Behavioral Responses of the Invasive Halyomorpha halys (Stål) to Traps Baited with Stereoisomeric Mixtures of 10,11-Epoxy-1-bisabolen-3-OL

The brown marmorated stink bug, Halyomorpha halys, is an invasive insect in the United States that is capable of inflicting significant yield losses for fruit, vegetable, and soybean growers. Recently, a male-produced aggregation pheromone of H. halys was identified as a 3.5:1 mixture of (3S,6S,7R,10S)-10,11-epoxy-1-bisabolen-3-ol and (3R,6S,7R,10S)-10,11-epoxy-1-bisabolen-3-ol, two stereoisomers of a natural sesquiterpene with a bisabolane skeleton, potentially existing in 16 stereoisomeric forms. In this study, we assessed attraction to pheromonal and non-pheromonal stereoisomeric mixtures of 10,11-epoxy-1-bisabolen-3-ol, which are easier to synthesize than single isomers, and evaluated dose-dependent responses to attractive mixtures in field trials. Some treatments not containing the natural pheromone components were moderately active in field-trapping studies, signifying that some stereoisomers of 10,11-epoxy-1-bisabolen-3-ol are sufficiently similar to the true pheromone in structure to trigger behavioral responses. Importantly, we found that mixtures of stereoisomers containing pheromone components were also highly attractive to H. halys, even in the presence of multiple “unnatural” stereoisomers. Further, adult and nymphal captures were dose-dependent, regardless of whether the lure contained pheromonal or non-pheromonal components. Our findings of attraction to pheromonal and non-pheromonal stereoisomers and lack of inhibition from non-pheromonal stereoisomers of 10,11-epoxy-1-bisabolen-3-ol increase the flexibility of developing pheromone-based products for H. halys.


Introduction
Halyomorpha halys (Stål), the brown marmorated stink bug (BMSB), is an invasive insect native to China, Taiwan, Korea, and Japan, which was accidently introduced into the United States sometime in the mid-late 1990s (Hoebeke and Carter 2003). Currently, as of June 2014 (www.stopbmsb.org for updates), H. halys is well established throughout the mid-Atlantic region (Leskey et al. 2012c), and has been officially detected in 43 states and the District of Columbia. Halyomorpha halys is a polyphagous pest of many crops in Asia (Lee et al. 2013a;Panizzi et al. 2000), North America, and Europe (Leskey et al. 2012a;Rice et al. 2014). In 2010, H. halys populations increased dramatically and attacked many crops in the mid-Atlantic region. Damage in commercial orchard crops reached critical levels with some growers experiencing serious losses in stone and pome fruit (Leskey et al. 2012a).
Native stink bugs have long been managed with broadspectrum insecticides, but since the passage of the Food Quality Protection Act in 1996, many broad-spectrum materials have been banned or severely curtailed through Electronic supplementary material The online version of this article (doi:10.1007/s10886-015-0566-x) contains supplementary material, which is available to authorized users. regulatory measures. This has caused populations of native stink bugs, long considered to be secondary pests in certain crops, to become more prevalent. Subsequently, as H. halys has become well established, populations have exerted tremendous season-long pest pressure, complicating the IPM programs for many crops. A number of broad spectrum insecticides have been identified as being effective against H. halys (Lee et al. 2013b;Leskey et al. 2012b;Nielsen et al. 2008); however, their use leads to profound changes in management and severe disruption to previously established IPM programs (Leskey et al. 2012a).
Native stink bug species typically have been monitored in cropping systems using sweep nets, beating samples, pheromone-baited traps, and/or blacklight traps (Borges et al. 2011;Kamminga et al. 2009;Krupke et al. 2001;Leskey and Hogmire 2005). For H. halys, pyramid traps baited with the aggregation pheromone [methyl (2E, 4E, 6Z)-decatrienoate] of Plautia stali Scott have been used to monitor populations (Joseph et al. 2013;Leskey et al. 2012d;Nielsen et al. 2011), because this compound was found to serve as a kairomone for H. halys (Aldrich et al. 2007;Khrimian et al. 2008). Although H. halys are attracted to methyl (2E, 4E, 6Z)-decatrienoate late in the season, adults do not respond to this chemical earlier, throughout much of the growing season (Leskey et al. 2012d), making it difficult to quantify the abundance and distribution of pest populations. Furthermore, some commercial lure formulations of this kairomone have had problems with release rates, contributing to low captures (Joseph et al. 2013). Black light traps also have been used to track the relative seasonal abundance (Nielsen et al. 2011), to monitor spread (Nielsen et al. 2013) and to identify landscape-level risk factors (Rice et al. 2014;Wallner et al. 2014). However, light traps are less specific and practical for growers to use to monitor populations on farms due to cost and labor.
Until recently, the pheromone of H. halys had not been identified, thereby limiting options for season-long monitoring at farm-scape levels. The two-component male-produced aggregation pheromone, consisting of (3S,6S,7R,10S)-10,11epoxy-1-bisabolen-3-ol and (3R,6S,7R,10S)-10,11-epoxy-1bisabolen-3-ol, of H. halys ) is attractive to males, females, and to nymphs and is synergized by the addition of methyl (2E, 4E, 6Z)-decatrienoate . While the two natural pheromone components are difficult to synthesize in pure form, they are relatively easy to produce as mixtures of stereoisomers. Thus, to develop and optimize commercially viable formulations of the pheromone, it is important to understand how other non-pheromonal stereoisomers of 10,11-epoxy-1-bisabolen-3-ol may affect attractiveness to H. halys, since they are byproducts of chemical syntheses. Therefore, the goal of this study was to assess field responses of H. halys to a range of stereoisomers of 10,11epoxy-1-bisabolen-3-ol, specifically to: 1) evaluate behavioral responses to pheromonal and non-pheromonal stereoisomers of H. halys; 2) determine if stereoisomeric mixtures containing pheromone and non-pheromone components are as attractive as mixtures containing only pheromone components; and 3) establish dose-dependent responses to pheromonal and non-pheromonal stereoisomers.

Methods and Materials
Chromatography Used in Synthetic Chemistry Gas chromatography (GC) analyses were performed on an Agilent Technologies 6890 N instrument equipped with a flame ionization detector and a DB-5 capillary column (30 m×0.32 mm i.d.×0.25 μm film). Hydrogen was used as carrier gas at 1 ml min −1 . Column temperature was maintained at 50°C for 3 min., and then raised to 270°C at 10°C min −1 . Electron impact ionization (EI) mass spectra were obtained at 70 eV with an Agilent Technologies 5973 mass selective detector (MS) interfaced with a 6890 N GC system equipped with either a 30 m×0.25 mm i.d.×0.25 μm film HP-5MS column, or one of the chiral columns described previously ). The HP-5MS column temperature was maintained at 50°C for 5 min, and then raised to 270°C at 10°C min −1 . Helium was used as a carrier gas at 1 ml min −1 . Thin layer chromatography (TLC) analyses were conducted on Whatman AL SIL G/UV plates using a 20 % ethanol solution of phosphomolybdic acid, and/or UV for visualization of compounds. Flash chromatography was carried out with 230-400 mesh silica gel (Fisher Scientific, Fair Lawn, NJ, USA).
Field Bioassays Pyramid traps of dimensions previously employed for native stink bugs (Hogmire and Leskey 2006;Leskey and Hogmire 2005) and for brown marmorated stink bugs (Leskey et al. 2012d) were used for all trials. Panels were constructed from Sintra® (partially extruded PVC) sheets (Laird Plastics, Pittsburgh, PA, USA) or plywood, and painted with flat black latex exterior paint, based on previous results indicating that adult and nymphal H. halys responded in greater numbers to this particular visual stimulus compared with other visual stimuli (Leskey et al. 2012d). Each panel was 1.22 m high, 52 cm wide at the base, and 7 cm wide at the top. Collection jars were constructed as per published dimensions (Hogmire and Leskey 2006) with an internal cone opening of 1.6 cm and trimmed wire edging to reduce escape. In addition, a Hercon Vaportape™ II (Hercon Environmental, Emigsville, PA, USA) was added as a killing agent to prevent escape of insects from traps. Traps were deployed~5 m from the border of apple and pear orchards at the Appalachian Fruit Research Station or in the border area between wood lots and row crops in Kearneysville, WV, USA, Shepherdstown, WV, USA, Keedysville, MD, USA, and at the Beltsville Agricultural Research Center, MD, USA. Traps were spaced 50 m apart and were baited with candidate compounds synthesized and impregnated into gray rubber septa or left unbaited as controls. In all trials, an unbaited trap was included and served as a control. Lures were changed every 2 weeks during a trial, as previous work showed that attraction to a septum decreases significantly after this point (D.W. unpublished data). Traps were checked once or twice per week during each trial, with all adults and nymphs counted and removed, and treatment positions rotated by weekly or twice weekly re-randomization. Trial location, timing and composition details are in Table 2.
Dose-dependent trials also were conducted with several of the lures, and these included an unbaited control trap in each set of replicates. Dose-dependent trials were conducted with non-pheromonal stereoisomer treatment #3 at 2, 8, and 50 mg during the mid-and late-season from 29 July -10 August and 12-16 September 2011, respectively, in Kearneysville, WV. Dose-dependent pheromone trials of stereoisomer treatment #6 were conducted at three locations: Shepherdstown, WV (17 June -12 July), Keedysville, MD (17 June -12 July), and Beltsville, MD (5 June -3 July) in 2012; loading rates included 0.1, 1, 10, and 100 mg. In 2013, #6 was combined with 66 mg of methyl (2E, 4E, 6Z)decatrienoate and tested at 10, 100, and 1000 mg of pheromone in soybean fields from 9-30 Aug 2013 in Jefferson County, WV. To achieve these higher doses, 10 of the 10 mg septa were combined by threading them with wire for the 100 mg treatment, and 100 of the 10 mg septa were combined for the 1000 mg #6 aggregation pheromone treatment. A range of doses above and below 200 mg was used, as this was the dose of pheromone used to trap native stink bugs in prior trapping studies (Krupke et al. 2001). Halyomorpha halys adults and nymphs were sampled only visually until 16 August, then with black pyramid traps (adults and nymphs), as described above, on a weekly basis. Nymphs and adults were pooled as a composite response variable from the visual sampling in soybean. The different replications were used as a random blocking factor. Because none of the response variables were normally distributed, the data were log-transformed to conform to the assumptions for an ANOVA. Residuals were inspected to ensure a mean variation around zero with homogenous variances. Pairwise comparisons between lures were conducted within a trial using Tukey's HSD. The dose-response trials for pheromonal and nonpheromonal stimuli were analyzed with linear regression, except for the soybean trials with #6+methyl (2E, 4E, 6Z)decatrienoate, which were analyzed as above. The capture of H. halys in traps was regressed against the concentration of chemical stimulus, with the latter as the independent variable. Because the data did not conform to a normal distribution, they were log-transformed prior to analysis. Regressions were performed separately across sites due to varying absolute numbers of H. halys, and adjusted R 2 -values and P-values were calculated for each regression. All statistical analyses were carried out in the program JMP Genomics v. 5.0 (SAS Institute, Inc. 2010), and we specified α=0.05 for all tests.
Dose-Dependent Trials For #3, a stereoisomer not produced by H. halys, a dose-dependent response for nymphs (Adjusted R 2 =0.94, P<0.001) and adults (Adj. R 2 =0.98, P<0.002; Fig. 5a) was observed in traps baited with 2, 8, and 50 mg compared with unbaited traps in a trial conducted during the mid-season. However, in the late season, although adults continued to exhibit a dose-dependent response to lure #3 (Adj. R 2 =0.88, P<0.05; Fig. 5b), very few nymphs were present in the field, and a dose-dependent response was not observed (Adj. R 2 =0.03, P=0.68).

Discussion
Thus, it appears that the presence of non-pheromonal isomers of 10,11-epoxy-1-bisabolen-3-ol were not antagonistic or inhibitory. The pheromone of Euschistus heros (F.) was identified as comprising three components, including methyl 2,6,10-trimethyldecanoate, the main component attracting females (Borges et al. 2011). There are eight stereoisomers of methyl 2,6,10-trimethyldecanoate, and bioassays revealed that the SRS isomer was the most attractive and that RSS alone appeared to be repellent. However, when presented as a racemic mixture, there appeared to be no antagonistic or inhibitory effect on overall attractiveness (Borges et al. 1998(Borges et al. , 2011. While no repellent isomers were identified for any of the compounds tested in this study, we likewise observed that racemic  Table 1 for identities of compounds  Table 1 for identities of compounds mixtures of 10,11-epoxy-1-bisabolen-3-ol were broadly attractive to both H. halys adults and nymphs. The lack of a behavioral response by H. halys to unnatural isomers is in contrast to other taxa, particularly Lepidoptera, where specific stereoisomers produced by one species can have a potent antagonistic effect on another (Leal et al. 2005;, thereby requiring high degrees of purity in pheromone formulations used for monitoring . Fortunately, this does not appear to be the case with H. halys. Indeed, when lures formulated with crude mixtures containing the two pheromone components in equal amounts and a range of other (7R)-10,11-epoxy-1bisabolen-3-ols were compared against lures with purified formulations, catches among the mixtures were similar, but greater than in unbaited controls. Indeed, traps baited with mixtures containing all 16 stereoisomers of 10,11-epoxy-1bisabolen-3-ol (#13, #14, produced from racemic citronellal) were also attractive to H. halys. Thus, our work has important practical implications for the commercial production of H. halys pheromone. It should be possible to start the synthesis with relatively cheap racemic citronellal instead of (R)citronellal. As a consequence, it may be possible to avoid a costly synthesis while retaining attractive potency. Certainly, these findings will enable companies to formulate pheromonebased products more easily for monitoring and management of H. halys.    Table 1 for identities of compounds Interestingly, some (S)-citronellal-derived isomers of 10, 11-epoxy-1-bisabolen-3-ol not produced by H. halys were attractive, signifying that some stereoisomers of (7S)-10,11-epoxy-1-bisabolen-3-ol are sufficiently similar to the true pheromone to trigger appropriate behavioral responses. In particular, captures in traps baited with #3 (comprised of four cisand four trans-epoxybisabolenols) were greater than unbaited traps in nearly all trials conducted throughout the season. To a lesser extent, traps with #2 were consistently more attractive to H. halys adults in WV, although this was not the case at MD field sites, likely due to the lower overall abundance of adults. In contrast, the harlequin bug, Murgantia histrionica (Hahn), a species with a male-produced pheromone comprising a 1.4:1 ratio of SSRS and SSRR stereoisomers of 10,11-epoxy-1bisabolen-3-ol , was not attracted to other non-pheromonal 10,11-epoxy-1-bisabolen-3-ol stereoisomers tested ). Attraction of insects to stereoisomers of the pheromone is rare. The sex pheromone of the pink hibiscus mealybug, MAconellicoccus hirsutus, comprises a 1:5 mixture of (R)-lavandulyl-(S)-2-methylbutanoate and (R)-maconellyl-(S)-2-methylbutanoate. However, it has been demonstrated (Zhang and Amalin 2005;Zhang et al. 2006) that a mixture of the two components with (S,S)-configurations also was somewhat active. Thus, the (S)-configuration of the acid moiety elicits attraction and the (R)-configuration induces inhibition, although there was some degree of tolerance toward chirality in the alcohol portion . Further research will focus on finding which stereoisomer(s) of 10,11-epoxy-1-bisabolen-3-ol with 7S configuration are attractive to H. halys.
Cross-attraction of H. halys to methyl (2E, 4E, 6Z)decatrienoate, the pheromone of Plautia stali has been well documented (Aldrich et al. 2007;Khrimian et al. 2008). Chinavia hilaris Say [formerly Acrosternum hilare (Say)], the green stink bug, is also cross-attracted to this compound (Aldrich et al. 2009). The purpose of cross-attraction in pheromones may include aiding the receiver to locate host plants, dispersing to overwintering sites, and/or benefiting from density-dependent protection from natural enemies (for a more extensive discussion, see Weber et al. 2014a). However, for H. halys, there are limitations in terms of the usefulness of this compound, as adults are only attracted to it late in the season, while nymphs are attracted all season (Leskey et al. 2012d). With this caveat, methyl (2E, 4E, 6Z)decatrienoate has proven to be a powerful synergist for the two-component H. halys aggregation pheromone throughout the season . Even in the limited trials conducted here, the effect of combining the pheromone and methyl (2E, 4E, 6Z)-decatrienoate resulted in adult captures that were at least 20 and 2-4 times greater than captures in unbaited traps and in traps baited with the pheromone alone, respectively. Thus, using the two-component aggregation pheromone in combination with methyl (2E, 4E, 6Z)-decatrienoate is one way to increase attractiveness and sensitivity of lures for monitoring H. halys populations.
A second method to enhance captures in traps is by increasing the amount of material or release rate of the lures. In trials conducted here, we observed increased responses to increasing amounts of the pheromone alone or in combination with methyl (2E, 4E, 6Z)-decatrienoate (Figs. 6 and 7). In previous experiments, a strong response to increasing amounts of methyl (2E, 4E, 6Z)-decatrienoate formulated into lures also was observed (Leskey et al. 2012d). In our study, we also reported increased dose-responses to #3, a chemical not produced by H. halys. Thus, there is potential for other non-pheromonal compounds to enhance monitoring and/or behavioral manipulation and management of H. halys.
Significant advances have been made toward effective pheromone-based monitoring of H. halys. These include the identification of the pheromone ) and methyl (2E, 4E, 6Z)-decatrienoate, a powerful synergist . Additionally, black pyramid traps provide a good visual stimulus and collection jars provisioned with a killing agent retain greater numbers of individuals (Leskey et al. 2012c). Previous research has shown that a combination of methyl (2E, 4E, 6Z)-decatrienoate and H. halys aggregation pheromone is able to detect H. halys populations reliably in both low and high density situations across the United States throughout the active growing season (Leskey et al. 2015). Overall, our work builds on the state-ofthe-art by contributing detailed information about the range of pheromonal and non-pheromonal cues to which H. halys responds, the purity required for significant behavioral responses, and documentation of dose-dependent responses to pheromonal and non-pheromonal stimuli. We hope this information will help lead to: 1) improved routes for synthesizing pheromones in a more cost-effective manner; 2) future research into identifying other non-pheromonal synergists to attract H. halys; and 3) investigation of attract-and-kill approaches for controlling H. halys populations while simultaneously reducing insecticide inputs.
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.