The preference choices of Conopomorpha sinensis Bradley (Lepidoptera: Gracilariidae) for litchi based on its host surface characteristics and volatiles

Conopomorpha sinensis Bradley is a host-specific pest of Litchi chinensis and Euphoria longan. Here, we demonstrated that C. sinensis has evolved special physical and chemical mechanisms for host plant location that enable it to survive and reproduce. Females favored laying their eggs on the convex surface of litchi fruit that had particular volatile characteristics. Experiments using a H-type olfactometer showed that female C. sinensis were attracted to litchi flowers, tender shoots, immature fruits, and mature fruits, with the highest attraction rate to mature fruits (74.67 ± 2.31%). There were no significant differences in the attraction of male C. sinensis to different litchi tissues. Further oviposition preference tests using the pericarp, pulp, and seeds of mature litchi fruits revealed that female C. sinensis prefer to lay their eggs on the pericarp. Litchi volatiles were found to be important in attracting C. sinensis to fruits for oviposition. Analysis of volatiles from different litchi tissues by HS-SPME-GC-MS revealed 31 similar volatiles, some of which may be important in the oviposition preference choices of C. sinensis on litchi fruit.

Conopomorpha sinensis Bradley (Lepidoptera: Gracilariidae) is an important pest of Litchi chinensis and Euphoria longan 1 . Its larvae can bore into litchi fruit, damaging flowers, tender shoots, and leaves. After hatching, the larvae immediately penetrate the fruits, feeding on the seed neck, and can transfer a variety of microorganisms that infect the fruits and eventually destroy the crop 2,3 . This has caused severe economic loss to the Chinese litchi industry, with disastrous consequences for the export of Litchi and longan fruit 4 . Due to its cryptic feeding behavior and overlapping of generations, C. sinensis is difficult to control. Difficulties experienced with chemical control have led researchers to explore the use of biological and molecular approaches to controlling this pest [5][6][7] .
Although chemicals and physical irradiation can control C. sinensis [8][9][10] , biological control, which is one of the most widely used environmentally benign approaches to controlling pests 11 , has focused on the natural enemies of C. sinensis. These include many species of spider, green lacewings, predatory mirids, ladybugs, ants, and parasitic wasps 6,12 . Plant protectants and biopesticides can also control C. sinensis 13,14 . Techniques based on chemical ecology have received less attention, except for primary screens of plant volatiles to repel or lure C. sinensis [15][16][17][18] . Whether the attraction of C. sinensis to litchi is mediated via behavior and chemoreception is unknown.
The selection and adaptation of insects to their hosts are the result of co-evolution [19][20][21][22] . Both the highly sensitive olfactory system of the insect and the odor of the host plant play vital roles in insect survival and reproduction 23 . Studying the mechanisms underlying chemical communication between insects and plants can provide a target for the development of highly efficient and specific regulators of insect behavior 24 . The aims of the present study were to identify the factors responsible for the selection of litchi as a suitable host by C. sinensis, and to screen for host volatiles capable of luring C. sinensis. This study will be useful in the development of insect attractants for field biological control and forecasting insect outbreaks. A better understanding of the chemical ecology of litchi and C. sinensis will help to provide a green biocontrol technology with the characteristics of non-toxicity, non-pollution, and sustainability.

Results
The choice behavior of C. sinensis depends on the surface characteristics and volatiles of host plants and non-host plants. The taxis behavior response of C. sinensis to the surface of host plants and differences in their responses to host plants and non-hosts were observed. There were significant differences in oviposition choice by female C. sinensis on different surfaces of host plants, which preferred to oviposit on the convex surface of host plants (F = 37.14, df = 3, 12, P < 0.05). The average number of eggs deposited on the convex surface of host plants was 30.33 ± 2.52, which was similar to that of the litchi fruit control (32.00 ± 2.00) (Fig. 1). The second favorite oviposition site was the concave surface of host plants. Female C. sinensis preferred host plants to non-host plants (F = 20.14, df = 3, 12, P < 0.05), which was similar to that of the control group (Fig. 2). Otherwise, there were fewer eggs, and even no difference in oviposition between non-host plants and non-host plastic balls. This demonstrates that the surface physical characteristics and volatiles of litchi had effects on the oviposition of C. sinensis.
Taxis reactions of C. sinensis to different tissues of litchi. In this study, the taxis behavior of adult C. sinensis to different litchi tissues was studied using a H-type olfactometer 13,25 . The results showed differences in the taxis behavior of female and male adults of C. sinensis to different tissues of the litchi host (Table 1), with female adults preferring the host more than male adults. The attraction rates of mature fruits to female adult C. sinensis was 74.67 ± 2.31%, which had a significant attraction than other tissues of host plants (F = 44.98, df = 3, 12, P < 0.05). The attraction rate of male adult C. sinensis to different litchi tissues was less than 27%, although this was not statistically significant (F = 4.44, df = 3, 12, P > 0.05).  Oviposition preference of female C. sinensis to different tissues of litchi fruit. The oviposition preference test showed that the number and proportion of laid eggs on pericarp tissues were the largest among all the treatments (Fig. 3). Female adults laid 175 ± 28 eggs on average and 88.83% in total on the pericarp, which is significantly different to the values obtained with pulps and seeds of litchi fruit (F = 10.93, df = 3, 9, P < 0.05).

Discussion
The relationship between insects and plants is mainly a reflection of insect feeding, oviposition, and choice of habitat 26 . The host range of C. sinensis is narrow, and there is high host specificity. To date, C. sinensis has only been reported in Litchi chinensis and Euphoria longan in South China and Southeast Asia 6,27,28 . Previous studies show that the cultivar of Litchi chinensis most seriously affected by C. sinensis is Feizixiao 15 . In the present study,  physical and chemical choice behavior tests confirmed that C. sinensis exhibited a preference for different tissues of the litchi cultivar Feizixiao. The host choice behavior of C. sinensis showed that female C. sinensis preferred to oviposit on a convex surface that resembled that of the surface of litchi fruit. Some researchers link this oviposition behavior to predator avoidance [29][30][31] . In addition, differences in the selection of host plants and non-hosts by C. sinensis confirmed that host volatiles attract C. sinensis to the oviposit. For many herbivorous insects, host plant selection is very important for progeny survival and fitness 32 .
The relationship between herbivorous insects and host plant compounds plays a vital role in the preference choices of insects to their host plants [33][34][35] . Different tissues and volatile cues from host plants are crucial for insect attraction, affecting choice of feeding sites 36 , oviposition behavior 37 , helping offspring to locate and recognize a suitable host plant for larval growth and movement 38,39 , and even finding a mating partner 40 . Behavioral assays in a H-type olfactometer showed that adult female C. sinensis preferred mature fruits to the other tissues of host plants. By contrast, there were no significant differences in the taxis reactions of male adult C. sinensis among the different litchi tissues. This suggests that host plant compounds and their nutritional quality might be key factors in offspring survival 30,[41][42][43] . Tests of the oviposition preference of female C. sinensis to different tissues of mature litchi fruits indicated that the greatest number and proportion of eggs were laid on the pericarp. This is consistent with the results of Xian et al. 13 , who reported that C. sinensis preferred the pericarp of host plant fruits after mating.
Plants can produce complicated odor blends 44 , with large numbers of volatiles comprising several compounds in particular ratios. These volatiles strongly influence the ecological interactions of plants with insects 45 . In the present study, 42 compounds, many of which were terpenes, were identified in the volatile blend emitted from different litchi tissues (Table 2). These findings provide a theoretical basis for studying chemical communication between C. sinensis and litchi. Previous studies revealed that zingiberene attracts C. sinensis to lay eggs on its host 18 . (E)-ß-farnesene and (Z)-ß-farnesene were two unique compounds from mature fruit pericarp of the litchi cultivar Feizixiao, and it was speculated that specific ratios of these compounds and six other volatiles that were mainly found in representative susceptible cultivars of litchi may contribute to the attraction of C. sinensis 46 . Our results suggest that the chemical composition of mature litchi fruit is most conducive to successful attraction of adult female C. sinensis. We found that silanol, trimethyl-, 2-isopropyl-5-methyl-9-methylenebicyclo[4.4.0] dec-1-ene, isocaryophillene, α-trans-bergamotene, (+)-cyclosativene, and cyclohexene were unique compounds of the volatiles of mature Feizixiao fruit. In addition, 7-epizingiberene, l-ß-bisabolene, and caryophyllene were detected as the main compounds of the volatiles emitted from immature and mature fruit, while (+)-aromadendrene is the common compound of the volatiles from different litchi tissues. Volatile blends comprising these constituents in specific proportions may be key signals for host identification in C. sinensis. How the mature fruit processes such volatile mixtures is unclear and requires further research.
In conclusion, our studies indicate that C. sinensis has evolved physical and chemical mechanisms to locate a suitable host, thereby enabling it to survive and reproduce. Host plant volatiles have a significant effect on the oviposition preference of adult female C. sinensis, and the analysis and identification of these volatiles afford many valuable basal data and candidate volatiles, which are key factors in the choice of C. sinensis to oviposit on litchi  5). To take advantage of the ability of host plant volatiles to influence directional selection in C. sinensis, it is necessary to determine the single and composite compounds of host volatiles that exert the greatest attractant activity to C. sinensis. Future work will require the implementation of electroantennogram (EAG) response measurements, wind tunnel tests, field tests, and molecular tools to tackle C. sinensis infestation effectively. Fresh fruits of the litchi cultivar Feizixiao were chosen as the basic material, since it is the cultivar most seriously affected by C. sinensis 15 . The litchi flower, tender shoot, fruit, and the fruit pericarp, pulp, seed were separated for experiments.

C. sinensis attraction to the surfaces of host plants and non-host plants Attraction of C. sinensis to the surfaces of host plants.
Ten mature litchi fruits were wrapped individually in either flat paper, convex paper, or concave paper; controls were not wrapped in paper. The fruit were then placed in the same insect rearing cage (45 cm × 45 cm × 50 cm). Twenty 2-day-old male and female adults of C. sinensis (♀:♂ = 1:1) were placed in one cage for the oviposition experiment with three replicates; every insect was used for once in the test. After 48 h, the egg numbers oviposited by females from each treatment was recorded.

Attraction of C. sinensis to host plants and non-host plants.
Ten each of the following were wrapped in convex paper: mature litchi fruits, fruits of a non-host plant (Prunus salicina Lindl, a kind of plum, the same  size as litchi fruits), and non-host (red plastic balls of similar dimensions to litchi fruits). The wrapped fruits and plastic balls were placed in the same insect rearing cage (dimensions as above). The control consisted of the same number of mature litchi fruits, unwrapped. The oviposition test with three replicates were recorded as described above.
Chemical choice behavior of C. sinensis to host plants Taxis reaction of C. sinensis to different tissues of litchi. The taxis reaction of C. sinensis to four different litchi tissues (flower, tender shoots, immature fruit, and mature fruit) was studied using the H-type olfactometer method of Kudon 25 and Xian et al. 13 . The olfactometer is in the form of letter H and made of plexiglass. It had two cylindrical channels (Height: 30 cm, Diameter:10 cm) on each side, and the channels plus a cylindrical bioassay chamber (Length: 30 cm, Diameter: 3 cm) composed the olfactometer proper. Insects were placed in the central entry of the bioassay chamber and their activity to which side were recorded in the olfactometer. One side of the H-type olfactometer contained tender shoots, immature fruit, or mature fruit; the other side of the olfactometer contained nothing and was a blank control. Host selection behaviors were observed and recorded 1 hour after an insect was placed in the H-type olfactometer. Fifty male and female adult C. sinensis were tested in the experiment, which was repeated three times. Every insect was used for once in the test. The number of C. sinensis showing a taxis reaction to the different litchi tissues was recorded and the attraction rate was calculated.
Oviposition preference of female C. sinensis on different tissues of litchi fruit. The oviposition preference of female C. sinensis was tested in an insect rearing cage (45 cm × 45 cm × 50 cm). Fifty 2-day-old male and female adult C. sinensis (♀:♂ = 1:1) were placed in the cage and fed with 5% honey, soaked into cotton. All tissues were wrapped in convex paper respectively. The cage was filled respectively with three litchi pericarps, pulps, seeds, and blank controls at random positions within the cage. Each experiment was performed in triplicate. The culture environment was at 26 ± 1 °C, 14 h:10 h (Light: Dark) photoperiod, and 60-80% RH. The number of ovipositing female C. sinensis was recorded after 48 h.
Analysis of volatiles from different litchi tissues using HS-SPME-GC-MS Headspace solid-phase microextraction (HS-SPME). Fresh litchi fruits were put into a 500 ml extraction bottle and sealed. The SPME conditions were further optimized by careful selection of the fibers, the extraction time, and the desorption time according to Bianchin et al. 47 . After 1 h microextraction, the extraction fiber was placed into the instrument for GC-MS analysis. The temperature for SPME fiber desorption was 200 °C with 3 min sample injection.
GC-MS analysis. A gas chromatograph (GC-8000 Top , Finnigan, USA) equipped with a gas chromatographic column (30 mm × 0.25 mm × 0.25 μm) was used during the study. The injector temperature was set at 240 °C and operated in split mode with a flow rate of 10 ml min −1 . Ultrapure helium at 1 ml min −1 was used as the carrier gas. The oven temperature program was as follows: 40 °C (held for 4 min), 10 °C min −1 to 240 °C (held for 3 min). MS detection was performed under the following conditions: ionization source at 230 °C, and the ionization mode was electron impact with electron energy of 70 eV, 350 v. The mass spectrometer was operated in total scan mode over a m/z range of 35-350 amu. The analytes were univocally characterized on the basis of their retention times and their mass-to-charge ratios.
Statistical analysis. The attraction rate formula of taxis reaction of C. sinensis to ifferent tissues of litchi = (number attracted by the treatment -number attracted by the control)/(n = 50) × 100%. Statistical analysis was performed using SPSS Statistics 22 software (IBM Corp., Armonk, NY, USA). One-way ANOVA with a post hoc LSD test was used. Results were considered statistically significant when p-values were <0.05.
The chemical compounds of different litchi tissue volatiles were determined by analyzing and comparing the mass spectra using the U.S. National Institute of Standards and Technology (NIST) spectrum library, and the retention times from standard mass spectrometry.