Volatile Emissions and Relative Attraction of the Fungal Symbionts of Tea Shot Hole Borer (Coleoptera: Curculionidae)

Euwallacea perbrevis is an ambrosia beetle that vectors fungal pathogens causing Fusarium dieback in Florida avocado trees. Current monitoring lures contain quercivorol, a fungus-produced volatile, but the exact attractant is unknown since lures contain a mixture of p-menth-2-en-1-ol isomers and both α- and β-phellandrene. This study used pure cultures of six symbiotic fungi isolated from E. perbrevis to document volatile emissions and determine the relative attraction of symbionts in binary choice assays. In a comparative test, headspace solid-phase microextraction followed by gas chromatography–mass spectroscopy was used to identify and quantify emissions from 3-week-old cultures. In a temporal study, Super-Q collection followed by gas chromatography–flame ionization detection was used to measure cis- and trans-p-menth-2-en-1-ol emissions for three months. A total of 15 compounds were detected, with monoterpene hydrocarbons and oxygenated monoterpenoids predominating. Only trans-p-menth-2-en-1-ol was common to all six symbionts. Peak levels of both isomers were observed at day 7, then gradually declined over a 90 day period. In choice tests, avocado sawdust disks inoculated with Fusarium sp. nov. were the most attractive. This symbiont produced only two volatiles, trans-p-menth-2-en-1-ol and limonene. The combined results indicate that trans-p-menth-2-en-1-ol is the primary female attractant emitted from symbiotic fungi, but limonene may be a secondary attractant of E. perbrevis.


Introduction
Shot hole borers in the genus Euwallacea (Coleoptera: Curculionidae: Scolytinae) are Asian ambrosia beetles that comprise a cryptic species complex newly invasive in the continental United States and in other countries. Species established in the USA include the tea shot hole borer (E. perbrevis Schedl) in Florida, and the polyphagous shot hole borer (E. fornicatus Eichhoff) and Kuroshio shot hole borer (E. kuroshio Gomez and Hulcr) in California [1]. Typical of ambrosia beetles, females are wood borers that colonize host trees in which they cultivate symbiotic fungi as a nutritional source for larvae and adults [2]. The exotic fungi vectored by Euwallacea spp. are phytopathogenic and induce Fusarium dieback, a destructive vascular disease of avocado, woody ornamentals, and numerous native American trees [3,4].

Volatile Extraction
Mycelial cultures on PDA Petri dishes of A. murorum, Acremonium sp., AF-6, AF-8, Fusarium sp. nov. and Graphium sp. were sampled for volatile emissions when 24-25 days old. Collection of volatile compounds was performed using a SPME fiber coated with 50/30 µm divinylbenzene/carboxen on polydimethylsiloxane (DVB/CAR/PDMS). Each fungal culture was placed in a Pyrex cylindrical jar then sealed with plastic film and left at 25 • C for 30 min for equilibration of volatiles in the headspace. After equilibration, the film was pierced with the SPME needle, and the fiber exposed to the sample headspace for 1 h at 25 • C. After the extraction, the fiber was immediately inserted into the injector port of the GC-MS for desorption 2 min in the splitless mode.

Gas Chromatography-Mass Spectrometry Analysis
Volatile emissions were analyzed by GC-MS Agilent 7890B GC coupled with 5977B mass selective detector (Agilent Technologies, Santa Clara, CA, USA). A DB-5 column (30 m × 0.25 mm inner diameter with 0.25 µm film thickness) was used and GC oven temperature program was 60 • C for 1.3 min, rose at 3 • C/min up to 246 • C. The injector and detector temperatures were kept at 220 • C and 230 • C. Helium was used as a carrier gas at a flow rate of 1.3 mL/min. Mass spectra were recorded at 70 eV. Mass range was m/z 35 to 450 Da and the scan rate was 2.8 scans/s. Mass Hunter B.07.06 software (Agilent Technologies) was used for data acquisition and processing. Volatile compounds emitted by symbionts were identified by comparing retention times, retention indices (RI) calculated using the Van den Dool and Kratz [20] equation in relation to a homologous series of n-alkanes (C 9 -C 21 ) and libraries MassFinder [21], Adams Library [22], Flavours and Fragrances of Natural and Synthetic Compounds 3 [23], and Wiley 12/NIST 2020 [24], and in-house library "SHRS Essential Oil Constituents-DB-5" which was built up from authentic standards and components of known essential oils [5,7,16]. Results expressed as the relative percentage of each compound peak area to the total GC-MS peak area. Compounds released from Petri plates without fungal culture and siloxanes in procedural blanks as background contamination produced by SPME fiber were subtracted. All sample preparations and analyses were independently performed in triplicate.

Collection and Quantification of p-Menth-2-en-1-ols
Volatile emissions of cis-p-menth-2-en-1-ol and trans-p-menth-2-en-1-ol from Graphium sp. cultures on PDA Petri plates (four replicates) were sampled over the course of three months every two or three days using Super-Q adsorbent (Analytical Research Systems, Gainesville, FL, USA). Cultures were stored for the duration of this study at 25 • C in an incubator (Shel lab, model SRI20PF, Sheldon Mfg. Inc., Cornelius, NC, USA). Super-Q volatile collections were conducted as described previously [16] but using a slightly modified version. Small oven bags (Reynolds, Richmond, VA, USA) were used for the collection of volatile emissions from each Petri plate, using two port holes to connect to the volatile collection system. Filtered air was introduced at 1 L/min through one port hole; a Super-Q trap collector attached to a vacuum line pulling at 1 L/min was set up on the other port hole (push-pull system). Collection of volatiles was timed for 15 min. Volatile emissions were eluted from the Super-Q adsorbent using 200 µL methylene chloride (#75-09-2, Sigma-Aldrich, St. Louis, MO, USA) and an internal standard equivalent to 5 µg of For quantitative analysis, samples were analyzed using a 7890B Agilent GC with a flame ionization detector (FID) and a split-splitless injector. The column was an Agilent DB-5 30 m × 0.25 mm × 0.25 µm (Agilent Technologies, Santa Clara, CA, USA). Injector temperature was 220 • C; split ratio was 20:1. The carrier gas was helium with a total flow of 27.2 mL/min and a pressure of 14.4 psi. The oven temperature was programmed at 60 • C for 1.3 min increasing at 3 • C/min to reach a final temperature of 246 • C. The FID was set to 250 • C. Extracts were quantified by GC-FID and analyzed by GC-MS for comparison purposes only.

Experimental Insects
Mature E. perbrevis females were obtained from laboratory colonies maintained at TREC, using rearing methods described previously by Cruz et al. [25]. Briefly, E. perbrevis females were excavated from avocado logs collected from an orchard in Miami-Dade County, Florida, USA (25 • 31 31" N, 80 • 29 7" W). An avocado sawdust rearing medium was prepared and poured into 50 mL conical centrifuge tubes [19]. Females were surfacedisinfested with 70% (v/v) ethanol for 5-7 s and individually placed into the tubes. Rearing tubes were capped with lids fitted with a 1 cm metallic screen opening for airflow and maintained at 25 • C and 75% relative humidity, with a 16:8 L:D photoperiod.

Laboratory Bioassays
Dual-choice bioassays were conducted in shallow plastic containers (21 cm long × 15 cm wide × 8 cm high) lined with white paper (Gordon Paper Co., Inc., Virginia Beach, VA, USA; BWKB18401000) to provide a rough surface upon which the beetles could walk, comparable to test arenas described previously [26]. Avocado sawdust disks inoculated with Fusarium sp. nov., AF-8 or Graphium sp. were placed on opposite sides of the container. A plastic vial containing six E. perbrevis females (≤7 days post-emergence) was placed at the center of each container. After a 5 min acclimation period, the vial was removed to allow the females to interact with the inoculated avocado sawdust disks. The container was then covered with a fitted lid with two screened ventilation holes and kept in a walk-in-rearing room at 25 • C in complete darkness. The number of beetles on, in or next to each avocado sawdust disk was recorded at one time point after 12 h. Beetles that did not choose a given treatment (not in, on or next to an avocado sawdust disk) were excluded from the analysis. Each combination of symbionts was replicated six times, reversing the position of the symbionts, and using new cohorts of beetles each time.

Statistical Analysis
The data obtained from GC-MS were subjected to multivariate analysis. Principal component analysis (PCA) was performed using XLSTAT 2021 (Addinsoft, New York, NY, USA). Pearson correlation was adopted. A hierarchical cluster analysis (HCA) was performed using JMP Pro. 16.0 software (SAS Institute Inc., Cary, NC, USA) using Ward's clustering method. Regression analysis with exponential decay models was used to analyze emissions of cisand trans-p-menth-2-en-1-ol from the Graphium sp. cultures; analysis by t-test was used to assess for differences in emissions of the two isomers at each sampling point (SigmaPlot 14.0, Systat Software Inc., San Jose, CA, USA). The results of the dualchoice bioassays were analyzed using a Kruskal-Wallis non-parametric test (SAS Institute Inc., Cary, NC, USA).
symbiont AF-8 than in Acremonium sp., AF-6 and Graphium sp. (Figure 2). Regarding the components present in the commercial quercivorol lure [16], trans-p-menth-2-en-1-ol was the only volatile detected across all six symbionts. cis-p-Menth-2-en-1-ol was detected in the emissions from Graphium sp. and both Acremonium spp. but was absent in the emissions from AF-6, AF-8, and Fusarium sp. nov. α-Phellandrene was detected in the emissions from A. murorum and Graphium sp., while β-phellandrene was emitted from cultures of A. murorum, AF-8, and Graphium sp. * RI exp: retention indices calculated on DB-5 column; ** RI lit: retention indices from Adams Library [22]; relative concentration was expressed as the peak area and data listed were the mean of three injection results ± standard error (SE); a compound identified using retention time of authentic standard, matching with MS libraries and comparison with reported RI; b compound identified by matching with MS library and comparison with reported RI.
symbiont AF-8 than in Acremonium sp., AF-6 and Graphium sp. (Figure 2). Regarding the components present in the commercial quercivorol lure [16], trans-p-menth-2-en-1-ol was the only volatile detected across all six symbionts. cis-p-Menth-2-en-1-ol was detected in the emissions from Graphium sp. and both Acremonium spp. but was absent in the emissions from AF-6, AF-8, and Fusarium sp. nov. α-Phellandrene was detected in the emissions from A. murorum and Graphium sp., while β-phellandrene was emitted from cultures of A. murorum, AF-8, and Graphium sp. symbiont AF-8 than in Acremonium sp., AF-6 and Graphium sp. (Figure 2). Regarding the components present in the commercial quercivorol lure [16], trans-p-menth-2-en-1-ol was the only volatile detected across all six symbionts. cis-p-Menth-2-en-1-ol was detected in the emissions from Graphium sp. and both Acremonium spp. but was absent in the emissions from AF-6, AF-8, and Fusarium sp. nov. α-Phellandrene was detected in the emissions from A. murorum and Graphium sp., while β-phellandrene was emitted from cultures of A. murorum, AF-8, and Graphium sp. * RI exp: retention indices calculated on DB-5 column; ** RI lit: retention indices from Adams Library [22]; relative concentration was expressed as the peak area and data listed were the mean of three injection results ± standard error (SE); a compound identified using retention time of authentic standard, matching with MS libraries and comparison with reported RI; b compound identified by matching with MS library and comparison with reported RI.
symbiont AF-8 than in Acremonium sp., AF-6 and Graphium sp. (Figure 2). Regarding the components present in the commercial quercivorol lure [16], trans-p-menth-2-en-1-ol was the only volatile detected across all six symbionts. cis-p-Menth-2-en-1-ol was detected in the emissions from Graphium sp. and both Acremonium spp. but was absent in the emissions from AF-6, AF-8, and Fusarium sp. nov. α-Phellandrene was detected in the emissions from A. murorum and Graphium sp., while β-phellandrene was emitted from cultures of A. murorum, AF-8, and Graphium sp. * RI exp: retention indices calculated on DB-5 column; ** RI lit: retention indices from Adams Library [22]; relative concentration was expressed as the peak area and data listed were the mean of three injection results ± standard error (SE); a compound identified using retention time of authentic standard, matching with MS libraries and comparison with reported RI; b compound identified by matching with MS library and comparison with reported RI. symbiont AF-8 than in Acremonium sp., AF-6 and Graphium sp. (Figure 2). Regarding the components present in the commercial quercivorol lure [16], trans-p-menth-2-en-1-ol was the only volatile detected across all six symbionts. cis-p-Menth-2-en-1-ol was detected in the emissions from Graphium sp. and both Acremonium spp. but was absent in the emissions from AF-6, AF-8, and Fusarium sp. nov. α-Phellandrene was detected in the emissions from A. murorum and Graphium sp., while β-phellandrene was emitted from cultures of A. murorum, AF-8, and Graphium sp. * RI exp: retention indices calculated on DB-5 column; ** RI lit: retention indices from Adams Library [22]; relative concentration was expressed as the peak area and data listed were the mean of three injection results ± standard error (SE); a compound identified using retention time of authentic standard, matching with MS libraries and comparison with reported RI; b compound identified by matching with MS library and comparison with reported RI.
* RI exp: retention indices calculated on DB-5 column; ** RI lit: retention indices from Adams Library [22]; relative concentration was expressed as the peak area and data listed were the mean of three injection results ± standard error (SE); a compound identified using retention time of authentic standard, matching with MS libraries and comparison with reported RI; b compound identified by matching with MS library and comparison with reported RI.
A hierarchical cluster analysis (HCA) dendrogram categorized the symbionts into three main clusters (Figure 4). Cluster I and II represented A. murorum and Graphium sp., respectively. Cluster III grouped Acremonium sp., AF-6, AF-8 and Fusarium sp. nov, confirming the results from PCA. PCA and HCA results revealed that cis-p-menth-2-en-1-ol and trans-p-menth-2-en-1-ol are the key components for discriminating effect of symbionts. This is the first report to compare chemical profiles of the volatile emissions in the six fungal species studied. tolochene) and 14 (valencene) are grouped in the lower left associated with Acremonium sp. and the three Fusarium species (AF-6, AF-8 and Fusarium sp. nov.). Aristolochene was first isolated from Aspergillus terreus, and the headspace profile of Penicillium roqueforti and P. digitatum cultures were found to be rich in (+)-aristolochene (up to 80%) and valencene (11%) [27].   Table 1.  Table 1.
A hierarchical cluster analysis (HCA) dendrogram categorized the symbionts into three main clusters (Figure 4). Cluster I and II represented A. murorum and Graphium sp., respectively. Cluster III grouped Acremonium sp., AF-6, AF-8 and Fusarium sp. nov, confirming the results from PCA. PCA and HCA results revealed that cis-p-menth-2-en-1-ol and trans-p-menth-2-en-1-ol are the key components for discriminating effect of symbionts. This is the first report to compare chemical profiles of the volatile emissions in the six fungal species studied. The temporal pattern of emissions for cis-and trans-p-menth-2-en-1-ol from the Graphium sp. cultures is depicted in Figure 5. Based on our experimental design, peak emissions of both compounds were detected at day 7 (although highest levels may have

cis-and trans-p-Menth-2-en-1-ol Emissions from Graphium sp.
The temporal pattern of emissions for cisand trans-p-menth-2-en-1-ol from the Graphium sp. cultures is depicted in Figure 5. Based on our experimental design, peak emissions of both compounds were detected at day 7 (although highest levels may have occurred slightly before or after that timepoint since measurements were not recorded daily). Following that peak, emissions began to decline gradually over a period of 90 days. The decrease in emissions over time was best fit by regression with exponential decay models (cis: y = 2.59e (−0.05x) , R 2 = 0.924; trans: y = 1.34e (−0.04x) , R 2 = 0.927). Emissions of cis-p-menth-2-en-1-ol were higher than trans-p-menth-2-en-1-ol for the first 30 days, but the difference was significant only on day 10 (t = 11.34, df = 2, p = 0.004) and day 20 (t = 3.24, df = 2, p = 0.042).

cis-and trans-p-Menth-2-en-1-ol Emissions from Graphium sp.
The temporal pattern of emissions for cis-and trans-p-menth-2-en-1-ol from the Graphium sp. cultures is depicted in Figure 5. Based on our experimental design, peak emissions of both compounds were detected at day 7 (although highest levels may have occurred slightly before or after that timepoint since measurements were not recorded daily). Following that peak, emissions began to decline gradually over a period of 90 days. The decrease in emissions over time was best fit by regression with exponential decay models (cis: y = 2.59e (−0.05x) , R 2 = 0.924; trans: y = 1.34e (−0.04x) , R 2 = 0.927). Emissions of cis-pmenth-2-en-1-ol were higher than trans-p-menth-2-en-1-ol for the first 30 days, but the difference was significant only on day 10 (t = 11.34, df = 2, p = 0.004) and day 20 (t = 3.24, df = 2, p = 0.042).

Laboratory Bioassays
Euwallacea perbrevis females were significantly more attracted to Fusarium sp. nov. than to AF-8. When given a choice between AF-8 and Graphium sp., females were significantly more attracted to AF-8, but no significant differences were observed when females were offered the Fusarium sp. nov. and Graphium sp. (p = 0.12) ( Figure 6).

Laboratory Bioassays
Euwallacea perbrevis females were significantly more attracted to Fusarium sp. nov. than to AF-8. When given a choice between AF-8 and Graphium sp., females were significantly more attracted to AF-8, but no significant differences were observed when females were offered the Fusarium sp. nov. and Graphium sp. (p = 0.12) ( Figure 6). trans-p-Menth-2-en-1-ol, detected in the volatiles of all three test fungi, is the main component of the quercivorol lure that is attractive to E. perbrevis [13,16] and other Euwallacea species [28]. The cis-isomer, also present in the quercivorol lure, was only detected in the emissions from Graphium sp. cis-p-Menth-2-en-1-ol did not increase the attraction of E. perbrevis to Graphium sp., suggesting that the trans-isomer is the primary attractant of E. perbrevis and possibly of other Euwallacea spp. Furthermore, trans-p-menth-2-en-1-ol was trans-p-Menth-2-en-1-ol, detected in the volatiles of all three test fungi, is the main component of the quercivorol lure that is attractive to E. perbrevis [13,16] and other Euwal-lacea species [28]. The cis-isomer, also present in the quercivorol lure, was only detected in the emissions from Graphium sp. cis-p-Menth-2-en-1-ol did not increase the attraction of E. perbrevis to Graphium sp., suggesting that the trans-isomer is the primary attractant of E. perbrevis and possibly of other Euwallacea spp. Furthermore, trans-p-menth-2-en-1-ol was the only volatile compound detected in the emissions from all six symbiont cultures (Table 1). Under natural conditions, the fungal gardens of E. perbrevis would consist of mixed cultures of all six symbionts, which would emit a strong 'bouquet' of trans-p-menth-2-en-1-ol.
The two minor components in the commercial lure [16], αand β-phellandrene, were also detected in the emissions from fungal symbionts. α-Phellandrene was present in emissions from the Graphium sp. cultures, and β-phellandrene was emitted by both Graphium sp. and AF-8; however, neither compound appeared to increase attraction of female E. perbrevis in choice assays. (−)-β-Phellandrene is the predominant monoterpene in the phloem tissue of lodgepole pine (Pinus contorta Douglas) and a known kairomone attractant for several species of bark beetle, including the economically important pine engraver, Ips pini (Say) [29].
Graphium sp. was the least attractive symbiont despite emissions of trans-p-menth-2-en-1-ol. This suggests that other volatiles from Graphium sp. may include potential repellents. For instance, cis-p-menth-2-en-1-ol and trans-piperitol are known to have repellant activity against Musca domestica L. and Haematobia irritans L. [30,31], and terpinen-4-ol has insecticidal and repellent activities against three stored-product insects: Tribolium castaneum Herbst, Lasioderma serricorne Fabricius, and Liposcelis bostrychophila Badonnel [32]. However, other volatiles produced by Graphium sp., like p-cymene and β-elemene, are involved in the attraction of X. glabratus to Lauraceous trees [33]. More research is needed to understand the effects of these and other volatiles from Graphium sp. as potential attractants or repellants of E. perbrevis. Identification of effective repellents would facilitate development of a pest management strategy using a 'push-pull' approach; a repellent applied to host trees would push pests away, while surrounding traps baited with attractant (i.e., the combination of trans-p-menth-2-en-1-ol and α-copaene [5]) would reduce numbers of host-seeking females [34].
Fusarium sp. nov. was the most attractive amongst the tested fungi. Besides trans-pmenth-2-en-1-ol, the only other volatile detected from Fusarium sp. nov. was limonene. Therefore, limonene is likely a secondary attractant of E. perbrevis, not present in the other two symbionts. Limonene is a known attractant kairomone of another scolytine species, the white pinecone beetle, Conophthorus coniperda Schwarz [35]. Limonene is also a precursor in the synthesis of trans-p-menth-2-en-1-ol [18]. Given the structural similarities, limonene and trans-p-menth-2-en-1-ol may be detected by the same antennal olfactory receptors, but additional electrophysiological studies [5,36] are needed to address that question.

Conclusions
Euwallacea perbrevis, a new invasive ambrosia beetle in Florida, maintains a complex association with multiple species of symbiotic fungi, some of which induce Fusarium dieback in host trees, including avocado. This study utilized pure cultures of six symbionts to identify and quantify their volatile emissions, and to determine the attractive components within those emissions. Chemical analyses indicated that trans-p-menth-2-en-1-ol was the only compound produced by all six symbionts; behavioral assays indicated that emissions of cis-p-menth-2-en-1-ol did not increase attraction. The most attractive symbiont, a new species of Fusarium, also emitted limonene, implicating this compound as a secondary attractant. Additional laboratory bioassays and electroantennographic analyses are needed with female E. perbrevis to assess limonene for behavioral responses and olfactory chemoreception, respectively. The results presented here may allow improving lures for field monitoring of E. perbrevis and possibly other invasive Euwallacea species.