Comparison of Trap Designs for Detection of Euwallacea Nr

applyparastyle "fig//caption/p[1]" parastyle "FigCapt" Copyedited by: OUP

Journal of Economic Entomology, XX(XX), 2019, 1–8 doi: 10.1093/jee/toz311 Sampling and Biostatistics Research Downloaded from https://academic.oup.com/jee/advance-article-abstract/doi/10.1093/jee/toz311/5632132 by U S Dept of Agriculture user on 20 November 2019 Comparison of Trap Designs for Detection of Euwallacea nr. fornicatus and Other Scolytinae (Coleoptera: Curculionidae) That Vector Fungal Pathogens of Avocado Trees in Florida

Paul E. Kendra,1,3 Wayne S. Montgomery,1 Teresa I. Narvaez,1 and Daniel Carrillo2

1USDA-ARS, Subtropical Horticulture Research Station, Miami, FL 33158, 2Tropical Research and Education Center, University of Florida, Homestead, FL 33031, and 3Corresponding author, e-mail: [email protected]

Subject Editor: Jana Lee

Received 11 September 2019; Editorial decision 23 October 2019

Abstract Laurel wilt and Fusarium dieback are vascular diseases caused by fungal symbionts of invasive ambrosia beetles (Coleoptera: Curculionidae: Scolytinae). Both diseases threaten avocado trees in Florida. Redbay ambrosia beetle, Xyleborus glabratus, is the primary vector of the laurel wilt pathogen, Raffaelea lauricola, but in recent years this symbiont has been transferred laterally to at least nine other species of ambrosia beetle, which now comprise a community of secondary vectors. Dieback disease, caused by Fusarium spp. fungi, is spread by shot hole borers in the Euwallacea fornicatus species complex. In this study, we conducted field tests in Florida avocado groves to compare efficacy of four trap designs for detection of Scolytinae. Treatments included an 8-funnel Lindgren trap, black 3-vane flight interception trap, green 3-vane interception trap, white sticky panel trap, and an unbaited sticky panel (control). In two tests targeting E. nr. fornicatus and X. glabratus, traps were baited with a two-component lure (α-copaene and quercivorol). In a test targeting other species, traps were baited with a low-release ethanol lure. For E. nr. fornicatus, sticky panels and black interception traps captured significantly more beetles than Lindgren traps; captures with green traps were intermediate. With ethanol- baited traps, 20 species of bark/ambrosia beetle were detected. Trap efficacy varied by species, but in general, sticky traps captured the highest number of beetles. Results indicate that sticky panel traps are more effective for monitoring ambrosia beetles than Lindgren funnel traps, the current standard, and may provide an economical alternative for pest detection in avocado groves.

Key words: ambrosia beetle, Fusarium dieback, laurel wilt, Persea americana, Xyleborus glabratus

Production of avocado (Persea americana Mill. [Laurales: Lauraceae]) [P. palustris (Raf.) Sarg. (Laurales: Lauraceae)], laurel wilt quickly in Florida is currently threatened by two vascular diseases—laurel spread southward into Florida where it became apparent that avo- wilt and Fusarium dieback—both caused by fungal symbionts of cado was also susceptible (reviewed by Kendra et al. 2013). Contrary invasive ambrosia beetles (Coleoptera: Curculionidae: Scolytinae: to predictions, populations of X. glabratus did not increase rapidly Xyleborini) endemic to Southeast Asia, but now established in the within commercial groves, a monoculture of apparently suitable host United States. The primary vector of laurel wilt is redbay ambrosia trees. Although avocado wood is highly attractive to females due to beetle, Xyleborus glabratus Eichhoff, first detected near Savannah, emissions of appropriate kairomones (Kendra et al. 2011a, 2012b, Georgia in 2002, and now found in 11 southeastern states (USDA-FS 2014a), and the species can develop successfully within cut avocado 2019). The predominant symbiont of X. glabratus is Raffaelea logs under laboratory conditions (Brar et al. 2013), avocado trees lauricola T. C. Harr., Fraedrich & Aghayeva (Ophiostomatales: in situ appear to be poor reproductive hosts (Carrillo et al. 2012). Ophiostomataceae) (Harrington et al. 2008), a fungus that triggers Despite very low numbers of X. glabratus in Florida groves (Kendra a host defensive response that includes extensive tylose formation in et al. 2017, Owens et al. 2019a), laurel wilt is still prevalent and con- xylem vessels, ultimately leading to compromised water transport, tinues to spread throughout the avocado production area of Miami- systemic wilt, and death of infected trees (Ploetz et al. 2012). Initially Dade County (Ploetz et al. 2017a). This is the result of a community regarded as an epidemic affecting native Lauraceae, primarily redbay of ambrosia beetles attracted to (Kendra et al. 2011b) and breeding [Persea borbonia (L.) Spreng. (Laurales: Lauraceae)] and swampbay in (Carrillo et al. 2012) avocado wood, as well as an unexpected,

Published by Oxford University Press on behalf of Entomological Society of America 2019. This work is written by (a) US Government employee(s) and is in 1 the public domain in the US. Copyedited by: OUP

2 Journal of Economic Entomology, 2019, Vol. XX, No. XX

unprecedented lateral transfer of R. lauricola to at least nine add- Materials and Methods itional species which now function as potential secondary vectors of Experimental Insects the pathogen (Carrillo et al. 2014, Ploetz et al. 2017b). Fusarium dieback is a more recent wilt disease that affects avo- Xyleborus glabratus used in behavioral bioassays were reared from cado, woody ornamentals, and native trees in California (Eskalen infested swampbay wood collected in Broward County, FL. Logs Downloaded from https://academic.oup.com/jee/advance-article-abstract/doi/10.1093/jee/toz311/5632132 by U S Dept of Agriculture user on 20 November 2019 et al. 2012, 2013), Florida (Carrillo et al. 2016; Owens et al. 2018), (8–12 cm diameter) were placed flat on moist paper towels in a and elsewhere (Mendel et al. 2012, García-Avila et al. 2016, Paap shallow plastic emergence chamber (122 cm long × 61 cm wide × et al. 2018). It is caused by a complex of fungal symbionts (pri- 18 cm high; Sterilite Corp., Townsend, MA) containing a tight-fitting marily in the genera Fusarium, Graphium, and Acremonium) vec- lid, which had two screened ventilation holes (12 cm diameter) at tored by shot hole borers in the Euwallacea fornicatus cryptic each end. The chamber was maintained at 24°C under natural light species complex (Freeman et al. 2013, Carrillo et al. 2016, Lynch and checked daily. Newly emerged females were collected by hand et al. 2016). There are seven described species within this complex with a camel hair brush and held in a small plastic storage box with (Smith et al. 2019), with two established in California and an- moist tissue until needed for testing. other in South Florida. The Florida species was recently reported as Euwallacea perbrevis (Schedl) (Smith et al. 2019), but due to a Laboratory Bioassays lack of consensus regarding nomenclature (O’Donnell et al. 2015, Test substrates consisted of two plastic bubble lures obtained from Stouthamer et al. 2017, Gomez et al. 2018), the conservative, col- Synergy Semiochemicals Corp. (Delta, BC, Canada). The first lure lective term E. near fornicatus will be retained in this report. Unlike contained 50% α-copaene oil (2.0 ml loaded into a 2.9-cm-diameter R. lauricola that spreads systemically throughout a host tree, the bubble; product #3302), with a release rate of 10 mg/d reported by Fusarium pathogen introduced by E. nr. fornicatus remains local- the manufacturer. The second lure contained quercivorol (97 mg ized near the beetle entrance holes and galleries, typically at the in a 1.2-cm-diameter bubble; product #3402), with a release rate base of a branch, which results in wilt and dieback of individual of 0.7 mg/d. An empty plastic bubble dispenser served as a nega- branches. tive control (blank treatment). [Note: the lure commonly referred Field lures have been developed for detection and monitoring to as ‘quercivorol’ contains a mixture of four isomers of p-menth- of both X. glabratus and E. nr. fornicatus. The chemical ecology 2-en-1-ol; only the 1S,4R-enantiomer has been termed quercivorol of host-seeking female X. glabratus has been the subject of re- (Kashiwagi et al. 2006), and it is not the major component in the lure search for over a decade, and a series of essential oils have been (Owens et al. 2019a).] utilized as kairomone attractants, culminating in a distilled es- Bioassays were conducted in rectangular plastic bins (iden- sential oil product highly enriched in (-)-α-copaene, which is the tical to the one described above for beetle rearing), using a modi- current standard lure (Kendra et al. 2016a, 2016b, 2018). In a fied protocol from that reported previously (Kendra et al. 2016a). field test targeting X. glabratus in avocado (Kendra et al. 2015), it A paper lining was placed on the bin floor to facilitate walking. Test was discovered that females of E. nr. fornicatus are also attracted lures were hung at opposite ends of the bin using wire hooks (paper to α-copaene. Subsequent research demonstrated that α-copaene clips) taped to the bin wall, and a piece of sticky card (4 × 8 cm) was is equal in attraction to quercivorol, a food-based attractant and placed directly beneath each lure to retain attracted beetles. Tests the standard lure for E. nr. fornicatus (Carrillo et al. 2015); more- were initiated in the early afternoon, prior to onset of the female over, a combination of α-copaene and quercivorol lures results flight window (~16:00 h EDST; Kendra et al. 2012a). For each rep- in synergistic attraction and improved pest detection (Kendra licate run, 10 females (≤48 h postemergence) were introduced at the et al. 2017). This two-component lure has field longevity of 3 mo center of the arena and immediately covered with an inverted glass (Owens et al. 2019a) and a sampling range of 30–35 m (Owens funnel (6 cm diameter). After 5 min acclimation, the funnel was lifted et al. 2019b). gently to release the beetles and start the bioassay. The bin was then The objective of the current study was to evaluate efficacy of covered and left undisturbed (18 h) until the next morning, at which four trap designs, baited with the best available attractants, for de- point the test was concluded and beetles counted on each card. tection of pest ambrosia beetles in Florida avocado groves. The two- Five replicated binary-choice tests were conducted (copaene vs component lure was used to target X. glabratus and E. nr. fornicatus blank, quercivorol vs blank, combination lure vs blank, copaene vs in two field tests. A third field test was conducted using low-release quercivorol, and combination lure vs copaene). Each choice test was ethanol lures, a general attractant for many ambrosia beetles (Miller replicated six times, using a new cohort of beetles and paper liner, and Rabaglia 2009), but not for X. glabratus (Kendra et al. 2014b) and reversing the position of test substrates between replicate runs. or E. nr. fornicatus (Carrillo et al. 2015). The latter test was intended to evaluate the effect of trap design on detection of the community of Traps ambrosia beetles in this agroecosystem, particularly species known Four trap designs were compared in this study. Three are available to be secondary vectors of the laurel wilt pathogen (Carrillo et al. commercially and specifically intended for monitoring wood-boring 2014, Ploetz et al 2017b). beetles: a black 8-unit Lindgren funnel trap (BioQuip Products, The two-component lure has been used previously in Florida Inc., Rancho Dominguez, CA; product #2853; Fig. 1A), a black grove surveys (Carrillo et al. 2016) and has been adopted by 3-vane flight interception trap (Synergy Semiochemicals Corp.; SAGARPA (Secretaría de Agricultura, Ganadería, Desarrollo Rural, product #4057; Fig. 1B), and a green 3-vane flight interception trap Pesca y Alimentación) in Mexico in surveillance programs for both (Synergy Semiochemicals Corp.; product #4058; Fig. 1C). With these pests, yet it has not been determined whether quercivorol functions ‘wet trap’ designs, the collection cups were filled with 300 ml of as an attractant or repellent for X. glabratus. Therefore, we also con- an aqueous solution of 10% propylene glycol (Low-Tox antifreeze; ducted laboratory bioassays with newly emerged female X. glabratus Prestone, Danbury, CT) to retain and preserve captures. The fourth to assess behavioral responses to α-copaene and quercivorol, pre- design consisted of a sticky panel trap (Fig. 1D) used previously in sented alone and in tandem, to test for a potential negative response our field evaluations of scolytine lures (Kendra et al. 2012c, 2014b) to quercivorol in the combination lure. and host preferences (Kendra et al. 2011a, 2014a). This latter trap Copyedited by: OUP

Journal of Economic Entomology, 2019, Vol. XX, No. XX 3 Downloaded from https://academic.oup.com/jee/advance-article-abstract/doi/10.1093/jee/toz311/5632132 by U S Dept of Agriculture user on 20 November 2019

Fig. 1. Trap designs evaluated for efficacy of detection of bark and ambrosia beetles in Florida avocado groves. (A) Black 8-unit Lindgren funnel trap. (B) Black 3-vane flight interception trap. (C) Green 3-vane flight interception trap. (D) White sticky panel trap. All traps baited with a low-release ethanol lure.

was assembled from two white sticky panels (23 × 28 cm Sentry panels were first cleared by briefly soaking them in Histo-Clear wing trap bottoms; Great Lakes IPM, Vestaburg, MI), oriented II (National Diagnostics, Atlanta, GA). With the exception of back-to-back, suspended from a wire hanger, and topped with an in- Hypothenemus spp., all scolytine and platypodine beetles were verted plastic plate (24 cm diameter) to provide a protective covering identified to species level according to Rabaglia et al. (2006) and comparable to that found on the three commercial trap designs. Atkinson et al. (2013).

Field Tests Statistical Analysis In two field trials designed for E. nr. fornicatus and X. glabratus, Data were analyzed using Systat Software (2017). Paired t-tests the four trap designs were baited with a combination of α-copaene were used to analyze results from the binary-choice bioassays. Field and quercivorol lures (described above). In addition, an unbaited captures were expressed as beetles/trap/week (captures pooled over white sticky trap was used as a negative control to assess passive time for each treatment within a replicate block) and results ana- baseline captures. Field test 1 was conducted from 10 October to lyzed with analysis of variance (ANOVA); significant ANOVAs 19 December 2017 (10 wk) in an avocado grove affected by laurel were then followed by mean separation with Tukey HSD test. wilt (25° 29′ 58.42″N, 80° 29′ 19.80″W); however, the grove was When necessary, data were square root (x + 0.5)-transformed to well managed with prompt removal of symptomatic trees. Field test stabilize variance prior to analysis. Results are presented as mean 2 was conducted from 21 December 2017 to 1 March 2018 (10 wk) ± SEM; probability was considered significant at a critical level of in a grove free of laurel wilt (25° 30′ 22.17″N, 80° 29′ 21.96″W). α = 0.05. A third 10-wk field trial was designed to sample the overall community of bark and ambrosia beetles. This test utilized low-release ethanol lures (Synergy Semiochemicals Corp.; product #3344; re- Results lease rate of 15 mg/d) to bait the four trap treatments and also included an unbaited sticky trap control. Field test 3 was conducted Laboratory Bioassays from 27 March to 5 June 2018 at another avocado grove affected by Five replicated binary-choice tests were conducted to assess behav- laurel wilt (25° 27′ 73.00″N, 80° 32′ 47.60″W). Unlike the site used ioral responses of female X. glabratus to α-copaene and quercivorol for test 1, this grove had an ample number of stressed and dying trees lures, presented alone and in combination (Table 1). Overall, the available as breeding substrates for beetles. response rate was very high, with an average of 91.68 (± 1.58) % All three tests followed a randomized complete block design of the beetles responding to one of the two choices. When given a with five replicate blocks; each block consisted of a row of traps choice between lures and a blank control, females were attracted to hung ~1.5 m above ground (Brar et al. 2012) in well-shaded lo- α-copaene alone and the combination of α-copaene plus quercivorol, cations in trees asymptomatic for laurel wilt or Fusarium die- but not to quercivorol alone. When presented with α-copaene versus back. There was a minimum spacing of 12 m between adjacent quercivorol, females were more attracted to α-copaene. When pre- traps within a row, and 15 m spacing between rows. At weekly sented with the combination of α-copaene and quercivorol versus intervals, traps were serviced to collect beetles and replace reten- α-copaene alone, females were equally distributed among the tion fluid/sticky cards. In addition, trap treatments were rotated two choices (the choice between α-copaene and quercivorol vs sequentially within each row. By the end of a 10-wk test, each quercivorol alone was not conducted since quercivorol displayed trap had completed two full rotations through each field position no attraction in all previous tests). These results provide no evi- within a row. All samples were processed in the laboratory (USDA- dence that quercivorol functions as an attractant or a repellent for ARS, Miami, FL) under a stereo microscope (Discovery.V12; Carl X. glabratus. Use of the two-component lure for dual detection of E. Zeiss Microscopy, LLC, White Plains, NY). Specimens from sticky nr. fornicatus and X. glabratus is valid. Copyedited by: OUP

4 Journal of Economic Entomology, 2019, Vol. XX, No. XX

Table 1. Response (mean ± SEM) of female Xyleborus glabratus to α-copaene (Cop) and quercivorol (Quer) lures presented in binary-choice bioassays

Choice 1/Choice 2 % Respondinga % at Choice 1b % at Choice 2b t df P

Cop/blank 95.0 89.5 ± 3.9 10.5 ± 3.9 14.43 10 <0.001* Downloaded from https://academic.oup.com/jee/advance-article-abstract/doi/10.1093/jee/toz311/5632132 by U S Dept of Agriculture user on 20 November 2019 Quer/blank 91.7 51.1 ± 3.8 48.9 ± 3.8 0.415 10 0.687 Cop+Quer/blank 95.0 85.9 ± 4.3 14.1 ± 4.3 11.85 10 <0.001* Cop/Quer 86.7 73.8 ± 4.7 26.2 ± 4.7 7.173 10 <0.001* Cop+Quer/Cop 90.0 52.6 ± 2.3 47.4 ± 2.3 1.625 10 0.135

aPercentage of beetles responding to either choice. bPercentage of responding beetles attracted by each choice. *Response to choice 1 significantly higher.

significantly higher than those obtained with the standard Lindgren trap. Captures with the green 3-vane trap were equivalent to those obtained with the standard Lindgren trap but significantly higher than those of the unbaited control. Captures with the Lindgren trap were statistically no different from those of the unbaited control. In field test 2 (Fig. 2B), there were similar differences in mean capture of E. nr. fornicatus among treatments (F = 21.488; df = 4,20; P < 0.001), but the population level was approximately four times higher at this site. The sticky panel trap and black 3-vane trap caught equally high numbers of beetles, followed by intermediate captures with the green 3-vane and Lindgren traps, and low captures with the unbaited control. As observed in field test 1, both the sticky trap and black 3-vane trap caught significantly more E. nr. fornicatus than the standard Lindgren trap. At this higher beetle density, captures with the Lindgren funnel trap were greater than those of the control trap. The combination of α-copaene and quercivorol lures resulted in good specificity for attraction of female E. nr. fornicatus, which comprised 74.3 and 75.9% of the trap captures in field tests 1 and 2, respectively (Table 2). Species with the next highest captures in test 1 were Ambrosiodmus devexulus (Wood), representing 9.3% of the total, and Hypothenemus spp., representing 6.9%. A single specimen of X. glabratus was caught in a baited sticky trap in test 1. In test 2, Hypothenemus spp. comprised 10.2% and Theoborus ricini (Eggers) 6.1% of the overall trap captures. In field test 3 (Fig. 3), there were significant differences in mean capture of bark and ambrosia beetles (all species combined) among the trap designs baited with ethanol lures (F = 30.031; df = 4,20; P < 0.001). Highest numbers of beetles were caught with the sticky panel trap, followed by intermediate numbers with the black 3-vane trap and Lindgren funnel trap, and lowest captures with the green 3-vane trap, which were statistically no different from those obtained

Fig. 2. Mean (± SEM) captures of female Euwallacea nr. fornicatus. (A) with the unbaited control trap. Field test 1: 10-wk trial in an avocado grove with laurel wilt. (B) Field test 2: Diversity of species and total numbers captured were greater 10-wk trial in a grove free of laurel wilt. Treatments consist of a white sticky in field test 3 than in the previous two tests (Table 2). Species of panel trap (Sticky), black 3-vane flight interception trap (Black), green 3-vane Hypothenemus made up the largest percentage of captures in flight interception trap (Green), black 8-unit Lindgren funnel trap (Lindgren), ethanol-baited traps, comprising 57.9% of the total, followed by and an unbaited sticky trap (Control). All traps except control baited with a Xyleborinus saxesenii (Ratzeburg) at 17.3%, Xyleborus bispinatus combination of α-copaene and quercivorol lures. Bars topped with the same letter are not significantly different (Tukey HSD mean separation,P < 0.05). Eichhoff at 9.4%, Xyleborus affinis Eichhoff at 4.9%, Xyleborus Please note differences in scale for the y-axis. volvulus (Fabricius) at 4.6%, and Premnobius cavipennis Eichhoff at 1.3%. Euwallacea nr. fornicatus represented less than 1% of the Field Tests captures, and three specimens of X. glabratus were trapped (two In field test 1 (Fig. 2A), there were significant differences in mean on baited sticky traps and one on an unbaited sticky trap). All 10 capture of female E. nr. fornicatus among the five treatments species known to harbor R. lauricola were detected at this site (indi- (F = 31.434; df = 4,20; P < 0.001). Highest numbers of beetles cated by an asterisk in Table 2). were caught with the sticky panel trap, followed by intermediate With the six most abundant species from test 3, numbers were numbers with the black and green 3-vane traps, low numbers with high enough to permit analysis of trap efficacy (Fig. 4). Sticky panel the Lindgren funnel trap, and no captures with the unbaited trap. traps consistently caught the highest number of beetles, but per- Captures with both the sticky trap and the black 3-vane trap were formance of other trap designs varied depending on species. For Copyedited by: OUP

Journal of Economic Entomology, 2019, Vol. XX, No. XX 5

Table 2. Bark and ambrosia beetles captured in three 10-wk field P < 0.001). For X. saxesenii (Fig. 4B), sticky traps caught sig- tests conducted in commercial avocado groves in Miami-Dade nificantly more beetles than all other trap designs (F = 33.728; County, FL df = 4,20; P < 0.001). For X. bispinatus (Fig. 4C), highest captures were obtained with sticky and Lindgren traps, but Lindgren captures Species Test 1a Test 2a Test 3b were not significantly greater than those of the black 3-vane trap Downloaded from https://academic.oup.com/jee/advance-article-abstract/doi/10.1093/jee/toz311/5632132 by U S Dept of Agriculture user on 20 November 2019 Subfamily Scolytinae (F = 17.782; df = 4,20; P < 0.001). For X. volvulus (Fig. 4D), sticky Tribe Dryocoetini traps captured more than all other traps (F = 32.352; df = 4,20; Coccotrypes carpophagus (Hornung) 0 18 49 P < 0.001). For X. affinis (Fig. 4E), highest captures were obtained Tribe Xyleborini with sticky and Lindgren traps, but Lindgren captures were not Ambrosiodmus devexulus (Wood) 120 75 5 significantly greater than those of the black or green 3-vane traps Ambrosiodmus lecontei Hopkins* 7 8 22 (F = 8.009; df = 4,20; P < 0.001). For P. cavipennis (Fig. 4F), sticky Euwallacea nr. fornicatus Eichhoff 960 3623 141 and black 3-vane traps caught the highest number of beetles, but Premnobius cavipennis Eichhoff 0 8 181 Theoborus ricini (Eggers) 31 293 26 mean captures with the black 3-vane trap overlapped with those of Xyleborinus andrewesi (Blandford)* 0 15 89 the green 3-vane trap and the Lindgren trap (F = 13.786; df = 4,20; Xyleborinus gracilis (Eichhoff)* 0 22 34 P < 0.001). Xyleborinus saxesenii (Ratzeburg)* 8 52 2499 Xyleborus affinis Eichhoff* 15 7 693 Xyleborus bispinatus Eichhoff* 16 83 1353 Discussion Xyleborus glabratus Eichhoff* 1 0 3 Xyleborus ferrugineus (Fabricius)* 0 0 75 Development of improved semiochemical-based detection systems Xyleborus volvulus (Fabricius)* 38 30 665 for ambrosia beetles will rely on a better understanding of the chem- Xylosandrus compactus (Eichhoff) 0 7 0 ical ecology and flight behavior of the adult female during that brief Xylosandrus crassiusculus (Motschulsky)* 0 45 97 stage when she leaves the natal tree and searches for a new host. Tribe Cryphalini For X. glabratus, the process has been viewed as a multistep pro- Cryptocarenus heveae (Hagedorn) 0 0 63 cess involving sequential input of pertinent environmental informa- Hypothenemus spp. 89 488 8347 tion, including long-range olfactory cues, mid-range visual cues, and Tribe Corthylini final integration of close-range/contact cues that result in a behav- Corthylus papulans Eichhoff 0 0 25 ioral switch from ‘host-seeking’ to ‘host-acceptance’ mode (Kendra Subfamily Platypodinae et al. 2014a). Optimization of pest detection requires discovery of Euplatypus parallelus (Fabricius) 7 0 52 effective attractants as well as identification of appropriate trap designs for your target species. Since effective attractants are known, aTraps baited with a combination of α-copaene and quercivorol lures. bTraps baited with low-release ethanol lures. the current study examined the latter requirement for ambrosia bee- *Species from which Raffaelea lauricola, etiological agent of laurel wilt, has tles that vector phytopathogenic fungi in Florida. been isolated (Ploetz et al. 2017b). The multiple funnel trap developed by Lindgren (1983) has long been the standard used by action agencies for detection and population monitoring of bark and ambrosia beetles. However, our results indicate that this trap design does not always provide ad- equate detection of critical ambrosia beetle species in Florida. In field tests 1 and 2, E. nr. fornicatus was trapped at significantly higher numbers with the sticky panel trap and the black 3-vane trap. The four X. glabratus captured (tests 1 and 3) were all intercepted with sticky traps, results consistent with a previous comparison between Lindgren and sticky traps conducted at a north Florida site with higher numbers of X. glabratus (Kendra et al. 2012c). In field test 3, mean cumulative captures of all beetle species were again significantly higher with the sticky panel trap than with the Lindgren trap, including captures of X. saxesenii and X. volvulus, potential carriers of the laurel wilt pathogen (Carrillo et al. 2014, Ploetz et al. 2017b). Brar et al. (2012) evaluated the effect of Lindgren trap size (4, 8, 12, and 16 funnels) on captures of X. glabratus. An increase from 4 to 8 funnels resulted in a significant increase in captures; however, further Fig. 3. Mean (± SEM) captures of bark and ambrosia beetles (all species increment in funnels did not improve the numbers captured. Based combined) in field test 3: 10-wk trial in an avocado grove with laurel wilt. on these results, our choice of the 8-unit funnel trap should have rep- Treatments consist of a white sticky panel trap (Sticky), black 3-vane flight resented the optimal Lindgren design for field comparison. interception trap (Black), green 3-vane flight interception trap (Green), black 8-unit Lindgren funnel trap (Lindgren), and an unbaited sticky trap In comparing the black and green versions of the 3-vane flight (Control). All traps except control baited with a low-release ethanol lure. Bars interception trap, captures of all species were less with the green topped with the same letter are not significantly different (Tukey HSD mean trap. Since construction was identical with these two traps, the dif- separation, P < 0.05). ference in captures indicates a strong influence of visual cues. As with the Lindgren funnel trap, the black color of the 3-vane trap, Hypothenemus spp. (Fig. 4A), equivalent captures were obtained in combination with its elongated cylindrical shape, is likely rep- with sticky, black 3-vane, and Lindgren traps, which were all higher resentative of a tree trunk or branch silhouette, providing an add- than captures with the green 3-vane trap (F = 36.339; df = 4,20; itional signal consistent with host location, as demonstrated for Copyedited by: OUP

6 Journal of Economic Entomology, 2019, Vol. XX, No. XX Downloaded from https://academic.oup.com/jee/advance-article-abstract/doi/10.1093/jee/toz311/5632132 by U S Dept of Agriculture user on 20 November 2019

Fig. 4. Mean (± SEM) captures of the six most abundant species in field test 3: 10-wk trial in an avocado grove with laurel wilt. Treatments consist of a white sticky panel trap (Sticky), black 3-vane flight interception trap (Black), green 3-vane flight interception trap (Green), black 8-unit Lindgren funnel trap (Lindgren), and an unbaited sticky trap (Control). All traps except control baited with a low-release ethanol lure. Bars topped with the same letter are not significantly different (Tukey HSD mean separation, P < 0.05). Please note differences in scale for the y-axis.

X. glabratus (Mayfield and Brownie 2013). It should be noted that, cues, and then fly off. Authors (P.E.K. and W.S.M.) have observed in the latter study, the visual cue was not sufficient on its own, but in- females of X. glabratus to hover near a lure but not land (during live creased female attraction only when presented in combination with collections reported in Kendra et al. 2012b) or land on the upper an appropriate chemical cue (a host-based kairomone). rim of a Lindgren funnel, but not fall into the trap (during field In all field tests, the white sticky panel trap consistently captured tests reported in Kendra et al. 2012c). Even with the sticky panel the highest number of beetles. This suggests that the lures utilized design, trapping efficacy has been estimated to be only 44.0 ± 2.3% were sufficiently attractive to bring females close enough to make for X. glabratus (field cage release-recapture tests withα -copaene contact with the adhesive, after which there was no opportunity for lure; Kendra et al. 2016b) and 31.2 ± 3.7% for E. nr. fornicatus escape. With the other three trap designs, it is possible for small bee- (open field release-recapture tests with two-component lure; Owens tles to land on the trap, fail to detect appropriate short-range/contact et al. 2019b). Therefore, trap captures are only indicative of relative Copyedited by: OUP

Journal of Economic Entomology, 2019, Vol. XX, No. XX 7

species abundance at an experimental site. The trapping period, Byers, J. A., Y. Maoz, and A. Levi-Zada. 2017. Attraction of the Euwallacea including length of study and time of year, may also affect the species sp. near fornicatus (Coleoptera: Curculionidae) to quercivorol and to in- abundance and composition detected, since population levels fluc- festations in avocado. J. Econ. Entomol. 110: 1512–1517. tuate seasonally (Kendra et al. 2012c, 2016a; Johnson et al. 2016). Carrillo, D., R. E. Duncan, and J. E. Peña. 2012. Ambrosia beetles (Coleoptera: Curculionidae: Scolytinae) that breed in avocado wood in Florida. Fla. Additional factors influencing trapping efficacy include vertical Entomol. 95: 573–579. Downloaded from https://academic.oup.com/jee/advance-article-abstract/doi/10.1093/jee/toz311/5632132 by U S Dept of Agriculture user on 20 November 2019 distribution and temporal flight periodicity of host-seeking beetles Carrillo, D., R. E. Duncan, J. N. Ploetz, A. Campbell, R. C. Ploetz, and (Kendra et al. 2012a, Menocal et al. 2018). The current field tests J. E. Peña. 2014. Lateral transfer of a phytopathogenic symbiont among were conducted with traps hung at a height of 1.5 m above ground, native and exotic ambrosia beetles. Plant Pathol. 63: 54–62. based on previous studies with X. glabratus (Brar et al. 2012) and Carrillo, D., T. Narvaez, A. A. Cossé, R. Stouthamer, and M. Cooperband. E. nr. fornicatus (Byers et al. 2017); however, a recent study with 2015. Attraction of Euwallacea nr. fornicatus (Coleoptera: Curculionidae: unbaited sticky cards indicated that many species fly at much greater Scolytinae) to lures containing quercivorol. Fla. Entomol. 98: 780–782. heights (Menocal et al. 2018). In addition, diurnal species like Carrillo, D., L. F. Cruz, P. E. Kendra, T. I. Narvaez, W. S. Montgomery, X. glabratus, E. nr. fornicatus, Xyleborinus andrewesi (Blandford), A. Monterroso, C. DeGrave, and M. F. Cooperband. 2016. Distribution, X. saxesenii, and Hypothenemus spp. (Kendra et al. 2012a, 2012b, pest status, and fungal associates of Euwallacea nr. fornicatus in Florida avocado groves. Insects 7(4): 55. 2017; Johnson et al. 2016; Menocal et al. 2018) are more likely to Eskalen, A., A. Gonzalez, D. H. Wang, M. Twizeyimana, J. S. Mayorquin, and utilize visual cues incorporated into the trap design when compared S. C. Lynch. 2012. First report of a Fusarium sp. and its vector tea shot with species that engage in flight near sunset and after dark, like hole borer (Euwallacea fornicatus) causing Fusarium dieback on avocado X. bispinatus and Xylosandrus crassiusculus (Motschulsky) (Kendra in California. Plant Dis. 96: 1070. et al. 2012b, Menocal et al. 2018). Eskalen, A., R. Stouthamer, S. C. Lynch, P. F. Rugman-Jones, M. Twizeyimana, In conclusion, the results of this study indicate that 1) quercivorol A. Gonzalez, and T. Thibault. 2013. Host range of Fusarium dieback and does not function as a repellent for X. glabratus and therefore the its ambrosia beetle (Coleoptera: Scolytinae) vector in Southern California. two-component lure (α-copaene and quercivorol) is appropriate Plant Dis. 97: 938–951. for dual detection of E. nr. fornicatus and X. glabratus; and 2) of Freeman, S., M. Sharon, M. Maymon, Z. Mendel, A. Protasov, T. Aoki, the four trap designs tested, the sticky panel trap provides the best A. Eskalen, and K. O’Donnell. 2013. Fusarium euwallaceae sp. nov.–a symbiotic fungus of Euwallacea sp., an invasive ambrosia beetle in Israel overall detection and monitoring of ambrosia beetles in Florida avo- and California. Mycologia. 105: 1595–1606. cado groves. Efficacy of other traps was variable and depended upon García-Avila, C. D. J., F. J. Trujillo-Arriaga, J. A. López-Buenfil, R. González- the target species. For E. nr. fornicatus, both the sticky panel trap Gómez, D. Carrillo, L. F. Cruz, I. Ruiz-Galván, A. Quezada-Salinas, and the black 3-vane trap provided more sensitive detection than and N. Acevedo-Reyes. 2016. First report of Euwallacea nr. fornicatus the standard Lindgren funnel trap. Previous field comparisons indi- (Coleoptera: Curculionidae) in Mexico. Fla. Entomol. 99: 555–556. cated that sticky traps also improve detection of X. glabratus, and Gomez, D. F., J. Skelton, M. S. Steininger, R. Stouthamer, P. Rugman-Jones, the current study suggests that sticky traps far exceed other designs W. Sittichaya, R. J. Rabaglia, and J. Hulcr. 2018. Species within the for detection of X. saxesenii and X. volvulus. In terms of costs as- Euwallacea fornicatus (Coleoptera: Curculionidae) complex revealed by sociated with the products evaluated, the current posted price of morphometric and phylogenetic analyses. Insect Syst. Divers. 2: 1–11. sticky panels is US$0.98 each, 8-unit funnel traps US$64.39 each, Harrington, T. C., S. W. Fraedrich, and D. N. Aghayeva. 2008. Raffaelea lauricola, a new ambrosia beetle symbiont and pathogen on the Lauraceae. and 3-vane traps US$30.12 each. Sticky panel traps may provide a Mycotaxon 104: 399–404. more effective and economical alternative for growers and regula- Johnson, A. J., P. E. Kendra, J. Skelton, and J. Hulcr. 2016. Species diver- tory agencies that monitor for pest ambrosia beetles. If a wet trap de- sity, phenology, and temporal flight patterns of Hypothenemus pygmy sign is highly preferred, use of the black 3-vane trap would result in borers (Coleoptera: Curculionidae: Scolytinae) in South Florida. Environ. a reduction of program costs over the currently used Lindgren trap. Entomol. 45: 627–632. Kashiwagi, T., T. Nakashima, S. I. Tebayashi, and C. S. Kim. 2006. Determination of the absolute configuration of quercivorol, (1S,4R)- Acknowledgments p-menth-2-en-1-ol, an aggregation pheromone of the ambrosia beetle We are grateful to Carlos de la Torre, John Barkett, and Mark Platypus quercivorus (Coleoptera: Platypodidae). Biosci. Biotechnol. Biochem. 70: 2544–2546. Philcox for granting permission to conduct research in their groves. Kendra, P. E., W. S. Montgomery, J. Niogret, J. E. Peña, J. L. Capinera, G. Brar, We thank Nur Tabanca (USDA-ARS, Miami, FL), Octavio Menocal N. D. Epsky, and R. R. Heath. 2011a. Attraction of the redbay ambrosia (University of Florida, Homestead), and the journal referees for pro- beetle, Xyleborus glabratus, to avocado, lychee, and essential oil Lures. J. viding critical reviews of an earlier version of this manuscript. We also Chem. Ecol. 37: 932–942. acknowledge Synergy Semiochemicals Corp. (Delta, BC, Canada) for Kendra, P. E., J. S. Sanchez, W. S. Montgomery, K. E. Okins, J. Niogret, providing the 3-vane flight interception traps. Use of a proprietary J. E. Peña, N. D. Epsky, and R. R. Heath. 2011b. Diversity of Scolytinae product does not constitute an endorsement by USDA-ARS. (Coleoptera: Curculionidae) attracted to avocado, lychee, and essential oil lures. Fla. Entomol. 94: 123–130. References Cited Kendra, P. E., W. S. Montgomery, J. Niogret, M. A. Deyrup, L. Guillén, and N. D. Epsky. 2012a. Xyleborus glabratus, X. affinis, and X. ferrugineus Atkinson, T. H., D. Carrillo, R. E. Duncan, and J. E. Peña. 2013. Occurrence (Coleoptera: Curculionidae: Scolytinae): electroantennogram responses of Xyleborus bispinatus (Coleoptera: Curculionidae: Scolytinae) Eichhoff to host-based attractants and temporal patterns in host-seeking flight. in southern Florida. Zootaxa. 3669: 96–100. Environ. Entomol. 41: 1597–1605. Brar, G. S., J. L. Capinera, S. McLean, P. E. Kendra, R. C. Ploetz, and Kendra, P. E., W. S. Montgomery, J. S. Sanchez, M. A. Deyrup, J. Niogret, J. E. Peña. 2012. Effect of trap size, trap height, and age of lure on sam- and N. D. Epsky. 2012b. Method for collection of live redbay ambrosia pling Xyleborus glabratus (Coleoptera: Curculionidae: Scolytinae), and its beetles, Xyleborus glabatus (Coleoptera: Curculionidae: Scolytinae). Fla. flight periodicity and seasonality. Fla. Entomol. 95: 1003–1011. Entomol. 95: 513–516. Brar, G. S., J. L. Capinera, P. E. Kendra, S. McLean, and J. E. Peña. 2013. Kendra, P. E., J. Niogret, W. S. Montgomery, J. S. Sanchez, M. A. Deyrup, Life cycle, development, and culture of Xyleborus glabratus (Coleoptera: G. E. Pruett, R. C. Ploetz, N. D. Epsky, and R. R. Heath. 2012c. Temporal Curculionidae: Scolytinae). Fla. Entomol. 96: 1158–1167. analysis of sesquiterpene emissions from manuka and phoebe oil lures and Copyedited by: OUP

8 Journal of Economic Entomology, 2019, Vol. XX, No. XX

efficacy for attraction of Xyleborus glabratus (Coleoptera: Curculionidae: (Coleoptera: Curculionidae) in Florida avocado orchards affected by Scolytinae). J. Econ. Entomol. 105: 659–669. laurel wilt. J. Econ. Entomol. 111: 1190–1196. Kendra, P. E., W. S. Montgomery, J. Niogret, and N. D. Epsky. 2013. An uncer- Miller, D. R., and R. J. Rabaglia. 2009. Ethanol and (-)-α-pinene: attractant tain future for American Lauraceae: a lethal threat from redbay ambrosia kairomones for bark and ambrosia beetles in the southeastern US. J. beetle and laurel wilt disease (A review.). Am. J. Plant Sci. 4: 727–738. Chem. Ecol. 35: 435–448.

Kendra, P. E., W. S. Montgomery, J. Niogret, G. E. Pruett, A. E. Mayfield, O’Donnell, K., S. Sink, R. Libeskind-Hadas, J. Hulcr, M. T. Kasson, Downloaded from https://academic.oup.com/jee/advance-article-abstract/doi/10.1093/jee/toz311/5632132 by U S Dept of Agriculture user on 20 November 2019 3rd, M. MacKenzie, M. A. Deyrup, G. R. Bauchan, R. C. Ploetz, and R. C. Ploetz, J. L. Konkol, J. N. Ploetz, D. Carrillo, A. Campbell, et al. N. D. Epsky. 2014a. North American Lauraceae: terpenoid emissions, rela- 2015. Discordant phylogenies suggest repeated host shifts in the Fusarium- tive attraction and boring preferences of redbay ambrosia beetle, Xyleborus Euwallacea ambrosia beetle mutualism. Fungal Genet. Biol. 82: 277–290. glabratus (Coleoptera: Curculionidae: Scolytinae). PLoS One. 9: e102086. Owens, D., L. F. Cruz, W. S. Montgomery, T. I. Narvaez, E. Q. Schnell, Kendra, P. E., W. S. Montgomery, J. Niogret, E. Q. Schnell, M. A. Deyrup, and N. Tabanca, R. E. Duncan, D. Carrillo, and P. E. Kendra. 2018. Host range N. D. Epsky. 2014b. Evaluation of seven essential oils identifies cubeb oil expansion and increasing damage potential of Euwallacea nr. fornicatus as most effective attractant for detection of Xyleborus glabratus. J. Pest (Coleoptera: Curculionidae) in Florida. Fla. Entomol. 101: 229–236. Sci. 87: 681–687. Owens, D., P. E. Kendra, N. Tabanca, T. I. Narvaez, W. S. Montgomery, Kendra, P. E., T. I. Narvaez, W. S. Montgomery, and D. Carrillo. 2015. E. Q. Schnell, and D. Carrillo. 2019a. Quantitative analysis of contents Ambrosia beetle communities in forest and agricultural ecosystems with and volatile emissions from α-copaene and quercivorol lures, and lon- laurel wilt disease (D3524). 53rd Annual Meeting of the Entomological gevity for attraction of Euwallacea nr. fornicatus in Florida. J. Pest Sci. Society of America, Minneapolis, MN. 15–18 Nov 2015. https://esa. 92: 237–252. confex.com/esa/2015/webprogram/Paper99702.html. Accessed 11 Owens, D., M. Seo, W. S. Montgomery, M. J. Rivera, L. L. Stelinski, and September 2019. P. E. Kendra. 2019b. Dispersal behavior of Euwallacea nr. fornicatus Kendra, P. E., W. S. Montgomery, M. A. Deyrup, and D. Wakarchuk. 2016a. (Coleoptera: Curculionidae: Scolytinae) in avocado groves and estimation Improved lure for redbay ambrosia beetle developed by enrichment of of lure sampling range. Agr. Forest Entomol. 21: 199–208. α-copaene content. J. Pest Sci. 89: 427–438. Paap, T., Z. W. de Beer, D. Migliorini, W. J. Nel, and M. J. Wingfield. 2018. Kendra, P. E., W. S. Montgomery, E. Q. Schnell, M. A. Deyrup, and N. D. Epsky. The polyphagous shot hole borer (PSHB) and its fungal symbiont 2016b. Efficacy ofα -copaene, cubeb, and eucalyptol lures for detection of Fusarium euwallaceae: a new invasion in South Africa. Australas. Plant redbay ambrosia beetle (Coleoptera: Curculionidae: Scolytinae). J. Econ. Pathol. 47: 231–237. Entomol. 109: 2428–2435. Ploetz, R. C., J. M. Pérez-Martínez, J. A. Smith, M. Hughes, T. J. Dreaden, Kendra, P. E., D. Owens, W. S. Montgomery, T. I. Narvaez, G. R. Bauchan, S. A. Inch, and Y. Fu. 2012. Responses of avocado to laurel wilt, caused by E. Q. Schnell, N. Tabanca, and D. Carrillo. 2017. α-Copaene is an at- Raffaelea lauricola. Plant Pathol. 61: 801–808. tractant, synergistic with quercivorol, for improved detection of Ploetz, R. C., P. E. Kendra, R. A. Choudhury, J. A. Rollins, A. Campbell, Euwallacea nr. fornicatus (Coleoptera: Curculionidae: Scolytinae). PLoS K. Garrett, M. Hughes, and T. Dreaden. 2017a. Laurel wilt in natural and One. 12: e0179416. agricultural ecosystems: understanding the drivers and scales of complex Kendra, P. E., W. S. Montgomery, J. Niogret, N. Tabanca, D. Owens, and pathosystems. Forests 8: 48. N. D. Epsky. 2018. Utility of essential oils for development of host-based Ploetz, R. C., J. L. Konkol, T. Narvaez, R. E. Duncan, R. J. Saucedo, lures for Xyleborus glabratus (Coleoptera: Curculionidae: Scolytinae), A. Campbell, J. Mantilla, D. Carrillo, and P. E. Kendra. 2017b. Presence vector of laurel wilt. Open Chem. 16: 393–400. and prevalence of Raffaelea lauricola, cause of laurel wilt, in different spe- Lindgren, B. S. 1983. A multiple funnel trap for scolytid beetles (Coleoptera). cies of ambrosia beetle in Florida, USA. J. Econ. Entomol. 110: 347–354. Can. Entomol. 115: 299–302. Rabaglia, R. J., S. A. Dole, and A. I. Cognato. 2006. Review of American Lynch, S. C., M. Twizeyimana, J. S. Mayorquin, D. H. Wang, F. Na, M. Kayim, Xyleborina (Coleoptera: Curculionidae: Scolytinae) occurring north of M. T. Kasson, P. Q. Thu, C. Bateman, P. Rugman-Jones, et al. 2016. Mexico, with an illustrated key. Ann. Entomol. Soc. Am. 99: 1034–1056. Identification, pathogenicity and abundance of Paracremonium pembeum Smith, S. M., D. E. Gomez, R. A. Beaver, J. Hulcr, and A. I. Cognato. 2019. sp. nov. and Graphium euwallaceae sp. nov.–two newly discovered Reassessment of the species in the Euwallacea fornicatus (Coleoptera: mycangial associates of the polyphagous shot hole borer (Euwallacea sp.) Curculionidae: Scolytinae) Complex after the rediscovery of the “lost” in California. Mycologia. 108: 313–329. type specimen. Insects 10(9): 261. Mayfield, A. E., 3rd, and C. Brownie. 2013. The redbay ambrosia beetle Stouthamer, R., P. Rugman-Jones, P. Q. Thu, A. Eskalen, T. Thibault, J. Hulcr, (Coleoptera: Curculionidae: Scolytinae) uses stem silhouette diameter as L. Wang, B. H. Jordal, C. Chen, M. Cooperband, et al. 2017. Tracing the a visual host-finding cue. Environ. Entomol. 42: 743–750. origin of a cryptic invader: phylogeography of the Euwallacea fornicatus Mendel, Z., A. Protasov, M. Sharon, A. Zveibil, S. B. Yehuda, K. O’Donnell, (Coleoptera: Curculionidae: Scolytinae) species complex. Agric. Forest R. Rabaglia, M. Wyosoki, and S. Freeman. 2012. An Asian ambrosia Entomol. 19: 366–375. beetle Euwallacea fornicatus and its novel symbiotic fungus Fusarium Systat Software. 2017. SigmaPlot v. 14.0. Systat Software, Inc., San Jose, CA. sp. pose a serious threat to the Israeli avocado industry. Phytoparasitica USDA-FS. 2019. United States Department of Agriculture, Forest Service, 40: 35–238. Southern Regional Extension Forestry, Forest Health. Distribution of Menocal, O., P. E. Kendra, W. S. Montgomery, J. H. Crane, and D. Carrillo. Counties with Laurel Wilt as of July 23, 2019. http://southernforesthealth. 2018. Vertical distribution and daily flight periodicity of ambrosia beetles net/diseases/laurel-wilt/distribution-map. Accessed 11 September 2019.