CHEMICAL Attraction of capitata (Diptera: ) and Endemic and Introduced Nontarget to BioLure Bait and Its Individual Components in

1 2 LUC LEBLANC, ROGER I. VARGAS, AND DANIEL RUBINOFF

Department of Plant and Environmental Protection Sciences, University of Hawaii, 3050 Maile Way, Room 310, Honolulu, HI 96822-2271

Environ. Entomol. 39(3): 989Ð998 (2010); DOI: 10.1603/EN09287 ABSTRACT BioLure, a synthetic food attractant for Mediterranean fruit ßy [Ceratitis capitata (Wiedemann)], is composed of three chemicals (ammonium acetate, trimethylamine hydrochloride, and putrescine). We deployed these components together and in separate MultiLure traps across predominantly native forests, non-native forests, farmlands, orchards, and residential areas on the islands of Hawaii and Maui, to evaluate attraction of C. capitata and nontarget insects. Large numbers (as many as 186 per trap per day) of mainly saprophagous nontarget ßies (primarily Drosophilidae, , Lonchaeidae, Neriidae, Otitidae, and Calliphoridae) were attracted to BioLure. Very few predators, parasitoids, or pollinators were attracted. Native species, predominantly drosophilid and calliphorid ßies, were attracted in large numbers in endemic forests, but mostly (at least 88%) introduced species were collected in orchards, backyards, and non-native forest. A comparison of attraction to the three separate components versus combined components in traps revealed that ammonium acetate and, to a lesser extent, putrescine are the key components attractive to nontarget species. Omitting the putrescine ingredient from BioLure did not drastically decrease C. capitata catches but reduced nontarget captures by 20%.

KEY WORDS attractant, nontarget, Drosophilidae, putrescine, ammonium

The Mediterranean fruit ßy [Ceratitis capitata leased from vials on sticky ball traps, proved to be as (Wiedemann)], a widespread fruit that breeds on attractive as torula yeast in orchards (Robacker 1995). Ͼ300 hosts (Liquido et al. 1991), does not respond An improved delivery system, using a two-component well to male-speciÞc lures. Therefore, quarantine sur- dry lure with ammonium acetate and putrescine inside veillance and population monitoring of fruit ßy species a vertical plastic trap with lateral entrance holes and like C. capitata have traditionally relied on glass a toxicant panel, was very attractive to both A. ludens McPhail traps with an aqueous protein solution as a and C. capitata (Heath et al. 1995). Replacing the food lure. However, these traps are bulky and hard to toxicant panel with a sticky insert in an open bottom service, and the protein solution is messy, must be trap (Heath et al. 1996) and adding trimethylamine to replaced regularly, does not preserve ßy specimens the other components in membranes (Heath et al. adequately, and attracts numerous nontarget insects 1997) increased C. capitata captures without reducing (Heath et al. 1995), especially when previously cap- A. ludens catches. The three-component lure is more tured insects start decaying inside the traps (Leblanc attractive to C. capitata than Nu-Lure or the two- et al. 2009b). component lure under widely varying environmental Improved monitoring systems developed in the last conditions in Mediterranean, Central American, and two decades use volatiles from the bacterial break- Indian Ocean countries (Epsky et al. 1999, Katsoyan- down of protein hydrolysate that are attractive to fruit nos et al. 1999). ßies. Among various combinations tested in the lab- Synthetic food lures have great potential for reduc- oratory, the mixture of ammonium bicarbonate, meth- ing C. capitata damage by mass trapping. This was ylamine hydrochloride, and putrescine was the most successfully shown by setting traps at the periphery of potent attractant for the Mexican fruit ßy [Anastrepha plum, , and persimmon orchards in Israel (Cohen ludens (Loew)] (Robacker and WarÞeld 1993). This and Yuval 2000). Combining food lures with weekly synthetic lure, when applied as a dry formulation re- GF-120 protein bait treatments in mass-trapping ef- forts prevented the buildup of C. capitata populations in coffee and colonization of adjacent persimmon or- 1 Corresponding author, e-mail: [email protected]. 2 U.S. PaciÞc Basin Agricultural Research Center, USDAÐARS, PO chards in Hawaii (McQuate et al. 2005). Mass-trapping Box 4459, Hilo, HI 96720. with BioLure is now a standard practice widely used 990 ENVIRONMENTAL ENTOMOLOGY Vol. 39, no. 3 in Spain for C. capitata areawide control on citrus pylene glycol (Sierra Antifreeze; Old World Indus- (Navarro-Llopis et al. 2008). tries, Northbrook, IL) was used in the bottom of the Nontarget attraction to liquid and sticky protein traps. All traps were hung on trees, 1.5Ð2 m above the bait traps has been reported in literature for decades ground, using 15-gauge aluminum tie wire. (Hardy 1952, Moore 1969, Neuenschwander et al. Experiment 1: Attraction to BioLure and Its Com- 1981, Asquith and Messing 1992, Thomas 2003). Al- ponents. At three sites, traps were baited with all three though most nontargets were saprophagous Diptera, components and the separated components to char- substantial numbers of beneÞcial Tachinidae (Moore acterize nontarget diversity and determine the com- 1969, Thomas 2003) and honeydew-feeding Chrysopi- ponents most attractive to nontargets. The forest site dae (Neuroptera) (Neuenschwander et al. 1981) were (Ͼ70 ha) was in a largely intact native wet montane captured. Nontarget captures were reduced through forest, dominated by ohia (Metrosideros polymorpha the use of synthetic food lures compared with protein Gaudich) (Myrtaceae), in the Upper Waiakea Forest lure in water. The three-component BioLure reduced Reserve, on Hawaii Island (site 2 on Fig. 1 in Leblanc nontarget ßy catches 15-fold compared with Naziman et al. 2009b) (900 m above sea level). The citrus site liquid bait (Gazit et al. 1998), 3-fold compared with (0.1 ha) was an (Citrus sinensis L. Osbeck) NuLure (Katsoyannos et al. 1999), and by one half orchard at the University of HawaiiÕs Waiakea Agri- (Martinez et al. 2007) and 4-fold (Thomas 2003) com- cultural Research Station, on Hawaii Island (site 8 on pared with torula yeast in McPhail traps. However, Fig. 1 in Leblanc et al. 2009b) (175 m). The coffee site lacewings (Chrysopidae) were equally (Thomas was an unmaintained, weedy coffee plantation (0.25 2003) or twice as attracted (Martinez et al. 2007, Con- ha) in Kula, Maui Island (site 42 on Fig. 3 in Leblanc way and Forrester 2007) in a two-component lure et al. 2009b) (910 m). At each site, traps were laid out compared with torula yeast traps. These nontarget in a randomized block design, with traps 7Ð10 m apart, risks associated with the use of BioLure must be con- as three rows (blocks) of Þve traps: one baited with sidered carefully in sensitive island environments such the three component membranes together, three as Hawaii, where a large diversity of endemic insects baited with the individual components, and one un- occurs, including 12 endangered species of Drosophi- baited control, with only the liquid preservative. Po- lidae (U.S. Fish and Wildlife Service 2007). sition assignment of traps within blocks was random. The objectives of this study were to document the Traps were maintained and emptied weekly for 9 wk range of nontarget species attracted to BioLure across in the citrus site (December 2005 to January 2006) and a range of habitats in Hawaii, from native forest hab- 10 wk in the coffee site (JuneÐAugust 2006). To avoid itats to non-native forests bordering agriculture to overcollecting endemic insects, traps in the forest site orchard environments. Additionally, we compared, were maintained for only 14 d and emptied every 2 d for the Þrst time, nontarget attraction of the individual in August 2005. Trap assignment to each support tree BioLure components and determined whether the was rerandomized within each block after each col- omission of the putrescine ingredient results in de- lection in all study sites. creased nontarget catches without compromising cap- Experiment 2: Habitat Effect and Omission of Pu- tures of target C. capitata. Results from this study will trescine. We also sought to characterize BioLure at- be useful in future recommendations for the safe use traction in orchard environments and their adjacent of BioLure that minimizes attraction of beneÞcial and forest habitats and the effects of putrescine omission endemic nontargets. on trap catches. To do this, we placed one set of three traps at each of 17 sites across eight locations (see Fig. 3 in Leblanc et al. 2009b for map of locations) in the Materials and Methods agricultural community of Kula (Maui island) (517Ð Traps and Lures. We used MultiLure traps (Better 1,138 m) over a 26-wk period. Eleven sets of these World Manufacturing, Fresno, CA) baited with Bio- traps were in persimmon (Diospyros kaki L.Þl.) Lure (Suterra, Bend, OR) fruit ßy food lure. The (Ebenaceae) orchards, four sets were in coffee or- plastic MultiLure trap (see Thomas 2003 for photo) is chards (including the experiment 1 site), and two sets similar in principle to the traditional open bottom glass were in the non-native forest adjacent to persimmon McPhail trap but consists of a transparent cover that orchards and coffee. Every site had one trap with the interlocks with an opaque bottom-half of varying three components, a second with two components (no color. The base of the trap is a collecting vessel for putrescine), and a third unbaited trap as a control. trapped insects. Yellow-bottom traps were used in this Traps were cleared weekly for the Þrst 13 wk (JuneÐ study. BioLure consists of three components (ammo- August 2006) and then monthly for the last 3 mo nium acetate, trimethylamine hydrochloride, and pu- (SeptemberÐNovember). Trap position at each site trescine) in separate packets with slow-release mem- was rerandomized every 3 wk or monthly in the last 3 branes that are attached to the inner surface of the trap mo. Membranes in all traps were replaced once after cover. This trapping system has been used by growers 13 wk, because they remain attractive to A. ludens and to reduce C. capitata populations on Maui since the C. capitata for up to 3Ð4 mo (Thomas et al. 2001, inception of the Area-Wide Fruit Integrated Pest Robacker and Czokajlo 2005, Navarro-Lopis et al. Management (IPM) Program in Hawaii (McQuate et 2008). al. 2005). To kill and preserve the captured Sample Processing and Data Analysis. All specimens for identiÞcation, 200 ml of a 20% solution of pro- were sorted, counted, sexed, and identiÞed to species June 2010 LEBLANC ET AL.: NONTARGET ATTRACTION TO BIOLURE IN HAWAII 991 whenever possible. For a few samples with large num- Results and Discussion bers of Drosophilidae from the citrus orchard on Ha- Overall Attraction. More than 302 species of non- waii island, numbers were estimated by counting ßies target arthropods were collected, in 217 genera, 88 in a 2-ml subsample (631 Ϯ 133 SEM; n ϭ 38). Pinned families, and 15 orders. We identiÞed 96.1% of the voucher specimens of all species were deposited at the specimens to species and a further 2.8% to only. University of Hawaii Museum (Manoa) and the Captures were numerically dominated by Diptera Bernice P. Bishop Museum (Honolulu). (94.3%), representing at least 161 species in 81 genera All counts were converted to number of insects per and 24 families, followed by Hymenoptera (1.7%), trap per day, and data were analyzed using analysis Lepidoptera (1.5%), and Coleoptera (1.2%). The of variance (ANOVA), with the minimum variance dominance of Diptera was also reported in Mexico unbiased quadratic estimation (PROC MIXED (79.6% of nontarget captures) (Thomas 2003) and MIVQUE0; SAS Institute 2004). MIVQUE0 provides (55.4%) (Martinez et al. 2007). Mean non- reliable estimates of parameters for data with a non- target captures in Kula (2.5 per trap per day) were normal distribution, large numbers of zero values, and comparable to other studies (2.3 and 18 per trap per unequal variances. There was no decrease in trap day by Thomas 2003 and Gazit et al. 1998, respec- catches over time, from lure or insect depletion, for tively), but were extremely high (186 per trap per any of the species. Weekly capture data from each trap day) at the experiment 1 citrus site. At most sites, are used as replicate data for individual traps in the captures were predominantly of nontargets: 100% in statistical analyses. Results on the tables and graphs native forest, 98.7% at the citrus site, and 68.6% among Ϯ are presented as means ( SEM) across all trapping the Kula sites, comparable to Ͼ90% reported by periods. Thomas (2003). However, target C. capitata numeri- To characterize general attraction to BioLure (ex- cally dominated trap captures (65.7%) at the experi- periment 1), we used data from the three-component ment 1 coffee site. The endemic forest site was dom- traps and the unbaited controls. Data were analyzed inated by endemic and introduced Drosophilidae separately by family (or order for less abundant (39.2 and 56.5% of all captures, respectively) and en- groups) and by site, using a mixed model with lure demic Calliphoridae (3.7%), whereas the great ma- treated as Þxed and block as random. Results are pre- jority (Ͼ90%) of captures were of introduced species Ϯ sented as mean ( SEM) captures per trap per day for in all the non-native sites. each group at each site. Groups signiÞcantly attracted Captures by three-component BioLure traps in ex- Ͻ (P 0.05) to BioLure are indicated, with F and P periment 1 (Table 1) were dominated by Drosophi- values in footnotes to the table. Capture data for un- lidae (92.6%) and Chloropidae (4.4%). BioLure cap- baited control traps, not signiÞcantly (P Ͻ 0.05) dif- tured signiÞcantly more (P Ͻ 0.05) individuals than ferent from zero, are not included on the table. unbaited controls for most Diptera and some of the Attraction to BioLure and to its separate compo- Hymenoptera and Coleoptera. Species attracted to nents (experiment 1) was compared for all species, or BioLure largely belong to families whose larvae are complexes of related species, represented by Ն200 scavengers on decaying plant or matter (Dros- captured specimens. A mixed model, with lure as Þxed ophilidae, Chloropidae, Lonchaeidae, Neriidae, Otiti- factor and block as random factor, was used. Least dae, Phoridae, Anthomyidae, Calliphoridae, Musci- square means estimates for habitat were compared dae, Sarcophagidae, and Nitidulidae). Most of these using TukeyÕs honest signiÞcant difference test. families were also dominant in BioLure (Thomas Habitat effect on nontarget captures for the Kula 2003) and protein hydrolysate (Hardy 1952, Asquith area (experiment 2) was analyzed by family or order. and Messing 1992) traps in other studies and are at- Capture data from the three-component traps was tracted to decaying fruit ßies that accumulate inside used, comparing three habitats: coffee orchards (4 male lure traps (Leblanc et al. 2009b). sites), persimmon orchards (11 sites), and non-native BioLure at the experiment 1 coffee site attracted forest adjacent to orchards (2 sites). Least square large numbers of C. capitata, whose larvae breed in means were compared using TukeyÕs honest signiÞ- ripe coffee berries, as well as smaller numbers of non- cant difference test. Results are shown only for groups target Tephritidae, mainly the ßower-breeding Di- that demonstrated a signiÞcant (P Ͻ 0.05) habitat oxyna sororcula (Wiedemann), but also a few speci- effect. mens of xanthochaeta Aldrich, introduced for The effect of putrescine omission (experiment 2) control of the weed Lantana camara L. (Verbenaceae) was determined by comparing captures of C. capitata, (Hardy and DelÞnado 1980). BioLure may have a local B. dorsalis, and all the nontarget species with at least impact on nontarget tephritid species, including those 100 trapped specimens, between traps with the three introduced for weed biological control. Therefore, it ingredients and traps without the putrescine. For C. should not be used in native habitats, where 25 en- capitata, data for female and male ßies were also sep- demic tephritids occur (Hardy and DelÞnado 1980). arately analyzed for sites with low (0.003Ð0.3 ßies per Drosophilidae is an exceptionally diverse group in trap per day), moderate (0.6Ð1.8 per trap per day), Hawaii, with 559 described endemic (OÕGrady et al. and high (10.99 per trap per day) ßy densities, based 2010) and 32 introduced species (Leblanc et al. on groupings in a frequency distribution histogram of 2009a), and it dominated nontarget captures, regard- mean captures by sites. less of trap location. Introduced drosophilids, mostly 992 ENVIRONMENTAL ENTOMOLOGY Vol. 39, no. 3

Table 1. Arthropods collected in three-component BioLure traps in a citrus orchard and an endemic forest on Hawaii Island and an abandoned coffee orchard in Kula, Maui island

MeanϮ SEM no. captured per trap per day Citrus Coffee Endemic forest Target species Tephritidae (I)a 2.364 Ϯ 0.341b 15.881 Ϯ 2.729b 0.0 Nontarget species Diptera Anthomyiidae (I) 0.0 0.117 Ϯ 0.026b 0.0 Calliphoridae (E-I)c 0.012 Ϯ 0.012 0.006 Ϯ 0.006 3.786 Ϯ 0.537b Cecidomyiidae (I) 0.231 Ϯ 0.067b 0.005 Ϯ 0.005 0.0 Chloropidae (I) 4.674 Ϯ 1.436b 4.031 Ϯ 0.790b 0.0 Drosophilidae (E) 0.0 0.010 Ϯ 0.010 40.048 Ϯ 8.309b Drosophilidae (I) 178.691 Ϯ 72.074b 2.574 Ϯ 0.394b 57.667 Ϯ 8.337b Lonchaeidae (I) 0.233 Ϯ 0.068b 0.016 Ϯ 0.012 0.0 Muscidae (I) 0.071 Ϯ 0.025b 0.125 Ϯ 0.032b 0.0 Neriidae (I) 1.440 Ϯ 0.318b 0.0 0.0 Otitidae (I) 0.434 Ϯ 0.085b 0.091 Ϯ 0.028b 0.0 Phoridae (E-I) 0.012 Ϯ 0.008 0.257 Ϯ 0.057b 0.405 Ϯ 0.132b Sarcophagidae (I) 0.058 Ϯ 0.018b 0.041 Ϯ 0.015 0.0 Tephritidae (nontarget) (I) 0.0 0.683 Ϯ 0.230b 0.0 Diptera misc (E-I) 0.161 Ϯ 0.030b 0.075 Ϯ 0.023 0.048 Ϯ 0.033 Hymenoptera Chalcidoidea (I) 0.011 Ϯ 0.007 0.197 Ϯ 0.060b 0.0 Formicidae (I) 0.060 Ϯ 0.020 0.006 Ϯ 0.006 0.0 Hymenoptera misc (E-I) 0.040 Ϯ 0.022 0.035 Ϯ 0.014 0.024 Ϯ 0.024 Coleoptera Coccinellidae (I) 0.012 Ϯ 0.007 0.013 Ϯ 0.009 0.024 Ϯ 0.024 Corylophidae (I) 0.012 Ϯ 0.008 0.020 Ϯ 0.010 0.0 Nitidulidae (E-I) 0.005 Ϯ 0.005 0.231 Ϯ 0.053b 0.0 Coleoptera misc (I) 0.010 Ϯ 0.007 0.006 Ϯ 0.006 0.024 Ϯ 0.024 Hemiptera (E-I) 0.171 Ϯ 0.083 0.019 Ϯ 0.011d 0.024 Ϯ 0.024 Lepidoptera (E-I) 0.005 Ϯ 0.005 0.100 Ϯ 0.023d 0.0 Araneae 0.0 0.012 Ϯ 0.008 0.0 Arthropoda misc 0.037 Ϯ 0.015 0.017 Ϯ 0.010 0.0 Total nontargets 186.376 Ϯ 71.799b 8.684 Ϯ 1.107b 102.048 Ϯ 16.165b Percent endemic (including Drosophilidae)e 0.07Ð0.09 2.6Ð4.4 43.0Ð43.4 Percent endemic (excluding Drosophilidae)e 3.9Ð5.0 12.9Ð22.4 88.5Ð97.8

a Target Tephritidae include Ceratitis capitata, Bactrocera cucurbitae (Coquillett), and B. dorsalis. b Mean captures are signiÞcantly higher (P Ͻ 0.05) in BioLure than control traps (means not displayed) within a site, ANOVA, PROC MIXED mivque0 (SAS Institute 2004). Degrees of freedom are 1,50 for citrus; 1,56 for coffee, and 1,38 for forest. P Ͻ 0.0001 for all signiÞcant differences, unless otherwise indicated. Tephritidae (target) cit.: F ϭ 50.59; cof.: F ϭ 42.66; Anthomyiidae cof.: F ϭ 9.92, P ϭ 0.0026; Calliphoridae for.: F ϭ 50.46; Cecidomyiidae cit.: F ϭ 11.37, P ϭ 0.0014; Chloropidae cit.: F ϭ 10.73, P ϭ 0.0019; cof.: F ϭ 26.95; Drosophilidae (endemic) for.: F ϭ 25.28; Drosophilidae (introduced) cit.: F ϭ 5.93, P ϭ 0.0185; cof.: F ϭ 32.82; for.: F ϭ 50.78; Lonchaeidae cit.: F ϭ 11.71, P ϭ 0.0012; Muscidae cit.: F ϭ 8.17, P ϭ 0.0062; cof.: F ϭ 12.20, P ϭ 0.0009; Neriidae cit.: F ϭ 20.54; Otitidae cit.: F ϭ 25.25; cof.: F ϭ 11.56, P ϭ 0.0012; Phoridae cof.: F ϭ 18.52; for.: F ϭ 9.52, P ϭ 0.0038; Sarcophagidae cit.: F ϭ 7.41, P ϭ 0.0089; Tephritidae (nontarget) cof.: F ϭ 10.77, P ϭ 0.0018; Diptera (misc) cit.: F ϭ 5.93, P ϭ 0.0185; Chalcidoidea cof.: F ϭ 8.88, P ϭ 0.0043; Nitidulidae cof.: F ϭ 11.20, P ϭ 0.0015; total nontargets cit.: F ϭ 6.47, P ϭ 0.0141; cof.: F ϭ 47.49; in for.: F ϭ 42.75. c Only endemic Calliphoridae (Dyscritomyia spp) at forest site, and only introduced Calliphoridae at other sites. d Mean captures are signiÞcantly higher (P Ͻ 0.05) in control (means not shown) than BioLure traps for Hemiptera (F ϭ 5.33, P ϭ 0.0247) and Lepidoptera (F ϭ 5.06, P ϭ 0.0285) in the coffee site (df ϭ 1,56), TukeyÕs test, PROC MIXED mivque0 (SAS Institute 2004). e % nontarget individuals of endemic species, when including and excluding Drosophilidae, presented as ranges, because of uncertainty of status for several species of Phoridae and Lepidoptera. E, endemic; I, introduced; E-I, family includes endemic and introduced species in samples.

Drosophila immigrans Sturtevant, D. sulfurigaster bil- common species of most of the other endemic groups imbata Bezzi, and D. suzukii (Matsumura), were abun- (haleakalae, picture wings, Nudidrosophila, split tar- dant at every site, including the endemic forest. En- sus, ciliated tarsus, bristle tarsus, modiÞed mouthparts, demic drosophilids were trapped in large numbers (40 and the genus ) are also attracted to Bio- per trap per day, 26 species) at the native forest site, Lure (L.L., unpublished data). The use of BioLure is dominated by the Antopocerus species group (two therefore not recommended in endemic forest, where species and 25% of endemic drosophilid captures) and the male attractant Trimedlure may be used instead to the “modiÞed tarsus” group (nine species, 71% of cap- monitor C. capitata, because its nontarget attraction is tures; mostly species of the “spoon tarsus” subgroup). minimal (Kido and Asquith 1995, Uchida et al. 2006). The larval breeding substrate for the majority of the Whereas introduced drosophilids were equally trapped endemic species is decaying leaves of Chei- abundant in habitats with native or introduced vege- rodendron trigynum (Gaud.) (Araliaceae) (Magnacca tation, only small numbers of endemic species were et al. 2008). Trapping at 57 other endemic forest sites trapped at four of the locations in Kula, several kilo- on Hawaii and Maui islands has shown that at least the meters away from endemic forest. They were slightly June 2010 LEBLANC ET AL.: NONTARGET ATTRACTION TO BIOLURE IN HAWAII 993 more common at the highest (1,138 m) location Corylophidae in protein bait traps. All these afore- (0.031 Ϯ 0.011 SEM per trap per day) than the three mentioned beetles are attracted to decaying ßies in other locations (0.012 Ϯ 0.004). The catches mainly traps (Leblanc et al. 2009b). consisted of three species in the haleakalae Very few beneÞcial insects were attracted to Bio- group, whose larvae breed on fungus (Magnacca et al. Lure in our survey, apart from small numbers of Ta- 2008), but also included nine species of Scaptomyza chinidae and one species of Encyrtidae. Only 0.97% of trapped in very small numbers (3.4 specimens per the BioLure captures were predators and parasitoids, species). Most of these endemic species were com- even lower than the 9.4 and 3.5Ð3.8% reported by mon in the Maui endemic forest traps (L.L., unpub- Thomas (2003) and Martinez et al. (2007), respec- lished data). The widespread occurrence of small tively. Two introduced tachinids, Archytas cirphis Cur- numbers of Scaptomyza (Bunostoma) anomala Hardy ran and Eucelatoria armigera (Coquillett), were at- in mixed forest sites, far from endemic habitats, was tracted to our BioLure traps in modest numbers previously documented on Kauai Island (Asquith (0.010 Ϯ 0.004 SEM per trap per day in Kula). Among 1995). Two species of the same subgenus (S. confusa parasitoid Hymenoptera, BioLure only attracted the Hardy and S. xanthopleura Hardy) were among the houseßy parasitoid Tachinaephagus zealandicus Ash- infrequent captures of endemic Scaptomyza in Kula. mead (Encyrtidae), a species also attracted to decay- This suggests that some endemic species, whereas ing fruit ßies (Leblanc et al. 2009b). Captures of much more common in native forest, do occur in small pollinators were negligible, again consistent with pre- numbers in other habitats where they may form a vious Þndings (Thomas 2003). small component of nontarget captures. Although very few Chrysopidae were trapped dur- Other families of plant scavenger ßies, predomi- ing the whole study, 18 of the 20 specimens of the two nantly Chloropidae, Lonchaeidae, Neriidae, and Otiti- nectar-feeding species [Mallada basalis (Walker) and dae, were drawn to BioLure at all sites except the Chrysoperla comanche (Banks)] collected in Kula endemic forest. Decaying fruit on the ground is prob- were in BioLure traps, rather than the controls. Ex- ably their primary larval breeding substrate at the perimental evidence amply supports the attraction of trapping sites (Hardy and DelÞnado 1980). Thomas several nectar-feeding adult chrysopids, but not the (2003) had predominantly captured decaying leaf- species with predatory adults, to protein hydrolysate breeding (35% of all Diptera), rather than (Neuenschwander et al. 1981). Chrysopid attraction Drosophilidae (11%) in Mexico. Although much less to BioLure and torula yeast were also reported by numerous (0.036 Ϯ 0.008 SEM per trap per day at the Thomas (2003) (only nectar-feeding species at- citrus and coffee sites), two species of introduced tracted), Martinez et al. (2007) (3.6% of all nontarget Lauxaniidae were clearly attracted to BioLure in our captures), and Conway and Forrester (2007). Nectar- study. feeding lacewings are also attracted to the male lure Humpbacked ßies (Phoridae), of which all species methyl eugenol (Leblanc et al. (2009b). We found no in Hawaii are believed to be scavengers (Hardy 1964), evidence of predatory brown lacewings (Hemerobi- were attracted to BioLure at all sites. One species idae) being attracted to BioLure or any of its compo- reported by Hardy as endemic (Megaselia furcatilis nents, with one species (Micromus timidus Hagen) Beyer) was regularly collected (0.12 Ϯ 0.03 SEM per collected in equal numbers (0.049 Ϯ 0.007 SEM per trap per day) in Kula. Because taxonomic knowledge trap per day) in all baited and control traps at the of Hawaiian Phoridae is incomplete, as reßected by experiment 1 citrus site. the predominance of undescribed native or unre- Habitat Effect. We compared captures of target ported introduced species in traps baited with decay- tephritids and all nontarget insects in the three-com- ing ßies (Leblanc et al. 2009b), some of the species ponent BioLure traps in Kula between traps set in assumed to be endemic, including M. furcatilis, may coffee (4 sites), persimmon orchards (11 sites), and actually be introduced. non-native forest (2 sites). In many cases, the sites Native Calliphoridae (Dyscritomyia) were strongly were Ͻ50m apart. Data for families that showed sig- attracted to BioLure at the endemic forest site. These niÞcant (P Ͻ 0.05) habitat-related differences are pre- slow-reproducing viviparous ßies, which bear one sented in Table 2. to maturity at a time (Hardy and DelÞnado Mediterranean fruit ßies were trapped in large num- 1980), would likely be threatened by the continuous bers only in the coffee sites. Pest fruit ßies generally use of BioLure in Hawaiian forests. The same spe- invade orchards from neighboring vegetation when cies of Dyscritomyia were equally drawn (2.7 per ripe fruit becomes available (Aluja 1996). In Kula, trap per day) to traps baited with decaying fruit ßies populations of C. capitata are maintained through a (Leblanc et al. 2009b). Outside native forest, only succession of bridge hosts, with loquats, Þgs, and introduced Calliphoridae were trapped, and in very peaches as the main reservoir, and the ßies invade small numbers (0.002Ð0.009 per trap per day), con- persimmon orchards only when they bear fruit (Wong sistent with results of Asquith and Messing (1992). et al. 1983). Although coffee is scarce in Kula, it is The only beetles attracted to BioLure were the heavily infested by C. capitata and can be a locally minute Corylophidae (0.055 Ϯ 0.010 SEM in Kula), important source of infestation when coffee is grown which feed on fungus spores, and the saprophagous adjacent to persimmon (McQuate et al. 2005). Nitidulidae. Asquith and Messing (1992) also reported The overall number and diversity of more than six the captures of introduced Staphylinidae (Atheta) and families of nontarget ßies and beetles was also signif- 994 ENVIRONMENTAL ENTOMOLOGY Vol. 39, no. 3

Table 2. Comparison of insect captures in three-component BioLure traps in coffee orchards (4 sites, 64 samples), persimmon orchards (11 sites, 158 samples), and nonnative forest adjacent to orchards (2 sites, 32 samples) in Kula, Maui Island

MeanϮ SEM no. captured per trap per day a F Nonnative Coffeeb Persimmonb forestb Target species Tephritidae (I)c 24.16 4.904 Ϯ 1.200a 0.0b 0.089 Ϯ 0.015b Nontarget species Diptera Anthomyiidae (I) 4.35 0.068 Ϯ 0.027a 0.0b 0.019 Ϯ 0.007b Drosophilidae (E) 40.97 0.004 Ϯ 0.002b 0.110 Ϯ 0.029a 0.004 Ϯ 0.002b Drosophilidae (I) 4.94 0.817 Ϯ 0.121b 1.634 Ϯ 0.539a 0.799 Ϯ 0.076b Neriidae (I) 8.80 0.027 Ϯ 0.011a 0.0b 0.0b Otitidae (I) 16.77 0.168 Ϯ 0.040a 0.0b 0.022 Ϯ 0.008b Sarcophagidae (I) 9.58 0.036 Ϯ 0.010a 0.0b 0.008 Ϯ 0.003b Tephritidae (I) 10.49 0.041 Ϯ 0.015a 0.0b 0.001 Ϯ 0.001b Diptera misc (E-I) 17.22 0.129 Ϯ 0.028a 0.0b 0.027 Ϯ 0.006b Coleoptera Corylophidae (I) 3.82 0.015 Ϯ 0.005b 0.118 Ϯ 0.032a 0.060 Ϯ 0.016ab Nitidulidae (E-I) 16.91 0.078 Ϯ 0.021a 0.004 Ϯ 0.004b 0.004 Ϯ 0.003b Total nontargets 14.80 2.454 Ϯ 0.329a 3.299 Ϯ 0.692b 2.030 Ϯ 0.263b Percent endemic (including Drosophilidae)d 7.8Ð12.2 1.2Ð3.2 4.6Ð10.2 Percent endemic (excluding Drosophilidae)d 12.0Ð18.8 2.4Ð6.4 8.2Ð18.0

Only data for groups that showed attraction to BioLure (Table 1) and a signiÞcant habitat effect on trap captures are included on the table. a Degrees of freedom for each F value are 2,251. P Ͻ 0.0001 for every groups, except for Anthomyiidae (P ϭ 0.0138), Drosophilidae (I) (P ϭ 0.0078), Neriidae (P ϭ 0.0002), and Corylophidae (P ϭ 0.0232). b Values in each row followed by the same letter are not signiÞcantly different at the 0.05 level, TukeyÕs test, PROC MIXED mivque0 (SAS Institute 2004). c The majority (97.3%) of captured target Tephritidae are Ceratitis capitata. d % nontarget individuals of endemic species, when including and excluding Drosophilidae, presented as ranges, because of uncertainty of status for several species of Phoridae and Lepidoptera. E, endemic; I, introduced; E-I, family includes endemic and introduced species in samples. icantly higher in coffee than in non-native forest or in synergy to increase attraction of not only fruit ßies, persimmon, likely because of the limited maintenance as shown previously (Heath et al. 1997, 2004), but also and sanitation at two of the four coffee orchards in the many of the nontargets. study, resulting in thick weed undergrowth and the Ammonium acetate and, to a lesser extent, pu- accumulation of decaying berries. Persimmon or- trescine, were the BioLure components most attrac- chards, however, have little or no undergrowth, and tive to nontargets (Table 3). Ammonium acetate was fallen fruits are gathered assiduously by growers, leav- the main attractant for at least 13 species, including C. ing little shelter or breeding substrate for sapropha- capitata, as determined previously (Heath et al. 2004), gous nontargets. Similarly, Thomas (2003) collected and all the endemic drosophilids and calliphorids. Am- higher numbers of a broader diversity of nontargets monium acetate and putrescine were equally attrac- in weedy abandoned citrus orchards and native veg- tive to four other species, including Bactrocera dorsa- etation than in well-maintained commercial citrus lis. For the majority of the above species, BioLure orchards. attracted more insects than the total of its three com- In contrast to the above results, higher numbers of ponents, suggesting a synergistic effect. Putrescine endemic and introduced drosophilids were trapped in was the chief attractant for four nontarget species the non-native forest than in either agricultural hab- (Table 3), which were frequently trapped in larger itat. Because target tephritids were not trapped in numbers than in BioLure traps. For another Þve non-native forest, there is no incentive to using species (Table 3), substantial attraction was BioLure traps in these forest refuges. Thus, growers achieved only when all components were used to- can avoid impacting populations of endemic drosophi- gether in BioLure. Although trimethylamine hydro- lids that persist in non-native forests habitats, with no chloride by itself was unattractive to fruit ßies or detriment to pest control. nontargets, we conÞrm that it is an important syn- Attraction to Individual Components. To elucidate ergistic component of BioLure for fruit ßies (Heath the role of individual components in fruit ßy and et al. 1997). nontarget attraction, we compared captures by traps Effects of Putrescine Omission. A comparison of with BioLure and its separate components for the 2 attraction to BioLure with and without putrescine at target and the 24 nontarget species for which Ն200 different densities of C. capitata in Kula (Fig. 1) specimens were collected (Table 3). The lure ratio showed that the two-component lure was as effective (captures by BioLure divided by the total of captures as the three-component lure in all situations, except by the separate components) was at least equal to that for females at low densities. These Þndings are mostly of C. capitata (1.4) for B. dorsalis and 17 nontarget consistent with Heath et al. (2004), who reported only species, suggesting that the combined components act a slight decrease in captures of males at high density ue21 L 2010 June

Table 3. Captures of target fruit flies and nontarget insects in traps baited with three-component BioLure and with the individual components in separate traps in native forest and a citrus orchard, on Hawaii Island, and an abandoned coffee plantation in Kula, Maui Island

MeanϮ SEM no. captured per trap per day Lure Lure Site a BioLure (three- Ammonium Trimethylamine Unbaited ratio F Putrescineb component)b acetateb hydrochlorideb controlb Target species Tephritidae: Ceratitis capitata (Wiedemann) Fem (I) Cof 1.4 24.95 8.929 Ϯ 1.405a 4.840 Ϯ 1.111b 1.635 Ϯ 0.311c 0.087 Ϯ 0.028c 0.054 Ϯ 0.019c

C. capitata Mal (I) Cof 1.4 18.48 6.868 Ϯ 1.367a 3.568 Ϯ 0.777b 1.126 Ϯ 0.289c 0.060 Ϯ 0.018c 0.018 Ϯ 0.010c AL ET EBLANC C. capitata (all) (I) Cof 1.4 22.64 15.797 Ϯ 2.726a 8.408 Ϯ 1.873b 2.761 Ϯ 0.576c 0.147 Ϯ 0.043c 0.071 Ϯ 0.023c Bactrocera dorsalis (Hendel) (I) Cit 1.6 28.28 2.198 Ϯ 0.336a 0.744 Ϯ 0.153b 0.153 Ϯ 0.409bc 0.164 Ϯ 0.032c 0.058 Ϯ 0.029c Nontarget species Calliphoridae: Dyscritomyia spp (E)c For 1.5 40.07 3.786 Ϯ 0.537a 2.357 Ϯ 0.324b 0.024 Ϯ 0.024c 0.071 Ϯ 0.039c 0.0c Chloropidae: Chloropsina citrivora Sabrosky (I) Cit 0.3 11.86 0.289 Ϯ 0.069b 0.553 Ϯ 0.116a 0.371 Ϯ 0.067ab 0.016 Ϯ 0.016c 0.009 Ϯ 0.006c

Gampsocera hardyi Kanmiya (I) Cit 3.4 5.50 1.044 Ϯ 0.441a 0.109 Ϯ 0.033b 0.171 Ϯ 0.055b 0.005 Ϯ 0.005b 0.0b N .: Gaurax bicoloripes (Malloch) (I) Cit 1.5 7.27 1.498 Ϯ 0.538a 0.697 Ϯ 0.202b 0.241 Ϯ 0.110bc 0.0c 0.0c Rhodesiella elegantula (Becker) (I) Cof 1.2 11.06 3.685 Ϯ 0.758a 1.744 Ϯ 0.436b 1.263 Ϯ 0.559bc 0.00c 0.014 Ϯ 0.008c ONTARGET R. scutellata (Meijere) (I) Cit 0.3 16.54 0.851 Ϯ 0.205bc 0.508 Ϯ 0.120cd 1.882 Ϯ 0.321a 0.078 Ϯ 0.022d 0.135 Ϯ 0.053d Drosophilidae: Drosophila (Antopocerus) spp (E)d For 4.1 22.22 10.024 Ϯ 2.075a 2.452 Ϯ 0.404ab 0.0b 0.0b 0.0b Drosophila (spoon tarsus) spp (E)e For 1.7 17.06 28.310 Ϯ 6.374a 15.286 Ϯ 2.683b 0.048 Ϯ 0.048c 0.857 Ϯ 0.213c 0.095 Ϯ 0.074c D. ananassae Doleschall (I) Cit 43.0 8.06 1.870 Ϯ 0.652a 0.042 Ϯ 0.023b 0.0b 0.005 Ϯ 0.005b 0.0b Ϯ Ϯ Ϯ A D. immigrans Sturtevant (I) Cit 47.1 5.87 7.170 2.929a 0.148 0.069b 0.0b 0.006 0.006b 0.0b TO TTRACTION D. sulfurigaster bilimbata Bezzi (I) Cit 115.2 5.54 158.751 Ϯ 66.846a 1.336 Ϯ 0.550b 0.022 Ϯ 0.011b 0.020 Ϯ 0.009b 0.042 Ϯ 0.018b D. suzukii (Matsumura) (I) For 6.4 46.64 54.143 Ϯ 7.710a 7.381 Ϯ 1.200b 0.119 Ϯ 0.084b 1.024 Ϯ 0.217b 0.0b D. suzukii Cit 4.4 20.18 10.752 Ϯ 2.306a 2.310 Ϯ 0.268b 0.061 Ϯ 0.018b 0.058 Ϯ 0.015b 0.011 Ϯ 0.008b D. suzukii Cof 1.7 25.61 1.874 Ϯ 0.284a 0.883 Ϯ 0.169b 0.070 Ϯ 0.019c 0.070 Ϯ 0.019c 0.247 Ϯ 0.056c Lauxaniidae: Homoneura unguiculata (Kertesz) (I) Cof Ͻ0.1 4.01 0.021 Ϯ 0.010b 0.135 Ϯ 0.050b 0.964 Ϯ 0.458a 0.033 Ϯ 0.019b 0.049 Ϯ 0.018b Muscidae: Atherigona orientalis Schiner (I) Cof 0.1 3.48 0.086 Ϯ 0.028b 0.009 Ϯ 0.007b 1.084 Ϯ 0.573a 0.0b 0.010 Ϯ 0.007b Neriidae: Telostylinus lineolatus (Wiedemann) (I) Cit 4.4 18.23 1.440 Ϯ 0.318a 0.310 Ϯ 0.068b 0.018 Ϯ 0.013b 0.0b 0.0b B IO Phoridae: Megaselia furcatilis Beyer (E) Cof 0.1 9.46 0.157 Ϯ 0.046b 0.020 Ϯ 0.009b 2.125 Ϯ 0.682a 0.005 Ϯ 0.005b 0.006 Ϯ 0.006b L

Tephritidae: Dioxyna sororcula (Wiedemann) (I) Cof 1.2 6.94 0.678 Ϯ 0.230a 0.547 Ϯ 0.166a 0.048 Ϯ 0.018b 0.006 Ϯ 0.006b 0.010 Ϯ 0.010b IN URE

Only species and sites with at least 200 captured specimens are included on table. All counts include females and males, unless otherwise speciÞed. a Lure ratio is the sum of captures by three-component BioLure traps divided by the sum of captures by all traps baited with the separate components. Synergism is assumed when the ratio is noticeably H higher than 1. AWAII b Values in each row followed by same letter are not signiÞcantly different at the 0.05 level, TukeyÕs test, PROC MIXED mivque0 (SAS Institute 2004). Degrees of freedom for F values of overall lure effect are 4,128 for citrus, 4,143 for coffee, and 4,98 for forest. For all species, P Ͻ 0.0001, except for D. immigrans in citrus (P ϭ 0.0002), D. sulfurigaster bilimbata (P ϭ 0.0004), G. hardyi (P ϭ 0.0004), H. unguiculata (P ϭ 0.0041), and A. orientalis (P ϭ 0.0096). c Includes Dyscritomyia lucilioides (Grimshaw) and D. obscura (Grimshaw). Both species are nearly indistinguishable. d Includes females and males of Drosophila tanythrix (Hardy) and D. cognata Grimshaw. Only males can be deÞnitely separated into species. e Includes females and males of Drosophila conformis Hardy, D. dasycnemia Hardy, D. incognita Hardy, D. neutralis Hardy, D. percnosoma Hardy, and D. sordidapex Grimshaw. Only males can be deÞnitely separated into species. E, endemic species; I, introduced species. 995 996 ENVIRONMENTAL ENTOMOLOGY Vol. 39, no. 3

Fig. 1. Comparison of captures (mean Ϯ SEM per trap per day) of Ceratitis capitata in three-component and two- component (without putrescine) BioLure traps. Data analyzed for all sites and separately for low (0.003Ð0.278 per trap per day), moderate (0.600Ð1.792 per trap per day), and high (10.992 per trap per day) ßy densities. Density ranges selected based on groupings in frequency distribution histogram. Values in each pair with the same letter are not signiÞcantly different at the P ϭ 0.05 level, ANOVA, PROC MIXED mivque0 (SAS Institute 2004). Numerator and denominator degrees of freedom for F values of overall lure effect are 1,338 for low density, 1,118 for moderate density, 1,30 for high density, and 1,490 for overall data. Females: low density: F ϭ 5.82, P ϭ 0.0064; moderate density: F ϭ 0.59, P ϭ 0.4436; high density: F ϭ 0.10, P ϭ 0.7500; all data: F ϭ 0.05, P ϭ 0.8486. Males: low density: F ϭ 0.62, P ϭ 0.4314; moderate density: F ϭ 0.01, P ϭ 0.9277; high density: F ϭ 0.02, P ϭ 0.8838; all data: F ϭ 0.01, P ϭ 0.9221. and females at very low density with the omission of etate was the most attractive to target C. capitata but putrescine. also to many species of nontargets. Fewer nontarget Elimination of putrescine resulted in signiÞcant species responded primarily to putrescine. Remov- capture reduction for 5 of the 12 nontarget species ing the putrescine ingredient from BioLure resulted in Fig. 2, and an increase for one species. The most in an overall 20% nontarget reduction, without com- pronounced reductions (by 53 and 75%) were co- promising captures of C. capitata, although removal incidentally for two species (M. furcatilis and A. did not reduce captures for more than one half of orientalis) for which putrescine was the key ingre- the nontargets, including the most common dros- dient responsible for their attraction (Table 3). ophilid species. Omission of putrescine resulted in an overall non- Finally, numbers of nontargets attracted to BioLure target capture reduction of Ϸ20%, from 2.542 Ϯ may have been overestimated in our study, because of 0.204 to 1.765 Ϯ 0.137 SEM per trap per day (F ϭ the liquid preservative in our traps. A synergistic effect 10.01, df ϭ 1,554, P ϭ 0.0016). Furthermore, reduc- of propylene glycol in BioLure traps results in in- ing nontarget captures increases trap efÞciency, re- creased captures of Anastrepha fruit ßies (Thomas et duces clogging of sticky boards used to retain ßies, al. 2001, Hall et al. 2005, Robacker and Czokajlo 2006, reduces workloads when clearing traps, and may Robacker and Thomas 2007, Thomas 2008). The same also contribute to reducing potential impacts on effect was shown for nontarget attraction to methyl endemic species, especially when trapping in or eugenol traps with an accumulation of dead oriental near areas with the potential for high endemic non- fruit ßies (Uchida et al. 2007). Growers in Hawaii target attraction. use a yellow sticky insert in BioLure traps, rather To summarize, the three-component BioLure than liquid, to retain the fruit ßies (McQuate et al. used in MultiLure traps was an effective attractant 2005). Captures of nontargets may be lower in their for female and male C. capitata but also attracted traps than they were in our study. Additional Þeld large numbers of saprophagous ßies, dominated in research is needed to determine whether, and to Hawaii by Drosophilidae. Because large numbers of what extent, propylene glycol contributes to the endemic drosophilids and calliphorids were at- attractiveness of BioLure to C. capitata and nontar- tracted in native Hawaiian forest, the use of BioLure get species. in indigenous habitats should be discouraged to avoid killing rare or endangered endemic species. BioLure use should be restricted to orchard and Acknowledgments backyard habitats, where mostly introduced non- N. W. Miller (USDAÐARS Hilo) and K. Kinney (Univer- targets occur, at a minimal safe distance of Ϸ300 m sity of Hawaii, Hilo Campus) provided Þeld assistance in from endemic forest (Leblanc et al. 2009b). Among setting up the tests on Hawaii Island and servicing the traps the individual BioLure components, ammonium ac- at the citrus site. M. Shipley and C. Kanada (University of June 2010 LEBLANC ET AL.: NONTARGET ATTRACTION TO BIOLURE IN HAWAII 997

Fig. 2. Comparison of nontarget captures (mean Ϯ SEM per trap per day) in three-component and two-component (without putrescine) BioLure traps in Kula, Maui island. Only species with at least 100 captured specimens are included. All counts include females and males, unless speciÞed below. All the species included are introduced. Values in each pair with the same letter are not signiÞcantly different at the P ϭ 0.05 level, ANOVA, PROC MIXED mivque0 (SAS Institute 2004). Numerator and denominator degrees of freedom for F values are 1,554 for each species. Chloropidae: Gampsocera hardyi Kanmiya: F ϭ 3.56, P ϭ 0.0597; Rhodesiella elegantula (Becker): F ϭ 12.11; P ϭ 0.0005. Drosophilidae: Coquillett: F ϭ 0.03, P ϭ 0.8524; D. immigrans Sturtevant: F ϭ 0.64, P ϭ 0.4246; D. suzukii (Matsumura): F ϭ 0.04, P ϭ 0.8416; maculata Coquillett: F ϭ 0.96, P ϭ 0.3272; misc Drosophilidae: F ϭ 6.96, P ϭ 0.0086. Muscidae: Atherigona orientalis Schiner: F ϭ 11.07, P ϭ 0.0009. Otitidae: Euxesta annonae (Fabricius): F ϭ 12.44, P ϭ 0.0005. Phoridae: Megaselia furcatilis Beyer: F ϭ 5.02, P ϭ 0.0255; Megaselia species one (females only): F ϭ 6.81, P ϭ 0.0093. Other nontarget Diptera: F ϭ 4.36, P ϭ 0.0373. Encyrtidae: Tachinaephagus zealandicus Ashmead: F ϭ 8.58, P ϭ 0.0035. Corylophidae: Corylophodes suturalis (Sharp): F ϭ 0.15, P ϭ 0.6952. Other arthropods: F ϭ 3.66, P ϭ 0.0563.

Hawaii, Manoa Campus) helped in the preliminary sorting of damage on different host species in Israel. J. Econ. En- the very large samples. IdentiÞcations of Drosophilidae were tomol. 93: 721Ð725. conÞrmed by P. OÕGrady (University of , Berkeley, Conway, H. E., and O. T. Forrester. 2007. Comparison of CA). M. G. Wright and J. C. Pin˜ ero have provided advice on Mexican fruit ßy (Diptera: Tephritidae) capture between data analysis and reviewed the manuscript. Comments by the McPhail traps with torula and Multilure traps with two anonymous external reviewers are also appreciated. Ac- Biolures in Southern . Fla. Entomol. 90: 579Ð580. cess to the endemic forest site was granted through a research Epsky, N. D., J. Hendrichs, B. I. Katsoyannos, L. A. Vasquez, J. P. permit from B. Gagne´ (Hawaii State Department of Land and Ros, A. Zumreoglu, R. Pereira, S. I. Seewooruthun, and R. R. Natural Resources). We thank the private landowners in Heath. 1999. Field evaluation of female-targeted trapping Kula for their trust and aloha spirit. This study was sponsored systems for Ceratitis capitata (Diptera: Tephritidae) in seven by USDAÐARS through a SpeciÞc Cooperative Agreement countries. J. Econ. Entomol. 92: 156Ð164. with University of HawaiiÕs CTAHR (0500-00044-016-07) ti- Gazit, Y., Y. Ro¨ssler, N. D. Epsky, and R. R. Heath. 1998. tled “Study of Attraction of Nontarget Organisms to Fruit Fly Trapping females of the Mediterranean fruit ßy (Diptera: Female Attractants and Male Lures in Hawaii.” Tephritidae) in Israel: comparison of lures and trap type. J. Econ. Entomol. 91: 1355Ð1359. References Cited Hall, D. G., R. E. Burns, C. C. Jenkins, K. L. Hibbard, D. L. Harris, J. M. Sivinski, and H. N. Nigg. 2005. Field com- Aluja, M. 1996. Future trends in fruit ßy management, pp. parison of chemical attractants and traps for Caribbean 309Ð320. In M. Aluja and P. Liedo (eds.), Fruit ßies fruit ßy (Diptera: Tephritidae) in citrus. J. Econ. biology and management. Springer, New York. Entomol. 98: 1641Ð1647. Asquith, A. 1995. Distribution, abundance and phenology of Hardy, D. E. 1952. collected in bait traps. Proc. Ha- Scaptomyza (Bunostoma) anomala Hardy (Diptera: Dros- waiian Entomol. Soc. 14: 407Ð409. ophilidae): a proposed representative species for moni- Hardy, D. E. 1964. Insects of Hawaii. Volume 11. Diptera: toring protein bait sprays in HawaiÕi. Proc. Hawaiian En- , Family Dolichopodidae. Cyclorrhapha, series tomol. Soc. 32: 69Ð81. . Families Lonchopteridae, Phoridae, Pipunculidae, Asquith, A., and R. H. Messing. 1992. Attraction of Hawaiian and Syrphidae. University of Hawaii Press, Honolulu, HI. ground litter invertebrates to protein hydrolysate bait. Hardy, D. E., and M. D. Delfinado. 1980. Insects of Hawaii. Environ. Entomol. 21: 1022Ð1028. Volume 13. Diptera: Cyclorrhapha III, series Schizo- Cohen, H., and B. Yuval. 2000. Perimeter trapping strategy phora, section Acalypterae, exclusive of family Drosophi- to reduce Mediterranean fruit ßy (Diptera: Tephritidae) lidae. University of Hawaii Press, Honolulu, HI. 998 ENVIRONMENTAL ENTOMOLOGY Vol. 39, no. 3

Heath, R. R., N. D. Epsky, A. Guzman, D. B. Dueben, A. (Neuroptera) in a Cretan olive orchard. Ann. Soc. En- Manukian, and W. L. Meyer. 1995. Development of a tomol. France 17: 213Ð220. dry plastic with food-based synthetic attract- O’Grady, P.M., K.N. Magnacca, and R.T. Lapoint. 2010. Tax- ant for the Mediterranean and Mexican fruit ßies onomic relationships within the endemic Hawaiian Dros- (Diptera: Tephritidae). J. Econ. Entomol. 88: 1307Ð1315. ophilidae. Bishop Museum Occasional Papers. 108: 1Ð34. Heath, R. R., N. D. Epsky, B. D. Dueben, and W. L. Meyer. Robacker, D. C. 1995. Attractiveness of a mixture of ammo- 1996. Systems to monitor and suppress Ceratitis capitata nia, methylamine and putrescine to Mexican fruit ßies (Diptera: Tephritidae) populations. Fla. Entomol. 79: (Diptera: Tephritidae) in a citrus orchard. Fla. Entomol. 144Ð153. 78: 571Ð578. Heath, R. R., N. D. Epsky, B. D. Dueben, J. Rizzo, and F. Robacker, D. C., and W. C. Warfield. 1993. Attraction of Jeronimo. 1997. Adding methyl-substituted ammonia both sexes of Mexican fruit ßy, Anastrepha ludens, to a derivatives to a food-based synthetic attractant on cap- mixture of ammonia, methylamine, and putrescine. ture of the Mediterranean and Mexican fruit ßies J. Chem. Ecol. 19: 2999Ð3016. (Diptera: Tephritidae). J. Econ. Entomol. 90: 1584Ð1589. Robacker, D. C., and D. Czokajlo. 2005. EfÞcacy of two Heath, R. R., N. D. Epsky, D. Midgarden, and B. I. Katsoyannos. synthetic food-odor lures for Mexican fruit ßies (Diptera: 2004. EfÞcacy of 1,4-Diaminobutane (putrescine) in a Tephritidae) is determined by trap type. J. Econ. Ento- food-based synthetic attractant for capture of Mediterra- mol. 98: 1517Ð1523. nean and Mexican fruit ßies (Diptera: Tephritidae). J. Econ. Robacker, D. C., and D. Czokajlo. 2006. Effect of propylene Entomol. 97: 1126Ð1131. glycol antifreeze on captures of Mexican fruit ßies Katsoyannos, B. I., R. R. Heath, N. T. Papadopoulos, N. D. (Diptera: Tephritidae) in traps baited with biolures and Epsky, and J. Hendrichs. 1999. Field evaluation of Medi- AFF lures. Fla. Entomol. 89: 286Ð287. terranean fruit ßy (Diptera: Tephritidae) female selective Robacker, D. C., and D. B. Thomas. 2007. Comparison of attractants for use in monitoring, mass trapping and sterile two synthetic food-odor lures for capture of feral Mexican insect technique programs. J. Econ. Entomol. 92: 583Ð589. fruit ßies (Diptera: Tephritidae) in Mexico and implica- Kido, M. H., and A. Asquith. 1995. Attraction of Hawaiian tions regarding use of irradiated ßies to assess lure efÞ- aquatic insects to pest tephritid parakairomone lures. cacy. J. Econ. Entomol. 100: 1147Ð1152. Environ. Entomol. 24: 810Ð816. SAS Institute. 2004. SAS/STAT 9.1 userÕs guide. SAS Insti- Leblanc, L., P. M. O’Grady, D. Rubinoff, and S. L. Mont- tute, Cary, NC. gomery. 2009a. New immigrant Drosophilidae in Ha- Thomas, D. B. 2003. Nontarget insects captured in fruit ßy waii, and a checklist of the established immigrant species. (Diptera: Tephritidae) surveillance traps. J. Econ. Ento- Proc. Hawaiian Entomol. Soc. 41: 121Ð127. mol. 96: 1732Ð1737. Leblanc, L., D. Rubinoff, and R. I. Vargas. 2009b. Attraction Thomas, D. B. 2008. A safe and effective propylene glycol of nontarget species to fruit ßy (Diptera: Tephritidae) based capture liquid for fruit ßy (Diptera: Tephritidae) traps male lures and decaying fruit ßies in Hawaii. Environ. baited with synthetic lures. Fla. Entomol. 91: 210Ð213. Entomol. 38: 1446Ð1461. Thomas, D. B., T. C. Holler, R. R. Heath, E. J. Salinas, and Liquido, N. J., L. A. Shinoda, and R. T. Cunningham. 1991. A. L. Moses. 2001. Trap-lure combinations for surveil- Host plants of the Mediterranean fruit ßy (Diptera: Te- lance of Anastrepha fruit ßies (Diptera: Tephritidae). Fla. phritidae): an annotated world review. Misc. Publications Entomol. 84: 344Ð351. 77. Entomology Society of America, Lanham, MD. Uchida, G. K., B. E. Mackey, R. I. Vargas, J. W. Beardsley, Magnacca, K. N., D. Foote, and P. M. O’Grady. 2008. A D. E. Hardy, M. L. Goff, and J. D. Stark. 2006. Response review of the endemic Hawaiian Drosophilidae and their of nontarget insects to methyl eugenol, cue-lure, trimed- host plants. Zootaxa 1728: 1Ð58. lure, and protein bait bucket traps on Kauai island, Ha- Martinez, A. J., E. J. Salinas, and P. Rendon. 2007. Capture waii, USA. Proc. Hawaiian Entomol. Soc. 38: 61Ð72. of Anastrepha species (Diptera: Tephritidae) with mul- Uchida, G. K., B. E. Mackey, D. O. McInnis, and R. I. Vargas. tilure traps and biolure attractants in Guatemala. Fla. 2007. Attraction of Bactrocera dorsalis (Diptera: Te- Entomol. 90: 258Ð263. phritidae) and nontarget insects to methyl eugenol McQuate, G. T., C. D. Sylva, and E. B. Jang. 2005. Mediter- bucket traps with different preservative ßuids on Oahu ranean fruit ßy (Dipt., Tephritidae) suppression in per- island, Hawaiian Islands. J. Econ. Entomol. 100: 723Ð729. simmon through bait spray in adjacent coffee plantings. U.S. Fish and Wildlife Service. 2007. Endangered and J. Appl. Entomol. 129: 110Ð117. threatened wildlife and plants: revised proposed desig- Moore, R. C. 1969. Attractiveness of baited and unbaited nation of critical habitats for 12 species of picture-wing lures to and beneÞcial ßies. J. Econ. Ento- ßies from the Hawaiian Islands: proposed rule. Fed. Reg. mol. 62: 1076Ð1078. 72: 67427Ð67522. Navarro-Lopis, V., F. Alfaro, J. Domı´nguez, J. Sanchis, and J. Wong, T. Y., J. I. Nishimoto, and N. Mochizuki. 1983. In- Primo. 2008. Evaluation of traps and lures for mass trap- festation patterns of Mediterranean fruit ßy and the ori- ping of Mediterranean fruit ßy in citrus groves. J. Econ. ental fruit ßy (Diptera: Tephritidae) in the Kula area of Entomol. 101: 126Ð131. Maui, Hawaii. Environ. Entomol. 12: 1031Ð1039. Neuenschwander, P., M. Canard, and S. Michelakis. 1981. The attractivity of protein hydrolysate baited McPhail traps to different chrysopid and hemerobiid species Received 3 October 2009; accepted 9 December 2009.