Plant Protection Quarterly Vol.24(1) 2009 27 (Merr.) Fosberg (Cucurbitaceae), an en- demic plant in Guam and Rota, and the Life-history, host preference and establishment introduction and establishment of M. oedi- status of oedipus (: ), pus in Guam and Saipan. The aim of this study is to obtain more precise estimates a biological control agent for and predict performance or generation (Cucurbitaceae) in the Mariana Islands times in the new range of M. oedipus. Materials and methods A A B G.V.P. Reddy , Z.T. Cruz and R. Muniappan All the laboratory studies were conducted A Agricultural Experiment Station, College of Natural and Applied Sciences, at the Western Pacifi c Biocontrol Quaran- University of Guam, Mangilao, Guam 96923, USA. tine Laboratory (WPBQL) of the Univer- B IPM CRSP, OIRED, Virginia Tech, 526 Prices Fork Road, Blacksburg, VA sity of Guam. 24061, USA. Importation of M. oedipus and laboratory conditions About 200 C. grandis cuttings (~20 cm long × 4–5 cm diameter) with immature stages Summary but it has become invasive in lowland of M. oedipus were brought from Kona Ivy gourd or scarlet gourd, Coccinia gran- areas of the Hawaiian Islands (Linney (Hawaii) to Guam and reared at the WP- dis (Violales: Cucurbitaceae), is a peren- 1986, Telford 1990, Chun 2001). This per- BQL. The controlled environmental condi- nial vine and a native of that has ennial vine grows best under conditions tions in the quarantine room were main- invaded lowland areas of the Hawaiian of moderate rainfall and high humidity tained at a 12:12 L:D photoperiod, with a and Mariana Islands. Failure of mechani- (Muniappan et al. in press). Accidental mean day and night temperature of 32 and cal or chemical control techniques to sup- introduction of this vine into Guam and 17°C , respectively, and a relative humid- press this weed led to the initiation of a Saipan in the 1980s resulted in the inva- ity of 65–80%. The quarantine room was biological control program in the Mari- sion of over 100 hectares in different parts illuminated with two plant growth lights ana Islands (Guam, Saipan, Rota and Ti- of Guam and almost 14 000 hectares in (Gualala Robotics Inc. Denver, Colorado) nian). This follows the success achieved Saipan (Raman et al. 2006, Bamba 2008). It controlled by automated timers. in Hawaii by introducing the natural en- is also invading the neighbouring islands emies, Acythopeus cocciniae (Coleoptera: of Rota and Tinian (USDA 2003, 2004). rearing Curculionidae), Acythopeus burkharto- Mechanical and chemical control meth- From the stock of 200 cuttings which con- rum (Coleoptera: Curculionidae) and ods have not proven effective, feasible or tained various immature stages of M. oedi- (Lepidoptera: Sesiidae). sustainable in the Mariana and Hawaiian pus, 20 cuttings were selected and placed Of these three biological control agents, Islands (Motooka et al. 2002, Muniappan in 10 identical plastic containers each 45 M. oedipus has proven to be effective in et al. in press). A biological control pro- cm long × 30 cm wide × 15 cm depth and suppressing C. grandis in Hawaii. gram was initiated in the Mariana Islands sealed with lids. Using a sharp blade, a 28 Life history of M. oedipus has been de- based on the success achieved in Hawaii cm × 13 cm incision made on the top of scribed. Females laid 144 eggs on average by introducing the natural enemies, Acy- the lid, which was covered with muslin during their lifetime. The mean duration thopeus cocciniae O’Brien and Pakaluk cloth and sealed with glue to allow aera- of the egg, larval and pupal stages were (Coleoptera: Curculionidae), Acythopeus tion. Subsequently, C. grandis vines about 9.4, 43.5 and 21.4 days, respectively. Fe- burkhartorum O’Brien and Pakaluk (Cole- 4–5 cm diameter were harvested from the males lived longer than the males with optera: Curculionidae) and Melittia oedipus fi eld and brought to the laboratory where a mean of 4.5 days and 3.4 days, respec- Oberthür (Lepidoptera: Sesiidae) (Reddy they were rinsed with a 10:1 (v:v) water/ tively. This study indicated that M. oed- and Muniappan 2005). Acythopeus cocciniae Clorox® (bleach) solution and dried off ipus is specifi c to C. grandis and larval grubs mine the leaves causing defoliation, with paper towels. Care was taken to avoid development did not occur on Zehneria A. burkhartorum grubs induce stem and any vines that might have been exposed to guamensis (Cucurbitaceae), an endemic petiole galls, and M. oedipus caterpillars herbicides or insecticides. Vines were cut plant of Guam and Rota. Field release burrow through the larger stems and oc- into pieces measuring 15 to 17 cm in length of M. oedipus in the Marianas resulted casionally the roots causing the vines to and eight to 20 pieces were placed in the in establishment in Guam and Saipan, wither (Chun 2001). Chun’s (2001) host containers atop the old cuttings so that the however the presence of pavement ant specifi city tests confi rmed that M. oedipus larvae could migrate to them as the old Tetramorium bicarinatum (Nylander) is monophagous and specifi c to C. grandis. cuttings deteriorated. This was essential in (Hymenoptera: Formicidae) on vines in In larval feeding tests, a few reached adult raising a successful culture. Olfactory cues the fi eld in Saipan has hindered estab- stage on cucumber (Cucumis sativus L. Cu- given off by the freshly cut vines stimu- lishment there. curbitaceae) vines but no further develop- lated the larvae to migrate to the newer Keywords: Melittia oedipus, Sesiidae, ment was observed. However, choice-tests cuttings and continue feeding. Coccinia grandis, Zehneria guamensis, for oviposition revealed that under fi eld When the cuttings had rotted, they Cucurbitaceae, life-history, host prefer- conditions, M. oedipus females would be were gently peeled apart and the pupae ence, Mariana Islands. highly unlikely to lay eggs on C. sativus. formed in them were removed. Weekly Since 1996, when M. oedipus was released collections of the rotten vine cuttings Introduction in the Hawaiian Islands, there has been no were placed in a double plastic garbage Ivy gourd or scarlet gourd, Coccinia gran- record of attack on C. sativus or any other bags and incinerated at the Guam Refi n- dis (L.) Voigt (Violales: Cucurbitaceae), a non-target plant (Chun 2001). ery and Environmental Services Company native of East Africa is widely distributed Chun (2001) carried out some prelimi- (GRESCO). The collected pupae were then in the Indo-Malayan region (Singh 1990). nary studies on the life history of M. oedi- placed in plastic boxes 14 cm long × 9.5 It has also become naturalized in parts pus. In this paper, we present details of the cm wide × 10 cm height until the adults of Australia, the Caribbean, the southern life history of M. oedipus, a host-specifi ci- emerged. Each box had an opening (8 cm United States, and several Pacifi c Islands ty test conducted on Zehneria guamensis × 4 cm) on the lid which was covered with 28 Plant Protection Quarterly Vol.24(1) 2009 muslin cloth and sealed with glue for aera- and life history under the laboratory Host preference studies tion purposes. Upon emergence from the conditions in Guam, and compared with The host specifi city tests on M. oedipus were pupa, new adults (<1 day old) were re- studies by Chun et al. (2001) carried out conducted at WPBQL of the University of leased into a cage with dimensions 91.4 in Hawaii. Guam (USDA 2007). Chun (2001) has al- cm wide × 45.7 cm deep × 63.5 cm tall. The ready carried out host specifi city tests on base of the cage was aluminium, while the Oviposition rate by M. oedipus several species of plants in the Cucurbita- top and back side was of Plexiglas which Three to six female and six to 12 newly ceae family and closely related ones using permitted light into the cage (light, par- emerged male were released into a the centrifugal phylogenic method recom- ticularly natural sunlight stimulates mat- cage (measuring 91.4 cm wide × 45.7 cm mended by Wapshere (1974). In consulta- ing and oviposition). The front door had deep × 63.5 cm tall) containing a single tions with the University of Guam, Guam two sleeves (of cheese cloth material) for potted C. grandis plant to ensure mating. Department of Agriculture, Department of release of new moths, collection of dead Moths were fed with a 1:1 (v:v) honey/wa- Land and Natural Resources of the North- ones and for feeding the moths. The other ter solution and left undisturbed for a 24 ern Marianas, the U.S. Fish and Wildlife two sides of the cage were covered and h period to allow for oviposition. Artifi cial Service, and APHIS, it was decided that sealed with muslin cloth for aeration. A C. sunlight, necessary to stimulate mating, the endemic species, Zehneria guamensis grandis plant was potted in a 3 L contain- was emitted via a mobile sun lamp which (Merr.) Fosberg (Cucurbitaceae), should er and placed inside the cage to emulate was maintained at a 12:12 L:D photope- be one of the candidates in a host spe- natural conditions as well as to allow for riod. After 24 h, the moths were removed cifi city test since it belongs to the same oviposition by the gravid females. and released into a separate cage and the family as C. grandis and is the only cucur- While adult male and female M. oedi- eggs collected, counted and recorded. Ad- bit species endemic to the region (Stone pus are similar in appearance, they can be ditionally, eggs were measured under a 1970). differentiated by closely examining the binocular microscope equipped with an These studies were carried out using ventral surface of the posterior tip of the ocular micrometer. This procedure con- no-choice and free-choice tests. In the no- abdomen. Whereas males have a small tinued until the female moths died. To choice tests, the response of M. oedipus was narrow abdomen, females have an en- obtain the mean number of eggs laid tested by placing individual host plants in larged abdomen and a pointed oviposi- per female, the total number of eggs col- separate cages. Six to seven week old C. tor. Moths were fed daily with a 1:1 (v:v) lected from the cage was divided by the grandis raised from cuttings and Z. gua- honey/water solution which was soaked number of female moths released into the mensis raised from seeds were placed sep- onto a clean yellow matting placed within cage. The experiment was replicated fi ve arately in a cage (measuring 91.4 cm wide a Petri dish (9 cm diameter). Our previ- times. × 45.7 cm deep × 63.5 cm tall) with cloth ous studies (Reddy et al. 2008) indicated sleeves on the sides for introducing adults. that yellow paper moistened with honey- Duration of various life stages of M. Adults were fed with a 1:1 (v:v) honey/ water increased the feeding rate of M. oedi- oedipus water solution. Four pairs of male and fe- pus, leading to the possibility of using it Three adult M. oedipus females and six male M. oedipus adults (<1 day old) were for mass culture. Moths laid eggs on the males (<1 day old) were released into a released into cages with host plants. Plants host plant as well as on the walls of the cage containing a single potted C. grandis were removed after 24 h, and the number cage. Eggs were carefully collected using plant. Moths were fed a 1:1 (v:v) honey/ of eggs laid on each plant was counted. a small moistened paintbrush or fi ngers, water solution and left for 24 h to allow for Once the laid eggs were counted and har- and were placed into plastic containers (20 oviposition by the gravid female. The eggs vested, the procedure was continued until cm long × 12 cm wide) with lids. The lids were then collected in a 9 cm Petri dish the moths had died. The host specifi city of had openings of 11 cm long × 5 cm wide and checked daily for larval emergence. M. oedipus to C. grandis and Z. guamensis covered with muslin cloth and sealed with Duration of the egg stage was calculated was also studied under free-choice of host glue for aeration. Coccinia grandis vines as the date of oviposition to the date of plants using the same cages. Six to seven (1.5–2 cm diameter) were harvested from eclosion (n = 36). week old C. grandis plants raised from the fi eld, cut into pieces 11–12 cm long and Newly emerged larvae were placed on cuttings and Z. guamensis plants raised 10 pieces, which were rinsed in 10:1 wa- fresh cuttings of C. grandis kept in plastic from seeds were placed together in a cage ter/Clorox solution and dried off, were containers. When the larvae migrated and with an inter-plant distance of about 8 cm. placed into the containers with M. oedipus bored into the cuttings, they were left un- and four pairs of males and females (<1 eggs. Upon eclosion the young larvae mi- disturbed. New cuttings of C. grandis were day old) were released into the cage. The grated to the new cuttings. placed into the containers weekly to en- plants were removed after 24 h, and the When bore holes and frass on the sure adequate food supply for the larvae. numbers of eggs laid on each host plant surface of the stems were observed, the The larvae migrated to the fresh cuttings were counted and harvested. This method smaller cuttings with young larvae were as the old cuttings started to deteriorate. was repeated until the moths had died. transferred into the plastic boxes contain- Newly formed pupae (n = 12) were re- Both no-choice and free-choice tests were ing older larvae and the culturing method moved and placed in separate containers replicated four times. was repeated. All the rearing boxes and for further observations. Adult emergence Separate tests were conducted to evalu- oviposition cages were placed on table- from the pupal case was monitored daily ate the infl uence of the two different host tops the bases of which were fi tted with and the duration of the pupal stage deter- plants, C. grandis and Z. guamensis, on the ant traps containing water-fi lled pans (18 mined. Since the larval stage is delicate survival rate of the immature stages and cm diameter and 5 cm depth). This cultur- and handling resulted in high mortality, adult emergence of M. oedipus. A small ing technique consistently produced vari- no attempt was made to determine the camel-hair paintbrush was used to attach ous stages of the M. oedipus with minimal duration of each instar. For the duration the M. oedipus eggs to the host plants. The mortality. of larval and pupal stages n = 12 was used. paintbrush was lightly dipped into a bowl Males and females were distinguished containing raw chicken egg albumen and Life-history of M. oedipus under from one another through morphological gently applied to a small portion of one laboratory conditions differences such as absence or presence of of the upper leaves of the host plants. The In order to develop a mass trapping tech- ovipositor and size of abdomen. eggs were then carefully adhered to the nique for M. oedipus, some basic studies leaf. A total of 10 M. oedipus eggs were at- were conducted on the oviposition rate tached to each live intact plant of C. grandis Plant Protection Quarterly Vol.24(1) 2009 29 and Z. guamensis, and these were placed Results and pupate in the vine (Eichlin 1995). The separately in a cage under no-choice con- Oviposition rate by M. oedipus mean size of the pupae (n = 10) was 2.0 × ditions. In free-choice tests, fi ve eggs were Mating by M. oedipus adults generally took 0.9 cm (length × width). The pupal case attached to each host plant in the same place on the day of emergence and after was brownish-grey (sometimes darker) in manner as the no-choice tests and repli- the wings of the moths had dried, which colour similar to that of the vine and the cated four times. Upon eclosion of larvae is typically when females begin to emit time to pupate was about three weeks. On from the eggs, larvae began feeding on pheromones (Chun 2001). The female ovi- the average, the overall duration from the the introduced plants. The larvae were positor which is at the posterior end of the egg to pupal stage was approximately 2.5 left undisturbed on the plants until pupa- abdomen is pointed, elongated and some- months. Adults of M. oedipus were similar tion. When a plant was severely damaged what narrow in appearance. Females were in overall appearance, particularly black by the larvae, a new plant was introduced generally observed to lay eggs within an and white from head to abdomen (with adjacent to the damaged plant, so that the hour of mating, and laid eggs individu- a thick single white stripe across the dor- larvae could migrate and feed on the new ally throughout the entire C. grandis plant, sal side of the abdomen), a tawny brown plant. The pupae were harvested from the including both upper and bottom portions thorax, fuzzy hair occurring near the hind different host plants and placed inside the of the leaves, tendrils, and stem. Eggs were wings, specks of orange on the forew- plastic boxes. Counts were made of adult also found laid all over the cage in which ings, and a white abdomen (ventral side). emergence from the pupae that formed on the moths were released, on the pot which Female moths have an ovipositor at the the respective plants. contained the host plant, and on the pan at anterior end and have a larger, fuller abdo- the base of the pot. During oviposition, the men than males. Although it was not sig- Field release female took on a peculiar behaviour. The nifi cantly different, females lived slightly An environmental-assessment statement female would hover over an area, as longer than the male moths (t = 1.84, df = 7, was prepared based on the host specifi - if looking for an ideal spot to lay her egg, P >0.05, Paired t-test, Analytical Software city tests conducted and submitted to Ani- would then lift up her ovipositor and tail 2003). The average lifespan of the adult mal and Plant Health Inspection Service end to initiate contact with either the plant male and female M. oedipus were 3 and 5 (APHIS) of US Department of Agriculture or part of the cage and then positioned days, respectively (Figure 1). (USDA-APHIS 2007). Field releases were her egg onto the desired area. There were made after securing permission from instances in which it took up to three to Host preference studies of M. oedipus on APHIS for the release of M. oedipus in fi ve times for the female moth to select an C. grandis and Z. guamensis Guam and the Northern Mariana Islands. area for oviposition. Eggs of M. oedipus ap- In both no-choice and free-choice tests, M. For fi eld release, C. grandis cuttings with peared brownish-red in colour and were oedipus females laid signifi cantly (P <0.05, lengths of 9–12 cm and 15–18 cm with vari- semi-spherical in shape with an inden- paired t-test) more eggs on C. grandis ous larval instars and adults were used. tation or dip at the top of the egg. Eggs (mean of 55.0 and 47.7, respectively) than The smaller sized cuttings had one to two appeared somewhat shimmery under the Z. guamensis. While a mean number of 0.6 larvae while the larger ones contained microscope light. Mean egg size (n = 10) eggs were oviposited on the Z. guamensis two to three. This was ascertained by oc- was 1.4 × 1.2 mm (height × diameter) and host plant in the free-choice test, no eggs currence of larval frass on the sides of the eggs hatched after 9.3 ± 0.1 days (mean ± were laid in the no-choice tests (Figure vines. The infested stems were tied with SEM) (n = 16) in the laboratory at 28–30°C. 2). When M. oedipus eggs were attached string onto live C. grandis vines naturally The mean number of eggs laid per female to leaves of C. grandis and Z. guamensis, grown in the fi eld, while adults were re- was 144 (Table 1). no larval development was observed in Z. leased on nearby C. grandis. The larvae guamensis. However, in both free-choice moved to the vines in the fi eld when the Duration of life stages of M. oedipus and no-choice tests, normal adults of M. cuttings brought in from the laboratory The newly emerged larva was creamy- oedipus emerged from larvae fed on C. started to deteriorate. From July to De- white in colour and the head capsule was grandis (Figure 3). cember 2007, 130 smaller and 368 larger of a dark-brown to black colour. It bored C. grandis cuttings with larvae were tied into the stem soon after emergence, and Field release and monitoring onto vines in the fi eld, in addition to 128 being a stem borer, the duration of each in- The survey conducted on 14 January 2008 (45♂ and 128♀) adult moths fi eld released star could not be observed as disturbance indicated that M. oedipus was established for the control of C. grandis in the areas and handling of larvae resulted in high in the original release sites of Marbo Caves of Marbo Caves (13.49°N, 144.87°E) and mortality. The larval feeding in the vines and Inarajan Bay in Guam. The survey Inarajan Bay (13.28°N, 144.75°E) on Guam. cause gall-like thickenings but these are conducted on 19 March 2008 in Saipan Similarly from August to November 2007, not true galls (Eichin 1995). Unlike its re- and Rota indicated that M. oedipus was 162 smaller and 600 larger C. grandis cut- lated species Melittia snowii, M. cucurbitae, established at one of the release sites in tings with M. oedipus larvae were tied onto M. calabaza, M. pulchripes and M. pauper Saipan but there were no signs of estab- vines occurring naturally in the fi eld in Ga- where in the larvae exit the vines and pu- lishment in Rota. It was also observed rapan (15.21°N, 145.72°E), Marpi (15.28°N, pate in the soil, M. oedipus larvae remain that pavement ant Tetramorium bicarinatum 145.81°E), and Kagman (15.19°N, 145.76°E) on Saipan. On 11 January 2007, 100 smaller and 80 larger C. grandis cuttings with M. Table 1. Studies on the oviposition rate by Melittia oedipus under oedipus larvae were tied onto vines to con- laboratory conditions. trol the C. grandis occurring naturally in Replication Number of Number of Mean number of the fi eld at the Rota Resort Golf Course females eggs laid eggs laid per female (14.19°N, 145.23°E) in Rota. 1 4 425.0 106.3 The data from the host specifi city tests were analyzed using a Paired t-test (SPSS 2 6 1066.0 177.7 15.0, SPSS Inc., Chicago, IL) based on the 3 4 625.0 156.3 proportion of eggs laid. 4 4 578.0 144.5 5 3 401.0 133.7 30 Plant Protection Quarterly Vol.24(1) 2009 (Nylander) (Hymenoptera: Formicidae) 60 had been attacking the immature stages n = 12 of M. oedipus. 50 Discussion The fi eld release of M. oedipus on Guam 40 and Saipan was meant to complement the effects of two biocontrol agents already 30 released, viz. A. cocciniae and A. burkharto- n = 12 rum, in efforts to control and suppress the invasive vine C. grandis on Guam and the 20 Northern Mariana Islands. While A. coc- n = 36 ciniae causes some damage by leaf-mining, (±SEM) of days Mean number 10 n = 8 A. burkhartorum has not established and n = 8 M. oedipus is expected to exert the most damaging effect on C. grandis among the 0 three (Muniappan et al. in press). M. oedi- Egg Larvae Pupae Male Female pus larvae bore into C. grandis stems, and even reach to the roots, causing withering Figure 1. Duration of the different life stages of Melittia oedipus. Each test and eventual death. In Hawaii, a few male was replicated four times. The bars with the number indicate the number parasitoids of Eupelmus sp (Hymenoptera: of immature stages/adults used in the test; n indicates the number of Eupelmidae) were recovered in the fi eld collected eggs of M. oedipus (Chun 2001). replications used in each test. However it seems they have had no sig- nifi cant effect on the spread of M. oedipus. On some occasions, rats were found to nib- 60 A ble on C. grandis vines and remove larvae from the stems. a In the present study the M. oedipus fe- 50 male laid 106–178 (mean: 144) eggs. Our results are somewhat different from Chun 40 (2001) who found females to contain 60–140 (mean: 110) eggs. Chun (2001) dis- sected the ovaries of unmated laborato- 30 ry-reared female M. oedipus to determine the average number of eggs per female, 20 whereas in our experiments, female moths under Guam quarantine laboratory condi- 10

tions were released in a cage with C. gran- (±SEM) of eggs laid Mean number dis, allowed to mate and oviposit to deter- B b mine the average number of eggs laid per 0 0 female. These adults were also fed with honey/water solution that might have en- C. grandis Z. guamensis C. grandis Z. guamensis hanced the reproductive potential of M. oedipus. While Chun (2001) reported the Free-choice No-choice overall life cycle to range 2–4 months, our Figure 2. Host specifi city test of Melittia oedipus. Mean number of eggs fi ndings indicated the overall life cycle to be around 2.5 months. obtained from four replications, each replicate consists of four pairs of Host specifi city tests were conducted in males and females. Bars with different letters are signifi cantly different Hawaii on various plants within and out- (P >0.05, paired t-test). side the family of Cucurbitaceae adopting the centrifugal phylogenic method (Chun 2001). Studies by Chun (2001) revealed no feeding by M. oedipus on plants outside agent for C. grandis on Guam. No larval rationale behind confi nement and prox- the Cucurbitaceae family, viz. Apiaceae, development was observed on Z. guamen- imity of the non-host Z. guamensis plant Brassicaceae, Fabaceae, Solanaceae, Fla- sis in both no-choice and free-choice tests. holds true in most cases in the Marianas. courtiaceae and Violaceae, and larval de- Sands and van Driesche (2000) report that In four months, 128 adults and ap- velopment into adults occurred only on the positive responses to the non-target proximately 1115 larvae of M. oedipus cucumber, and were few in number. Be- host plant in no-choice tests may develop were released on Guam. Since its field cause Guam has an endemic cucurbit spe- only after an extended period (e.g. sev- release on Guam (both larvae and adults cies, Z. guamensis, it was determined that eral days) of deprivation due to the des- of M. oedipus) at Inarajan Bay (July 2007) host specifi city studies be conducted on peration ‘effect’. However, in the present and Marbo Caves (September 2007), M. Guam to verify whether M. oedipus would study this did not happen as the larvae oedipus has established in the fi eld. Female have a negative impact on the endemic that emerged on Z. guamensis died. These moths have been observed under Guam species. Both free-choice and no-choice authors also suggest that in the case of ar- laboratory conditions to delay oviposition host specifi city studies on Z. guamensis in thropod agents, major problems may arise in the presence of ants. No parasitism or Guam showed that M. oedipus preferred to if agents are held in confi nement with predation was observed on M. oedipus in oviposit, consume and develop only on C. other organisms that would rarely if ever, the fi elds of Guam, whereas the pavement grandis, thereby making it a safe biocontrol be found naturally in close proximity. The ant has been observed on C. grandis vines. Plant Protection Quarterly Vol.24(1) 2009 31 4.0 Sands, D.P.A. and van Driesche, R.G. a (2000). Evaluating the host range for 3.5 biological control of : ration- A ale, methodology and interpretation. In 3.0 ‘Host specifi city testing of exotic arthro- pod biological control agents. The bio- 2.5 logical basis for improvement in safety’, 2.0 eds R.G. van Driesche, T. Heard, A. Mc- Clay and R. Reardon. FHTET-99-1, pp. 1.5 69-83. (Forest Services, Morgantown, West Virginia). 1.0 Singh, A.K. (1990). Cytogenetics and evo- lution in Cucurbitaceae. In ‘Biology 0.5 B b and utilization of the Cucurbitaceae’,

Mean number (±SEM) of adults emerged Mean number 0 0 eds D.M. Bates, E.W. Robinson and C. 0 Jeffrey, pp. 10-28. (Cornell University Press, Ithaca and London) C. grandis Z. guamensis C. grandis Z. guamensis Stone, B.C. (1970). The flora of Guam. Free-choice No-choice Micronesica 6, 1-659. Telford, I.R.H. (1990). Cucurbitaceae. Figure 3. Host specifi city test of Melittia oedipus. In free-choice test, 15 eggs In ‘Manual of the fl owering plants of and in no-choice test, 10 eggs of M. oedipus were attached to each test plant. Hawai’i, vol. 1’, eds W.L. Wagner, D.R. Mean number was obtained from four replications. Bars with different Herbst and S.H. Sohmer, pp. 568-81. (University of Hawaii Press/Bishop letters are signifi cantly different (P >0.05, paired t-test). Museum Press, Honolulu). USDA (United States Department of Ag- riculture) (2003). Field release of Acy- Ant repellents were used in the M. oedipus Linney, G. (1986) Coccinia grandis (L.) Voigt: thopeus cocciniae (Coleoptera: Curculio- release areas in an effort to alleviate the in- a new cucurbitaceous weed in Hawai’i. nidae), a non-indigenous leaf-mining terference. The establishment and spread Hawaii Botanical Society Newsletter 25, weevil for control of ivy gourd, Coccinia of M. oedipus, its possible recruitment of 3-5. grandis (Cucurbitaceae), in Guam and local natural enemies, and its impact on Motooka, P., Ching, L. and Nagai, G. (2002). Saipan – environmental assessment the population of C. grandis on Guam are Herbicidal weed control methods for 2003, and Plant Health Inspec- being monitored. pasture and natural Areas of Hawaii. tion Service (APHIS), Riverdale, Mary- CTAHR publication WC-8,Cooperative land, 10 p. Acknowledgments Extension Service, College of Tropical USDA (United States Department of Ag- This research was supported by grant 07- Agriculture and Human Resources, riculture) (2004). Field release of Acy- DG-11052021-219 Invasive Plants from University of Hawai‘i. thopeus burkhartorum (Coleoptera: Cur- the USDA Forest Service and also Grant Muniappan, R., Reddy, G.V.P. and Raman, culionidae) a non-indigenous weevil #98-34135-6786 from the Tropical and Sub- A. (in press). Coccinia grandis (L.) Voigt for the control of ivy gourd, Coccinia tropical Agricultural Research (TSTAR), (Cucurbitaceae). In ‘Biological control grandis (Cucurbitaceae), in Guam and Special Grants, CSREES, USDA. We thank of tropical weeds using arthropods’, Northern Mariana Islands – envi- Mr Raymond Gumataotao and Ms Nakita eds R. Muniappan, G.V.P. Reddy and A. ronmental assessment 2004, Animal Braganza for their help in rearing of the Raman. (Cambridge University Press, and Plant Health Inspection Service . We also thank Dr Jack Tenorio Cambridge, UK). (APHIS), Riverdale, Maryland, 12 p. for helping us in releasing the biocontrol Raman, A., Cruz, Z.T., Muniappan, R. USDA (United States Department of Ag- agent in the Northern Marianas Islands. and Reddy, G.V.P. (2006). Biology, host riculture) (2008). Field release of Melit- specifi city, and fi eld release of gall-in- tia oedipus (Lepidoptera: Sesiidae), a References ducing Acythopeus burkhartorum (Cole- non-indigenous moth for control of ivy Analytical Software (2003). Statistix 8 optera: Curculionidae: Baridae), a bio- gourd, Coccinia grandis (Cucurbitaceae), (Analytical Software, Tallahassee, FL, logical control agent for the invasive in Guam and the Northern Mariana Is- USA) 396 p. weed, Coccinia grandis (Cucurbitaceae) lands – environmental assessment 2007, Bamba, J. (2008). Studies on the biology in Guam and Saipan. Tijdschrift voor En- Animal and Plant Health Inspection and host range of Acythopeus cocciniae, tomologie 150, 181-91. Service (APHIS), Riverdale, Maryland, a biological control agent for Coccinia Reddy, G.V.P., Cruz, Z.T., Braganza, N. and 13 p. grandis, M.S. thesis (Environmental sci- Muniappan, R. (2008). The response of Wapshere, A.J. (1974). A strategy for evalu- ence), University of Guam, Mangilao, Melittia oedipus (Lepidoptera: Sesiidae) ating the safety of organisms for bio- 55 p. to visual cues is increased by the pres- logical weed control. Annals of Applied Chun, M.E. (2001). Biology and host spe- ence of food source. Journal of Economic Biology 77, 201-11. cifi city of Melittia oedipus (Lepidoptera: Entomology (in review). Sesiidae), a biological control agent of Reddy, G.V.P. and Muniappan, R. (2005). Coccinia grandis (Cucurbitaceae). Pro- Biological control of the ivy gourd, ceedings of the Hawaiian Entomological Coccinia grandis (Cucurbitaceae) in the Society 35, 85-93. Marianas Islands. Paper presented at Eichlin, T.D. (1995). New data and a rede- the Annual Meeting of the Entomologi- scription for Melittia oedipus, an Afri- cal Society of America, Fort Lauderdale, can vine borer (Lepidoptera: Sesiidae). Florida. Tropical Lepidoptera 6, 47-51.