Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
0$/$&2/2*,$144 FEEDING PREFERENCES OF TWO PREDATORY SNAILS INTRODUCED TO HAWAII AND THEIR CONSERVATION IMPLICATIONS Wallace M. Meyer III1, 2* & Robert H. Cowie1 ABSTRACT Two introduced predatory land snails, Euglandina rosea and Oxychilus alliarius, have been LPSOLFDWHGLQWKHGHFOLQHRIQDWLYH+DZDLLDQDQG3DFL¿FLVODQGODQGVQDLOV:HH[DPLQHGWKH feeding ecology of E. rosea and O. alliariusIRFXVLQJ¿UVWRQSUH\VL]HDQGVSHFLHVSUHIHUHQFHV and second on quantifying consumption rates, which is necessary to address the possible OHYHOVRILPSDFWRQQDWXUDOSRSXODWLRQV,QSUH\VL]HSUHIHUHQFHH[SHULPHQWVE. rosea always FRQVXPHGPRUHVQDLOVLQWKHVPDOOHUVL]HFODVV,QSUH\VSHFLHVSUHIHUHQFHH[SHULPHQWVE. roseaSUHIHUUHGVQDLOVWRVOXJVEXWVKRZHGQRSUHIHUHQFHDPRQJVQDLOVSHFLHV6L]HSUHIHUHQFHH[SHULPHQWVLQGLFDWHGWKDWO. alliarius will only consume prey less than 3 mm in shell OHQJWK,QSUH\VSHFLHVSUHIHUHQFHH[SHULPHQWVO. alliarius consumed every snail species offered but did not consume either of the slug species. Consumption rates were positively UHODWHGWRWKHVL]HRIWKHSUHGDWRUZLWKWKHODUJHUSUHGDWRUE. rosea, having a much higher consumption rate than O. alliarius, and with larger E. rosea consuming more than smaller E. rosea7KHVHUHVXOWVKDYHXQIRUWXQDWHFRQVHUYDWLRQUDPL¿FDWLRQVEHFDXVHVLQFHWKH\LQGLFDWH that both E. rosea and O. alliarius are generalist snail predators feeding preferentially on the VPDOOHVWVQDLOVDOWKRXJKWKH\UDUHO\HDWVOXJV6LQFHPDQ\H[WDQW+DZDLLDQDQG3DFL¿FLVODQG snail species are small, they are at risk of predation by these two introduced predators, while invasive slugs – there are no native slugs in Hawaii – may not be impacted by them. Key words: consumption rate, Euglandina rosea, Gastropoda, Oxychilus alliarius, slug. INTRODUCTION /DQGVQDLOIDXQDVZRUOGZLGHDUHXQGHUH[treme threat, and have the dubious honor of KDYLQJWKHKLJKHVWQXPEHURIGRFXPHQWHGH[WLQFWLRQVRIDQ\PDMRUWD[RQRPLFJURXS&RZLH 2004; Nash, 2004; Lydeard et al., 2004). The 2008 IUCN Red List of Threatened Species ,8&1LQFOXGHGJDVWURSRGVVQDLOV DQGVOXJVRIZKLFKZHUHOLVWHGDVH[WLQFW 1DWLYH3DFL¿FLVODQGODQGVQDLOVLQSDUWLFXODU DUH UHFRJQL]HG IRU WKHLU GLYHUVLW\ KLJK OHYHOV of endemism and the fact that they are under H[WUHPHWKUHDWZLWKSHUKDSVDVPDQ\DV RI WKH VSHFLHV KDYLQJ GLVDSSHDUHG /\GHDUG HW DO )RU H[DPSOH H[WLQFWLRQ RI WKH WUHHVQDLO VSHFLHV $FKDWLQHOOLQDH LQ +DZDLL DQG 3DUWXOLGDH RQ RWKHU 3DFL¿F LVODQGV KDV EHHQFDWDVWURSKLF&ODUNHHWDO+DG¿HOG 0XUUD\HWDO&RZLH+DG¿HOGHWDO&RRWH/RqYH In Hawaii, the situation may be especially critical. The native Hawaiian land-snail fauna XVHGWREHH[WUHPHO\GLYHUVHRYHUVSHFLHVDQGH[KLELWHGH[WUHPHO\KLJKHQGHPLVP RYHU&RZLHHWDOEXWHVWLPDWHV RIH[WLQFWLRQRIWKHVHXQLTXHVSHFLHVUDQJHIURP ±6ROHPWR&RZLH /\GHDUGHWDO)RUH[DPSOHWKH$PDVWULdae, an endemic Hawaiian family of more than VSHFLHV&RZLHHWDOPD\QRZEH UHGXFHGWRDVIHZDVWHQRUVRVSHFLHVH[LVWLQJ LQWLQ\KLJKO\ORFDOL]HGUHPQDQWSRSXODWLRQV The Endodontidae in Hawaii and throughout WKH3DFL¿FDSSHDUUHGXFHGWRVSDUVHSRSXODtions on those islands they formerly inhabited /\GHDUGHWDO0H\HU Two introduced predatory land snails, Euglandina rosea and Oxychilus alliarius, have been implicated in the decline of the native land-snail faunas. Euglandina rosea was purposely introGXFHGWR+DZDLLLQWRFRQWUROSRSXODWLRQV 1&HQWHUIRU&RQVHUYDWLRQ5HVHDUFKDQG7UDLQLQJ3DFL¿F%LRVFLHQFHV5HVHDUFK&HQWHU8QLYHUVLW\RI+DZDLLDW0DQRD 0DLOH:D\*LOPRUH+RQROXOX+DZDLL86$ 2'HSDUWPHQWRI=RRORJ\8QLYHUVLW\RI+DZDLLDW0DQRD0F&DUWK\0DOO(GPRQVRQ+RQROXOX+DZDLL86$ *Corresponding author: [email protected] 135 MEYER & COWIE of another introduced snail, Achatina fulica, WKHJLDQW$IULFDQVQDLO'DYLV%XWOHU 6LPEHUORII&LYH\UHO6LPEHUORII Cowie, 2001). Subsequently, it was introduced HOVHZKHUH 0HDG DQG LV QRZ ZLGHO\ GLVWULEXWHG WKURXJKRXW WKH 3DFLILF &RZLH 2001; Cowie et al., 2008). Unfortunately, the introduction of E. rosea has not reduced A. fulicaSRSXODWLRQV&LYH\UHO6LPEHUORII Cowie, 2001), but has been associated with the decline of many of the tree-snail species, QRWRQO\LQ+DZDLLEXWHOVHZKHUHLQWKH3DFL¿F &ODUNHHWDO+DG¿HOG+DG¿HOGHW DO0XUUD\*HUODFK&RRWH /RqYHDQGPD\KDYHFDXVHGWKHH[WLQFWLRQRIRWKHUQDWLYHODQGVQDLOVSHFLHV&RZLH 1DWLYHWR(XURSHO. alliarius was ¿UVWUHFRUGHGLQWKH+DZDLLDQ,VODQGVLQ &RZLH,JQRUHGXQWLOPRUHUHFHQWO\O. alliarius KDV EHFRPH H[WUHPHO\ DEXQGDQW RQ WKHLVODQGVRI0DXL6HYHUQVDQG+DZDLL 0H\HU &RZLH LQ SUHVV DQG QRZ RFFXUV DOVRRQ0RORNDLDQG.DXDL&RZLHDQG RQ 2DKX : 0 0 SHUVRQDO REVHUYDWLRQ Oxychilus alliarius is omnivorous, but is known as a snail predator; it may have played a role in the decline of the native Hawaiian land snail fauna, as it is thought to have done in New =HDODQG%DUNHU(IIRUGDQGLWFRXOG EH D WKUHDW WR WKH UHPDLQLQJ H[WDQW VSHFLHV 6HYHUQV Certain aspects of the feeding ecology of E. rosea have been studied, including feeding EHKDYLRUDQGSUH\WUDLOIROORZLQJ&RRND E*HUODFKSUH\VL]HSUHIHUHQFHV&KLX &KRX'DYLV%XWOHU1LVKLGD 1DSRPSHWK&RRNDE*HUODFK DQGSUH\VSHFLHVVHOHFWLRQLQWKHZLOG *ULI¿WKVHWDO+RZHYHUH[WUDSRODWLQJ from these results to determine the effect E. rosea is having on land snail faunas as a whole LV GLI¿FXOW HVSHFLDOO\ EHFDXVH FRQVXPSWLRQ rates and the reasons for prey preferences have not been adequately addressed. For LQVWDQFHH[SHULPHQWVWKDWDGGUHVVHGVSHFLHV SUHIHUHQFHVGLGQRWFRQWUROIRUSUH\VL]HDPRQJ SUH\VSHFLHVRIIHUHG&RRNDE,QVWHDG WRWDOZHLJKWRIHDFKVSHFLHVZDVVWDQGDUGL]HG +RZHYHULISUH\VL]HLVDVLPSRUWDQWDVVXJJHVWHG &KLX &KRX 'DYLV %XWOHU 1LVKLGD 1DSRPSHWK &RRN DEWKHIDFWWKDWWKHVPDOOHUVSHFLHVZHUH SUHIHUUHG LQ FKRLFH H[SHULPHQWV WHOOV XV OLWWOH DERXW SUHIHUHQFHV IRU VSHFL¿F VSHFLHV$OVR WKH PLFURFRVPV XVHG LQ WKHVH H[SHULPHQWV FOHDUJODVVGLVKHVGLGQRWSURYLGHFRQGLWLRQV permitting all possible predator avoidance behaviors. We know of no reports of feeding SUHIHUHQFHH[SHULPHQWVRQO. alliarius. Despite the two predators’ reputations, relatively little attention has been paid to the ELRORJ\RIHLWKHU2XUREMHFWLYHZDVWRH[DPLQH the feeding ecology of E. rosea and O. alliarius, IRFXVLQJ¿UVWRQSUH\VL]HDQGVSHFLHVSUHIHUences, and second on quantifying consumption rates in order to address the possible levels of impact on natural populations. METHODS $OOH[SHULPHQWVZHUHFRQGXFWHGZLWKVQDLOV collected from the wild less than one month previously. They were allowed to acclimate to laboratory conditions for at least three days SULRUWRDQ\H[SHULPHQWV%RWKSUHGDWRUVSHcies were fed a diet of Bradybaena similaris, a common invasive snail in Hawaii. Bradybaena similaris was not used as a prey species in WKHVXEVHTXHQWH[SHULPHQWV$OODQLPDOVZHUH maintained in the same kind of containers as WKRVHXVHGLQWKHH[SHULPHQWV[[ cm for E. rosea[[FPIRUO. alliarius. /DERUDWRU\WHPSHUDWXUHVZHUH±&GXULQJ WKHH[SHULPHQWV Different methods were used to assess the feeding behavior of each predator, since E. rosea has a much higher consumption rate than O. alliarius. The primary difference EHWZHHQWKHPHWKRGVZHUHWKDWH[SHULPHQWV H[DPLQLQJ E. rosea SUHIHUHQFHV SUH\ VL]H DQG VSHFLHV ZHUH FKRLFH H[SHULPHQWV ZLWK individual predators given a choice between two prey options, while O. alliariusH[SHULPHQWV were no-choice, as each predator was only offered one prey option. A different individual was used as a predator for every replicate across DOOH[SHULPHQWV)RUWKHE. roseaH[SHULPHQWV because hundreds of prey individuals were required, we used only non-native species, IURP ¿YH GLVWLQFW IDPLOLHV WZR VOXJ VSHFLHV Veronicella cubensis, Deroceras laeve) and WKUHHVQDLOVSHFLHVAchatina fulica, Paropeas achatinaceum, Succinea tenella). These species are phylogenetically diverse, such that the results should be generalisable. Inclusion of a species of non-native Succineidae, a family WKDW KDV SURSRUWLRQDWHO\ PRUH H[WDQW QDWLYH VSHFLHVLQ+DZDLLWKDQDQ\RWKHUIDPLO\5XQGHOOHWDODOVRSURYLGHVPRUHVSHFL¿F relevance to the local fauna. Given the much lower consumption rates of O. alliarius, it was PREDATORY SNAILS IN HAWAII reasonable to include native species as prey Succinea spp., Philonesia sp., Tornatellides spp.) in addition to common invasive species Arion intermedius, Deroceras laeve, Paropeas achatinaceum), all of which occur in areas in Hawaii with abundant O. alliarius. 137 We have only seen O. alliarius feed via the aperture, as opposed to consuming the entire prey snail including its shell. Differences in PRUWDOLW\ EHWZHHQ FRQWURO DQG H[SHULPHQWDO WUHDWPHQWV IRU HDFK VL]H FODVV ZHUH WHVWHG using the log odds ratio, l(o)6RNDO5RKOI DVIROORZV 3UH\6L]H3UHIHUHQFH Euglandina rosea Seventeen adult E. rosea±PPVKHOO OHQJWKZHUHVWDUYHGIRUWZRGD\VSULRUWRH[SHULPHQWDWLRQWRVWDQGDUGL]HKXQJHUOHYHOV(DFK H[SHULPHQWDOFRQWDLQHUZDV¿OOHGZLWKFPRI VRLODFDOFLXPVRXUFHFXWWOHERQHDQGVRPH lettuce OHDYHVDVIRRGIRUWKHSUH\7KHSUH\A. fulicaZHUHGLYLGHGLQWRIRXUVL]HFODVVHV± 1–2, 2–3, and 3–4 cm shell length). Ten snails IURPWZRRIWKHIRXUVL]HFODVVHVZHUHRIIHUHG to one E. rosea after the prey had been allowed to acclimate to the container for one day. Euglandina rosea consumes small snails whole as well as eating larger snails via the shell aperture. Every 12 h consumed individuals were replaced with new individuals of the same VL]HFODVV1RPRUHWKDQIRXUVQDLOVRIRQHVL]H class were replaced at one interval, so a choice RI SUH\ ZDV DOZD\V DYDLODEOH ([SHULPHQWV were run for four to ten days depending on consumption rates and prey availability. At the HQGRIWKHH[SHULPHQWVWKHQXPEHUVRIVPDOO prey and large prey consumed were added up for each container. Differences in preference between small and large prey were assessed by comparing the number of containers with more small prey consumed to containers with more large prey consumed using a sign test 6RNDO5RKOI Oxychilus alliarius Adult O. alliarius±PPVKHOOZLGWKZHUH RIIHUHGSUH\QDWLYHSuccinea spp. from areas where O. alliarius is abundant) of various VL]HV2QHSUH\LWHPZDVSODFHGLQDFRQWDLQHU with 2 cm of soil and one leaf of Broussaisia arguta to provide it with habitat and food. Prey ZHUHGLYLGHGLQWRVL[VL]HFODVVHV±± ±±±DQG±PPVKHOOOHQJWK 6HYHQW\IRXU H[SHULPHQWDO UHSOLFDWHV HDFK contained one adult O. alliarius in addition to the prey snail, while 48 control replicates each contained only the prey snail. Each trial was run for seven days at the end of which prey were recorded as alive or dead. which is useful because it permits standard errors, sl(o), to be calculated from binary data, as follows: where N 11 LV WKH QXPEHU RI H[SHULPHQWDO replicates in which the prey survived; N12 is the number of control replicates in which the species survived; N21LVWKHQXPEHURIH[SHULmental replicates in which the prey died; N22 is the number of control replicates in which the prey died. ,IPRUWDOLW\LVHTXDOLQFRQWURODQGH[SHULPHQWDOWUHDWPHQWVWKHORJRGGVUDWLRLV]HUR6LJQL¿cant differences in mortality were determined ZKHQDORJRGGVUDWLRZDVIXUWKHUWKDQ VWDQGDUGHUURUVDZD\IURP]HURSDQG IXUWKHUWKDQVWDQGDUGHUURUVDZD\IURP ]HURS Prey-Species Preference Euglandina rosea Prey-species preference of E. rosea was assessed by offering a choice between two of the ¿YHSUH\VSHFLHVOLVWHGDERYH3UH\ZHUH± PPLQH[WHQGHGIRRWOHQJWK7KHVL]HGLVWULEXtion, that is, the numbers of snails/slugs of variRXVVL]HVLQHDFKWULDOZDVPDWFKHGEHWZHHQ WKHWZRSUH\VSHFLHVLQRUGHUWRPLQLPL]HDQ\ HIIHFWRIVL]HLQGHWHUPLQLQJVSHFLHVSUHIHUHQFH 8VLQJWKHOHQJWKRIWKHH[WHQGHGIRRWUDWKHU WKDQDVKHOOVL]HPHDVXUHZDVQHFHVVDU\VLQFH VRPHRIWKHSUH\FKRLFHVZHUHVOXJVV. cubensis and D. laeve). Of the ten possible pair-wise FRPELQDWLRQVHLJKWZHUHWHVWHGD. laeve vs. S. tenella and D. laeve vs. A. fulicaZHUHH[cluded because of lack of adequate numbers of D. laeve). There were ten replicates for each pair-wise combination. A different E. rosea and SUH\WHQLQGLYLGXDOVRIWZRVSHFLHVZHUH used in each replicate. The number of each W\SHRISUH\FRQVXPHGQRUHSODFHPHQWRYHU a 24 h period was recorded. 138 MEYER & COWIE $SUHIHUHQFHLQGH[ZDVFDOFXODWHGIRUHDFK predator in each replicate/container using a YHUVLRQ RI 0DQO\¶V Į LQGH[ WKDW DFFRXQWV IRU variable prey populations, that is where prey DUHQRWUHSODFHG.UHEVDVIROORZV ZKHUHĮiLV0DQO\¶VĮSUHIHUHQFHLQGH[IRUSUH\ type i; pi, and pj are the proportions of prey i RUMUHPDLQLQJDWWKHHQGRIWKHH[SHULPHQWL «PM «PWKDWLVHi/ni in which ei is the number of prey type i remaining XQHDWHQDWWKHHQGRIWKHH[SHULPHQWQi is the LQLWLDOQXPEHURISUH\W\SHLLQWKHH[SHULPHQW DQGPLVWKHQXPEHURISUH\W\SHVVLPLODUO\IRU ej/nj9DOXHVIRUWKLVLQGH[UDQJHEHWZHHQDQG 1; values > 0.5 indicate preference and values DYRLGDQFH 6LJQL¿FDQFH RI GLYHUJHQFH IURP QR SUHIHUHQFH ZDV DVVHVVHG ZLWK single sample t-tests. LQWKHVL]HUDQJH±PPDQGLQWKHUDQJH 5–12 mm shell length) were then offered to an individual E. rosea for a period of 24 h. Each prey snail was weighed to the nearest 0.0001 JDQGLQGLYLGXDOO\PDUNHG$WWKHHQGRIWKHH[periment, the number of snails remaining alive was counted, and any shells remaining from consumed snails were weighed to the nearest 0.0001 g, these weights being subtracted from the weights of those snails at the beginning of WKH H[SHULPHQW WR JLYH D YDOXH RI WKH ZHLJKW consumed. Adding this amount to the known weight of snails consumed whole provided a value for the total weight consumed. Growth RISUH\VQDLOVRYHUWKHKRIWKHH[SHULPHQW was assumed to be negligible. Wet weight consumed was regressed on E. roseaVL]H0DQO\¶V ĮLQGH[GHVFULEHGDERYHZDVXVHGWRGH¿QH preference, followed by a Mann-Whitney U-test WRWHVWIRUVLJQL¿FDQWGLIIHUHQFHVLQSUHIHUHQFH IRUVPDOOSUH\±PPYVODUJHSUH\± mm) between adult and juvenile E. rosea. Oxychilus alliarius Prey-species preference of O. alliarius ZDV DVVHVVHG E\ RIIHULQJ DGXOWV ± PP PD[LPXPVKHOOZLGWKDVLQJOHSUH\LQGLYLGXDO RI YDULRXV VSHFLHV PP PD[LPXP VKHOO GLPHQVLRQIRUVQDLOVDQGVLPLODUO\VL]HGVOXJV GHWHUPLQHG E\ PDWFKLQJ IRRW VL]H EHWZHHQ slugs and snails) following the same procedure and analyses as in the O. alliariusSUH\VL]HH[periments. Prey species used were: the native snail species Succinea VSSH[SHULPHQWDO trials, 18 controls), TornatellidesVSS and PhilonesiaVSWKHLQYDVLYHVQDLO species P. achatinaceumDQGWKHLQYDsive slug species A. intermediusDQG D. leave Consumption Rate TABLE 1. Number of Achatina fulica eaten by Euglandina rosea when given a choice between WZRVL]HFODVVHV(DFKH[SHULPHQWXVHGDGLIIHUHQW E. rosea individual. Number Eaten 6L]H&ODVVHV 'XUDWLRQGD\V Smaller Larger 0–1 vs. 1–2 cm 0–1 vs. 2–3 cm 0–1 vs. 3–4 cm 1–2 vs. 2–3 cm Euglandina rosea Twelve E. rosea were used in consumption UDWH H[SHULPHQWV VHYHQ DGXOWV ± PP VKHOO OHQJWK ¿YH MXYHQLOHV ± PP VKHOO OHQJWK)RUDZHHNSULRUWRWKHH[SHULPHQWV they ZHUHSURYLGHGZLWKSUH\VQDLOVA.fulica, P. achatinaceum, B. similarisRIYDULRXVVL]HV ad libitum. Snails were not starved, so that the consumption rate measured would be a more QDWXUDO UDWKHU WKDQ D PD[LPXP UDWH 7KLUW\ SUH\ VQDLOV QRQQDWLYH Succinea tenella, 15 1–2 vs. 3–4 cm 2–3 vs. 3–4 cm Total 4 4 4 4 5 5 5 5 5 5 5 4 4 10 10 10 10 10 13 2 11 4 10 4 4 3 0 3 1 0 2 1 0 77 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 PREDATORY SNAILS IN HAWAII TABLE 2. Predation of Oxychilus alliarius on native SuccineaVSSRIGLIIHUHQWVL]HFODVVHVVKHOOOHQJWK P-Value based on log odds ratio tests. NS indicates P-Values > 0.05. Control ([SHULPHQWDO 6L]H&ODVV No. Survived No. Died 0–3 mm ±PP ±PP !PP 20 17 12 0 0 0 1 No. Survived No. Died P-value 33 0 0 0 NS NS NS 3 17 12 RESULTS Oxychilus alliarius Twelve O. alliarius were used in consumptionUDWHH[SHULPHQWV7KH\ZHUHFROOHFWHGIURPWKH ¿HOGDQGSODFHGGLUHFWO\LQWRLQGLYLGXDOFRQWDLQHUVHDFKZLWK¿YHSUH\VQDLOV Succinea spp., 1–3 mm shell length) for seven days. Each prey item was weighed to the nearest 0.0001 g and was individually marked prior to being placed LQWKHFRQWDLQHU$WWKHHQGRIWKHH[SHULPHQW the snails remaining alive were counted and any shells remaining from consumed snails were weighed to the nearest 0.0001 g, these weights being subtracted from the weights of WKRVHVQDLOVDWWKHEHJLQQLQJRIWKHH[SHULPHQW to give a value of the weight consumed. 3UH\6L]H3UHIHUHQFH Regardless of which snail combination E. rosea ZDV RIIHUHG 7DEOH LQ DOO FDVHV LQ ZKLFKDWOHDVWRQHSUH\ZDVFRQVXPHGRI 17) more of the smaller snails were consumed VLJQ WHVW 3 (YHQ EHLQJ H[WUHPHO\ conservative and considering the three instances in which no prey were eaten and the one in which two of the larger prey were eaten LQWKH±YV±FPWULDOVDVVXSSRUWLQJWKH null hypothesis of no preference, gives a twoVLGHG39DOXHRI3RROLQJWKHGDWD VQDLOVLQWKHVPDOOHUVL]HFODVVZHUHFRQVXPHG 7$%/(5HVXOWVRISDLUZLVHFRPSDULVRQRISUHIHUHQFHV0DQO\¶VĮRIE. roseaDPRQJ¿YHSUH\VSHFLHV Vc Veronicella cubensis, Dl Deroceras laeveWKHWZRVOXJVSHFLHVPa Paropeas achatinaceum, Af Achatina fulica, St Succinea tenella9DOXHVRIĮ!LQGLFDWHSUHIHUHQFH9DOXHVLQGLFDWH DYRLGDQFH$VLQJOHVDPSOHWWHVWZDVXVHGWRWHVWLIWKHSUHIHUHQFHLQGLFHVGLIIHUHGVLJQL¿FDQWO\IURP 6WDWLVWLFDOO\VLJQL¿FDQWGLIIHUHQFHVDUHLQGLFDWHGE\DVWHULVNV3YDOXHVDUHLQDOOFDVHV Treatment 0DQO\¶VĮ3UHIHUHQFH,QGH[6( Snails vs. Snails Pa vs. St Pa vs. Af Af vs. St Pa Pa Af S Af S Slugs vs. Snails Vc vs. Pa Vc vs. St Vc vs. Af Dl vs. Pa Vc Vc Vc Dl Pa S Af Pa Slugs vs. Slugs Dl vs. Vc Dl Vc 140 MEYER & COWIE ),*&RQVXPSWLRQUDWHZHWZHLJKWFRQVXPHGGXULQJDKSHULRGYVZHLJKW of Euglandina rosea. ),*0DQO\¶VĮLQGLFHVIRUDGXOWDQGMXYHQLOHE. rosea preference for smaller prey PPVKHOOOHQJWKRUODUJHUSUH\12 mm shell length). Error bars represent ± 1 standard error. PREDATORY SNAILS IN HAWAII 141 TABLE 4. Predation of Oxychilus alliarius RQQDWLYHDQGLQWURGXFHGVQDLOVSuccinea spp., Tornatellides spp., Philonesia sp., P. achatinaceumDQGVOXJVD. laeve and A. intermedius). P-Value based on log odds ratio tests. NS indicates P-values > 0.05. *Native species. Control ([SHULPHQWDO No. Survived No. Died No. Survived No. Died P-value Snails Succinea spp.* Tornatellides spp.* Philonesia sp.* P. achatinaceum 18 14 12 8 0 1 0 0 1 4 1 25 12 8 Slugs D. laeve A. intermedius 10 0 1 10 17 1 0 NS NS 7D[D FRPSDUHGWRRQO\WZRIURPWKHODUJHUVL]HFODVV Euglandina rosea thus prefers smaller snails when given a choice. In the O. alliarius H[SHULPHQWV QR VQDLO above 3 mm in shell length was preyed upon FRPSDUHGWRRIWKHWKDWZHUH mm in shell length or less; hence, there was QRVLJQL¿FDQWGLIIHUHQFHEHWZHHQFRQWURODQG H[SHULPHQWDOWUHDWPHQWVH[FHSWLQWKH±PP SUH\VL]HFDWHJRU\7DEOH Prey-Species Preference Euglandina rosea always preferred the snail over the slug species but showed no preferHQFHDPRQJWKHWKUHHVQDLOVSHFLHV7DEOH Veronicella cubensis was the least preferred prey, and D. laeve seemed to be preferred over V. cubensis but was less attractive than P. achatinaceum and therefore probably also OHVVDWWUDFWLYHWKDQWKHRWKHUVQDLOVSHFLHVS. tenella and A. fulica). Oxychilus alliarius readily consumed all snail species offered, as indicated by the much KLJKHUPRUWDOLW\LQWKHH[SHULPHQWDOWUHDWPHQWV but there was no difference in mortality between FRQWURODQGH[SHULPHQWDOWUHDWPHQWVIRUWKHWZR VOXJVSHFLHV7DEOH Consumption Rate Wet weight consumed increased as E. rosea ZHLJKW LQFUHDVHG )LJ 7KH KLJK U2 value )LJVXJJHVWVWKDWWKLVPDWKHPDWLFDOPRGHO describes the relationship well. The regression line was not forced to go through the origin but the y-intercept was nevertheless close to it. Adult E. rosea preferred smaller snails to larger VQDLOVĮ 6( 7 GI 3EXWMXYHQLOHE. rosea showed no SUHIHUHQFHĮ 6( 7 GI 3 7KHGLIIHUHQFHLQSUHIHUHQFH ZDVVLJQL¿FDQW8 3 )LJ The consumption rate for O. alliarius was 6( J RI ZHW ZHLJKW SHU GD\ZKLFKHTXDWHVWRDERXWRQHVQDLO±PP PD[LPXPVKHOOGLPHQVLRQFRQVXPHGE\HDFK LQGLYLGXDOGXULQJWKHH[SHULPHQWGD\V DISCUSSION Small snails were more vulnerable to predation by both E. rosea and O. alliarius. While O. alliarius can only consume small snails OHVV WKDQ PP PD[LPXP VKHOO GLPHQVLRQ E. rosea can consume larger prey. However, ZKHQ JLYHQ D FKRLFH RI SUH\ VL]HV E. rosea ZLOO HDW VLJQL¿FDQWO\ PRUH VPDOO VQDLOV WKDQ ODUJHVQDLOV7DEOH)LJ7KHVHUHVXOWVDUH FRQVLVWHQWZLWKRWKHUVWXGLHVRISUH\VL]HSUHIerence in E. rosea&KLX&KRX'DYLV %XWOHU1LVKLGD1DSRPSHWK &RRNDE*ULI¿WKVHWDO*HUODFK +RZHYHURXUUHVXOWVLQGLFDWHWKDWWKLV VL]HSUHIHUHQFHLVUHODWLYHQRWDEVROXWHWKDWLV PLGVL]HGSUH\±RU±FPLQVKHOOOHQJWK are not eaten if presented along with smaller prey, but are eaten if presented with larger SUH\7DEOH,WLVQRWVLPSO\WKDWODUJHUSUH\ FDQQRWEHHDWHQ7KLVEHKDYLRUDOÀH[LELOLW\LV interesting, and perhaps important for the dy- 142 MEYER & COWIE namics of the interaction, although the data are too few to evaluate further. Nonetheless, our results and those of other studies indicate that GLIIHUHQFHVLQVL]HDPRQJSUH\FDQLQÀXHQFH the predatory behavior of E. rosea, and thus confound conclusions of prey species preferHQFHLISUH\VL]HLVQRWDSSURSULDWHO\FRQWUROOHG LQH[SHULPHQWV 7KH VL]H RI E. rosea is also important in XQGHUVWDQGLQJSUH\VL]HSUHIHUHQFH$OWKRXJK counter-intuitive, larger E. rosea had a stronJHUSUHIHUHQFHIRUVPDOOHUSUH\)LJ7KLV may be because large E. rosea have a greater ability to consume snails whole. Thus, the impact of large E. rosea on small snails may EHHYHQJUHDWHU,QWKH¿HOGDODUJHPDMRULW\ of the species consumed by E. rosea were VZDOORZHGZKROH*ULI¿WKVHWDO2QH reason usually proposed for E. rosea’s preference for smaller snails is calcium intake. &RRNDVXJJHVWHGWKDWWKHUHPD\EHD compromise between the input of important nutrients, primarily calcium, when the prey is FRQVXPHGZKROHVKHOOLQFOXGHGDQGWKHFDloric intake that comes from the consumption of just body tissue. %RWKSUHGDWRUVH[KLELWHGDFOHDUSUHIHUHQFH for the snail species over the slug species in all WHVWV7DEOHV,QDOOH[SHULPHQWVH[DPLQing E. rosea’s preference between snails and slugs, it preferred the snails, while showing no preference among snail species in snail vs. VQDLOH[SHULPHQWV7DEOH+RZHYHUEHWZHHQ the slugs, E. rosea did prefer D. laeve over V. cubensis7DEOHOxychillus alliarius showed similar preferences, consuming all snail speFLHVRIIHUHGEXWQRVOXJV7DEOH7KHUHDUH no native slugs in Hawaii, but non-native slugs DUH DEXQGDQW DQG ZLGHVSUHDG &RZLH HW DO 2008; Meyer & Cowie, in press) and negatively LQÀXHQFH +DZDLLDQ HFRV\VWHPV E\ UHGXFLQJ VXUYLYRUVKLS RI \RXQJ QDWLYH SODQWV -RH Daehler, 2008). Thus, predation on slugs could a priori be seen as a positive counter balance to the introduction of these predatory snails, but unfortunately, our results show that both O. alliarius and E. rosea prefer snails to slugs, DQGPDQ\RIWKHVOXJVUHDFKVL]HVZHOODERYH WKH SUHIHUUHG VL]H UDQJH RI ERWK O. alliarius and E. rosea. Consumption rate was related to predator VL]H7KDWRIO. alliarus was much lower than that of E. rosea UHÀHFWLQJ WKH PXFK VPDOOHU VL]H RI O. alliarius; and within E. rosea conVXPSWLRQ UDWH LQFUHDVHG ZLWK SUHGDWRU VL]H )LJ+RZHYHUWKHLPSDFWRIO. alliarius on native snails may still be considerable because of the high densities this species can reach, notably at higher elevations on the islands of Maui and Hawaii. 0DQ\RIWKHH[WDQWQDWLYH3DFL¿F,VODQGVQDLOV DUH VPDOO HJ QXPHURXV HQGRGRQWLGV KHOLFLQLGVYHUWLJLQLGVDQGDFKDWLQHOOLGVH[FOXGLQJ $FKDWLQHOOLQDHDUHPPLQPD[LPXPVKHOO dimension for their entire life) compared to some of the abundant introduced snails/slugs WKDWKDYHEHFRPHHVWDEOLVKHGHJAchatina fulica, Cornu aspersum, Limax maximus, Veronicella cubensis, and Meghimatium striatum) &RZLHHWDO-RH'DHKOHU0H\HU & Cowie, in press). Most survive only in sparse SRSXODWLRQV /\GHDUG HW DO 0H\HU 7KHUHVXOWVLQGLFDWHWKDWO. alliarius and E. rosea would readily consume these native snails. Some native succineid species are also relatively small during the early stages of their life history, being less than 1 mm in shell length DWKDWFKLQJDQGUDUHO\H[FHHGLQJPPLQVKHOO OHQJWK%URZQHWDO5XQGHOO&RZLH 2003; W. M. M., personal observations), making them vulnerable to O. alliariusIRUWKH¿UVWIHZ months of their lives and to E. rosea for their entire lives. Achatinellids are more diverse in VL]H)RULQVWDQFHPDQ\$FKDWLQHOOLGDHZLWK WKHH[FHSWLRQRI$FKDWLQHOOLQDHDUHZLGHO\EXW SDWFKLO\GLVWULEXWHGDQGH[WUHPHO\VPDOOPDQ\ PPLQVKHOOOHQJWKDQGDUHWKXVYXOQHUDEOH to predation by both predators. Conversely, Achatinellinae, the Hawaiian tree snails, are larger. For instance, Achatinella mustelina are !PPDWELUWKDQGUHDFKVL]HVJUHDWHUWKDQ PP LQ VKHOO OHQJWK +DG¿HOG 0RXQWDLQ +DG¿HOGHWDO7KXVZKLOHWKH\ are not likely to be consumed by O. alliarius, their populations have declined following the introduction of E. rosea+DG¿HOG0RXQWDLQ +DG¿HOG+DG¿HOGHWDODV KDYHVLPLODUVL]HG3DUWXOLGDH3DFL¿FLVODQGWUHH snails) following the introduction of E. rosea HOVHZKHUH&ODUNHHWDO0XUUD\HWDO 0XUUD\*HUODFK&RRWH /RqYH Although consumption rates of O. alliarius are much lower than those of E. rosea, O. alliarius may be having a large impact on native land snails. Populations of O. alliarius are more dense than those of E. rosea in native rain forests, notably on the islands of Hawaii and Maui, and are becoming more prevalent on other islands at higher elevations where most of the H[WDQWQDWLYHVQDLOVDUHIRXQG6HYHUQV Meyer & Cowie, in press). Endangered Hawaiian tree snails are probably not impacted by O. alliarius, as they are too large even at birth, PREDATORY SNAILS IN HAWAII but much of the remaining Hawaiian land snail diversity may well be affected. Nevertheless, to ascertain the real individual and combined impacts of these predators on the land snail community as a whole and on populations of species of particular interest will require data RQWKHSRSXODWLRQGHQVLWLHVDQGVL]HGLVWULEXtions of these predators as well as on the lifehistories of the prey species. Unfortunately, data on population densities and life-histories for both prey and predators are few. However, investigations of these aspects of their ecology, combined with the preference and consumption rate data presented here, would permit models to be developed describing the dynamics of this system, which could help biodiversity managers protect native populations and control alien species. $OWKRXJK PDQ\ 3DFL¿F ODQGVQDLO VSHFLHV DUHH[WLQFWPDQ\XQLTXHVSHFLHVUHPDLQUHSresenting all of the native families. Knowledge of the feeding ecology of these predators is therefore essential if natural resources managers are to design better conservation strategies targeting these remaining native species. The results of this study are relevant beyond Hawaii as both predatory snails and the introduced VSHFLHVXVHGDVSUH\VSHFLHVLQWKHVHH[SHULments have been widely distributed by human DFWLYLWLHV 6ROHP &RZLH &RRWH /RqYH%DUNHU(IIRUG&iGL] & Gallardo, 2007). This may be especially pertinent for understanding the threat of these SUHGDWRU\ VQDLO VSHFLHV LQ WKH 3DFL¿F VLQFH PDQ\ QDWLYH 3DFL¿F LVODQG VQDLO IDXQDV DUH composed of species belonging to many of WKHVDPHIDPLOLHV&RZLHHWDO&RZLH :HVWLOOKDYHWKHRSSRUWXQLW\WRSURWHFW WKHVHXQLTXHOLQHDJHVIURPH[WLQFWLRQEXWWKLV is dependent on monitoring the presence of these predators and evaluating their impacts. ACKNOWLEDGEMENTS We thank the following for funding and support: U.S. Army Environmental, Oahu, the University of Hawaii Ecology, Evolution and &RQVHUYDWLRQ %LRORJ\ SURJUDP 16) JUDQW DGE05-38550 to K. Y. Kaneshiro). We thank Alvin Yoshinaga and staff at Lyon Arboretum, and the U.S. Army Environmental staff, especially 9LQFH&RVWHOORDQG6WHSKDQLH-RHIRUDFFHVV and help in collecting snails and slugs. We also thank Brian Bowen, Tad Fukami, Rebecca Ostertag and especially Andy Taylor for their helpful comments on a draft of this paper. 143 LITERATURE CITED BARKER, G. M. & M. G. EFFORD, 2004, Predatory gastropods as natural enemies of terrestrial gastropods and other invertebrates. Pp. ±LQG. M. BARKER, ed., Natural enemies of terrestrial molluscs. CABI Publishing, WallingIRUG[SS BROWN, S. G., B. K. SPAIN & K. CROWELL, $¿HOGVWXG\RIWKHOLIHKLVWRU\RIDQHQdemic Hawaiian succineid land snail. Malacologia, 45: 175–178. CÁDIZ, F. & C. S. GALLARDO, 2007, Arion intermedius *DVWURSRGD 6W\ORPPDWRSKRUD ¿UVWUHFRUGRIWKLVLQWURGXFHGVOXJLQ&KLOHZLWK notes on its anatomy and natural history. Revista Chilena de Historia Natural± CHIU, S. C. & K. C. CHOU2EVHUYDWLRQV on the biology of the carnivorous snail Euglandina rosea. Bulletin of the Institute of Zoology, Academia Sinica 7DLSHL± CIVEYREL, L. & D. SIMBERLOFF$WDOHRI two snails: is the cure worse than the disease? Biodiversity and Conservation, 5: 1231–1252. &/$5.(%&--0855$<06-2+1SON7KHH[WLQFWLRQRIHQGHPLFVSHFLHV by a program of biological control. 3DFL¿F6FLence± COOK, A.,D7KHRUJDQL]DWLRQRIIHHGLQJLQ the carnivorous snail Euglandina rosea. Malacologia± COOK, A., E )XQFWLRQDO DVSHFWV RI WUDLO following by the carnivorous snail Euglandina rosea. Malacologia, ± COOK, AD)DFWRUVDIIHFWLQJSUH\FKRLFH and feeding technique in the carnivorous snail Euglandina rosea Ferussac. Journal of Molluscan Studies± COOK, A. E7KH EDVLV RI IRRG FKRLFH E\ the carnivorous snail, Euglandina rosea. British Council for Crop Protection Monograph, ± COOTE, T. & É. LOÈVE)URPVSHFLHV WR ¿YH HQGHPLF WUHH VQDLOV RI WKH 6RFLHW\ ,Vlands fall prey to an ill-judged biological control progamme. Oryx± COWIE, R. H.,(YROXWLRQDQGH[WLQFWLRQRI 3DUWXOLGDHHQGHPLF3DFL¿FLVODQGODQGVQDLOV Philosophical Transactions of the Royal Society of London, %± COWIE, R. H3DFL¿FLVODQGODQGVQDLOV relationships, origins, and determinants of diversity. Pp. 347–372, in: A. KEAST & S. E. MILLER, eds. The origin and evolution of 3DFL¿F LVODQG ELRWDV 1HZ *XLQHD WR HDVWern Polynesia: patterns and processes. SPB Academic Publishing, Amsterdam, vi SS COWIE, R. H.,&DWDORJDQGELEOLRJUDSK\RI the nonindiginous nonmarine snails and slugs of the Hawaiian Islands. Bishop Museum Occasional Papers± COWIE, R. H3DWWHUQVRILQWURGXFWLRQRI non-indigenous non-marine snails and slugs in the Hawaiian Islands. Biodiversity and Conservation± 144 MEYER & COWIE COWIE, R. H., 2001, Can snails ever be effective and safe biocontrol agents? International Journal of Pest Management, 47: 23–40. COWIE, R. H., 2002, Invertebrate invasions on PaFL¿FLVODQGVDQGWKHUHSODFHPHQWRIXQLTXHQDWLYH faunas: a synthesis of the land and freshwater snails. Biological Invasions>@± COWIE, R. H., 2004, Disappearing snails and alien invasions: the biodiversity/conservation LQWHUIDFHLQWKH3DFL¿FJournal of Conchology, Special Publication, 3: 23–37. COWIE, R. H., N. L. EVENHUIS & C. C. CHRISTENSENCatalog of the native land and IUHVKZDWHU PROOXVFV RI WKH +DZDLLDQ ,VODQGV. %DFNKX\V3XEOLVKHUV/HLGHQYLSS COWIE, R. H., K. A. HAYES, C. T. TRAN & W. M. MEYER, III, 2008, The horticultural industry as a vector of alien snails and slugs: widespread invasions in Hawaii. International Journal of Pest Management± '$9,6&-*'%87/(5,QWURGXFHG enemies of the giant African snail Achatina fulica %RZGLFKLQ+DZDLL3XOPRQDWD$FKDWLQLGDH 3URFHHGLQJV RI WKH +DZDLLDQ (QWRPRORJLFDO Society± *(5/$&+-The ecology of western Indian Ocean carnivorous land snails. Phelsuma, 7: 14–24. *(5/$&+-Predator, prey and pathogen interactions in introduced snail populations. Animal Conservation± GRIFFITHS, O., A. COOK & S. M. WELLS The diet of the introduced carnivorous snail Euglandina rosea in Mauritius and its implications for threatened island gastropod faunas. Journal of Zoology± HADFIELD, M. G.([WLQFWLRQLQ+DZDLLDQ Achatinelline snails. Malacologia± HADFIELD, M. G. & B. S. MOUNTAIN$ ¿HOGVWXG\RIDYDQLVKLQJVSHFLHVAchatinella mustelina, in the Waianae Mountains of Oahu. 3DFL¿F6FLHQFH, 34: 345–358. HADFIELD, M. G., S. E. MILLER & A. H. CARWILE 7KH GHFLPDWLRQ RI WKH HQGHPLF Hawai‘ian [sic] tree snails by alien predators. American Zoologist± IUCN, 2008, The IUCN Red List of Threatened Species. International Union for Conservation of Nature and Natural Resources, Gland. http://www. LXFQUHGOLVWRUJ>DFFHVVHG)HEUXDU\@ -2( 6 0 & & '$(+/(5, 2008, Invasive slugs as an under-appreciated obstacle for rare plant restoration: evidence from the Hawaiian Islands. Biological Invasions, 10: 245–255. .5(%6&-Ecological methodology, 2nd ed. Addison-Welsey Educational Publishers, 0HQOR3DUN&DOLIRUQLD[LLSS LYDEARD, C., R. H. COWIE, W. F. PONDER, A. E. BOGAN, P. BOUCHET, S. A. CLARK, K. S. &800,1*67-)5(672*$5*20,1<' G. HERBERT, R. HERSHLER, K. E. PEREZ, B. ROTH, M. SEDDON, E. E. STRONG & F. G. THOMPSON, 2004, The global decline of nonmarine mollusks. BioScience, 54: 321–330. MEAD, A. R The giant African snail: a problem in economic malacology. University of &KLFDJR3UHVV&KLFDJR[YLLSS MEYER, W. M., III5HFRUGVRIUDUHJURXQG dwelling land snails on O‘ahu. Bishop Museum Occasional Papers, 88: 57–58. MEYER, W. M., III. & R. H. COWIE, in press, Temperate species are a threat to tropical island biodiversity. Biotropica. 0855$<(7KHVLQLVWHUVQDLOEndeavour, 17: 78–83. 0855$< - ( 0855$< 0 6 -2+1621 B. CLARKE7KHH[WLQFWLRQRIPartula on Moorea. 3DFL¿F6FLHQFH, 42: 150–153. NASH, S., 2004, Desperately seeking charisma: improving the status of invertebrates. BioScience± 1,6+,'$7%1$3203(7+(IIHFWRI DJHVSHFL¿FSUHGDWLRQRQDJHGLVWULEXWLRQDQG survival of the giant African snail, Achatina fulica. 3URFHHGLQJV RI WKH +DZDLLDQ (QWRPRORJLFDO Society± 581'(//5-5+&2:,(*URZWK and reproduction in Hawaiian succineid land snails. Journal of Molluscan Studies ± 581'(//5-%6+2//$1'5+&2:,( 2004, Molecular phylogeny and biogeography RIHQGHPLF+DZDLLDQVXFFLQHLGODQGVQDLOV3XOmonata: Grastropoda). Molecular Phylogenetics and Evolution± 6(9(5160$QRWKHUWKUHDWWR+DZDLL¶V endemics. +DZDLLDQ6KHOO1HZV 6,0%(5/2)) ' :K\ GR LQWURGXFHG species appear to devastate islands more than mainland areas. 3DFL¿F6FLHQFH± 62.$/55)-52+/), Biometry: the principles and practice of statistics. W. H. FreePDQ&RPSDQ\1HZ<RUN[L[SS 62/(0$1HZUHFRUGVRI1HZ&DOHGRnian nonmarine mollusks and an analysis of the introduced mollusks. 3DFL¿F 6FLHQFH, 18: 130–137. 62/(0 $ +RZ PDQ\ +DZDLLDQ ODQG snail species are left? And what we can do for them. Bishop Museum Occasional Papers, 30: 27–40. 5HYLVHGPVDFFHSWHG)HEUXDU\