Newly Discovered Components of Dendrolimus pini Sex Pheromone

Simple Summary Larvae of the pine-tree lappet moth (Dendrolimus pini) feed on needles of pine trees in Europe and Asia. During outbreaks, they can massively defoliate pine forests. The discovery of unknown components of D. pini sex pheromone opens possibilities for optimizing the lures for trapping D. pini males and increasing their use in insect monitoring systems. Abstract The pine-tree lappet moth, D. pini, is a harmful defoliator of pine forests in Europe and Asia and a potentially invasive species in North America. The lures for trapping D. pini males based on two known components of its sex pheromone appeared weakly attractive to male moths. Identification of all components of the sex pheromone might allow for the development of more effective lures. The pheromone was sampled from virgin females using SPME and analyzed using gas chromatography coupled with mass spectrometry. Four new likely components ((Z5)-dodecenal, (Z5)-dodecen-1-ol, (Z5)-decen-1-yl acetate, (Z5)-tetradecen-1-yl acetate) and two known components ((Z5,E7)-dodecadienal, (Z5,E7)-dodecadien-1-ol) were identified based on comparison against authentic standards, Kováts indices and spectra libraries. The samples also contained several sesquiterpenes. Wind tunnel and field experiments showed that some blends of synthetic pheromone components alone or enriched with Scots pine essential oil (SPEO) were attractive to D. pini males. One component—(Z5)-decen-1-yl acetate—had a repelling effect. The presented knowledge of D. pini sex pheromone provides a basis for developing optimal lures for monitoring or controlling insect populations.


Introduction
The pine-tree lappet moth, Dendrolimus pini L. (Lepidoptera, Lasiocampidae), inhabits pine forests of Europe and Asia, from the United Kingdom to Northern China and Far-East Russia [1,2]. It is considered a potential invasive insect in North America [3]. The insect prefers Scots pine Pinus sylvestris L. as a host tree but has also occurred on other pine species, firs, spruces, cedars, junipers, and hemlocks [1,3,4]. It is one of the significant Scots pine defoliators in many countries, including Poland and Germany [2]. Periodic large outbreaks of D. pini have occurred over centuries [5]. In Poland, the outbreak in 2011-2014 was the largest one in the previous 70 years and covered about 184,000 hectares. Trees highly defoliated by D. pini larvae become less vital and more vulnerable to secondary pests and environmental stress like rising groundwater. The early detection of increasing populations of insects is crucial for effective outbreak prevention. The control measures are usually conducted by applying available registered insecticides.
Dendrolimus pini moths fly and mate mainly from mid-July to mid-August [6]. Larvae start feeding at the turn of summer and autumn in the same year. In the late autumn, usually in November, they descend to the forest litter for overwintering. Larvae resume feeding in early spring and pupate in tree crowns or on trunks at the end of June. The primary method of assessing the D. pini population density in Poland is the autumnal counting of overwintering larvae in forest litter [7]. The dominance structure of hibernating larval instars may depend on the outbreak phase. For instance, young instars prevailed at the outbreak's beginning, exceeding 80% of the total larval population (Sukovata L., unpublished data). Unfortunately, the cryptic coloration makes it hard to recognize young larvae in the Scots pine forest litter. Therefore, the counters often overlook the larvae and underestimate the population densities at the initial outbreak phases. Thus, a new method for the early detection of D. pini at low and increasing population densities is urgently needed. A promising solution is applying sex-pheromone traps to estimate the abundance of male moths [8].
Early studies of the pheromone composition and corresponding field experiments with baited traps showed that (Z5,E7)-dodecadienal ((Z5,E7)-12:Ald) [9,10] and (Z5,E7)dodecadien-1-ol ((Z5,E7)-12:OH) occurred in the sex pheromone of D. pini [11]. The catches achieved in various countries with traps baited either with (Z5,E7)-12:Ald or with mixtures of (Z5,E7)-12:Ald and (Z5,E7)-12:OH were rather low and did not exceed one male per day and trap [10,[12][13][14][15]. Light traps of the Jalas type, used without the pheromones, caught up to seven males per day per trap [14]. Low catches in the pheromone traps could result from several factors, including the incomplete composition of sex-pheromone lures, which were insufficiently attractive to the insects, too low or too high loads of the pheromone in lures, and low densities of the D. pini populations studied. Therefore, we aimed to (a) unveil the complete composition of the sex pheromone of D. pini, (b) check if its synthetic equivalent is effective in trapping males, and (c) explore the importance of each pheromone component in mate finding. The work scope ranged from rearing the insects in the laboratory and chemical analyses of emissions of calling females through wind tunnel evaluation of male response to various compounds and their blends to final field testing of selected combinations in forest stands inhabited by D. pini.

Insects
Larvae of D. pini were either collected in the field or obtained from eggs laid by females in a laboratory at the Forest Research Institute (FRI). The larvae were placed on fresh Scots pine twigs in ventilated rectangular boxes 45 × 30 × 23 cm in size and fed until the pupal stage. The cut ends of the twigs were dipped in plastic cups filled with water. When necessary, the twigs were replaced with new ones.
Pupae were extracted from cocoons, sexed by visual inspection of genital regions, moved to a laboratory at the Institute of Physical Chemistry (IPC), and housed individually in 150 mL glass beakers. The beakers were closed with perforated aluminum foil covers and equipped with cylindrical mesh supports made of acid-resistant steel to aid the emerging moths. Male and female pupae were kept separately in ventilated glass terraria shaded with black cardboard and equipped with automatic LED lighting, which simulated day and night (L17:D7) with one-hour-long dawn and dusk periods. For convenience, we inverted the diel cycles of the insects so that the scotophases started at 9 a.m. When handling and sampling the insects, we used red diode light within the wavelength range of 620-630 nm, invisible to D. pini moths [16]. Terraria with male and female insects were stored in separate rooms at a controlled temperature (23 ± 2 • C) and relative humidity (50-70%).

Collection of Female Volatiles
We collected the volatile emissions from single, calling virgin females 0 to 3 days old under the red light directly in the beakers they were kept in. Usually, the females started to call about half an hour into the scotophase and continued with short breaks until dawn. We used the Solid Phase MicroExtraction (SPME) fibers with polydimethylsiloxane (PDMS) or polydimethylsiloxane-divinylbenzene (PDMS-DVB) coating, mounted in standard holders (Supelco, Sigma Aldrich, St. Louis, MO, USA). The fibers were 1 cm long and consisted of fused silica supports and PDMS coatings 100 µm thick or PDMS-DVB coatings 65 µm thick. Before sampling, the SPME fibers were conditioned for 30 min at 250 • C in a gas chromatograph inlet with a gentle stream of helium. The holders with conditioned or sample-loaded fibers were stored separately in glass containers filled with argon. When a female extended its abdominal tip, we inserted an SPME needle into the beaker and positioned the fiber as close as possible to the tip, usually 3-5 mm away (Supplementary Materials, Figure S1). We determined that the sampling time of 15 min was sufficient to obtain good chromatograms. We also sampled a few non-calling females and males for comparison.

Chemical Analyses of Female Emissions
The volatiles collected were analyzed using a Trace 1300 gas chromatograph coupled to an ITQ700 ion-trap mass spectrometer with an EI ion source (Thermo Scientific, Waltham, MA, USA). A polar ZBWAX capillary column (30 m × 0.25 mm ID, 0.25 µm film thickness, Zebron, Anaheim, CA, USA, Phenomenex, Torrance, CA, USA) was used. Helium (99.9999%, Air Liquide) with a constant flow of 1 mL/min was a carrier gas. The temperature program started at 60 • C, held for 5 min, then increased by 10 • C/min to 250 • C and stayed at 250 • C for 5 min. The temperature of the transfer line was 250 • C. Each SPME fiber was inserted directly into the inlet of the gas chromatograph, where the analytes were desorbed for 3 min in a splitless mode. The ion source and the injector operated at a constant temperature of 250 • C. The ionization energy was always 70 eV. The scanning covered the entire available range (50-650 amu). Components of female emissions were identified by their retention times and Kováts indices and by matching their mass spectra with those of authentic standards (either purchased or synthesized with the exception of unavailable α-muurolene standard). The isomers of each analyzed compound eluted with different retention times ( Figure S5, Supplementary Materials). The structural elucidation of unknown emission components was also supported by applying the NIST and Wiley mass spectra libraries.
Kováts retention indices were calculated following the gradient GC/MS conditions used [17,18]. A mixture of C 7 -C 30 straight-chain hydrocarbons was used as a reference.
The Scots pine essential oil (SPEO) was obtained by steam distillation, described in the next section. A mixture of C 7 -C 30 straight-chain hydrocarbons used to determine Kováts retention indices were purchased from Sigma Aldrich (St. Louis, MO, USA). n-Hexane was used as a solvent to prepare lures (99% analytically pure, Sigma Aldrich, St. Louis, MO, USA). Mass spectra and nuclear magnetic resonance (NMR) data, which finger-printed the identity and purity of the likely components of sex pheromone, are available in the Supplementary Materials file.

Steam Distillation of SPEO in a Deryng Apparatus
Fresh Scots pine needles collected at the sites inhabited by D. pini were put into a round-bottom flask, then filled with distilled water and connected to a Deryng apparatus. The steam distillation lasted 3 h. The essential oil collected was dried using anhydrous sodium sulfate, filtered, and analyzed by GC/MS. The efficiency of the SPEO distillation was 1 mL per 180 g of fresh plant material.

Identification of SPEO Components by GC/MS
SPEO was analyzed using the GC/ITQ700 ion trap mass spectrometer. The analytical procedure was the same as for the analyses of female emissions except for the injection mode. The split injection ratio was 1:100. 1 µL (~0.8-0.9 µg) of pure SPEO was injected in GC/MS. The mass spectra of SPEO components were compared to the mass spectra of the authentic standards (except unavailable α-muurolene standard) and, additionally, to the NIST and Wiley libraries. The results were also compared with the literature data on the elution order of terpene-derived compounds [21][22][23][24]. The components firmly identified in the SPEO samples are listed in the Supplementary Materials.

Preparation of Lures and Dispensers
Stock solutions of lures of defined composition were prepared by mixing n-hexane solutions of selected components of the D. pini sex pheromone in glass bottles. The lures were dispensed from 1 mL polyethylene vials with push-on hinged caps and an outer diameter of 8 mm (Kartell, Italy), which appeared optimal for luring D. pini with mixtures of (Z5,E7)-12:Ald and (Z5,E7)-12:OH [25]. The vials were filled with various lure solutions, capped, and hung in a gentle stream of clean air at room temperature until no liquid was visible inside. Then, they were packed into alu-foil bags, thermally sealed, and stored at −22 • C in a freezer for further use. Table 1 shows the chemical compositions of the lures prepared for testing. We used the components in proportions similar to those in SPME samples, which, however, were not equivalent to the proportions in sampled emissions. For (Z5,E7)-12:Ald and (Z5,E7)-12:OH, we used the proportions published in the literature [11].
Field Expt 4 a Last row shows the sets of lures compared in four field experiments (Expt) described in the corresponding subsection; b Proportion of (Z5,E7)-12:Ald and (Z5,E7)-12:OH was adopted from Kovalev et al. [11].

Wind Tunnel
The experiments were carried out in a rectangular polycarbonate wind tunnel 2.5 m long × 0.6 m wide × 0.6 m high, constructed in-house (Supplementary Materials, Figure S2). A turbine blower provided airflow up to 1 m/s in an open or closed-loop set-up. Two activated-carbon filters mounted at the inlet and outlet of the tunnel cleaned the airstream. The tunnel was shaded with white foam boards and illuminated inside with red LED light (620-630 nm). The computer software continuously recorded air temperature, relative humidity, and airflow rate during each experiment. The test insects were placed on a take-off platform located at the outlet of the tunnel. The lure dispensers were hung on the polycarbonate mesh at the air-inlet. Three black-and-white video cameras recorded the behavior of insects at a speed of 35 frames per second.
We conducted all experiments during the insects' scotophase at room temperature (25-30 • C), relative humidity 30-40%, and airflow varied between 0.4-0.6 m/s. In most cases, the males were flying even when no bait was placed in the air stream. Interestingly, when we were preparing the traps for hanging in the forest in daylight, many moths appeared flying around.
In each experiment, we placed three males in the tunnel and offered pure airflow or a single lure in three vials hung across the tunnel inlet. If used, SPEO was applied in separate vials hung next to the lure vials. We recorded the insect behavior both as video films and in hand-written notes. The strongest reaction observed was attributed to one of the behavior categories R i defined in Table 2, along with arbitrary numerical weights w i promoting the active behavior of moths (see also Supplementary Materials, Figure S3). We chose the weights equal to (i − 1) 2 , but any ascending values, for instance, i, would also work. The SM includes a wmv file with a video recording illustrating the behavior category 7.  For each lure, the experiment was repeated several times (Tables S1 and S2 in the Supplementary Materials). The effectiveness factor f of the lure was calculated using Equation (1): where: n i is the number of experiments in which the behavior R i was observed, and Σ i n i is the total number of experiments for the given lure. The lures were compared directly by their effectiveness factors. In all experiments, the vial dispensers loaded with the tested mixtures of compounds were mounted beneath the lids of white polyethylene cross-vane traps IBL-5 supplied by the Chemipan R&D Laboratories, Poland (Supplementary Materials, Figure S4). Those traps were found to be the most effective ones for capturing D. pini males [25]. Square vanes (20 cm × 20 cm) made of polypropylene cellular board (Tekpol ® ) were mounted between a trap's lid, and a funnel screwed onto a polyethylene collector. Each collector had a drainage hole with a metal mesh in the bottom. A cardboard strip (3.5 cm × 3 cm) impregnated with the insecticide (7% transfluthrin) (Bros Co., Ltd., Poznań, Poland) was added to each collector to kill the trapped moths.

Field Experiments
All experiments were set in a randomized complete block design. The distances between the blocks and between the traps in the blocks were 20-30 m. The traps were hung on pine branches, 4-6 m above the ground, using a telescopic entomological pole that ended with a hook. The traps were checked and emptied twice in experiments 1-3 (with a rotation after the first check) and once in experiment 4. Experiment 1 lasted seven days, from 24 to 31 July, and aimed to compare the effectiveness of two lures, STD and MD10 (Table 1), with eight replicates per lure. The STD lure contained two pheromone components known to date ((Z5,E7)-12:Ald, (Z5,E7)-12:OH), and MD10 was the complete lure, which included two known and three newly discovered components of the pheromone ((Z5)-12:Ald, (Z5)-12:OH, (Z5)-14:OAc), (Z5)-10:OH, and SPEO.
Experiment 2 was carried out from 31 July to 6 August and included the MD10 and MD12 lures (Table 1) to evaluate the effect of SPEO on the effectiveness of the complete lure. Each lure was tested in 10 replicates.
In Experiment 3, which lasted from 6 to 12 August, we compared the MD12 lure (containing six components) with five lures (MD14, MD15, MD16, MD17, and MD18- Table 1) lacking one or two components. Each lure was tested in eight replicates. Experiment 4 was set up on 27 July 2016 and lasted until 30 July 2016. It aimed to evaluate the effect of adding (Z5)-10:OAc to the MD17 lure found most effective in experiment 3. Thus, we tested the MD17 and MD17OAc lures, each lure in 10 replicates.

Statistical Analyses
The effectiveness of the lures tested in the wind-tunnel experiments was compared using the Kruskal-Wallis ANOVA for ranks in Sigma Plot 12.5 package (academic license, Systat Software, USA), separately for the experiments with and without SPEO.
Before the statistical analyses of the results from the field experiments, we pooled the data from each experiment across the time scale to calculate the total number of moths per trap during the entire exposure period. The effect of the lure type on the total moth catches was evaluated using a generalized linear mixed model with a negative binomial distribution (NB_GLMM) of the dependent variable, in which the lure type and block were used as a fixed and random covariate, respectively. A Wald χ 2 test was used to test the significance of fixed variables [26]. In experiment 3, the Dunnett test was applied to compare the tested lures to the control lure, i.e., MD12. The goodness of fit provided by the model was estimated by residual diagnostics [27,28].
All analyses were done using the R environment [29], version 3.5.1, with rStudio [30], version 1.1.463. The following R packages were used: glmmTMB [31] for NB_GLMM, emmeans [32] for the Dunnett test, and DHARMa [33] for the model diagnostics. The significance level was α = 0.05 for all analyses.

Identification of Female Volatiles
We collected and analyzed genuine samples of volatiles from calling D. pini females. The upper panel in Figure 1 shows a total ion current (TIC) GC/MS chromatogram of a sample collected from a one-day-old female in the 3rd h of the scotophase. The lower panel in Figure 1 shows a TIC chromatogram of a mixture of authentic compounds with a composition equivalent to the MD17 lure (Table 1). (E5,E7)-12:Ald was included in that composition as an impurity in the isomeric (Z5,E7)-12:Ald component. For consistency, the liquid mixture was sampled using SPME and introduced to the GC/MS inlet in the same way as the pheromone samples. The peaks assigned in Figure 1 to the compounds identified are also listed in Table 3 Table 1 (b). Chromatograms were obtained using a polar ZBWAX column.
Furthermore, five sesquiterpenes-β-caryophyllene, β-selinene, α-muurolene, δ-cadinene and γ-cadinene-were identified by the m/z 204 molecular ions and characteristic patterns of sesquiterpene-specific fragment ions (m/z 189, 175, 161, 147, 133, 119, 105, 91 and 79). Comparative GC/MS analysis showed that retention times and fragmentation patterns of all compounds identified clearly matched those of purchased or synthesized standards. Thus, two alcohols, three aldehydes, and two acetates were the candidate components of D. pini sex pheromone, while sesquiterpenes were instead linked to plant material consumed by caterpillars [38][39][40][41]. We excluded the (E5,E7)-12:Ald from further experiments because its contents in the samples (assumed proportional to the peak areas) were negligible compared to the contents of its (Z5,E7)-12:Ald isomer. However, 0.3% of the (E5,E7) isomer was present in the (Z5,E7) isomer as an impurity (Figure 1, bottom panel). The proportion of those isomers was similar in SPME samples of female emissions (Figure 1, top panel). Besides, we sampled male moths and non-calling females to find that their emissions contained sesquiterpenes but not the other compounds observed in calling females ( Figure S6 in Supplementary Materials).

Wind-Tunnel Experiments
The raw results of the wind tunnel experiments are shown in Tables S1 and S2 in the Supplementary Materials (lures without and with SPEO, respectively). Three fivecomponent blends without SPEO (MD14, MD16, and MD17) and a four-component blend (MD18) appeared to be attractive to D. pini males (Figure 2a), but the differences between all blends were not statistically significant (Kruskal-Wallis test: H = 4.726, df = 5, p = 0.450). (Z5)-10:OH was a plausible guess-addition as a homolog of (Z5)-12:OH and a component of the sex pheromone of many moths like larch casebearer, Coleoptera laricella [46,47], turnip moth Agrotis segetum [48,49], and millet stem borer Coniesta ignefusalis [50]. The flow of pure air through the tunnel made the insects fly almost in all experiments and could serve as a reference for the lures tested.
The addition of the essential oil distilled from the Scots pine needles (SPEO) had no significant effect on the effectiveness of the lures in the wind tunnel tests (Figure 2) since the differences observed were not statistically significant (H = 3.906, df = 5, p = 0.563). Interestingly, the presence of SPEO in the lures slightly changed the pattern in male responses. The effectiveness of MD17 and MD12 lures with SPEO were the highest of all lures tested.   Table S1, Supplementary Materials). The differences between the lures were not significant in all cases. For the composition of the lures, see Table 1.

Field Experiments
In Experiment 1, the STD and MD10 lures differed significantly in the attractiveness to D. pini males (Wald test: χ 2 = 8.68, df = 13, p =0.0032). During seven days of the experiment, the traps with the MD10 lure containing six components and SPEO caught 17.3 ± 2.8 moths/trap, while the STD lure attracted 7.6 ± 1.5 moths/trap (Figure 3a). Experiment 2 showed that SPEO had a strong significant effect on catches of D. pini (after excluding one outlier, χ 2 = 8.95, df = 1, p = 0.0028). The traps with the MD12 lure (without SPEO) caught significantly fewer males (77.8 ± 7.2 moths/trap) during the six-  Table S1, Supplementary Materials). The differences between the lures were not significant in all cases. For the composition of the lures, see Table 1.
In addition to the main observations, we also confirmed that the red LED light (wavelength range 620-630 nm) does not distract D. pini in the scotophase period, which significantly facilitates the work with these insects in a laboratory framework.
In summary, we identified (Z5)-12:Ald, (Z5)-12:OH, (Z5)-10:OAc, and (Z5)-14:OAc as new likely components of the sex pheromone emitted by D. pini female moths, which accompanied the previously discovered components (Z5,E7)-12:Ald and (Z5,E7)-12:OH. Besides, we identified trace amounts of (E5,E7)-12:Ald, which may play an important role in insect communication, particularly in distinguishing species of the same genus inhabiting the same environment; thus, it is worth further research. We proved that (Z5)-12:Ald significantly attracted the male moths while (Z5)-10:OAc had a repelling effect. Thus, further work is warranted to unveil the exact roles of other discovered compounds. The mixture of all components identified but (Z5)-10:OAc, which has a repellent effect, is an effective lure for D. pini males and provides a base for further work to optimize the lure composition for use in the trap-based monitoring of D. pini populations. For instance, the lures lacking some components may be sufficiently effective while easier and cheaper to prepare. Our results indicate that the optimization may start from the MD17 lure lacking (Z5)-10:OH. The lure should be applied either in cross-vane traps or Unitraps with polyethylene vial dispensers [25,77]. The addition of SPEO enhances the performance of the lure. Thus, further work is required to reveal its composition across Scots pine chemotypes and understand its role in D. pini communication.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/insects13111063/s1, Figure S1: Sampling of volatile emission from D. pini females; Figure S2: Wind tunnel; Figure S3: The behavior of D. pini males in the wind tunnel; Figure S4: The IBL-5 trap used in the field experiments; 70eV EI ION TRAP MASS spectra of compounds identified in this work; NMR spectra of likely components od D. pini sex pheromone; List of compounds identified in Scots pine essential oil; Figure S5: Chromatographic separation of isomers of compounds in D. pini emissions; Figure S6: Comparison of emissions from a male and a noncalling female; Table S1: Tunnel experiments without SPEO-raw observations; and Table S2: Tunnel experiments with SPEO-raw observations), and a video file behavior_7.wmv containing a recording of the utmost behavior of males in the wind tunnel. Reference [78] is cited in the supplementary materials.