BIOLOGICAL AND MICROBIAL CONTROL Releases of Psyttalia fletcheri (Hymenoptera: ) and Sterile to Suppress (Diptera: ) in Hawaii

ROGER I. VARGAS,1 JAY LONG, NEIL W. MILLER,1 KATHLEEN DELATE,2 3 1 1 1 CHARLES G. JACKSON, GRANT K. UCHIDA, RENATO C. BAUTISTA, AND ERNIE J. HARRIS

J. Econ. Entomol. 97(5): 1531Ð1539 (2004) ABSTRACT Ivy gourd, Coccinia grandis (L.) Voigt, patches throughout Kailua-Kona, Hawaii Island, HI, were identiÞed as persistent sources of melon ßy, cucurbitae (Coquillett). These patches had a low incidence of Psyttalia fletcheri (Silvestri), its major braconid natural enemy in Hawaii, and were used to evaluate augmentative releases of P. fletcheri against melon ßy. In Þeld cage studies of releases, numbers of melon ßies emerging from ivy gourd placed inside treatment cages were reduced up to 21-fold, and numbers of were increased 11-fold. In open Þeld releases of P. fletcheri into ivy gourd patches, parasitization rates were increased 4.7 times in release plots compared with those in control plots. However, there was no signiÞcant reduction in emergence of melon ßies from fruit. In subsequent cage tests with sterile melon ßies and P. fletcheri, combinations of sterile ßies and P. fletcheri produced the greatest reduction (9-fold) in melon ßy emergence from , pepo L. Reductions obtained with sterile ßies alone or in combination with parasitoids were signiÞcantly greater than those in the control, whereas those for parasitoids alone were not. Although these results suggest that the effects of sterile ßies were greater than those for parasitoids, from a multitactic melon ßy management strategy, sterile ßies would complement the effects of P. fletcheri. Cost and sustainability of these nonchemical approaches will be examined further in an ongoing areawide pest management program for melon ßy in Hawaii.

KEY WORDS Bactrocera cucurbitae, augmentative parasitoid releases, sterile ßy releases

Melon ßy, Bactrocera cucurbitae (Coquillett), is a se- ongena L.; and bellpepper, annuum L.), Ru- rious agricultural pest of cucurbits. It has been re- taceae (e.g., citrus, Citrus spp.), Myrtaceae (e.g., com- corded from Ͼ125 plant species (Weems 1964) and is mon guava, Psidium guajava L.; and strawberry guava, found in India, Myanmar, Malaysia, Thailand, the Phil- Psidium cattleianum Sabine), Rosaceae [e.g., loquat, ippines, southern China, Taiwan, eastern Africa, Eriobotrya japonica (Thunb.) Lindl. and peach, Guam, the Commonwealth of the Northern Mariana Prunus persica (L.) Batsch.], and Passißoraceae (e.g., Islands, New Guinea (Papua), Solomon Islands, passionfruit, Passiflora edulis Sims) (White and Elson- Nauru, and the Hawaiian Islands (Nishida 1953, White Harris 1992, Ramadan and Messing 2002). Two major and Elson-Harris 1992). In 1895, it was discovered in feral hosts are wild bitter melon and ivy gourd, Coc- Hawaii (Back and Pemberton 1917), where it causes cinia grandis (L.) Voigt (Liquido et al. 1990, Uchida et serious economic damage to cultivated species of Cu- al. 1990). curbitaceae [e.g., , Cucumis sativus L.; wa- Establishment of melon ßy in Hawaii resulted in the termelon, Citrullus lanatus (Thunb.) Matsum. & Na- introduction of many natural enemies. Nishida (1955) kai; cantaloupe, Cucumis melo L.; bitter melon listed a total of eight species of hymenopterous para- L.; and zucchini, Cu- sitoids and six predators found in Hawaii; however, all curbita pepo L.]. When populations are high and cu- of the parasitoids, except Psyttalia fletcheri (Silvestri), curbits scarce, melon ßies also attack, but less fre- were of little importance from the standpoint of bio- quently, species of Solanaceae (e.g., , logical control, primarily because of their scarcity and Lycopersicon esculentum Mill.; , Solanum mel- nonspeciÞcity to melon ßy. P. fletcheri, a widespread larval-pupal parasitoid of melon ßy in India, was in- troduced into Hawaii in 1916 (Willard 1920, Clausen This article reports the results of research only. Mention of pro- prietary product does not constitute an endorsement or recommen- et al. 1965). The kind of host fruit infested by melon dation by the USDA. ßy seems to inßuence P. fletcheri parasitization rate. 1 U.S. PaciÞc Basin Agricultural Research Center, USDAÐARS, P.O. For example, Nishida (1955) found little or no para- Box 4459, Hilo, HI 96720. sitization of larvae in papaya, Carica papaya L., 2 Department of Horticulture, Iowa State University, Ames, IA 50011. bellpepper, and tomato. Willard (1920) reported that 3 USDAÐARS, Western Cotton Research Laboratory; 4135 East parasitization of larvae in cucumber fruit ranged from Broadway Rd., Phoenix, AZ 85040. 7.3 to 29.8%, whereas parasitization on wild bitter 1532 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 5 melon fruit was as high as 96.9%. Host larvae location Flies captured in traps were collected and counted in different parts of the same plant also inßuenced the monthly. Presence of ripe fruit in patches varied sea- parasitization by P. fletcheri. For example, in , sonally. Approximately, 100 fruit (three-fourths to where the larvae may be found in both vines and fruit, fully ripe) were collected randomly from each plot consistently higher parasitization was obtained in monthly. Fruit were weighed and placed in batches of vines (Nishida 1955). Apparently, in vines, the larvae 100 on a wood-framed metal screen (43 by 28 by 6 cm) are situated just beneath the epidermis throughout the inside Þberglass holding boxes (50by 32 by 15 cm) that developmental period and are within reach of the contained 1.5 cm of sand. Fruit were held for 3 wk. parasitoidÕs ovipostior for a longer time. However, Boxes were provided with screened ventilation holes larvae during later stages of development have a ten- on the sides and a pair of plastic drip pans (25 by 19 dency to burrow deeply into the ßesh of the fruit and by 3 cm) between the screen and the sand-lined bot- are less accessible to parasitoids. tom to collect fruit juice. A Þne mesh nylon screen was Ivy gourd is native to Africa, but it also occurs wild placed over pans to prevent larvae from falling into in the Indo-Malayan region (Singh 1990) and is nat- juice. Sand from fruit holding boxes was sifted weekly. uralized in parts of Australia, the Carribbean, the Pupae were transferred to smaller plastic containers southern mainland United States, and several PaciÞc (9.2 cm in diameter by 4.7 cm in height, Highland Islands (Linney 1986, Telford 1990). During the 1960s, Plastics) with screened lids and sand and held until it was accidentally introduced into Hawaii, where it is emergence of ßies or parasitoids. Fruit were held in a considered a noxious weed and a target for biological room maintained at 22 Ϯ 5ЊC, ambient (40Ð90%) RH, control (OÕBrien and Pakaluk 1998, Chun 2001). Ivy and a photoperiod of 12:12 (L:D), and recovered pu- gourd frequently blankets trees, understory vegeta- pae were held in an environmental cabinet maintained tion, fences, and other artiÞcial structures in residen- at 25 Ϯ 2ЊC, 60 Ϯ 10% RH, and a photoperiod of 12:12 tial and agricultural areas (Chun 2001). Presently, ivy (L:D) h. Numbers of melon ßy and P. fletcheri that gourd is restricted in Hawaii to Oahu Island and to the emerged were recorded. Dead pupae were dissected leeward side of Hawaii Island, speciÞcally to the to determine whether parasitization had occurred. To Kailua-Kona area. It is a perennial climbing vine with determine individual infestation of fruit, four lots of a tuberous rootstock producing annual stems up to 250fruit were collected from four different sites dur- several meters long. Main stems may be 5Ð8 cm in ing April 1997. These fruit were held individually in diameter at ground level, and the plants start new small plastic cups with screened lids to determine growth rapidly after rainfall. Fruit are green with lon- infestation and parasitization rates per individual fruit. gitudinal white stripes when immature, but change to These results were pooled (n ϭ 1000). scarlet red at maturity. They are 25Ð60mm in length P. fletcheri Field Cage Studies. P. fletcheri wasps and 15Ð30mm in diameter (Telford 1990).Seeds are were obtained from a colony maintained for 50gen- spread by birds, rodents, and humans (Uchida et al. erations in the laboratory at the PaciÞc Basin Agri- 1990). Both Uchida et al. (1990) and Jackson et al. cultural Research Center (PBARC) facility in Hono- (2003) found ivy gourd to be a major source of melon lulu, HI. Parasitoids were shipped inside melon ßy ßies. pupae to the PBARC facility in Hilo, HI, and allowed Reported here are studies in Kailua-Kona, HI, on P. to emerge inside plastic buckets (15 cm in depth by 20 fletcheri and melon ßy ecology in ivy gourd and on cm in diameter). Pupae were placed inside buckets at effects of augmentative releases of P. fletcheri. Fur- densities of 1, 2, or 3 g (30males and 80females, 80 thermore, we examined the potential effects of sterile males and 230females, and 160males and 450females, melon ßy and P. fletcheri releases on wild melon ßy respectively). Numbers of parasitoids to emerge were abundance. SpeciÞcally we studied (1) annual P. determined from emergence inside quality control fletcheri parasitization of melon ßy breeding in ivy buckets to determine the number of wasps (males and gourd patches, (2) Þeld cage releases of P. fletcheri females) released into cages during experiments. Two against melon ßy infesting ivy gourd fruit, (3) open quality control buckets were examined per density per Þeld releases of P. fletcheri against melon ßy infesting replicate. Buckets had screen lids and were small patches of ivy gourd, and (4) Þeld cage releases provided with honey (Sioux Honey Association, Sioux of P. fletcheri and sterile ßies against melon ßy infesting City, IA) and water. Pupae were held in a room at zucchini fruit. temperatures of 22 Ϯ 5ЊC, with a relative humidity of 40Ð90%, and a photoperiod of 12:12 (L:D) h until eclosion. Parasitoids 4Ð5 d-old, were used in the 24 h Materials and Methods exposure tests with fruit collected from wild ivy gourd Melon Fly and P. fletcheri Ecology. Ten patches of plants. Freshly picked fruit, 90Ð100% ripe, and natu- ivy gourd throughout the Kailua-Kona, HI, area (Fig. rally infested with melon ßy and P. fletcheri, were 1) were surveyed from April 1996 to March 1997 to randomly assigned to treatment and control cages. determine melon ßy abundance. A male melon ßy Field cage tests were conducted during 1996 at the bucket trap, constructed from plastic containers University of Hawaii Þeld station at Kainaliu, HI, on 22 (Highland Plastics, Pasadena, CA) and baited with a January 29 January, 5 February, and 5 March (four mixture of cue-lure and naled (Dibrom Concentrate, replicates). Tests were conducted inside nylon screen Valent USA Corp., Walnut Creek, CA), was main- Þeld cages (3 m in height by3mindiameter) set up tained in each patch (Vargas et al. 1989, 1990, 2003b). under the roof of an open-air shadehouse (Prokopy October 2004 VARGAS ET AL.: MELON FLY PARASITOID AND STERILE FLY RELEASES 1533

Fig. 1. Map of Kailua-Kona area of Hawaii Island showing study and release sites (triangles). and Vargas 1996). Four evenly spaced Þeld cages were paper buckets held in the laboratory to estimate the erected along a north-south transect. A 1.5-m-tall pot- number of wasps released into plots during experi- ted guava tree was placed inside each cage to provide ments. Buckets with screened lids and insects were a plant canopy for parasitoids. Inside each cage were provided with honey and water. On the basis of pre- two Þberglass tray (50by 32 cm) platforms, mounted vious surveys, four pairs of release and control plots on top of a 1-m length of 1.9-cm PVC pipe coated with (four replicates) with ivy gourd plants and fruit, a 2.5-cm band of Tangle-Trap (Tanglefoot Company, Ϸ0.5 ha, were selected throughout the town of Kailua- Grand Rapids, MI) and anchored to the ground. Test- Kona for evaluation of parasite releases. Four separate ing began at 0900 hours and ended 24 h later. Fifty ripe release tests were conducted during 1996 at Alii Dr. fruit were placed on each tray at the beginning of each (19Њ 37.9716 N, 155Њ 59.344Ј W, 35.4-m elevation) on trial for a total of 100 fruit per cage. Parasitoids were 6 June, Hualalai Rd. (19Њ 38.240Ј N, 155Њ 59.220Ј W, released into three cages at the densities described 54.3-m elevation) on 3 July, Middle Keei Dr. previously and allowed to oviposit for 24 h. Test fruit (19Њ 28.077Ј N, 155Њ 54.458Ј W, 122.6-m elevation) on were removed from trays, weighed, placed in boxes, 1 August, and Kamehameha III Dr. (19Њ 35.071Ј N, and handled as described previously. Pupae were re- 155Њ 57.357Ј W, 185.9-m elevation) on 10October. covered from fruit and held as described previously. Corresponding control plots were located at Henry St. Numbers of melon ßy and P. fletcheri that emerged (19Њ 38.573Ј N, 155Њ 59.423Ј W, 59.5-m elevation), Kua- were recorded, and dead pupae were dissected to kini Highway (19Њ 37.799Ј N, 155Њ 59.144Ј W, 48-m determine whether parasitization had occurred. elevation), Middle Keei Dr. (19Њ 28.239Ј N, 155Њ P. fletcheri Open Field Release Studies. P. fletcheri 54.383Ј W 141.1-m elevation), and Kamehameha III wasps were obtained and handled as described pre- Dr. (19Њ 34.565Ј N, 155Њ 57.564Ј W, 104.2-m elevation), viously. Approximately 1,000 pupae were placed and respectively. Mean (ϮSEM) numbers of parasitoids held inside 1.95-liter paper buckets (Sweetheart Cup released from buckets (range 30Ð33) per location per Co. Inc., Owings Mills, MD). Percentage of parasitoids week were 32,983 Ϯ 3,439, 32,729 Ϯ 5,696, 33,999 Ϯ to emerge was estimated from three quality control 2,810, and 37,197 Ϯ 3,337, respectively. There was no 1534 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 5 signiÞcant difference in the numbers of parasitoids ßies and P. fletcheri recovered from fruit for all sites released per site (analysis of variance [ANOVA]: F ϭ were pooled and summarized (melon ßies per gram) 0.22; df ϭ 3, 11; P ϭ 0.8817) (SAS Institute 1999). by month. Numbers of B. cucurbitae and P. fletcheri Wasps were6dofageandwere released for four recovered from fruit in the cage parasitoid tests were consecutive weeks. Two days after the release, 10fruit analyzed by Proc GLM (SAS Institute 1999). Numbers were randomly collected from each of 10sites within of B. cucurbitae and P. fletcheri recovered from fruit a plot for a total of 100 fruit per plot per sample date. were transformed to ln(x ϩ 1) to stabilize variances. Equal numbers of fruit were collected and compared For open Þeld releases, melon ßy and P. fletcheri data from release and control plots. Fruit were held and were analyzed with a repeated measures ANOVA. A processed as described previously. split-plot design was used with main plots arranged in Field Cage Studies with P. fletcheri and Sterile a completely randomized design. The main plot treat- Melon Flies. Tests were conducted at the University ments were release and the subplot factors were date. of Hawaii Þeld station at Kainaliu, HI, inside the four A GLIMMIX.SAS macro was used. It allowed for Pois- cages described previously on 30 October 2001, 10 son distributed count data as the response and Þts a May 2002, and 12 December 2002 (three replicates). split-plot, mixed model (Littell et al. 1996). A proba- Each cage contained Þve potted guava trees arranged bility level of 0.05 was used as the signiÞcance criterion to provide a single plant canopy (1.25 m in diameter). for all statistical tests (SAS Institute 1999).

Wild melon ßies were the F1 generation obtained from P1 stock recovered from infested fruit from papaya orchards at Kapoho, HI. Parasitoids were from the Results same colony described previously. Sterile melon ßies (1:1 male:female) were obtained from the PBARC Melon Fly and P. fletcheri Ecology. During a 1-yr rearing laboratory in Honolulu. Melon ßy pupae survey of ivy gourd patches in Kailua-Kona from April were reared from a 400 generation-old laboratory col- 1996 to March 1997, highest numbers of male melon ony and irradiated with a Gammacell 220Excel irra- ßies were captured during late summer with a monthly diator (MDS Norton, Kanata, Canada) at 10KR. mean (ϮSEM, n ϭ 12) of 29.33 Ϯ 14.94 ßies per trap The three treatments (sterile ßies, P. fletcheri and per day (Fig. 2). Approximately 32.4% of the fruit were sterile ßies, and P. fletcheri) and a control were ran- infested with a mean (ϮSEM) emergence of 2.05 Ϯ domly assigned to four cages. All experiments began 0.13 melon ßies per fruit; with 5.7% of the fruit con- Ϯ Ϯ with wild F1 melon ßies (200 males and 200 females) taining P. fletcheri (mean SEM, 0.22 0.04 wasps per being introduced into each cage. Wild ßies were re- fruit). Numbers of ßies recovered from fruit ranged leased 2 d after eclosion. Cages with sterile ßies con- from 0.05 to 0.27 ßies per gram (Fig. 3). Parasitization tained Ϸ8,000 irradiated melon ßy pupae placed inside of melon ßy by P. fletcheri ranged from 1 to 14%. a small plastic tray on one of the platforms. Sterile ßies P. fletcheri Field Cage Studies. Mean (ϮSEM) num- were allowed to emerge and roost in the guava tree bers of melon ßies emerging from fruit placed inside canopy. Ratio of sterile males to wild females was a control and three treatment cages (1-, 2-, and 3-g Ϸ20:1, at the beginning of the experiment. In treat- densities of parasitized pupae) were 131.8 Ϯ 19.4, ments with P. fletcheri, 200 5-d-old male and female 29.8 Ϯ 8.8, 35.6 Ϯ 8.3, and 6.4 Ϯ 2.9 melon ßies per 100 wasps were released into the cages 5 wk after the fruit, respectively (Table 1). All treatment densities beginning of the test, when wild ßies had matured. All differed signiÞcantly (F ϭ 6.31; df ϭ 3, 16; P ϭ 0.0005) insects were provided with water contained in 3.8-liter from the control. Although there was a numerical plastic buckets with lids Þtted with a 45.7-cm strand of difference, there was no signiÞcant difference (P Ͼ 0.9-cm-diameter cotton dental wick and a plastic tray 0.05) among the treatments. Mean (ϮSEM) numbers (17.5 by 12.5 by 4 cm) with 300 ml of a 3:1 volumetric of P. fletcheri to emerge from fruit in the control, and mixture of sugar and enzymatic yeast hydrolysate the three treatment cages were 5.2 Ϯ 1.8, 48.0 Ϯ 11.0, (U.S. Biochemical Corp., Cleveland, OH). Water and 59.2 Ϯ 12.4, and 32.4 Ϯ 8.4 P. fletcheri per 100 fruit, food were placed on the platform inside the cage. respectively. All treatments differed signiÞcantly (F ϭ Honey was also streaked on pieces of lumite screen 20.8; df ϭ 3, 16; P Ͻ 0.0001) from the control. There attached to the ant-resistant platforms. was no signiÞcant (P Ͼ 0.05) difference among treat- Five weeks from the start of the test, four zucchini ments. fruit were placed on each of the platforms on top of a P. fletcheri Field Releases. In Þeld releases, the ef- 1.25-cm layer of sand. Fruit were weighed and washed fects of treatment (F ϭ 7.40; df ϭ 1, 8.57; P ϭ 0.0247) before placement on platforms. Larvae from wild and week (F ϭ 6.93; df ϭ 4, 17.9; P ϭ 0.0015) were melon ßies were allowed to develop inside fruit for signiÞcant in determining parasitization rate (15.2% 12 d. Test fruit were weighed, placed in Þberglass for release and 3.3% for control plots) (Table 2); how- boxes and handled as described previously. Pupae ever, the week ϫ treatment interaction was not sig- were recovered from fruit and held as described pre- niÞcant (F ϭ 0.78; df ϭ 4, 18.1; P ϭ 0.5552). The effects viously. Numbers of emerged melon ßies per gram of of parasitoid releases (F ϭ 1.00; df ϭ 1, 4.08; P ϭ fruit were calculated. 0.3735) and week (F ϭ 0.43; df ϭ 4, 17.7; P ϭ 0.7880) Statistical Methods. For trap survey data, melon ßy did not have a signiÞcant effect on emergence of ßies captures were summarized monthly (melon ßies per from fruit. The week ϫ treatment interaction was also trap per day). For fruit survey data, numbers of melon not signiÞcant (F ϭ 1.01; df ϭ 4, 17.6; P ϭ 0.4284). October 2004 VARGAS ET AL.: MELON FLY PARASITOID AND STERILE FLY RELEASES 1535

Fig. 2. Captures of male melon ßies (mean Ϯ SEM) in cue-lure traps maintained in patches of ivy gourd throughout Kailua-Kona, HI, from April 1996 to March 1997.

Sterile Melon Flies and P. fletcheri Releases. In cage of fruit and only 5Ð6% of the melon ßies were para- tests with sterile melon ßies and P. fletcheri, treatment sitized by P. fletcheri (Jackson et al. 2003). In studies had a signiÞcant (F ϭ 4.00; df ϭ 3, 9; P ϭ 0.0459) effect of P. fletcheri by Nishida (1955), parasitization rates on emergence of ßies from fruit. Mean (ϮSEM) num- were highest in winter and lowest in summer. Our trap bers of melon ßies emerging per gram of fruit from and fruit surveys of ivy gourd patches during 1996 and cages with parasitoids alone, sterile ßies alone, and 1997 in Kailua-Kona indicated highest melon ßy cap- sterile ßies and parasitoids together were 0.68 Ϯ 0.34, tures during late summer, a range of 0.05Ð0.27 adult 0.16 Ϯ 0.08, and 0.10 Ϯ 0.09, respectively, compared melon ßies obtained per gram of fruit, 1Ð14% parasit- with 0.84 Ϯ 0.14 melon ßies per gram for a control cage ization by P. fletcheri, and lower parasitization rates (Table 3). during summer than in winter. All studies in Hawaii to date (Uchida et al. 1990, Jackson et al. 2003) suggest ivy gourd is an excellent host of melon ßy, with P. Discussion fletcheri being the major parasitoid, albeit in low num- Melon Fly and P. fletcheri Ecology. In a previous bers. Our Þndings have implications not only for Ha- study of ivy gourd in Kailua-Kona, HI, during 1994 and waii but also for the PaciÞc region where ivy gourd is 1995, 0.5Ð0.8 adult melon ßies were obtained per gram being spread. For example, in the northern Mariana

Fig. 3. Population dynamics of melon ßy and P. fletcheri as determined from fruit samples collected from March 1996 until February 1997 throughout Kailua-Kona, HI. 1536 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 5

Table 1. Number of melon flies and parasitoids emerging from ivy gourd fruit placed inside cages with three densities of P. fletcheri

Treatment (No. ͓mean Ϯ SEM͔ emerging/100 fruit exposed 24 h) Species P. fletcheri parasitized pupae placed inside cages (g) 0123 Melon ßy 131.8 Ϯ 19.4a 29.8 Ϯ 8.8b 35.6 Ϯ 8.3b 6.4 Ϯ 2.9b P. fletcheri 5.2 Ϯ 1.8a 48.0 Ϯ 11.0b 59.2 Ϯ 12.4b 32.4 Ϯ 8.4b

Values in the same row followed by the same letter are not signiÞcantly different according to TukeyÕs studentized range (honestly signiÞcant difference) test at the P ϭ 0.05 level (SAS Institute 1999). islands of Rota, Tinian, and Saipan where melon ßy ivy gourd by melon ßy and low P. fletcheri parasitiza- was eradicated in 1965 (Steiner et al. 1968, Mitchell tion rates provided an ideal opportunity to test the 1980), it has become reestablished from Guam effectiveness of augmentative parasitoid releases of P. (Mitchell 1980), and due to an abundance of ivy fletcheri against melon ßy. gourd, has now become a very serious pest of local P. fletcheri Releases. Several studies have demon- (McGregor 2002). strated the feasibility of parasite augmentation with Newell et al. (1952) and Nishida (1955) found that fruit ßies. In Hawaii, release of Diachasmimorpha P. fletcheri attained high levels of parasitization in wild tryoni (Cameron) (at 20,000 per square kilometer per bitter melon fruit, but it was scarce in cultivated fruit week over a 14-km2 area) more than tripled Mediter- (Nishida 1953), even though wild and cultivated areas ranean fruit ßy, Ceratitis capitata (Wiedemann), par- were contiguous. Initially, Ͼ50% parasitization of asitization rates (Wong et al. 1991). In Florida, release melon ßy was reported from collections of infested of 20,000Ð60,000 Diachasmimorpha longicaudata cucurbits (Willard 1920); however, subsequent stud- (Ashmead) wasps per week (in 5- and 13-km2 areas) ies indicated considerably lower parasitization rates, reduced populations of Caribbean fruit ßy, Anastrepha particularly in cultivated areas (Nishida 1955). suspensa (Loew), by 95% (Sivinski et al. 1996). In Nishida (1955) also observed that P. fletcheri, like the Mexico, aerial releases of D. longicaudata resulted in melon ßy, preferred weedy Þelds, but it was inactive increased parasitization rates in mango orchards and under intense light and high temperatures. In com- a 2.7-fold suppression of Anastrepha spp. populations paring parasitization data obtained in Hawaii and In- in backyard orchards (Montoya et al. 2000). In the dia, it seemed that in both localities the parasitoid is current study with releases of P. fletcheri against melon most active during the fall and winter, but the per- ßy, numbers of melon ßies emerging from fruit placed centage of parasitized larvae was considerably less in inside treatment cages were reduced up to 21-fold, and India than in Hawaii. Duration of the period in which numbers of parasitoids were increased 11-fold. In the parasitoid showed its maximum activity through- open Þeld releases of P. fletcheri into ivy gourd patches out the year was also shorter in India (Nishida 1955). throughout the Kailua-Kona area, parasitization rates Purcell and Messing (1996) suggested that parasitiza- were increased 4.7 times in release plots compared tion was higher in rotting fruit on the ground than in with those in control plots. However, this increase was commercially ripe fruit. In the current study, P. fletch- not high enough to signiÞcantly (P Ͼ 0.05) affect the eri was detected year-round in wild ivy gourd patches. emergence of ßies from fruit. When our results ob- However, unlike with wild bitter melon, parasitization tained in Þeld cages were compared with those from rates of P. fletcheri infesting melon ßy in ivy gourd open Þeld releases, our conclusion is that in patches of were low. Nonetheless, the persistent infestation of ivy gourd, parasitoids had more difÞculty Þnding highly dispersed melon ßy larvae in fruit often hidden Table 2. Number of melon flies emerging from fruit and per- in vegetation, compared with Þeld cage tests where centage of P. fletcheri parasitization based on fruit sampled from infested ivy gourd fruit on platforms were more easily control and release plots located. Comparison of our results from cage and Þeld tests suggest that higher parasitization rates (32.4Ð Test Wk Treatment Estimate Upper Lower 48%) are easily obtained in the cage situation where Flies/g 1 C 0.042 0.008 0.212 1 R 0.036 0.008 0.161 2 C 0.089 0.020 0.400 Table 3. Mean melon flies recovered from fruit placed inside 2 R 0.027 0.006 0.120 one of four treatment cages 3 C 0.050 0.011 0.235 3 R 0.025 0.005 0.116 Melon ßy/g Treatment 4 C 0.063 0.013 0.316 (mean Ϯ SEM) 4 R 0.033 0.007 0.156 5 C 0.026 0.005 0.137 Control 0.84 Ϯ 0.14 A 5 R 0.045 0.007 0.271 P. fletcheri 0.68 Ϯ 0.34 AB Parasitization (%) C 3.3a 1.014.9 Sterile melon ßies 0.16 Ϯ 0.08 B R 15.2b 3.2 36.0 P. fletcheri/sterile melon ßies 0.10 Ϯ 0.09 B

C, control site; R, release site. Least squares means and 95% CI. Values in the same column followed by the same letter are not Values in each category followed by the same letter are not signiÞ- signiÞcantly different at the P ϭ 0.05 level (least signiÞcant difference, cantly different at the 0.0025 level (PROC MIXED, Littell et al. 1996). PROC GLM, SAS Institute 1999). October 2004 VARGAS ET AL.: MELON FLY PARASITOID AND STERILE FLY RELEASES 1537 parasitoids had easy access to larvae but was much ßy suppression, although in many habitats and from an more difÞcult in the weedy Þeld situation (15.2%) IPM perspective, they would be useful. In demo- where melon ßy larvae were highly dispersed graphic projections by Vargas et al. (2002), P. fletcheri throughout the Þeld inside fruit. was shown to possess an intrinsic rate of increase 25% Sterile Melon Fly and P. fletcheri Releases. Eradi- less than that of melon ßy. This suggests that even if cation of melon ßy by overßooding with sterile males P. fletcheri could locate most of the melon ßy larvae, was Þrst demonstrated in the northern Mariana Islands it would not be able to keep pace with the melon ßy, by Steiner et al. (1968). Melon ßy was subsequently on the basis of a lower reproductive rate. Further- eradicated from the Okinawa Islands with sterile more, our results showing modest P. fletcheri parasit- melon ßies at ratios of Ͻ10:1 (sterile males:wild fe- ization rates compared with other Hawaiian fruit ßy males) (Koyama 1996). In the present Þeld cage stud- parasitoids such as Fopius arisanus (Sonan) (Vargas et ies, sterile ßies had a signiÞcant effect on melon ßy al. 2001) suggest that a search for more effective melon infestation of zucchini fruit. Releases of sterile ßies at ßy parasitoids for introduction into Hawaii might be ratios of 20:1 were found to be effective and rapid at considered. Based on our cage studies, what may sup- reducing melon ßy infestation in zucchini. Knipling plement the effectiveness of these parasitoids would (1979, 1992) proposed the use of sterile insects and be the addition of sterile ßy releases to an IPM system. parasitoids for eradication of insect populations. Com- Our data suggest that releases of small numbers of binations of sterile insect and parasitoid releases were sterile ßies would have an immediate effect on fruit used to suppress Mediterranean fruit ßy populations in infestation over a single generation. For example, if Kula, HI (Wong et al. 1992). In Guatemala, reports the net reproductive rate (2.6-fold increase per gen- suggest successful control of Mediterranean fruit ßy eration) of melon ßy was reduced by 50% over one on coffee, Coffea arabica L., farms by augmentative generation, there would be a 1.3-fold decrease in the release of D. longicaudata and sterile ßies (Cancino- number of females produced per female in the next Diaz et al. 1996). Our results in Þeld cages conÞrm the generation (Vargas et al. 1997). These Þndings will be compatability of parasitoid releases followed by sterile further tested in an area-wide demonstration site at ßy releases. In the present cage tests with sterile melon Kamuela, HI, where both P. fletcheri and sterile melon ßies and P. fletcheri, combinations of sterile ßies and P. ßies will be released. It remains to be determined fletcheri produced the greatest numerical reduction whether sterile ßies and augmentative parasitoid re- (9-fold) in melon ßy emergence from fruit, although leases will be cost-effective and sustainable in area- the reduction was not statistically different than sterile wide IPM systems in Hawaii. ßies or parasitoids alone. However, reductions ob- tained with sterile ßies alone or in combination with parasitoids were signiÞcantly greater than (P Ͻ 0.05) Acknowledgments the control, whereas those for parasitoids alone were We are grateful to P. Barr, T. Urago, R. Gibbons, and L. not. These results suggest that the effect of sterile ßies Oride (USPBARC, USDAÐARS, Hilo and R. Kurashima, D. on adult reproduction may be greater than parasitoid Ichimura, C. Brown, and C. Albrecht, USPBARC, USDAÐ mortality on larvae for applications in an integrated ARS, Honolulu) for assistance with these experiments. We pest management (IPM) system. Nonetheless, the ef- thank Harold Steen and Mark Meisner for valuable assistance fects of the sterile melon ßies and the parasitoids were in setting up experiments at the UH Agricultural Experiment on different stages of the insect, the adult and larval Station at Kainaliu. stages, respectively, and would seem to be compatible from an IPM perspective, when multiple strategies are References Cited desirable. Melon ßy is a severe pest on small melon and veg- Back, E. A., and C. E. Pemberton. 1917. The melon ßy in etable farms throughout Hawaii. In 1999, USDAÐARS Hawaii. U.S. Dep. Agric. Bull. 491. funded the Hawaii Fruit Fly Areawide Pest Manage- Cancino-Diaz, J., L. S. Ruiz, and E. Aguilar. 1996. Evalu- ment program to suppress melon ßies below economic acion de liberaciones inundativas de parasitoids Diachas- thresholds, while reducing the use of organophos- mimorpha longicaudata sobre povlaciones de Ceratitis capitata en Þncas cafetaleras en Guatemala C. A. p. 68. In phate on farms (Vargas et al. 2003a). The Proceedings, Second Meeting of the Working Group on program includes developing and integrating biolog- Fruit Flies of the Second Meeting of the Working Group ically based pest control technology into a compre- on Fruit Flies of the Western Hemisphere, 3Ð8 November hensive management package that will be economi- 1996, Vina del Mar, Chile. cally viable, environmentally friendly, and sustainable. Chun, M. E. 2001. Biology and host speciÞcity of Melittia Components includes 1) Þeld sanitation; 2) protein oedipus (Lepidoptera: Sesiidae), a biological control bait sprays (Vargas et al. 2001); 3) male annihilation agent of Coccinia grandis (Cucurbitaceae). Proc. Hawai- (Vargas et al. 2000); and if needed, 4) augmentative ian Entomol. Soc. 35: 85Ð93. parasitoid releases and 5) sterile insect releases. The Clausen, C. P., D. W. Clancy, and Q. C. Chock. 1965. Bio- logical control of the oriental fruit ßy (Dacus dorsalis present small plot and Þeld cage results provide pos- Hendel) and other fruit ßies in Hawaii. U.S. Dep. Agric. itive evidence on the use of sterile ßies and parasitoids Tech. Bull. 1322. for suppression of melon ßy. However, the modest Jackson, C. G., R. I. Vargas, and D. Y. Suda. 2003. Popula- reductions obtained with parasitoids, suggest that tions of Bactrocera cucurbitae (Diptera: Tephritidae) and parasitoid releases are not the entire solution to melon its parasitoid, Psyttalia fletcheri (Hymenoptera: Bra- 1538 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 5

conidae) in Coccinia grandis (Cucurbitaceae) or ivy (eds.), Biology and utilization of the Cucurbitaceae. Cor- gourd on the island of Hawaii. Proc. Hawaiian Entomol. nell University Press, Ithaca, NY. Soc. 36: 39Ð46. Sivinski, J. M., C. O. Calkins, R. Baranowsky, D. Harris, Knipling, E. F. 1979. The basic principles of insect popula- J. Brambila, J. Diaz, R. E. Burns, T. Holler, and D. Dod- tion suppression and management. U.S. Dep. Agric., Ag- son. 1996. Suppression of Caribbean fruit ßy (Anas- ric. Handb. No. 512. trepha suspensa (Loew) Diptera: Tephritidae) population Knipling, E. F. 1992. Principles of insect ana- through augmented releases of the parasitoid Diachas- lyzed from new perspectives, practical implications for mimorpha longicaudata (Ashmead) (Hymenoptera: Bra- regulating insect populations by biological means. U.S. conidae). Biol. Control 6: 177Ð185. Dep. Agric., Agric. Res. Serv., Agric. Handb. No. 693. Steiner, L. F., E. J. Harris, W. C. Mitchell, M. S. Fujimoto, and Koyama, J. 1996. Eradication of the melon ßy, Bactrocera L. D. Christenson. 1968. Melon ßy eradication by over- cucurbitae, by the in Japan. IAEA ßooding with sterile ßies. J. Econ. Entomol. 58: 519Ð522. training manual. June 19. Telford, I.R.H. 1990. Cruciferae. Cucurbitaceae, pp. 569Ð Linney, G. 1986. Coccinia grandis (L.) Voigt: a new cucur- 581. In W. L. Wagner, D. R. Herbst, and S. H. Sohmer bitaceous weed in Hawaii. Hawaiian Bot. Soc. Newsl. 25: (eds.), Manual of the ßowering plants of Hawaii. Bishop 3Ð5. Museum Special Publication 83. University of Hawaii Liquido, N. J., R. T. Cunningham, S. Nakagawa, and G. Press, Honolulu, HI. Uchida. 1990. Survey of Dacus cucurbitae Coquillett Uchida, G. K., R. I. Vargas, J. W. Beardsley, and N. J. Liquido. (Diptera: Tephritidae) infestations in the cultivated and 1990. Host suitability of wild cucurbits for melon ßy, weedy forms of Momordica charantia L. (Cucurbitaceae). Dacus cucurbitae Coquillett, in Hawaii, with notes on Proc. Hawaiian Entomol. Soc. 30: 31Ð36. their distribution and taxonomic status. Proc. Hawaiian Littell, R. C., G. A. Milliken, W. W. Stroup, and R. D. Entomol. Soc. 30: 37Ð52. Wolfinger. 1996. SAS system for mixed models. SAS In- Vargas, R. I., J. D. Stark, and T. Nishida. 1989. Abundance, stitute, Cary, NC. distribution, and dispersion indices of the oriental fruit ßy McGregor, A. 2002. Eradication of melon ßy from Guam and melon ßy on Kauai. Hawaiian Islands. J. Econ. En- and the Commonwealth of the Northern Mariana Islands: tomol. 82: l609Ðl6l5. a beneÞt analysis. Secretariat of the PaciÞc Community. Vargas, R. I., J. D. Stark, and T. Nishida. 1990. Population Suva, Fiji. dynamics, habitat preference, and seasonal distribution Mitchell, W. C. 1980. VeriÞcation of the absence of oriental patterns of oriental fruit ßy and melon ßy in an agricul- fruit and melon fruit ßy following an eradication program tural area. Environ. Entomol. l9: l820Ðl828. in the Mariana Islands. Proc. Hawaiian Entomol. Soc. 23: Vargas, R. I., W. A. Walsh, D. Kanehisa, E. B. Jang, and J. W. 239Ð241. Armstrong. 1997. Demography of four Hawaiian fruit Montoya, P., P. Liedo, B. Benrey, J. Cancino, J. F. Barrera, ßies (Diptera: Tephritidae) reared at Þve constant tem- J. Sivinski, and M. Aluja. 2000. Biological control of peratures. Ann. Entomol. Soc. Am. 90: 162Ð168. Anastrepha spp. (Diptera: Tephritidae in mango orchards Vargas, R. I., J. D. Stark, M. H. Kido, H. M. Ketter, and L. C. through augmentative releases of Diachasimimorpha lon- Whitehand. 2000. Methyl eugenol and cue-lure traps for gicaudata (Ashmead) (Hymenoptera: Braconidae). Biol. suppression of male oriental fruit ßies and melon ßies Control 18: 216Ð224. (Diptera: Tephritidae) in Hawaii: effects of lure mixtures Newell, I. M., W. C. Mitchell, and F. L. Rathburn. 1952. and weathering. J. Econ. Entomol. 93: 81Ð87. Infestation norms for Dacus curcubitae in Momordica bal- Vargas, R. I., S. L. Peck, G. T. McQuate, C. G. Jackson, J. D. samina, and seasonal differences in activity of the para- Stark, and J. W. Armstrong. 2001. Potential for area- site, Opius fletcheri. Proc. Hawaiian Entomol. Soc. 14: wide integrated management of Mediterranean fruit ßy 497Ð508. with a braconid parasitoid and a novel bait spray. J. Econ. Nishida, T. 1953. Ecological study of the melon ßy, Dacus Entomol. 94: 817Ð825. cucurbitae Coquillett, in the Hawaiian Islands. Ph.D. dis- Vargas, R.I.M. Ramadan, T. Hussain, N. Mochizuki, R. C. sertation, University of California, Berkeley, CA. Bautista, and J. D. Stark. 2002. Comparative demogra- Nishida, T. 1955. Natural enemies of the melon ßy, Dacus phy of six fruit ßy (Diptera: Tephritidae) parasitoids curcubitae Coq. in Hawaii. Ann. Entomol. Soc. Am. 48: (Hymenoptera: Braconidae). Biol. Control 25: 30Ð40. 171Ð178. Vargas, R. I., E. B. Jang, and L. M. Klungness. 2003a. Area- O’Brien, C. W., and J. Pakaluk. 1998. Two new species of wide pest management of fruit ßies in Hawaiian and Acythopeus Pascoe (Coleoptera: Curculionidae: Baridi- vegetables, pp. 37Ð46. In Recent Trends on Sterile Insect nae) from Coccinia grandis (L.) Voight (Cucurbitaceae) Technique and Area-wide Integrated Pest Management. in Kenya. Proc. Entomol. Wash. 100: 764Ð774. Research Institute for Subtropics. Prokopy, R. J., and R. I. Vargas. 1996. Attraction of Ceratitis Vargas, R. I., N. W. Miller, and J. D. Stark. 2003b. Field trials capitata (Diptera: Tephritidae) ßies to odor of coffee of spinosad as a replacement for naled, ddvp, and mala- fruit. J. Chem. Ecol. 22: 807Ð820. thion in methyl eugenol and cue-lure bucket traps to Purcell, M. F., and R. H. Messing. 1996. Ripeness effects of attract and kill male oriental fruit ßies and melon ßies three vegetable crops on abundance of augmentatively (Diptera: Tephritidae) in Hawaii. J. Econ. Entomol. 96: released Psyttalia fletcheri (Hym.: Braconidae): improved 1780Ð1785. sampling and release methods. Entomophaga 41: 105Ð115. Weems. 1964. Melon ßy (Dacus cucurbitae Coquillett) Ramadan, M. M., and R. H. Messing. 2002. A survey for (Diptera: Tephritidae). Entomology Circular, Division of potential biocontrol agents of Bactrocera cucurbitae Plant Industry, Florida Department of Agriculture and (Diptera: Tephritidae) in Thailand. Proc. Hawaiian En- Consumer Services 21: 1Ð2. tomol. Soc. 36: 115Ð122. Willard, H. F. 1920. Opius fletcheri as a parasite of the Melon SAS Institute. 1999. SAS/STAT userÕs guide, version 6. SAS ßy in Hawaii. J. Agric. Res. 20: 434Ð438. Institute, Cary, NC. White, I.M. and M. M. Elson-Harris. 1992. Fruit ßies of Singh, A. K. 1990. Cytogenetics and evolution in Cucurbi- economic signiÞcance: their identiÞcation and bionom- taceae. In D. M. Bates, R. W. Robinson, and C. Jeffrey ics. CAB International, Wallingford, United Kingdom. October 2004 VARGAS ET AL.: MELON FLY PARASITOID AND STERILE FLY RELEASES 1539

Wong, T.T.Y., M. M. Ramadan, D. O. McInnis, N. Mochizuki, Tephritidae) population with concurrent parasitoid and J. I. Nishimoto, and J. C. Herr. 1991. Augmentative re- sterile ßy releases in Kula, Maui, Hawaii. J. Econ. Ento- leases of Diachasmimorpha tryoni (Hymenoptera: Bra- mol. 85: 1671Ð1681. conidae) to suppress a Mediterranean fruit ßy (Diptera: Tephritidae) population in Kula, Maui, Hawaii. Biol. Con- trol 1: 2Ð7. Wong, T.T.Y., M. M. Ramadan, J. C. Herr, and D. O. McInnis. 1992. Suppression of a Mediterranean fruit ßy (Diptera: Received 23 March 2004; accepted 9 July 2004.