Novel lures and COI sequences reveal cryptic new species of Bactrocera fruit flies in the Solomon Islands (Diptera, Tephritidae, Dacini)

Abstract Results from a snap-shot survey of Dacine fruit flies carried out on three of the Solomon Islands in April 2018 are reported. Using traps baited with the male lures cue-lure, methyl eugenol, and zingerone, 30 of the 48 species previously known to occur in the Solomon Islands were collected. Six species are newly described here: Bactroceraallodistinctasp. nov., B.geminosimulatasp. nov., B.kolombangaraesp. nov., B.quasienochrasp. nov., B.tsatsiaisp. nov., and B.vargasisp. nov., all authored by Leblanc & Doorenweerd. An illustrated key to the 54 species now known to be present in the country is provided.


Introduction
Dacine fruit flies (Diptera: Tephritidae: Dacini), a species-rich Old World tropical group, is composed of 947 currently known species, including 83 crop pests (White and Elson-Harris 1992;Vargas et al. 2015;Doorenweerd et al. 2018). Diversity is particularly high in Australasia, with 332 species described and an imminent publication of 65 new species from Papua New Guinea (R.A.I. Drew, pers. comm.). Many more new species are being discovered, especially cryptic species, with ever improving molecular diagnostic tools and the emergence of new generation male lures Manrakhan et al. 2017;Royer et al. 2018Royer et al. , 2019Doorenweerd et al. 2020).
The earliest Dacine fruit fly record in the Solomon Islands was the description of Bactrocera longicornis Macquart, in 1835. By 1939, eleven species were known (Malloch 1939), growing to 26 five decades later (Drew 1989). Extensive survey efforts through trapping and host fruit surveys during the Regional Fruit Fly Projects in the Pacific (Allwood and Drew 1997;Allwood 2000;Lidner and McLeod 2008) nearly doubled the number of species to 48 (Drew and Romig 2001). Two decades later, we carried out a snap-shot survey on three islands (Guadalcanal, Kolombangara, Gizo), with the inclusion of zingerone lure, to collect fresh material and develop molecular diagnostic tools to help further characterize the species found in the Solomon Islands. In just a couple of weeks, we discovered six new species, including cryptic species that would not have been detectable without molecular characterization. We herein describe these new species and provide a key to the 54 species now present in the Solomon Islands.

Collecting and curation
We maintained 79 sets of three traps separately baited with male lures (cue-lure, methyl eugenol and zingerone) in the Solomon Islands in April 2018. We used commercially available cue-lure and methyl eugenol plugs (Scentry Biologicals, Billings, Montana). Zingerone (= vanillylacetone) lure was prepared by dipping dental cotton wicks in zingerone powder (Sigma-Aldrich) melted over a hot plate and allowed to solidify in the wicks. Small vertical bucket traps (Leblanc et al. 2015: fig. 1) were made of 5-oz urine sample cups (Stockwell Scientific, Scottsdale, Arizona) with two 20 mm wide lateral circular openings on opposite sides, 12 mm below the top, with a hole drilled in the lid center, through which a 30-cm-long, 15-gauge, aluminum tie wire was inserted, and bent into a hook below the lid. The male lure unit and a 10 × 10 mm piece of dichlorvos (DVVP) strip (Vaportape II; Hercon Environmental, Emingsville, PA) were attached to the hook below the lid. A 10-cm-wide black square plastic food plate (Waddington North America) was placed on top of the trap to prevent flooding by frequent rain. A solution of 25% propylene glycol (Better World Manufacturing, Fresno, CA) was used in the trap to preserve captured flies, until they were transferred to 95% ethanol during trap servicing. The 79 sets of traps in agricultural areas and endemic forest on the islands of Guadalcanal and Kolombangara, and agricultural areas on Gizo Island (Fig. 1) were maintained for 12, four, and six days, respectively. Forest trapping sites were ca. 50 meters apart along transects that followed trails. Sampled flies were stored in 95% ethanol in a -20 °C freezer to preserve DNA. All flies were identified to species using available keys (Drew 1989;Drew and Romig 2001). We pulled one or two legs from specimens that were selected for DNA extraction (for further details on DNA extraction methods see Doorenweerd et al. 2020). All holotypes and all, or a subset of, the paratypes were double-mounted to be stored as dry specimens in collections for permanent future reference. Before drying flies for double-mounting (White and Elson-Harris 1992), we pinned them through the scutum with a minuten pin and soaked them in diethyl-ether for 3 -12 hours to fix and preserve their natural coloration. We photographed specimens using a Nikon D7100 camera attached to an Olympus SZX10 microscope and used Helicon Focus pro v6.7.1 to merge pictures taken at a range of focal planes. To measure specimens (all available or up to 10 specimens measured per species), we used an ocular grid mounted on an Olympus SZ30 dissecting microscope.

Morphological terms and taxonomic assignment
Morphological terminology used in the descriptions follows  and assignment of species to genera follows Doorenweerd et al. (2018). We treat Zeugodacus as a distinct genus from Bactrocera and Dacus (Krosch et al. 2012;Virgilio et al. 2015;Dupuis et al. 2017;San Jose et al. 2018). Subgenus assignment for each species follows reclassifications recently published by Hancock and Drew (Hancock 2015;Drew and Hancock 2016;Hancock andDrew 2015, 2018a, b). The host plant records included in the key follow the compilation published by Leblanc et al. (2012). For accurate taxonomic application of host plant records from the literature we used the World Flora Online (WFO 2021).

COI sequence analysis
Representatives of all species newly described here were also included in the cytochrome c oxidase I (COI) study of Doorenweerd et al. (2020), under tentative species names. For that study, 1493 base pairs of the COI gene were sequenced and comparatively analyzed in a dataset that included 163 species of Bactrocera. We include here the maximum likelihood gene tree from that study and the summary Bactrocera species statistics as supplementary material (Suppl. material 1: Fig. S1, Suppl. material 2: Table S1). For the methods for DNA extraction, sequencing and analyses we refer to Doorenweerd et al. (2020). Collecting information as well as COI sequences are available on BOLD (Ratnasingham and Hebert 2007) dataset (DOI: http://dx.doi.org/10.5883/DS-DACCOI), as well as NCBI GenBank (accessions MZ196488-MZ196507). Each specimen for which DNA was extracted was assigned a unique code in the format "UHIM.ms00000", physically labelled as such, and this number forms the 'Sample ID' in BOLD.

Estimating biodiversity
We used EstimateS software (Colwell 2019) to generate species accumulation curves and estimate species diversity, using the incidence-based Chao 2 algorithm. We generated accumulation curves, with 100 randomizations without replacement for confidence intervals for all sites collectively, separately for each island, and comparing agricultural and forest sites in Guadalcanal and Kolombangara.  (Drew) (Fig. 4) in the presence of orange-brown lateral and posterior markings on the predominantly black scutum, abdominal tergites III-V with a narrower medial black stripe, the lateral black markings on tergite IV narrowed posteriorly, and the rather diffuse fuscous crossband on the wing. It differs from B. distincta (Malloch) in that the costal band is diffuse orange-brown and the crossband is sinuous, with a bend along vein M (Fig. 3E), whereas the entire costal band, including in the basicostal and costal cells, is dark fuscous and the crossband is broad and straight in B. distincta (Fig. 5E).
Legs (Fig. 3F). All legs entirely fulvous with apical 2⁄ 5 of hind tibia fuscous. Fore femur with a row of long pale dorsal setae. Mid-tibia with apical black spur.
Wing (Fig. 3E). Length 5.6 ± 0.2 (5.3-5.9) mm; basal costal and costal cells fuscous with microtrichia in posterodistal corner of costal cell; broad fuscous costal band confluent with R 4+5 , remaining broad at apex and ending at apex of medial vein; a diffuse orange-brown crossband along crossvein r-m, continuing along M and dmcu to reach posterior wing margin, and a broad fuscous anal streak over cell bcu and basal margin of cu 1 ; remainder of wing light fuscous; dense aggregation of microtrichia around A 1 + CuA 2 ; supernumerary lobe weakly developed.
Abdomen (Fig. 3C, D). Oval with tergites not fused; pecten present on tergite III; posterior lobe of surstylus short; abdominal sternite V with a deep concavity on posterior margin. Base of syntergite I+II wider than long. Syntergite I+II orange-brown with base black and a narrow sub-basal transverse medial black band. Tergites III-V orangebrown with moderately broad medial black stripe reaching apex, and large lateral black markings on tergite III and anterolateral corners of tergites IV and V. Ceromata on tergite V indistinct from abdomen orange-brown color. Sternite I dark fuscous, sternite II fulvous, and sternites III-V fulvous tending fuscous medially.
Female. Unknown Male attractant. Cue-lure. Etymology. The specific name is a noun in apposition, derived from the Greek allos (another) and the species resembles B. distincta (Malloch). Previously, B. pseudodistincta (Drew) had been described as a species with similar appearance to B. distincta. All three are present in Oceania.
Notes. Bactrocera allodistincta was included as B. spnSol01 in Doorenweerd et al. (2020).  (Malloch), only distinguished by a subtle difference in wing infuscation in the presence of a light fuscous tinge as a broad, somewhat triangular area covering much of the middle of the wing, including the areas bordering r-m and dm-cu ( Fig. 9E-G); the latter is absent in B. simulata ( Fig. 9A-D). The new species can be distinguished from B. bryoniae (Tryon) by the lighter fuscous tinge of the costal band, a narrower anal streak and the largely to entirely black abdomen, whereas the abdomen in B. bryoniae is orange-brown with a narrow black 'T'-shaped pattern (Fig.  8). Bactrocera bryoniae is widespread in Australia and New Guinea but is absent from the Solomon Islands.

Bactrocera
Molecular diagnosis. The COI sequences of B. geminosimulata [N = 4] are similar to those of B. bryoniae [N = 5], but with a minimum of 1.47% pairwise distance. The reference COI dataset only includes B. bryoniae from Australia. The COI sequences suggest no close relationship with B. simulata, and can be used to reliably distinguish B. geminosimulata from B. simulata.
Legs (Fig. 6E). Coxae and trochanters black. Remainder of legs fulvous with hind tibia tending fuscous to dark fuscous. Fore femur with a row of long dark dorsal setae. Mid-tibia with an apical black spur.
Wing (Fig. 9E-G). Length 6.4 ± 0.4 (5.9-6.9) mm; basal costal and costal cells fulvous with microtrichia in posterodistal corner of costal cell; broad dark fuscous costal band confluent with R 4+5 , ending between R 4+5 and medial vein; light fuscous tinge as a broad, somewhat triangular area covering much of the middle of the wing, including the areas bordering r-m and dm-cu (absent in B. simulata); broad dark fuscous anal streak; dense aggregation of microtrichia around A 1 + CuA 2 ; supernumerary lobe moderately developed.
Abdomen (Fig. 6C, D). Oval with tergites not fused; pecten present on tergite III; posterior lobe of surstylus short; abdominal sternite V with a deep concavity on posterior margin. Base of syntergite I+II wider than long. Syntergite I+II black except for yellow along posterior half of and narrowly orange-brown along anterior margin of tergite II. Tergites III-V entirely black or with two broad longitudinal orange-brown areas running from center of tergite IV to posterior margin of tergite V, each side of a broad medial longitudinal dull black stripe. Ceromata on tergite V black. Abdominal sternites black.
Female. Unknown Male attractant. Cue-lure. Etymology. The specific name is a noun in apposition, derived from the Latin noun geminus (twins) and the epithet of the sympatric and morphologically nearly identical B. simulata (Malloch).
Notes. Bactrocera geminosimulata was included as B. spSol12 in Doorenweerd et al. (2020).  Differential diagnosis. Bactrocera kolombangarae appears similar to B. morula ( Fig. 11), but has two pairs of setae on the scutellum, a narrow anepisternal stripe, and the costal band very narrow and faint beyond the apex of R 2+3 (Fig. 10). It is also similar to B. (Parazeugodacus) abbreviata (Hardy), a species from Southeast Asia. Unlike B. kolombangarae, B. abbreviata has yellow femora, very short lateral postsutural vitta, and orange-brown medially on abdomen tergites III-V.
Legs (Fig. 10F). Legs black with yellow at basal 2/5 of fore and hind femora and basal 1/6 of mid femur, and yellow fore basitarsus and mid and hind tarsi. Fore femur with a row of long pale dorsal setae. Mid-tibia with an apical black spur.
Abdomen (Fig. 10C, D). Oval with tergites not fused; pecten present on tergite III; posterior lobe of surstylus short; abdominal sternite V with a shallow concavity on posterior margin. Base of syntergite I+II wider than long. Tergites entirely black except for elongate creamy yellow short sublateral bands along posterior margin of tergite II. Ceromata on tergite V black. Abdominal sternites dark except for yellow sternite II.
Female. Unknown Male attractant. Zingerone. Etymology. This species epithet is a noun in genitive case, derived from the locality where the majority of the specimens were collected; Kolombangara Island.
Notes. This species belongs to the subgenus Parazeugodacus as defined by Hancock and Drew (2015), based on morphological characters (shallow posterior concavity on male sternite V, posterior lobe of surstylus short, postpronotal seta absent, postsutural supra-alar, prescutellar acrostichal and two pairs of scutellar setae present, costal band very narrow and nearly indistinct). Its COI sequences also suggest closest affinity with other members of Parazeugodacus (Suppl. material 1: Fig. S1). Bactrocera kolombangarae was included as B. spnSol06 in Doorenweerd et al. (2020). Differential diagnosis. Bactrocera quasienochra (Fig. 12) is similar to B. enochra (Drew) (Fig. 13). It differs by the absence of broad black lateral markings on abdomen tergites III-V, and the narrower lateral postsutural vitta, ending before intra-alar seta.

Bactrocera (Bactrocera) quasienochra
Molecular diagnosis. We sequenced the holotype for COI, and its sequence is closest to an undescribed species from Malaysia (B. spMalaysia11 in Doorenweerd et al. (2020)) at 11.19% pairwise distance. The B. quasienochra sequence has an even greater distance to those of B. enochra [N = 6]. (Fig. 12A). Height 1.83 mm. Frons, of even width, 0.93 mm long and 1.56 times as long as broad; dark fulvous and narrowly fulvous anterolaterally; anteromedial hump covered by short red-brown microtrichia; three pairs of dark fuscous frontal setae present; lunule fulvous. Ocellar triangle black. Vertex fulvous with two pairs of dark fuscous vertical setae. Face fulvous with a pair of large oval black spots in antennal furrows; length 0.53 mm. Gena fulvous, with large dark fuscous subocular spot and a red-brown seta. Occiput fulvous and dark fulvous behind vertex; row of postocular setae weakly developed, with ca. four nearly indistinct setae. Antenna with scape and pedicel dark fulvous and flagellum fulvous with lateral surface and inner apical half dark fuscous; a strong fulvous dorsal seta on pedicel; arista fulvous basally and black distally; length of segments: 0.27 mm; 0.30 mm; 0.87 mm.
Legs (Fig. 12F). Legs entirely fulvous with hind tibia tending fuscous on dorsal surface. Fore femur with a row of long fulvous dorsal setae. Mid-tibia with an apical black spur.
Wing (Fig. 12E). Length 6.7 mm; basal costal and costal cells fuscous with microtrichia in posterodistal corner of costal cell; narrow fuscous costal band confluent with R 2+3 , not expanded at apex, and ending mid distance between apex of R 4+5 and medial vein, and broad fuscous anal streak; remainder of wing hyaline; dense aggregation of microtrichia around A 1 + CuA 2 ; supernumerary lobe weakly developed.
Abdomen (Fig. 12C, D). Elongate-oval with tergites not fused; pecten present on tergite III; posterior lobe of surstylus short; abdominal sternite V with a deep concavity on posterior margin. Base of syntergite I+II wider than long. All tergites orange-brown with a medial longitudinal black stripe gradually broadened from base of tergite III and extended apically along the entire lateral margins of tergite V except their bases. Ceromata on tergite V indistinct from abdomen orange-brown color. Abdominal sternites fulvous.
Female. Unknown Male attractant. Cue-lure. Etymology. The species name is a noun in apposition, derived from the Latin adverb quasi (just as if ) used in conjunction with the epithet of the species it closely resembles; B. enochra.
Legs (Fig. 15). All legs entirely fulvous with hind femur and fore tarsomeres II-IV fuscous. Fore femur with a row of long pale dorsal setae. Mid-tibia with an apical black spur.
Wing (Fig. 14I). Length 7.1 ± 0.3 (6.6-7.5) mm; basal costal and costal cells fuscous with microtrichia in posterodistal corner of costal cell; light fuscous costal band confluent with R 2+3 , not expanded at apex and ending mid distance between apex of R 4+5 and medial vein, a diffuse broad fuscous cross band along r-m crossvein, continuing in straight line through discal medial (dm) cell and reaching wing margin at level of CuA 1 , and a broad fuscous anal streak; remainder of wing hyaline; dense aggregation of microtrichia around A 1 + CuA 2 ; supernumerary lobe moderately developed. Abdomen (Fig. 14C-H). Oval with tergites not fused; pecten present on tergite III; posterior lobe of surstylus short; abdominal sternite V with a deep concavity on posterior margin. Base of syntergite I+II wider than long. Syntergite I+II with tergite I black and tergite II orange-brown with or without a small basal black triangular and two small sublateral black markings. Tergites III-V orange-brown with broad medial longitudinal black stripe reaching apex of tergite V and extended apically along entire lateral margins of tergite V, and two broad sublateral stripes covering tergite III (may be interrupted on that tergite) and continuing on tergite IV and along lateral margins on tergite V. Dark marking variable and may cover almost all of tergites III-V ( Fig. 14C-G). Ceromata on tergite V dark fuscous. Abdominal sternites fulvous.
Female. Unknown Male attractant. Zingerone. Etymology. The epithet tsatsiai is a noun in genitive case, referring to the personal name Francis Tsatsia, a long-time colleague, friend, co-author of the present publication, and currently the director of Biosecurity Solomon Islands.
Molecular diagnosis. We sequenced three specimens which have COI sequences closest to B. quasiinfulata Drew & Romig at 7.24% minimum pairwise distance. The maximum intraspecific distance is 1.2%. Sequences of the morphologically similar B. frauenfeldi and B. trilineola were also included in the reference dataset but are highly dissimilar to B. vargasi with >8% pairwise distance. Bactrocera parafrauenfeldi was not included in the reference set but is presumed to be closely related to B. trilineola (Drew 1989).
Legs (Fig. 16F). Legs black with yellow fore femur, basal 2⁄ 5 of mid and hind femur, and mid and hind tarsi. Fore femur with a row of long pale dorsal setae. Mid-tibia with an apical black spur.
Wing (Fig. 16E). Length 6.2 ± 0.6 (5.3-6.9) mm; basal costal and costal cells dark fuscous with microtrichia covering both cells; faint narrow fuscous costal band confluent with R 2+3 , remaining narrow to end shortly past the apex of R 2+3 ; dark fuscous straight band across r-m and dm-cu veins and reaching wing margin; broad dark fuscous anal streak; remainder of wing hyaline; dense aggregation of microtrichia around A 1 + CuA 2 ; supernumerary lobe weakly developed. Abdomen (Fig. 16C). Oval with tergites not fused; pecten present on tergite III; posterior lobe of surstylus short; abdominal sternum V with a deep concavity on posterior margin. Base of syntergite I+II wider than long. Tergites entirely black with yellow lateral bands along posterior margin of tergite II. Ceromata on tergite V black. Abdominal sternites black.
Female. Unknown. Etymology. We proudly name this species to honor the famous fruit fly ecologist Roger I. Vargas (1947Vargas ( -2018 (Stark et al. 2018). The species name vargasi is a noun in genitive case. Roger and LL collaborated extensively on projects during years spent in the South Pacific Islands. Roger brought LL to Hawaii in 2003 to continue working on fruit flies, and he secured funding and provided guidance that allowed LL to obtain a PhD title in 2010.

Key to Dacine fruit fly species of Solomon Islands
This is a modified version of the key published by Drew and Romig (2001). We include for each species subgenus assignment and information on male lure attraction and host fruit (after Leblanc et al. 2012), whenever known.

Discussion
The snap-shot survey yielded 16,843 Dacine flies, belonging to 30 known and six new species, described herein, increasing the number of species known from the Solomon Islands from 48 to 54 (Table 1). Twenty-eight species were represented by at least ten specimens and the five most collected species were Bactrocera frauenfeldi (Schiner) (43.0% of all specimens), B. froggatti (Bezzi) (13.4%), B. umbrosa (Fabricius) (9.6%), B. morula Drew (7.1%), and B. pagdeni (Malloch) (7.0%). Our sampling effort was very fruitful, yielding 29 of the 37 species previously collected by trapping over eight years, plus we found six new species, and 18 new island records (Table 1). We collected 31 of 48, 22 of 31, and 8 of 14 species found on Guadalcanal, Kolombangara, and Gizo, respectively. The species accumulation curves (Fig. 2) demonstrate the highest species diversity to be in the forests of Guadalcanal, with twice as many species as in agricultural sites (Fig. 2C). Despite the deployment of 36 sets of traps in the rich protected forests of Kolombangara (688 km 2 ), the number of collected and projected species was still half as many as on Guadalcanal (5,302 km 2 ) (Fig. 2B, C), consistent with previously published accounts (Drew and Romig 2001;Hollingsworth et al. 2003). The difference is related to island size, with number of fruit fly species clearly correlated (r 2 = 60.9%) to island size in the Solomon Islands (Suppl. material 3: Fig. S2).
In addition to collecting three new species, the use of zingerone lure revealed that Bactrocera pagdeni, formerly known only by its female holotype (Drew 1974), one specimen at the Bishop Museum collections (BPBM), and a few males recently captured in zingerone traps (Hancock and Drew 2018b), is actually a common and widespread species, with 1,174 specimens collected during our survey (Table 1). Likewise, the recent discovery of the attraction of Bactrocera quadrisetosa (Bezzi) to dihydroeugenol and isoeugenol lures in Vanuatu (Leblanc, unpublished) will likely reveal that this species is also common and widespread in the Solomon Islands. Clearly, many new species are awaiting discovery with the increasing availability of new generation lures (Manrakhan et al. 2017;Royer et al. 2018Royer et al. , 2019. Several rare species that require further attention in future surveys include: B. aithogaster Drew (known by only two specimens), B. bancroftii (Tryon) (one specimen of this Australian species from Guadalcanal), B. furvescens Drew, a Papua New Guinea species of which a single specimen was collected in 1971 in Honiara, and B. unipunctata (Malloch) known from a single specimen collected on Florida Island (Malloch 1939). The species from Mount Austen (Guadalcanal), identified as B. musae (Tryon) (Drew and Romig 2001), is likely a nonpest species member of the B. musae complex (Drew et al. 2011). Bactrocera musae is a major pest of banana, and no fruit fly infestations have been observed on banana in the Solomon Islands, even in recent years (FT, pers. obs.).
The COI sequences we obtained for the new species typically have large minimum pairwise distances to their nearest congeners, up to 12%, whereas the average minimum distance between species for Bactrocera is 6.09% (Suppl. material 2: Table  S1; Doorenweerd et al. 2020). This is likely due to a lack of species from Papua New Guinea represented in the reference dataset, where there is a large, mostly unstudied, diversity of Bactrocera (White and Evenhuis 1999;Drew pers. comm.). As a consequence, the currently available reference data suggests that COI reliably distinguishes all newly described species, but further sampling of species in New Guinea may reduce the pairwise distance resolution (Doorenweerd et al. 2020). There is one potentially new species (B. spnSol08; molecular voucher UHIM.ms08767), for which we have one specimen, that we leave undescribed. Although its COI sequence is highly divergent, closest to B. hantanae Tsuruta & White at 9.89%, there is only a single specimen and its morphology has no apparent differences with that of B. dorsalis (Hendel). Future sampling will hopefully bring in a larger series of this potentially new species to enable further examination of the morphological characters.    In addition to the data from this survey, we summarized trapping data in the Solomon Islands generated during the Regional Fruit Fly Projects in the Pacific, as a further indication of the relative abundance and to update the distribution of each species (Table 1). Over 1.8 million flies were collected from 180 sites maintained throughout the archipelago between 1994 and 2001 Drew and Romig 2001;Hollingsworth et al. 2003;Leblanc et al. 2012). A few specimens of then undescribed B. geminosimulata and B. quasienochra may have been included among these records.        Table 1. Checklist of Dacine fruit flies of Solomon Islands, including number of specimens collected during the Regional Fruit Fly Projects (1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001) and the 2018 survey. References to earliest record for each group on islands are: A: Macquart 1835, B: Froggatt 1910, C: Bezzi 1919, D: Malloch 1939, E: Hardy 1954, F: Drew 1972, G: Drew 1974, H: Eta 1985, I: Drew 1989, J: Waterhouse 1993K: Hollingsworth et. al. 1997, L: Drew andRomig 2001, new: previously unpublished (Drew, 1972) Cue-lure 1,226 208 I L L L L L L L L L L B. enochra (Drew, 1972) Cue  (Bezzi, 1928) Methyl eugenol 33,514 36 983 1236 I L L G L L L C D D L L L B. furvescens Drew, 1989 Cue-lure I