking water. Contaminated drinkingwater is a potential source of A. cantonensisinfection for humans (Wallace and Rosen1969b). Cheng and Alicata (1964) found thatinfective-stage A. cantonensis were releasedinto water from both damaged and undam-aged Achatina fulica and Subulina octona (Bru-guiere, 1792) and survived in water for up to72 hr. Similar results have been found for An-giostrongylus costaricensis Morera & Cespedes,1971, released from the freshwater snailBiomphalaria glabrata (Say) (Ubelaker et al.1980). None of our survey respondents re-ported finding P. cf. martensi inside theirwater tanks, but we received such a reportsubsequently. Many Puna residents rely onwater collected from roofs for household use,including more than one-half of the respon-dents in our survey. Several respondents ex-pressed concerns about the possible healthrisks associated with P. cf. martensi climbing

on their water tanks. However, the majorityof survey respondents using catchment watersaid they did not use this water for drinking.Four respondents mentioned that they dranktheir catchment water, but all said they usedparticle filtering systems.

The association of P. cf. martensi withplastic and other smooth surfaces may indi-cate a feeding preference for surface-growingalgae or fungi, although semi-slugs were fre-quently found on plastic apparently free ofmicroorganism growth. Alternatively, the as-sociation could be related to water conserva-tion or could represent an adaptation foravoiding disease-causing organisms. Asatoet al. (2004) hypothesized that P. martensi ismore susceptible to infection with A. canton-ensis than ‘‘Veronicella alte’’ [an apparent mis-naming of Laevicaulis alte (Ferussac, 1821)]because of the lower density of muscle tissuein the former species. Artificial infection byA. cantonensis was found to increase mortalityrates in the aquatic snail Physa elliptica (Lea,1834) relative to snails in the experimentalcontrol (Wallace and Rosen 1969a). How-ever, it is not clear whether infection by A.cantonensis increases mortality of mollusksgenerally.

Working in Malaysia, Lim and Heyneman(1965) studied Microparmarion malayanus(Collinge) a semi-slug species in the samefamily as P. martensi. They noted that M. ma-layanus climbed trees at night, was particu-larly common under heaps of dead palmleaves during the day, and was an importantcarrier of A. cantonensis. We observed thesesame characteristics in P. cf. martensi in Ha-wai‘i. In Malaysia, M. malayanus was observedto feed on rat feces, and rats were observed tofeed on M. malayanus. These aspects of thebiology of. P. cf. martensi in Hawai‘i havenot been studied. However, their elucidationwould be an important step toward under-standing the importance of P. cf. martensi forthe epidemiology of angiostrongyliasis inHawai‘i.

acknowledgments

We thank Robert Cowie and Kenneth Hayes(University of Hawai‘i at Manoa, Honolulu)


for providing additional information aboutthe distribution of P. cf. martensi in Hawai‘iand for critically reviewing early drafts of themanuscript.

Literature Cited

Alicata, J. E. 1991. The discovery of Angio-strongylus cantonensis as a cause of humaneosinophilic meningitis. Parasitol. Today7:151–153.

Alicata, J. E., and K. Jindrak. 1970. Angio-strongylosis in the Pacific and SoutheastAsia. Charles C. Thomas, Springfield, Illi-nois.

Asato, R., K. Taira, M. Nakamura, J. Kudaka,K. Itokazu, and M. Kawanaka. 2004.Laboratory and epidemiology communica-tions: Changing epidemiology of Angio-strongylus cantonensis in Okinawa Prefecture,Japan. Jpn. J. Infect. Dis. 57:184–186.

Ash, L. R. 1976. Observations on the role ofmollusks and planarians in the transmis-sion of Angiostrongylus cantonensis infectionto man in New Caledonia. Rev. Biol. Trop.24:163–174.

Cheng, T. C., and J. E. Alicata. 1964.Possible role of water in the transmissionof Angiostrongylus cantonensis (Nematoda:Metastrongylidae). J. Parasitol. 50:39–40.

Cowie, R. H. 1997. Catalog and bibliographyof the nonindigenous nonmarine snailsand slugs of the Hawaiian Islands. BishopMus. Occas. Pap. 50:1–66.

———. 1998. Catalog of the nonmarinesnails and slugs of the Samoan Islands.Bishop Mus. Bull. Zool. (3):1–122.

da Silva, A. J., F. J. Bornay-Llinares, I. N.Moura, S. B. Slemenda, J. L. Tuttle, andN. J. Pieniazek. 1999. Fast and reliableextraction of protozoan parasite DNAfrom fecal specimens. Mol. Diagn. 4:57–64.

Ebeling, W. 2002. Chapter 12, Miscellaneouspests, urban entomology (Web site). Uni-versity of California, Riverside, Entomol-ogy Department. http://www.entomology.ucr.edu/ebeling/ (accessed 4 January2006).

Furutani, S. C., and L. Arita-Tsutsumi. 1998.Feeding preference of the two-striped

slug, Veronicella cubensis (Pfeiffer) on taro.J. Hawaii. Pac. Agric. 9:1–6.

Graeff-Teixeira, C., and P. Morera. 1995.Metodo de digestao de moluscos em acidoclorıdrico para isolamento de larvas demetastrongilıdeos. Biociencias 3:85–89.

Ho, W. H. 1995. A review of the land-snailfauna of Singapore. Raffles Bull. Zool.43:91–113.

Hollingsworth, R. G., and R. H. Cowie.2006. Apple snails as disease vectors. Pages121–132 in R. Joshi and L. Sebastian, eds.Global advances in ecology and manage-ment of golden apple snails. PhilippineRice Research Institute (PhilRice), Munoz,Nueva Ecija.

Kliks, M. M., and N. E. Palumbo. 1992. Eo-sinophilic meningitis beyond the Pacificbasin: The global dispersal of a peridomes-tic zoonosis caused by Angiostrongylus can-tonensis, the nematode lungworm of rats.Soc. Sci. Med. 34:199–212.

Lim, B.-L., and D. Heyneman. 1965. Host-parasite studies of Angiostrongylus canton-ensis (Nematoda, Metastrongylidae) inMalaysian rodents: Natural infection ofrodents and molluscs in urban and ruralareas of central Malaya. Ann. Trop. Med.Parasitol. 59:425–433.

Mackerras, M. J., and D. F. Sandars. 1955.The life history of the rat lungworm,Angiostrongylus cantonensis (Chen) (Nema-toda: Metastrongylidae). Aust. J. Zool.3:1–21.

Minato, H. 1975. A new record of Parmarionmartensi from Ishigaki Island, the SouthernRyukyus, Japan. Venus 34 (3–4): 109–111.

Minato, H., and K. Okubo. 1991. A record ofParmarion martensi Simroth, 1893 (Pulmo-nata: Helicarionidae) collected from Tai-wan. The Chiribotan 22:3–4.

Qvarnstrom, Y., J. J. Sullivan, H. S. Bishop,R. G. Hollingsworth, and A. J. da Silva.2007. PCR-based detection of Angiostron-gylus cantonensis in tissue and mucus secre-tions from molluscan hosts. Appl. Environ.Microbiol. 73:1415–1419.

SAS Institute, Inc. 2000–2004. SAS onlineversion 9.1. Cary, North Carolina.

Simroth, H. 1893. Ueber einige Parmarion-Arten. Pages 100–110, pl. 7, 8 in Weber,


M., ed. Zoologische Ergebnisse einer Re-ise in Niederlandisch Ost-Indien heraus-gegeben von Dr. Max Weber. Vol. 2. E. J.Brill, Leiden.

Ubelaker, J. E., G. R. Bullick, and J. Caruso.1980. Emergence of third-stage larvae ofAngiostrongylus costaricensis Morera andCespedes 1971 from Biomphalaria glabrata(Say). J. Parasitol. 66:856–856.

van Benthem Jutting, W. S. S. 1950. System-atic studies on the non-marine Molluscaof the Indo-Australian Archipelago. II.

Critical revision of the Javanese pulmonateland-snails of the families Helicarionidae,Pleurodontidae, Fruticicolidae & Streptax-idae. Treubia 20:381–505.

Wallace, G. D., and L. Rosen. 1969a. Exper-imental infection of Pacific island molluskswith Angiostrongylus cantonensis. Am. J.Trop. Med. Hyg. 18:13–19.

———. 1969b. Studies on eosinophilic men-ingitis. V. Molluscan hosts of Angiostrongy-lus cantonensis on Pacific islands. Am. J.Trop. Med. Hyg. 18:206–216.


Documents

Distribution of Parmarion cf. martensi (Pulmonata ... · The semi-slugParmarion cf. martensi Sim-roth, 1893, is a recent introduction to the is-land of Hawai‘i. The ﬁrst record