Invasive Fruit Flies

Study of Attraction of Nontarget Organisms to Fruit Fly Female Attractants and Male Lures in Hawaii.

Luc Leblanc, Daniel Rubinoff, and Mike San-Jose

(Cooperative Agreement No. 58-5320-4-501 between USDA-ARS and the University of Hawaii at Manoa CTAHR-PEPS)

The main purpose of this project was to assess environmental risks of using fruit fly (Tephritidae) attractants for their control or eradication in the Hawaiian environment. More specific focus was on the use of male lures (cue-lure, methyl eugenol) and female synthetic food attractants (BioLure, torula yeast and solulys).

Study sites

Traps baited with female attractants and male lures were setup in native, mixed native and non-native forests, farmlands, orchards and residential areas. Traps were maintained for 10-24 weeks and emptied weekly at 35 sites on Hawaii Island (2005) (Figure 1) and 46 sites on Maui (2006) (Figure 2). Trap catches were compared against catches from unbaited control traps. On Hawaii island, 9 sites formed a 20 km transect along the Stainback Highway (138-1,045 m above sea level), 15 sites were maintained in a 35 km transect along the Saddle Road (439-2,012 m), 6 sites were along the upper Hamakua Ditch Trail in wet native forest (North Kohala Forest Reserve) (906-1,019 m), and 5 sites were in the agricultural community of Waimea (744-872 m). On Maui, we maintained 14 sets of traps in 9 sites in the Kula agricultural community (517-1,138 m), primarily in persimmon and coffee orchards and their adjacent non-native forests, and 37 sites in mostly endemic forest on the northern slope of Haleakala Mountain (1,184-1,583 m) in the Makawao, Waikamoi and Koolau Forest reserves.

Fig 1: Trapping sites on Hawaii island
Fig 2: Trapping sites on Maui island

Attraction to male Lures

Male lures, combined with insecticides, are commonly used in traps to monitor male Bactrocera fruit fly populations and detect incursion of exotic species. They also constitute a powerful tool for fruit fly suppression and eradication through male annihilation. The large-scale use of male lures for control and eradication has raised concern of possible nontarget impacts on insects other than fruit flies, especially beneficial species and the numerous endemic Hawaiian insects.

Past attempts in Hawaii to characterize the range of nontargets attracted to male lures, by comparing captures in lure-baited traps with unbaited control traps, have resulted in an impressive list of 36 insect species, in 16 families of Diptera, Coleoptera, Hemiptera and Hymenoptera, assumed to be attracted to methyl eugenol. Among them were 11 species of endemic Hawaiian Drosophilidae (Figures 3 to 7), a very diverse group with 559 described species, of which 12 are recognized as endangered by the US Fish and Wildlife. However, at least 26 of these purportedly lure-attracted species are scavengers, and authors cautioned that some of them may have been actually secondarily attracted to decaying fruit flies that could not be excluded from entering male lure traps.

Fig 3 Fig 4 Fig 5 Fig 6 Fig 7
Figures 3 to 7: Various representatives of endemic Hawaiian Drosophilidae

Our study aimed to further investigate male lure nontarget effects by carefully discriminating between actual attraction to male lures and secondary attraction to decaying insects. Trapping was carried out across a broad range of environments, including intact native forest, mixed native/invasive forest, invasive forest, agricultural and residential areas.

Bucket traps made from drinking cups and baited with cue-lure or methyl eugenol (ME) (Scentry lure plugs) were maintained and emptied weekly at every trapping sites on Hawaii and Maui islands. To control for the possible attraction to dead insects, bucket traps artificially baited with decaying oriental fruit flies were maintained at all sites, except in the Maui endemic forest. Unbaited bucket traps were also maintained at all sites to control for the random entry of insects into traps. Vapor tape and a 20% solution of propylene glycol were included in all traps to kill and preserve trapped insects, respectively. In addition to the bucket traps, MultiLure traps charged with the 3-component BioLure food attractant and, in Maui forest, bucket traps baited with fermented mushroom, were maintained to ensure that the apparent lack of attraction to male lures is not due to the absence of potential nontarget species at trapping sites.

Traps with male lures and decaying flies captured 401 recognized arthropod species, in 17 orders and 93 families, dominated by Diptera (94.9% of all captures, 248 species), primarily in the families Drosophilidae, Phoridae and Milichiidae.

Cue-lure did not significantly attract any nontarget insects, and melon flies were usually not numerous enough in traps to cause secondary attract scavengers. These results are fully consistent with conclusions from previous studies.

Seven nontarget species in five insect orders were significantly attracted to ME-baited traps, regardless of the presence or absence of decaying fruit flies. Five of these are closely associated with flowers, feeding on pollen or nectar. Honeybees (Apis mellifera L.) (Apidae) and the flower fly Allograpta obliqua (Say) (Syrphidae) were attracted in rather small numbers (0.04-0.09 per trap per day). Honeybee attraction to ME in Hawaii was previously documented in literature, and orchid bees (Apidae: Euglossinae) are similarly drawn to ME in South America. Two endemic species of Crambid moths [Mestolobes minuscula (Butler) (Figure 8) and Orthomecyna exigua (Butler) (Figure 9)] were also attracted to ME in Kula (Maui) orchards. Although endemic, these two species are common on Maui in non-native habitats at lower altitude. The introduced sap beetle Carpophilus marginellus Motsch. (Nitidulidae), a common flower visitor that contributes to fruit tree pollination in Japan, is attracted to decaying fruit flies, but also to ME in traps. The attraction of flower insects is no surprise, since ME or some of its related compounds have been detected in the flower blossoms of a diversity of plant families.

Although our traps captured limited numbers of green lacewings (Chrysopidae), they were previously reported as attracted to ME in Hawaii, the Philippines, and Taiwan. Adult chrysopids, depending on species, feed either on live insects, or on flower pollen and nectar and honeydew from Hemiptera. At least of two of the three species attracted to ME are flower feeders, strongly suggesting an attraction to ME as emulation of floral compounds, further supported by the lacewing attraction to the natural flower fragrance compound eugenol in Malaysian rainforest.

The endemic plant bug Orthotylus coprosmae Polhemus (Hemiptera: Miridae) (Figure 10) is attracted to ME in Maui endemic forests, consistent with previous ME attraction records of three other endemic mirids on Kauai. At least one of these feeds on a host plant known to contain ME. A similar plant kairomone relationship may also explain published records of endemic anobiid beetles attraction to ME traps on Kauai.

Methyl eugenol attracts females of the endemic fungus gnat Bradysia setigera (Hardy)  (Sciaridae) (Figure 11). Conspecific males and other common sciarids are not attracted. In this case it likely acts as a pheromone analogue rather than a kairomone.

Fig 8 Fig 9 Fig 10 Fig 11
Figures 8 to 11: Endemic Hawaiian nontarget insects attracted to methyl eugenol

At least 56 species in 21 families of Diptera, Hymenoptera and Coleoptera were significantly drawn to decaying fruit flies rather than male lure. They were abundant in traps with decaying flies, and were collected in ME traps only when enough dead trapped flies had accumulated. We demonstrated that 8 of the 36 species previously reported as attracted to ME are actually drawn to dead flies, and that most other species belonged to families attracted to decaying flies, rather than ME, most commonly the Drosophilidae, Phoridae, Chloropidae, Lonchaeidae, Milichiidae, Neriidae, Otitidae, Psychodidae, Sphaeroceridae, Calliphoridae, Muscidae, and Sarcophagidae. Most species in these families are indeed scavengers. Aside from the ME-attracted C. marginellus, other endemic and introduced Nitidulidae were attracted only to decaying flies.

Drosophilidae were the most numerous and diverse nontargets attracted to decaying flies at all sites. Our results confirm or strongly support that most or all of the 11 endemic and 5 introduced drosophilids reported in literature as ME-attracted were actually attracted to dead flies. Nearly half (143 of 306) of the drosophilids expected to occur at the trapping sites were collected using BioLure and mushroom bait traps, but not in male lure traps. Invaluable data was generated on drosophilid distribution in endemic and disturbed ecosystems in Hawaii. It is an unforeseen extra benefit from this research project that is the focus of a publication soon to be submitted to an invertebrate conservation-oriented journal. Fragrant leaves of Cheirodendron trigynum (Gaud.) Heller (Araliaceae), the most common larval host species for endemic drosophilids, were demonstrated through Solid Phase Microextraction (SPME) analysis not to contain ME.

A few species of predators and parasitoids were attracted to decaying flies. Almost all are associated with decaying matter (Staphylinidae) or parasitoids of houseflies (Encyrtidae, Braconidae, Pteromalidae).

Precautionary suggestions are provided to minimize the undesirable nontarget impact of the use of ME for control or eradication on flower-associated insects, endemic plant bugs and fungus gnats and scavenger insects. Flower insects were attracted in small numbers (0.03-0.15 per trap per day) in our study, consistent with previously published data, suggesting that attraction is likely to be short-ranged, and can be further minimized if a ME trap or dispenser is hung to a tree past its flowering stage. This was confirmed by the much lower honeybee and moth captures in non-flowering persimmon trees that in adjacent flowering coffee trees in Kula. A comparison of captures of endemic saprophagous insects in decaying fly and BioLure traps in orchards and backyards, native forest and ecotone forest adjacent (< 100 m) to native forest shows that very small numbers of a few endemic species are captured in non-native sites.
Fig 12: Endemic Hawaiian viviparous blowflies (Dyscritomyia)

This is not the case in endemic forest and its adjacent ecotones, where a broad diversity of endemic drosophilids and calliphorids (the larviparous Dyscritomyia) (Figure 12) are trapped. Based on studies of range of dispersal of 14 common North American Drosophilid species by Donald McInnis (USDA-ARS), who estimated their maximal dispersal distance to be 300 meters at most, we concluded that using traps at least 300 meters from native forest will minimize possible nontarget effects, if dead target flies accumulate inside traps.

Attraction to BioLure

The dry food lure BioLure in the MultiLure trap (Figure 13) was developed as an alternative to the traditionally used liquid protein lures in glass McPhail traps (Figure 14) by identifying and using the individual volatiles from the bacterial breakdown of protein hydrolysate that were most attractive to fruit flies. The 3-component BioLure, used in our study, is composed of ammonium acetate, trimethylamine hydrochloride and putrescine. Although originally developed as a monitoring tool, BioLure has become a common tool to control the Mediterranean fruit fly [Ceratitis capitata (Wiedemann)], through mass-trapping, in fruit tree orchards in Spain and Israel. In Hawaii it is used, in combination with protein bait sprays, to suppress C. capitata in Maui orchards.

Fig 13 Fig 14

Nontarget attraction to BioLure, mainly saprophagous Diptera, has been reported in literature, but not studied systematically in a diversity of habitats. We have therefore investigated nontarget attraction to BioLure across a range of endemic and agricultural habitats in Hawaii, compared nontarget attraction of the individual BioLure components, and determined whether the omission of the putrescine ingredient results in decreased nontarget catches without reducing captures of target C. capitata.

MultiLure traps baited with 3-component BioLure, as well as unbaited controls, were maintained continuously, or intermittently in endemic forest, at every trapping site on Hawaii and Maui islands. A solution of 20% propylene glycol was added to each trap to preserve and facilitate identification of captured insects. Trap contents was emptied weekly. These traps served the dual purpose of characterizing nontarget attraction to BioLure and confirming the presence of nontargets that may potentially be attracted to the male lure traps also present at each site. Additionally, three sites were selected, one in endemic forest and two in orchards, to study fruit fly and nontarget attraction to the separated individual components of BioLure. A third trap, with only two components, was also maintained at each site in Kula (Maui), to determine if the omission of the putrescine ingredient results in a decrease in nontarget captures without compromising captures of C. capitata.

Captures in BioLure traps were numerically dominated by Diptera (94.3%). The endemic forest site samples were dominated by endemic and introduced Drosophilidae and endemic Calliphoridae, while most of the captured nontargets in the nonnative sites were of introduced species. The majority of nontarget species belonged to families whose larvae are scavengers on decaying plant or animal matter (Drosophilidae, Chloropidae, Lonchaeidae, Neriidae, Otitidae, Phoridae, Anthomyiidae, Calliphoridae, Muscidae, Sarcophagidae and Nitidulidae). These same families and species of were also strongly attracted to the bucket traps baited with decaying fruit flies.

BioLure attracted few beneficial predators or parasitoids, except for a few parasitoids of house flies and moderate numbers of tachinids flies. Pollinators were not attracted to BioLure. Although it attracted very few green lacewings (Chrysopidae), there is literature evidence that protein hydrolysate and BioLure can attract the pollen and nectar-feeding species.

A comparison of attraction to the components in separate traps and the three components together inside a trap in a randomized block design has shown that ammonium acetate or, to a lesser extent, putrescine, are the main components attractive to nontargets, depending on the species. It was also demonstrated, as is the case for target fruit flies, that the three components act in synergy, attracting larger numbers of nontargets together in a trap than in three traps baited with the separate components.

The elimination of the putrescine ingredient from BioLure traps resulted in significant capture reduction for five nontarget species and an overall 20% reduction in the number of nontargets attracted, with no reduction in target C. capitata captures. I can therefore be omitted from BioLure when used for Mediterranean fruit fly monitoring or control in Hawaii.


Leblanc, L., Rubinoff, D., and R.I. Vargas. 2009. Attraction of nontarget species to fruit fly (Diptera: Tephritidae) male lures and decaying fruit flies in Hawaii. Environ. Entomol. 38: 1446-1461.

Leblanc, L., O’Grady, P.M., Rubinoff, D., and S.L. Montgomery. 2009. New immigrant Drosophilidae in Hawaii, and a checklist of the established immigrant species. Proceedings of the Hawaiian Entomological Society. 41: 121-127.

Leblanc, L., Vargas, R.I., and D. Rubinoff. 2010. Attraction of Ceratitis capitata (Diptera: Tephritidae) and endemic and introduced nontarget insects to BioLure bait and its individual components in Hawaii. Environ. Entomol. IN PRESS.

Leblanc, L., Vargas, R.I., and D. Rubinoff. 2010. A comparison of nontarget attraction to BioLure and liquid protein food lures in Hawaii. IN PREPARATION.

Vargas, R.I., Shelly, T.E., Leblanc, L. and J.C. Piñero. 2010. Recent advances in methyl eugenol and cue-lure technologies for fruit fly detection, monitoring, and control. Vitamins and Hormones. Section: Pheromones, vol. 83. Academic Press. SUBMITTED.

There are four species of pest fruit flies (Tephritidae) in Hawai'i. All are invasive, and most of them are from tropical Asia.
For more information on fruit fly diversity and management, consult the Pacific Fruit Fly Website and the Hawaii Area-Wide Fruit Fly Pest management Program site.
Melon Fly (Bactrocera cucurbitae) was the first pest fruit fly detected in Hawai'i in 1895. Its larvae breed on cultivated and wild Cucurbitaceae, as well as papaya.
Mediterranean Fruit Fly (Ceratitis capitata) has been the dominant polyphagous fruit pest since from its discovery in 1907, until the introduction of Oriental fruit fly.

Oriental Fruit Fly (Bactrocera dorsalis) was detected in Hawai'i in 1945, and has become the main fruit pest of edible and wild fruits in Hawai'i. Following its introduction, it has displaced the Mediterranean fly to become the dominant species, limiting the Mediterranean fly to attack coffee and fruit at higher elevations.
Malaysian fruit fly (Bactrocera latifrons) is the most recent migrant, known in Hawai'i since 1983. Its host range is restricted to cultivated and wild solanaceous plants. It is not a serious crop pest.

Photos are courtesy of Jari Sugano (University of Hawai'i at Manoa)