The pheromone verbenone and its function in Dendroctonus armandi ( Coleoptera : Curculionidae : Scolytinae

The Chinese white pine beetle, Dendroctonus armandi Tsai and Li is a native species of bark beetle and one of the most destructive in much of western China. Little is known about the characterization of the pheromones trans-verbenol and verbenone, and their functions in D. armandi are unknown. Electroantennogram tests (EAG) and olfactory assays (Y-tube assays) in the laboratory revealed that (1) trans-verbenol may be an anti-aggregation pheromone for male and an aggregation pheromone for female D. armandi and (2) female beetles are more attracted to controls (hexane) than low concentrations of verbenone and male beetles more attracted to controls than high concentrations of verbenone. Field trials indicated that the addition of verbenone to bait used to trap D. armandi remarkably decreased the effi ciency of fi eld trapping. These results indicate that verbenone is an anti-aggregation pheromone for male D. armandi. This research provides evidence of the role of verbenone among the different types of pheromones. The pheromone verbenone clearly could be used to protect healthy Chinese white pines. * Correspondin author; e-mail: chenhui@nwsuaf.edu.cn. INTRODUCTION The Chinese white pine beetle, Dendroctonus armandi Tsai and Li, is considered to be a serious pest in the Qinling and Bashan Mountains in China as it kills mature Chinese white pines (Pinus armandi Franch), resulting in huge economic losses. In China, over 3×108 m3 of Chinese white pines have been killed by D. armandi since the 1970s (Xie & Lv, 2012). As D. armandi is a major threat to the health of P. armandi, it may pose a signifi cant threat to the development of Qinling and Bashan Mountain forest ecosystems. Management measures in recent years include forest tending, chemical control and semiochemical-based trapping. D. armandi mainly attacks healthy P. armandi trees that are over 30 years old (Chen & Tang, 2007). In recent years, D. armandi begun to attack much younger P. armandi (Chen et al., 2015). D. armandi completes its life cycle in the phloem of P. armandi and the adult stage leave this host and attacks new host trees. D. armandi females attack a tree and then use some kind of semiochemical to attract both males and females, which is similar to what is recorded for Dendroctonus valens (Liu et al., 2006; Pureswaran et al., 2008a). Semiochemical communication in bark beetles “enables host and mate location, aggregation and resource partitioning” (Borden et al., 1986; Liu et al., 2013), and aggregation pheromones ensure successful Eur. J. Entomol. 114: 53–60, 2017 doi: 10.14411/eje.2017.008


INTRODUCTION
The Chinese white pine beetle, Dendroctonus armandi Tsai and Li, is considered to be a serious pest in the Qinling and Bashan Mountains in China as it kills mature Chinese white pines (Pinus armandi Franch), resulting in huge economic losses.In China, over 3×10 8 m 3 of Chinese white pines have been killed by D. armandi since the 1970s (Xie & Lv, 2012).As D. armandi is a major threat to the health of P. armandi, it may pose a signifi cant threat to the development of Qinling and Bashan Mountain forest ecosystems.Management measures in recent years include forest tending, chemical control and semiochemical-based trapping.D. armandi mainly attacks healthy P. armandi trees that are over 30 years old (Chen & Tang, 2007).In recent years, D. armandi begun to attack much younger P. armandi (Chen et al., 2015).D. armandi completes its life cycle in the phloem of P. armandi and the adult stage leave this host and attacks new host trees.D. armandi females attack a tree and then use some kind of semiochemical to attract both males and females, which is similar to what is recorded for Dendroctonus valens (Liu et al., 2006;Pureswaran et al., 2008a).Semiochemical communication in bark beetles "enables host and mate location, aggregation and resource partitioning" (Borden et al., 1986;Liu et al., 2013), and aggregation pheromones ensure successful rate of 40 mL/min.The open end of the metal tube was positioned 1.2 cm away from the antenna.In the experiment, the air controller delivered 0.2 s fl ows of air to the antenna.To ensure that the test antenna full recovered from a stimulus, there was a 60 s interval between tests.The test of each pheromone was from a low concentration to a high concentration (Zhang et al., 2010).A hexane-only control and standard solution (1-hexanol at 1 μg/μL in hexane) was presented before and after each test pheromone.Each pheromone was tested on fi ve male beetles and fi ve female beetles, and the trials were repeated at least fi ve times for each EAG preparation.
Although trans-verbenol and verbenone has been detected using gas chromatographic and mass spectral (GC-MS) analyses of the hindguts of female beetles and the fumes emanating from P. armandi logs naturally attacked by D. armandi (Xie & Lv, 2012;Chen et al., 2015), the characterization and function of trans-verbenol and verbenone in D. armandi are unknown.Moreover, there is little information on the release and function of semiochemicals.We identifi ed the characteristics of trans-verbenol and verbenone produced by D. armandi based on electrophysiological responses and laboratory olfactometer trials.Verbenone was further studied in a fi eld trial.The results of these studies could provide a basis for future studies and might be used in the biocontrol of these beetles.

Insects
Chinese white pines infested with D. armandi, were collected from the Qinling Mountains, Shaanxi, China (33°26´53.0̋ N, 108°28´48.3̋ E, at a mean altitude of 1500 m) in November 2015.Logs cut from these Chinese white pines were placed in a nylon plastic net (of mesh size ≤ 0.8 mm) in the laboratory and watered twice every day to keep them moist.The beetles that fl ew from the logs were collected everyday, sorted by sex and stored at 4°C in a refrigerator.Only active beetles were used in the EAG experiment and laboratory olfactometer trials (Wood, 1982;Light, 1983).

EAG assays
The test method used was that described by Zhang et al. (2010).The antennae of D. armandi were dissected under a microscope and were connected between two electrodes.The recording electrode was connected to the distal edge of the club and the indifferent electrode to the scape of the antenna using a conductive adhesive.Filter paper (5 by 50 mm) was used to carry 20 μL of a test solution of the odour, the solvent was then allowed to evaporate for 30 s and then the fi lter paper was inserted into a Pasteur pipette (10 mm diameter by 15 cm long).
Filter paper treated with 20 μL hexane alone were used as controls.One end of the Pasteur pipette was inserted into a hole in a metal tube (15 mm diameter by 15 cm long), the other end of which was connected to an air controller (model CS-05b, Syntech, the Netherlands) that delivered humidifi ed air at a constant trials, this experiment ended on August 31, 2014.Multiple funnel traps were used in all the fi eld trials, which were obtained from Sino-Czech Trading Co. Ltd., Beijing, China.
In 2013, the fi eld trapping experiment was done at Huoditang and Pingheliang Forestry Stations from May 20 to August 31.Four treatments ( A, B, C, D) were compared (Table 5).Treatment A was the control.In 2014, the fi eld trapping experiment was done at Huoditang and Pingheliang Forestry Stations from May 20 to August 31.Four treatments (E, F, G, H) were tested (Table 5).Treatment E was the control.The method used is that described by Xie & Lv (2012).The mixed reagent treatments were released from a 15 mL slow-release plas tic vial at a release speed of 200 mg a day.Eight multiple-funnel traps were set up at random with 10 replicates per treatment.This fi eld equipment was installed in early May 2013 and 2014 before the fl ight period of D. armandi.The traps were spaced at least 30 m apart and were checked every three days.Live D. armandi were sexed using the fact that males stridulate (McGhehey, 1968).

Statistical analysis
The results of EAG assays were corrected for solvent and systematic bias by subtracting the mean response to the solvent-only controls before and after exposure to each sample from the response to the test compound.In order to compensate for the effect of the decline in activity of the test antennae and individual differences between experiments, the EAG assays were standardized by calculating the EAGs as percentages relative to the response to the standard solution (1-hexanol at 1 μg/μL in hexane).Mann-Whitney tests using SPSS (1999) were used to determine significant differences between sexes in the relative EAG responses.In the olfactometer bioassays, chi-square tests using SPSS (1999) were used to compare the responses to the different compounds.For the fi eld trapping experiments, statistical analysis was conducted using one-way ANOVA to detect if the numbers of female and male D. armandi trapped were signifi cantly affected by the treatments and LSD multiple comparison used to determine signifi cant differences among treatments.

DISCUSSION
This is the fi rst behavioral test of the response of D. armandi to trans-verbenol and verbenone.We identifi ed an electrophysiological response to trans-verbenol and verbenone by the antennae of both sexes as well as a dosedependent behavioral response to trans-verbenol and verbenone in both sexes in olfactometer bioassays.Furthermore, the characterization and function of verbenone was verifi ed in fi eld trials at Huoditang and Pingheliang Forestry Stations in 2013 and 2014.Dendroctonus spp.are known to use semiochemicals and host volatiles when attacking host trees and attracting partners (Schlyter & Birgersson, 1999;Byers & Zhang, 2012).Semiochemicals might be a key factor determining the successful mass colonization by Dendroctonus spp., which can overcome the defense system of host trees (Chen et al., 2015).trans- Verbenol [4,7,hept-3-en-2-ol], is the important component of the aggregation pheromone pro-Table 3. Attraction of male and female D. armandi to trans-verbenol-X and verbenone-X blend in Y-tube assays a .
Verbenone is the auto-oxidation product of verbenol and has an anti-aggregation role in the majority of Dendroctonus spp.(Lindgren & Miller, 2002).Our results not only demonstrate that trans-verbenol may be an aggregation pheromone for females and an anti-aggregation pheromone for males, but also demonstrate that verbenone is an antiaggregation pheromone in D. armandi.Based on gas chromatographic and mass spectrometry (GC-MS) analyses of volatiles collected from live invading unmated females, paired females and paired males of D. armandi (relevant data has not been published), the level of trans-verbenol produced by paired females was signifi cantly higher than by unmated females and paired males.The effect of transverbenol was clearer.After a female mates, the concentration of trans-verbenol increases to that of paired females.High concentrations of trans-verbenol were released by paired females in order to attract other females to attack the same pine and inhibit males from releasing the signal "I have mated".trans-verbenol played its role after the mating of female and male.This accords with the GC-MS analyses, the behavioral experiment and life history of D. armandi.Verbenone is produced when the beetles attacking a pine tree start competing for space and nutrition.Furthermore, the amount of verbenone produced by paired males was signifi cantly higher in the GC-MS analysis (relevant data not presented).So verbenone also acts as antiaggregation pheromone after mating.The aim of this study regarding verbenone was to determine whether it was an anti-aggregation pheromone and a further study is planned to determine how and when.The electroantennographic and behavioral responses of Dendroctonus armandi to trichloroethylene are reported by Wang et al. (2011b).
Trichloroethylene is signifi cantly more attractive for males than females.This was taken into consideration in the fi eld experiment, but the reason for the lure effect is still not clear, but it may be acting as a synergist.
The EAG dose-response curve of D. armandi to transverbenol revealed that male beetles showed a signifi cantly greater response to the control (hexane) than trans-verbenol and that female beetles a signifi cantly greater response to trans-verbenol (10 μg/μL) than the control (Fig. 1).The laboratory olfactometer trials showed that male beetles were more attracted to the controls (hexane) than to transverbenol and female beetles more attracted to trans-verbenol (10 μg/μL) than the control (Table 1).trans-Verbenol may act as an anti-aggregation pheromone for male and an aggregation pheromone for female D. armandi.The characterization and function of trans-verbenol in female D. armandi are similar to that of some other tree-killing Dendroctonus spp., such as D. ponderosae (Borden et al., 1987), D. pseudotsugae (Rudinsky et al., 1972) and D. frontalis (Payne et al., 1978;Pureswaran et al., 2008b).However, the characterization and function of trans-verbenol were not clear when X [(R)-(+)-α-pinene: (-)-β-pinene: (+)-3-carene = 1 : 1 : 1] was used as a control (Table 3).More work is needed on the characterization and function of trans-verbenol, especially its role as an anti-aggregation or aggregation pheromone.
The EAG dose-response curve of D. armandi to verbenone revealed that the response of female beetles was signifi cantly greater to the control (hexane) than to low concentrations of verbenone and of male beetles signifi cantly greater to the control (hexane) than high concentrations of verbenone.The laboratory olfactometer trials revealed that female beetles were more attracted to controls (hexane) than low concentrations of verbenone and male beetles more attracted to controls (hexane) than high concentrations of verbenone.Both the trapping at Huoditang and Pingheliang in 2014 revealed the same consistent trend of a decreasing female and male response to added verbenone, although the result for treatment F at Pingheliang was not signifi cant (Fig. 2).At high concentrations of verbenone, the decreased response was more signifi cant for males than for females.At low concentrations of verbenone, the decreased response was more signifi cant for females than for males.The addition verbenone to bait greatly decreased the effi ciency of fi eld trapping of D. armandi.Verbenone has the potential to be used to control beetle populations and so protect healthy Chinese white pine.Collectively, this evidence supports our conclusion that for female and male D. armandi verbenone is an anti-aggregation pheromone.In bark beetles, the pheromone verbenone is produced from dietary α-pinene in the fat body (Blomquist et al., 2010) and is an anti-aggregation pheromone in most Dendroctonus spp.D. armandi infestations appear to be increasing throughout the Qinling Mountains, and it is often diffi cult to use physicochemical methods and semiochemical traps to protect healthy Chinese white pines a gainst attack by D.

Group
Chemical regent (R)-(+)-α-pinene (-)-β-pinene (+)-3-carene (S)-(-)-α-pinene trichloroethylene verbenone armandi.It might be possible to use the anti-aggregation pheromone verbenone to reduce the mortality of Chinese white pines and thereby protect healthy pines.There are fi eld trapping and tree protection studies on D. brevicomis, D. ponderosae and D. valens (Gillette et al., 2006(Gillette et al., , 2012;;Fettig et al., 2009).Although, a GC-MS analysis of hindgut extracts from female D. armandi detected verbenone (Chen et al., 2015;Xie & Lv, 2012), little is known about the function of verbenone as an anti-aggregation pheromone in D. armandi.In this study, verbenone signifi cantly reduced the number of D. armandi trapped compared with the control at two sites by approximately 30%.This indicates that verbenone might be used to protect healthy Chinese white pines.The effi ciency of verbenone pouch release devices is especially high when some Dendroctonus spp.beetles are abundant (Progar, 2003(Progar, , 2005;;Borden et al., 2006).We plan to use verbenone pouch release devices in future studies.In summary, this is the fi rst biological and behavioral analysis of trans-verbenol and ve rbenone as possible semiochemicals of D. armandi that ha s demonstrated the biological activities and characteristics of trans-verbenol and verbenone in electrophysiological, laboratory olfactometer experiments and fi eld trials.Furthermore, studies are needed to confi rm the attraction of both sexes of D. armandi to trans-verbenol and evaluate the use of verbenone for protecting trees.

Fig. 1 .
Fig. 1.EAG response (mean % ± SE) of D. armandi to trans-verbenol and verbenone measured relative to the standard solution (1-hexanol at 1 μg/μL in hexane).Error bars are standard errors based on fi ve replicates.Asterisk denote a signifi cant difference between males and females in their response to the same concentration (t test; * P ≤ 0.05; ** P ≤ 0.01).

Fig. 2 .
Fig. 2. Annual mean captures of adult D. armandi by eight kinds of reagent.Bars indicate mean response with standard errors and different letters above bars indicate signifi cant differences at P ≤ 0.05 based on LSD multiple comparison (ANOVA).The regent ratios in groups A-H are in table 5. Reagents were released as a mixture at 250 mg/day.

Table 1 .
Attraction of male and female D. armandi to various concentrations of trans-verbenol (T-V) in Y-tube assays a .
a Table entries are the numbers of D. armandi of each sex that responded to either the control or treatment stimulus.Asterisks mean signifi cant differences (* P ≤ 0.05, ** P ≤ 0.01, chi-squared test).

Table 2 .
Attraction of male and female D. armandi to various concentrations of verbenone (V) in Y-tube assays.
a Table entries are the numbers of D. armandi of each sex that responded to either the control or treatment stimulus.Asterisks mean signifi cant differences (* P ≤ 0.05, ** P ≤ 0.01, chi-squared test).

Table 4 .
ANOVA of the numbers of males and females of Dendroctonus armandi captured by multiple funnel traps during fi eld-trapping at Huoditang and Pingheliang Forestry Stations, Shaanxi Province, China (20 May to 31 August 2013-2014) (n = 10 traps per treatment).

Table 5 .
The regent ratios used in the fi eld trapping experiment at Huoditang and Pingheliang Forestry Stations in 2013 (group A, B, C, D) and 2014 (group E, F, G, H).