Electrophysiological Responses and Reproductive Behavior of Fall Webworm Moths (Hyphantria cunea Drury) are Influenced by Volatile Compounds from Its Mulberry Host (Morus alba L.)

Hyphantria cunea (Drury) is an invasive pest of Morus alba L. in China. β-ocimene and cis-2-penten-1-ol among eleven electro-physiologically active leaf volatiles from M. alba have been reported to influence captures of Hyphantria cunea moths when added into sex pheromone traps. This study further investigated influences of volatile types and their dosages on the electro-physiological responses in the antennae of male and female moths, as well as on mating and oviposition behaviors. Females were, regardless of dosages, more sensitive to β-ocimene and cis-2-penten-1-ol in electro-physiological response tests than males. For males, a dose response was detected, i.e., a dosage of 10 μg and 100 μg of either chemical stimulated higher electric response in their antennae than 1 μg. Moth pairs either exposed respectively to a herbivore-induced M. alba volatile blend (HIPV), to a mechanically-damaged M. alba volatile blend (MDV), to β-ocimene, to cis-2-penten-1-ol, or to pentane as a control showed that pairs exposed to β-ocimene most likely mated, followed by HIPV blends and least by the other volatiles or the control. In contrast, β-ocimene induced about 70% of the female oviposition behaviors and was nearly 4.5 times the oviposition rate than cis-2-penten-1-ol and 2 times than the control. However, none of the chemicals had any effect on the 48 h fecundity or on egg sizes. In conclusion, β-ocimene from mulberry plants alone could promote mating and oviposition in H. cunea at a dosage of 1 mg. The results indicate that reproductive behaviors of H. cunea moths can be enhanced through HIPV blends and β-ocimene induced by feeding of larvae. This contra phenomenon has revealed a different ecology in this moth during colonizing China as local pests would commonly be repelled by herbivore induced chemicals. These chemicals can be used for the development of biological control approaches such as being used together with sex pheromone traps.


Introduction
The fall webworm, Hyphantria cunea Drury (Lepidoptera: Arctiidae), is native to North America but spread into central Europe and eastern Asia in the 1940s [1,2]. In 1979, this insect was first recorded in China in Liaoning Province, and it has continued to spread through much of the country [3]. Hyphantria cunea has caused extensive damage during its 30-year invasion because it feeds on 175 species of host plants [4]. However, it prefers to eat mulberry (Morus alba L., Moraceae), rock Table 1. Chemicals used in electro-physiological tests as well as in mating and oviposition bioassays. β-ocimene and cis-2-penten-1-ol represent host volatiles, and other chemicals were used as controls, solvents and washing. We conducted two mulberry treatments, each with five samplings. (1) Mechanically-damaged leaf volatile compounds: Volatiles were collected from mechanically-damaged leaves. A total 100 leaves were chosen and 25 leaves were randomly cut in half. Then the chosen mulberry leaves were encased in oven bags (482 mmˆ596 mm, Reynolds, Richmond, VA, USA) on the trees. (2) For collecting herbivore-induced leaf volatile compounds (HIPVs), ten H. cunea larvae were placed on mulberry leaves encased in oven bags, and allowed to feed for 1 h. Then larvae and their excrement were removed.

Chemical
For all treatments, air was collected via an air sampler (QC-1, Beijing Municipal Institute of Labor Protection, Beijing, China) and 0.3 g of Super Q adsorbent connected to the oven bags. Air flowing into the bags was purified by an activated carbon sphere on the bag inlet and air flows at 50 mL/min. Volatile samples were collected for 24 h before elution with 2 mL of solvent (70% pentane, 30% diethyl ether) and 72 L air was captured via absorbent in total. Eluted samples were frozen at´20˝C prior to the moth experiments.

Assessing Effects of Morus alba Volatiles β-ocimene and cis-2-penten-1-ol on EAG Responses of Hyphantria cunea
The electroantennogram (EAG) method was adopted from [31] to test differences of male and female responses to β-ocimene and cis-2-penten-1-ol. Each antenna was prepared by cutting both extremes and immediately mounted between two Ag/AgCl electrodes filled with Kaissling saline (glucose 354 mmol/L, KCl 6.4 mmol/L, KH 2 PO 4 20 mmol/L, MgCl 2 12 mmol/L, CaCl 2 1 mmol/L, NaCl 12 mmol/L, KOH 9.6 mmol/L, pH 6.5). The electrode at the distal end of the antenna was connected via an interface box to a signal acquisition interface board (IDAC; Syntech, Kirchzarten, Germany) connected to a computer. EAG signals were recorded by a computer using AutoSpike software (Syntech, Irvine, CA, USA).
Each of these compounds was purchased in pure form and diluted to 10 mg/mL with solvent (70% pentane and 30% diethyl ether). A gas control unit (Syntech CS-05) gave an air flow rate of 30 mL/s with the gas outlet facing the antenna and 1 cm distance from it. For each solution, 10 µL was dropped on a 0.5ˆ5 cm filter paper in a Pasteur pipette. The Pasteur tube was placed between the gas control unit and the gas outlet.
Antennae from both sexes of H. cunea were stimulated for 0.1 s with an interval time of 1 min. Nonanal was used as a standard compound and solvent was used as the control. Tests were run in the following sequence: control, nonanal, sample, control, nonanal. Each antenna was tested twice and each compound was tested with 10 different antennae per gender. Electroantennogram data was adjusted to eliminate the interference of solvent and environment using the formula: where R C is the raw electroantennogram value of the sample, R C-1 is the value of the control before the sample, and R C+1 is the value of control after the sample [32].
The adjusted EAG value was expressed as a specific value by dividing by the standard compound value. Mean values were tested using one-way ANOVA, and means were compared with Tukey HSD at p = 0.05. Each compound's mean difference between male and female moths was tested using General Linear Model.
Chemical concentrations for dose responses were: 0.1, 1, 10 and 100 mg/mL. Dosages of each stimulation were 10 µLˆ(0.1, 1, 10 and 100 mg/mL) = 1, 10, 100 and 1000 µg. Test and data analysis methods were same as above, and at least 6 antennae were tested for each dosage of each chemical compound. Newly-emerged moths were kept at 25˘1˝C under a 16 h light/8 h dark photoperiod in environmental chambers, and 1-3 day old adult moths were used in experiments.
Gray rubber septa (The West Company, Phoenixville, PA, USA) loaded with 1000 µg (100 µL at 10 mg/mL) of either β-ocimene or cis-2-penten-1-ol were used in these experiments for mating and ovipostion tests. This dosage was chosen according to previous electroantennogram results [26]. HIPV and MDV samples were added 100 µL per septum. Concentrations of total volatile compounds were at an average of 10.96 µg/µL. The average of total mass of volatile organic compounds captured by each absorbent was 21.92 mg by calculating with external standard method [27]. The solvent alone was used as control.
A pair (one male, one female) of virgin moths was placed in a beaker (2000 mL, d = 130 mm, h = 185 mm) together with one loaded septum. Experiments began at the 6th h of the dark phase to cover calling peaks of female moths [3], and the number of pairs that mated within 2 h was observed and counted.
Mating rate was defined as the proportion of pairs who conducted mating behaviors within every group (five independent pairs, so the value was intervals of 0, 0.2, 0.4, 0.6, 0.8 and 1.0; this set-up was to reduce variation among replicates since 1 pair per sample size would lead to an interval value between 0 and 1, which would have resulted in larger standard deviation). Oviposition rate was similarly defined as the proportion of oviposition observed within every group.
Egg numbers were counted after 48 h in every pair with oviposition observed, and average egg sizes were calculated by random measurement of egg masses. The effects of the compounds on moth behavior were analyzed using one-way ANOVA. Tukey HSD multiple-comparison test was used to compare mean differences between groups at P = 0.05. All analyses were performed in SPSS 20.0 (IBM, Chicago, IL, USA).

Discussion
Changes in chemical volatile profiles have been repeatedly reported throughout many host plants when they are fed on by pest insects [28] or mechanically damaged. Oral secretion-induced plant volatile changes have also been reported having essential functions towards female moth oviposition [30]. For example, β-ocimene and cis-2-penten-1-ol have been reported as volatiles from mulberry leaves and their existences and emission volume changed dramatically during feeding by H. cunea larvae [27]. These botanical volatile chemical blends showed bioactivities on H. cunea adults in laboratory experiments such as electroantennogram tests and in wind tunnel experiments. Furthermore, the two compounds seem to also affect long distance host location of H. cunea in the field and were suggested to be synergists for H. cunea sex pheromones [25,28]. However, we did not know their short distance functions especially when they are presented together with sex pheromones.
In this study, we found that HIPVs from mulberry could significantly increase the mating rate compared with the solvent control (treatment No. 1). This result implies that mulberry HIPVs may act differently on H. cunea than on most regional pests, which are usually repelled by the HIPVs [30,33]. Although we showed an increased mating as a result of HIPV in our study, the oviposition rate was not enhanced, and this might be the result of potential bioactive compounds in the blends of HIPV compounds other than β-ocimene. HIPV blends may have complex effects on moth behaviors because of their multiple constituents, and this complexity can be more variable in the environment than in the lab [27]. The enhancement of H. cunea mating by mulberry HIPVs may make this moth more adaptable to this plant than local insect herbivores. In futures studies, we will assess which parts/compounds and dosages may have which effects on their own. Higher EAG amplitudes stimulated by both chemicals implies that female moths are more sensitive to host volatile compounds than males. In addition, this may result of the female interest to find suitable sites for their offspring. It might be evolutionarily risky for females to prefer chemicals, which are also preferred by their natural enemies. β-Ocimene is a monoterpene, a natural 10-carbon member of the terpenoid family, which are common constituents of floral scents that attract insects including natural enemies or pollinators [34]. Ocimene activates defensive responses in Arabidopsis thaliana [35,36] and functions as a volatile brood pheromone involved in social regulation in honeybee colonies [37]. For example, z-ocimene was also reported to have a function in inducing herbivory in a number of plant species [35,[38][39][40]. Interestingly, in some of the reported works, β-ocimene was considered neutral for chemical preferences of natural enemies like parasitoids [41]. Therefore, oral secretion induction and attractiveness to this chemical would benefit H. cunea since the chemical would not attract their natural enemies. This oviposition preference to herbivore-induced volatiles of H. cunea moths might also reveal a potential novel invading model which is different from local pest species. This model has similarities with other invasive pests in China, such as red turpentine beetle Dendroctonus valens [42], and this might be a result of a lack in local natural enemies with the same chemical odorant preferences.
Our experiments showed that β-ocimene increased both the mating and oviposition rates in H. cunea. This implies that HIPVs and β-ocimene may affect the moths when presented together with sex pheromone compounds because sex pheromones were emitted by females during the calling when we conducted the experiments in dark phase. Similar conclusions have been reported that volatile organic compounds could enhance the ability of antennal sensors to better recognize pheromones [15,43]. Thus, β-ocimene could increase the efficiency of sex pheromone traps for H. cunea moths [26]. This indicates that β-ocimene may function in both long distance (trapping) and short distance (mating). As HIPVs enhanced mating behaviors in short distance as shown in the lab, but did not affect long distance location of male moths as shown in [26], we suggest that the short distance function of those chemicals is primarily for mating, as also suggested by [44].
Cis-2-penten-1-ol was reported to be present in peppermint oil [45], which repels many insects. Moreover, cis-2-penten-1-ol has been discovered in tea leaves [46]. Interestingly, tea tree is not a host plant for H. cunea [47]. This implies that at least cis-2-penten-1-ol would not be preferred by this moth for host location, or like it has been reported, would even repel this moth [27]. In addition, the moth may overcome this problem by changing the volatile chemical profile of host plants during the feeding process. When testing mating and oviposition, cis-2-penten-1-ol showed no effect on the moths' behavior compared with solvent control. This indicates that cis-2-penten-1-ol is more likely a long distance function compound than β-ocimene, and it acts maybe more on male than on female moths. We also speculate that this compound may affect a metabolic process instead of the olfactory based behavior reflection itself. Influences of metabolic targeting compounds are often expressed in the following generation of insects [48,49]. Therefore, the influences of cis-2-penten-1-ol might also occur after a long term exposure and reveal a tradeoff between egg number and egg size [50].

Conclusions
In conclusion, the tested mulberry compounds, β-ocimene and cis-2-penten-1-ol, were not essential for H. cunea moths to mate and oviposit. However, mulberry volatile compounds did affect reproduction behaviors in H. cunea by increasing the mating rate and oviposition rate. As β-ocimene can also work as synergist with sex pheromone compounds for helping in the location of male moths [26], we suggest that the blend of HIPVs can function both in long distances (e.g., for female or host location) and short distances (mating and oviposition). In addition, β-ocimene was only induced by H. cunea larvae feeding on the mulberry host plants, which indicates that these moths prefer to feed and reproduce on previously-damaged plants. This aggregated attacking model has also been reported for explaining the colonization of other invasive species in China, such as red turpentine beetle. Our study has revealed a conflict that invasive H. cunea would treat HIPVs differently from local pests as they preferred the herbivore-induced component β-ocimene. However, the olfactory perception behind this phenomenon still needs to be discovered.