Identification and Behavioral Evaluation of Sex Pheromone Components of the Chinese Pine Caterpillar Moth, Dendrolimus tabulaeformis

Background The Chinese pine caterpillar moth, Dendrolimus tabulaeformis Tsai and Liu (Lepidoptera: Lasiocampidae) is the most important defoliator of coniferous trees in northern China. Outbreaks occur over enormous areas and often lead to the death of forests during 2–3 successive years of defoliation. The sex pheromone of D. tabulaeformis was investigated to define its chemistry and behavioral activity. Methodology/Principal Findings Sex pheromone was collected from calling female D. tabulaeformis by headspace solid phase microextraction (SPME) and by solvent extraction of pheromone glands. Extracts were analyzed by coupled gas chromatography/mass spectrometry (GC-MS) and coupled GC-electroantennographic detection (GC-EAD), using antennae from male moths. Five components from the extracts elicited antennal responses. These compounds were identified by a combination of retention indices, electron impact mass spectral matches, and derivatization as (Z)-5-dodecenyl acetate (Z5-12:OAc), (Z)-5-dodecenyl alcohol (Z5-12:OH), (5Z,7E)-5,7-dodecadien-1-yl acetate (Z5,E7-12:OAc), (5Z,7E)-5,7-dodecadien-1-yl propionate (Z5,E7-12:OPr), and (5Z,7E)-5,7-dodecadien-1-ol (Z5,E7-12:OH). Behavioral assays showed that male D. tabulaeformis strongly discriminated against incomplete and aberrant blend ratios. The correct ratio of Z5,E7-12:OAc, Z5,E7-12:OH, and Z5,E7-12:OPr was essential for optimal upwind flight and source contact. The two monoenes, Z5-12:OAc and Z5-12:OH, alone or binary mixtures, had no effect on behavioral responses when added to the optimal three-component blend. Conclusions/Significance The fact that deviations from the optimal ratio of 100∶100∶4.5 of Z5,E7-12:OAc, Z5,EZ7-12:OH, and Z5,E7-12:OPr resulted in marked decreases in male responses suggests that biosynthesis of the pheromone components is precisely controlled. The optimal blend of the sex pheromone components of D. tabulaeformis worked out in this study should find immediate use in monitoring this pest in Chinese forests.


Introduction
The Chinese pine caterpillar moth, Dendrolimus tabulaeformis is native to northern China, where it is an economically important pest of conifers. It is mainly distributed in the provinces of He'bei, He'nan, Shān'xi, Shȃn'xi, Shan'dong, Liao'ning, Si'chuan, and Inner Mongolia autonomous region, where it attacks Chinese pine (Pinus tabulaeformis Carr.), Scotch pine (P. sylvestris var. mongolica Litv.), Armand pine (P. armandii Franch.), Masson pine (P. massoniana Lamb.), and Bunge pine (P. bungeana Zucc. ex Endl.) [1]. Dendrolimus tabulaeformis has one to three generations per year depending on climatic conditions. In the Beijing region, larvae that eclose from the end of August to mid-September feed until late autumn and then spend the winter in forest litter or soil close to the trunk. In spring, caterpillars climb up to the crown to feed intensively, and pupate in May/June in cocoons. Flights of adults begin in June/July and last until August. Copulation usually lasts ca. 24 h and then females lay eggs on the needles. One egg mass may contain up to several hundred eggs. Control efforts against D. tabulaeformis consist mainly of aerial treatment with chemical and bacterial insecticides, and even collection of cocoons during outbreak years. Pheromone-baited traps are used to monitor the population dynamics of several Dendrolimus species so that pesticide applications can be optimally timed.
So far, the sex pheromones of six Dendrolimus species have been identified, five of which consist of isomers of (5Z,7E)-dodecadien-1-ol, and/or the corresponding acetates, propionates, or aldehyde derivatives [2][3][4][5][6][7][8]. The sixth species, D. houi uses (5E,7Z)dodecadien-1-ol and the corresponding acetate and aldehyde as sex pheromone components [9]. Although preliminary field screening tests showed that a lure impregnated with a 1:1:1 ratio of Z5,E7-12:OH, Z5,E7-12:OAc, and Z5,E7-12:OPr attracted a few male D. tabulaeformis [10] and the responses of antennae of male D. tabulaeformis to these and similar compounds were measured [11], the actual pheromone components of this species and their behavioral roles have not been properly identified. The purpose of this study was to properly identify the sex pheromone of D. tabulaeformis. Field and wind tunnel experiments then were carried out to optimize the pheromone blend.

GC-MS analyses of pheromone extracts and MTAD and DMDS adducts
GC-MS analyses of SPME collections on a DB-5MS column confirmed the identities of the five components that had elicited EAD responses, including Z5,E7-12: OAc (Table 2). GC-MS analysis of the DMDS adducts (30 FE) showed a molecular ion at m/z 320, a base peak at m/z 145 [H-(CH 2 ) 6 CH = S + CH 3 ] and a peak at m/z 175 [CH 3 S + = CH(CH 2 ) 4 OOCCH 3 ] for EAD1, and a molecular ion at m/z 278, a base peak at m/z 145 [H-(CH 2 ) 6 CH = S + CH 3 ] and a peak at m/z 133 [CH 3 S + = CH(CH 2 ) 4 OH] for EAD2, indicating that the double bond was located at position 5 in both compounds ( Table 2). The retention times also matched those of the adducts from authentic Z5-12:OAc (21.59 min) and Z5-12:OH (20.43 min), confirming the previous identifications.
In wind tunnel bioassays, all males tested took flight in response to all synthetic blends tested except the solvent control, indicating that even the incomplete blend of Z5,E7-12:OAc and Z5,E7-12:OH in a 1:1 ratio was sufficient to elicit flight initiation (

Discussion
The five components from the pheromone glands that elicited EAD responses were unambiguously identified as Z5-12:OAc, Z5-12:OH, Z5,E7-12:OAc, Z5,E7-12:OPr, and Z5,E7-12:OH by GC, GC-MS, GC-EAD analyses, and microscale derivatizations. Moreover, antennae of males responded to the Z5,E7-isomer of the conjugated dienes and the Z5-isomer of the monoenes acetate, alcohol, and/or propionate in EAD analyses, and the other isomers were not detected in the gland extracts. In wind tunnel bioassays, a mixture of the two major components Z5,E7-12:OAc and Z5,E7-12:OH in a 1:1 ratio was sufficient to induce flight by males, but the minor component, Z5,E7-12:OPr was essential for optimal upwind flight and source contact. In total, our data suggest that the female-produced sex pheromone of D. tabulaeformis is a three-component blend of Z5,E7-12:OAc, Z5,E7-12:OH, and Z5,E7-12:OPr, and the optimum ratio has been determined to be ca. 100:100:4.5, which is close to the ratio found in analyses of the volatiles emitted by calling females (100:100:8). Furthermore, our Table 1. Compounds, Kováts Indices, EAD responses, detected rate, and absolute amounts of potential pheromone components collected by CAR/DVB SPME fiber (70 mm) from live calling virgin females of Dendrolimus tabulaeformis or extracted from pheromone glands with hexane.   data showed that male D. tabulaeformis strongly discriminate against off-ratio pheromone blends, which would impose strong stabilizing selection on the pheromone signal. Species-specific pheromone blends can be generated from different sets of constituents, or by mixing the same set of constituents in different ratios [12]. The identified sex pheromones of Dendrolimus species share a common chemical theme of being C 12 5,7-dienes with alcohol, acetate, propionate, or aldehyde functional groups, with species-specific blends being generated by using different subsets of components and/or ratios of components. For example, in the species studied here, a specific ratio of sex pheromone components is required in order to achieve optimal male flights. However, the best attractant for male D. spectabilis, which is sympatric with D. tabulaeformis in Shan'dong,  He'bei, Liao'ning, and Inner Mongolia, consists of a 100:3:25 blend of Z5,E7-12:OH/Z5,E7-12:OAc/Z5,E7-12:OPr [6], with Z5,E7-12:OH as the principal sex pheromone component [2]. In contrast, comparative studies of the sex pheromones of the most closely related species, D. punctatus, which is sympatric with D. tabulaeformis in Shȃn'xi, He'nan, and Chong'qing, suggest that male moths of this species respond to a broad range of ratios of Z5,E7-12:OH/Z5,E7-12:OAc/Z5,E7-12:OPr blend (median blend 100:82:30) [13]. Dendrolimus punctatus, D. spectabilis, and D. tabulaeformis all share the same three pheromone components, Z5,E7-12:OH, Z5,E7-12:OAc, and Z5,E7-12:OPr, but the differences in the ratios of the three-component blends are essential for maintaining separate communication channels. Furthermore, the bioassay data suggests that Z5-12:OAc and Z5-12:OH are not part of the attractive pheromone blend of D. tabulaeformis, although they elicit relatively strong EAD responses from male antennae, because neither of these compounds enhanced the attractiveness of three-component blend in field test. The same phenomenon was also observed in D. punctatus [13]. However, D. tabulaeformis is partially sympatric with D. kikuchii in He'nan province, and a major difference between the two pheromone systems is that the pheromone blend of D. kikuchii substitutes Z5-12:OAc for Z5,E7-12:OPr in an optimal ratio of 100:20:25 Z5,E7-12:OAc/Z5,E7-12:OH/Z5-12:OAc [8]. Thus, the minor components Z5-12:OAc and Z5,E7-12:OPr may serve to enhance the specificity of the signals in D. tabulaeformis and D. kikuchii. In addition, the antennae of male D. tabulaeformis responded to synthetic Z5,E7-12:Ald and E5,Z7-12:Ald in our GC-EAD analyses, but the two compounds were not detected in pheromone gland extracts. It has been reported that the sex pheromones of D. superans and D. houi consist of Z5,E7-12:Ald [7] and E5,Z7-12:Ald [9], respectively. Dendrolimus tabulaeformis is broadly sympatric with D. superans in Inner Mongolia, He'bei, and Liao'ning, and sympatric with D. houi in Shȃn'xi and Chongqing. Addition of Z5,E7-12:Ald (ca. 5% of Z5,E7-12:OAc) to the threecomponent blend of D. tabulaeformis resulted in decreased captures of males. Thus, these two aldehydes might play an important role in chemical communication of sympatric, closely related species, either as essential minor components of pheromone blends, or conversely as interspecific inhibitors.
The optimal blend of the sex pheromone components of D. tabulaeformis worked out in this study should find immediate use in monitoring this pest in Chinese forests.

Ethics statement
All necessary permits were obtained for the described field studies. The forestry bureau of Ping'quan city, responsible for Da'shi lake forest center, issued the permission for our field studies in the Chinese pine forest at this site. These field studies did not involve endangered or protected species.

Insects
Cocoons of D. tabulaeformis were collected from the host tree, Pinus tabulaeformis at Da'shi lake forestry center near Ping'quan city, He'bei province on 30 July 2011 and maintained in a rearing room at 2561uC, 60-80% relative humidity (RH) with a photoperiod of 16:8 h light:dark until adult emergence. Newly emerged males and females were housed separately under the same conditions. Virgin females were used for preparation of pheromone extracts, whereas males were used for EAG studies.

Sample collection
Sex pheromone glands were excised from females during the peak period of calling (0-2 d old, 4-7 h into the scotophase) and extracted for 30 min with hexane (10 ml per gland). The glands were removed before sealing the extracts in glass tubes, which were stored at 220uC until bioassay and chemical analyses.
Pheromones emitted by 1-2 d old calling virgin females held in 0.02 m 3 metal screen cages were collected by solid phase microextraction (SPME, 70 mm CarboWax divinylbenzene (CW/DVB) coated fiber, Supelco Inc., Bellefonte, PA, USA) under ambient laboratory conditions. The SPME fiber was placed 1-2 mm from the extruded gland, sampling from the onset of calling (about 3 h into the scotophase) to the end of scotophase.
The positions of double bonds in conjugated dienes were determined by means of MTAD derivatization [14,15], and in monounsaturated alkenes by the iodine-catalysed addition of dimethyl-disulfide across the double bond [16]. In this study, 30 female equivalents (FE) of extract were derivatized by MTAD and DMDS reagents respectively, as previously described [8,16]. The derivatives were analyzed by GC-MS.

Gas chromatography-electroantennographic detection (GC-EAD) analyses
The coupled GC-EAD system used was based on that described by Kong et al. (2011). The HP6890 GC (Agilent, Palo Alto, CA, USA) was equipped with a split/splitless injector, a DB-WAX column (J&W Scientific, Folsom, CA, USA, 30 m60.25 mm i.d. 60.25 mm) with the column effluent split between the FID and EAD detectors (FID/EAD, 1:2). The effluent exited into a purified and humidified airstream (450 ml/min) directed over the antennal preparation. A detached antenna from a 1-to 3-d-old male with the terminal segment removed, was mounted between two EAG probes (PRG-2, Syntech, Kirchzarten, Germany) with electrically conductive gel (Spectra 360 electrode gel, Parker Laboratories Inc. Orange, NJ, USA). Synthetic compounds and SPME fibers with loaded samples were injected splitless using a temperature program starting at 60uC for 1 min, then programmed at 6uC/ min to 230uC (15 min hold), using nitrogen carrier gas (33 cm/ sec). After adsorption, the CW/DVB fiber was treated with a 1 ml aliquot of hexadecane, eicosane, and tetracosane (5 ng/ml of each in dichloromethane) with a syringe, and then immediately introduced into the HP 6890 GC injector where it was desorbed for 2 min (injector temperature 220uC). The purpose of these standards was to assign retention indices to the components of the crude extract [17]. Acquisition and analysis of the EAD signals were performed by GC-EAD software (16-bit version, Syntech, Kirchzarten, Germany).

GC and GC-MS analyses
Quantitative analyses of 21 glands extracted individually were conducted on an HP7890 GC with an FID detector using a DB-WAX column (J&W Scientific, Folsom, CA, USA, 30 m60.25 mm i.d. 60.25 mm film) in splitless mode, temperature program, 60uC for 1 min, then 6uC/min to 240uC (hold 10 min). Injector and detector temperatures were set at 230uC and 250uC, respectively. A 1 ml aliquot of E5,Z7-12:OAc (5 ng/ml) was used as an external standard to quantify the pheromone components. Absolute amounts were calculated from the integrated peak area of each component relative to that of the external standard.
Pheromone gland solvent extracts, SPME collections (desorbed for 2 min in the injector), and references also were analyzed with a Finnigan Trace DSQ GC-MS (EI mode, 70 eV, mass range 41-560 amu) on DB-WAX and DB-5MS capillary columns (splitless mode, 30 m60.25 mm i.d. 60.25 mm film) under the following conditions: injector, ion source, and transfer line temperatures were held at 220uC, 250uC, and 240uC, respectively; helium as carrier gas at 1.0 ml/min; scan time 0.2 sec; oven temperature program: 60uC for 1 min, then 6uC/min to 260uC, hold for 20 min. When analyzed on the DB-5MS column, a 1 ml aliquot of a series of straight-chain aliphatic hydrocarbons (from C 14 -C 18 , 5 ng/ml of each in dichloromethane) was added directly to the sample-loaded fiber or solvent extracts with a syringe in order to assign retention indices to the components of crude extracts. DMDS and MTAD derivatives of pheromone gland extracts were analyzed with the same GC-MS (EI mode, 70 eV, mass range 41-560 amu) on a DB-5MS capillary column (30 m60.25 mm i.d. 60.5 mm film) in splitless mode. The column temperature was programmed from 80uC for 1 min, then 8uC/min to 280uC (10 min hold). Injector, ion source, and transfer line temperatures were set at 280uC, 250uC, and 250uC, respectively. In GC-MS analysis, compounds were identified by comparison of retention indices and mass spectra with those of synthetic standards or their derivatives.

Behavioral assays
Field bioassays were conducted in the Da'shi lake forestry center near Ping'quan city, He'bei province. At 20-30 m intervals, deltashaped traps made from adhesive cardboard were suspended from trees 1.5-1.7 m above ground in randomized complete blocks. In each of four blocks (replicates), traps were baited with a gray sleeve septum (The West Company, Phoenixville, PA, USA) impregnated with test chemicals in hexane, and 200 mg of BHT in 100 ml CH 2 Cl 2 were added subsequently as an antioxidant. The experiment started on 1 August 2011 and was terminated two weeks later.
Synthetic compounds were selected for more detailed examination in a wind tunnel (flight section 22061006120 cm) based on data from the field tests. Air purified by filtration through activated charcoal was blown by a centrifugal fan at 40 cm/s into the tunnel and the outgoing air was exhausted with a second fan. The flight section was lit diffusely from the top at 2-3 lx with a red light, and the room was kept at 2562uC and 50-65% RH. Synthetic compounds in different ratios were released from the centre of the upwind end of the tunnel by means of gray rubber septa. Males 1-3 days of age were placed in the tunnel room at the initiation of scotophase, 3 h prior to testing, to acclimate to the conditions of the assay room. Males were scored for three key behaviors in the flight sequence: taking flight (TF), initiation of upwind flight (UP), and source contact (SC) [18]. During each 4 h test period, two treatments were tested, with twenty males tested individually per treatment. The two-and three-component blends were prepared in hexane and applied in 50 ml aliquots to gray rubber septa to achieve increasing dosages of 0, 30, 60, 80, 90, 100, and 150 ng/ septum for Z5,E7-12:OPr while keeping the major components, Z5,E7-12:OH and Z5,E7-12:OAc in a 1:1 fixed ratio (2000 ng/ septum for each) in order to determine the behavioral effect of variable amounts of the minor component Z5,E7-12:OPr.

Statistical analysis
Trap-catch data from field bioassays were subjected to transformation (log(x+1)) to ensure normal distribution and homogeneity of variance and then analyzed by one-way ANOVA followed by Tukey's test with a significance level of 0.05, for comparisons of means (SPSS13.0). The zero trap-catch data were not included in the analyses because of their lack of variance. For wind tunnel bioassays, the percentage of males responding in the flight tunnel was arcsin(x) transformed and then submitted to a one-way ANOVA, followed by Tukey's test with a significance level of 0.05, for comparisons of means.