Fall Webworm Host Plant Preferences Generate a Reduced Predation Enemy-Free Space in Its Interaction with Parasitoids

Simple Summary

The tri-trophic interactions among the fall webworm (Hyphantria cunea), its host plants, and its parasitoid wasp (Chouioia cunea) were investigated, with particular focus on the correlation between H. cunea’s host plant preference and the attractiveness of these plants to C. cunea. It was found that Morus alba, despite providing suboptimal nutritional quality for H. cunea larvae, became the most preferred host plant due to its significantly reduced attraction to C. cunea compared to alternative host species. Through volatile analysis, M. alba was identified to lack tridecane emission, a key semiochemical that attracts C. cunea. These results demonstrate adaptive host selection behavior in H. cunea that minimizes parasitism risk, providing novel insights for refining biological control strategies against this invasive species.

Abstract

Plants and insects are developing strategies to avoid each other’s defense systems. Host plants may release volatile compounds to attract the natural enemies of herbivores; insect pests may also select host plants that are deterrent to natural enemies to avoid such predation. Here we investigated whether the host plant preference of Hyphantria cunea correlates with the attractiveness of these plants to Chouioia cunea, a parasitoid wasp that serves as the primary natural enemy of H. cunea. We found Morus alba was the preferred host plant for female H. cunea. Although M. alba provided suboptimal nutritional value for H. cunea growth and development compared to other plants, it attracted fewer C. cunea relative to alternative host plants. Gas chromatography–mass spectrometry (GC–MS) coupled with gas chromatography–electroantennographic detection (GC-EAD) analysis identified six distinct compounds among the herbivore-induced plant volatiles (HIPVs) produced following H. cunea feeding. Notably, M. alba was the sole plant species that did not emit tridecane. These results suggest that H. cunea utilizes M. alba as a reduced predation enemy-free space, thereby minimizing parasitization by C. cunea. Our research emphasizes the importance of considering adaptive responses of herbivores within the context of multi-trophic relationships, rather than solely focusing on optimizing herbivore growth on the most nutritionally suitable plant host.

1. Introduction

Plants and insects have coexisted for over 350 million years, during which both have developed strategies to avoid each other’s defense systems. For instance, host plants may release volatile compounds that attract the natural enemies of herbivores by offering sustenance and shelter to these natural enemies, and insect pests may also select host plants that are deterrent to natural enemies to avoid such predation (i.e., enemy-free space). There are numerous physical, physiological, and behavioral mechanisms that can support ‘enemy-free space’ [1,2,3]. For instance, gall wasps are defended against natural enemies by rapidly changing gall morphology [4,5,6]. Herbivores may also develop behavioral preferences to consume or oviposit on host plants that are nutritionally sub-optimal but are less frequented by their natural enemies. Consequently, herbivores preferring these less optimal plant species may experience enhanced survival compared to more nutritious hosts on which the herbivore is more susceptible to parasitization [2,7,8,9,10].

Research suggests that the influence of host plants on the risk of attack by parasitoids is an important selective force in the evolution of herbivore diet breadth [11]. For example, larvae of the generalist arctiid moth Platyprepia virginalis experienced 83% parasitism by tachinid flies on hemlock (Conium maculatum), whereas <50% of the caterpillars collected from lupin (Lupinus albus) were parasitized [12]. Another example of a generalist herbivore feeding on suboptimal plants to reduce parasitization is the arctiid Grammia geneura: its larvae tend to feed on plant species that, although nutritionally suboptimal, may offer chemical compounds that enhance their survival following parasitization [4]. All these examples suggest that host plant preferences are governed not only by nutritional quality but also by the third trophic level, natural enemies [2,3,13,14].

The fall webworm Hyphantria cunea (Drury 1773) (Lepidoptera: Erebidae) is a notorious invasive species. In China, since its introduction in 1979, the moth has spread into numerous provinces and cities, including Liaoning, Shandong, Anhui, Shanxi, Hebei, Henan, Shanghai, Tianjin, and Beijing, and continues to expand its range [15,16]. The larvae could feed on >600 host plant species, with about 100 found in China [17]. The parasitoid wasp Chouioia cunea Yang 1989 (Hymenoptera: Eulophidae) [18] is an endoparasitic chalcid wasp indigenous to China that parasitizes H. cunea. A single H. cunea pupa may yield between 145 and 365 adult wasps, with a remarkably high percentage (98–99%) of emerging wasps being female [19]. In China, C. cunea is widely used in H. cunea biological control programs [20,21,22,23].

Although H. cunea can subsist on hundreds of plant species, the larvae exhibit preferences for certain host plants [24,25,26]. It remains unclear whether the nutritionally optimal plants for H. cunea correspond to its most preferred hosts; it is also not clear whether the preferred host plants are most attractive to both H. cunea and its parasitoid C. cunea. Infochemicals provide important cues for insects to locate and recognize their hosts and avoid enemies [27,28,29]. Some work has focused on plant–herbivorous insect interactions, such as the role of stored plant volatiles on pests [30]; for example, the phenolic acids in the leaves of hazel (Corylus spp.) have an allelopathic effect on Myzocallis coryli [31], and granary weevils are attracted or repelled by different concentrations of cereal volatiles [32].

Other researchers have shown that chemicals from H. cunea pupae or its by-products (e.g., frass) could induce the arrestment of C. cunea females [33], and the response of parasitoid Holepyris sylvanidis to host-associated volatiles can also improve parasitization success [34]. However, these works focused on the paired interactions between H. cunea and its parasitoids but rarely analyzed the tritropic interactions among host plants, insects, and natural enemies from the perspective of infochemical communication. In this study, the adaptive responses of herbivores were considered in the context of multi-trophic relationships, and the chemical ecology and behavior of both the herbivore and the parasitoid were investigated. Our research could contribute to our understanding of H. cunea invasion spreading, and results provided here could also improve biological control efficiency by promoting C. cunea parasitization.

2. Materials and Methods

2.1. Insects Rearing

C. cunea parasitoid wasps were obtained from the Natural Enemy Breeding Center of Luohe Central South Forestry (Henan, China) in 2012. Antheraea pernyi (Lepidoptera: Saturniidae) was the substitute host of C. cunea, obtained from the Benxi Tussah Breeding Base (Liaoning, China). The H. cunea eggs were obtained from the Chinese Academy of Forestry in 2020, where they were reared on an artificial diet after hatching. The hatched larvae were then separately reared till pupation on the leaves of mulberry (Morus alba L.), peach (Amygdalus persica L. var. persica f. duplex Rehd), empress tree (Paulownia tomentosa), tree of heaven (Ailanthus altissima (Mill.) Swingle), weeping willow (Salix babylonica), Korean aspen (Populus davidiana), Chinese ash (Fraxinus chinensis), and Old World sycamore (Platanus orientalis Linn.). The leaves of these eight plant species were collected on the day of the experiment at the Tianjin Normal University campus.

2.2. Larvae Performance on Host Plants

To test larvae performances on different host plants, three sets of 100 hatched larvae were fed with eight plant species’ leaves separately. As the body length and weight of H. cunea varied very little before the fourth instar, only the data after the 4th instar were collected. When the larvae reached the 4th instar (from the 12th day after hatching), 10 larvae were randomly selected from each cage every 2 days, and their body length and body weight were measured. The pupation rate was calculated after all the larvae pupated. The cages were kept in an incubator at 25 °C and 70% relative humidity (RH) with a 14:10 light: dark cycle.

2.3. Volatiles Collection

Before collecting volatiles, the leaves were immersed in water for two hours to remove biological contaminants and pesticides, after which they were allowed to air-dry naturally. Two groups were considered for dynamic volatile collections: the first group remained unexposed to herbivores prior to headspace collection; and the second was exposed to 5th instar larvae that had been deprived of food for 12 h. Following 6 h of leaf consumption by the larvae, the insects and their frass were removed, and headspace collection commenced.

For collecting volatiles for chromatography analysis, twenty grams of leaves were placed in each specialized 1 L glass conical flask. Purified room air, filtered through activated charcoal, was continuously passed through the system at a flow rate of 600 mL-min⁻¹. Volatile compounds were trapped using glass tubes containing 100 mg of Porapak™ Q (Waters Corporation, Shanghai, China), retained between silanized glass wool plugs, with an effluent flow rate of 500 mL-min⁻¹. The consistent positive pressure within the chamber ensured that no ambient air infiltrated the chamber. Volatile compounds were collected in the Porapak Q traps for 6 h and subsequently eluted with 1 mL of freshly distilled hexane. Through nitrogen evaporation, the sample was concentrated to 100 µL and preserved at −20 °C for subsequent chromatographic analysis.

2.4. Gas Chromatography and GC-MS

Volatiles were separated and quantified by Agilent GC 7890A, equipped with a DB-5MS column (30 m × 250 μm × 0.25 μm, Agilent Technologies, Santa Clara, CA, USA), with splitless injection (250 °C). The oven temperature was maintained at 50 °C for 1 min, then programmed at 4 °C min⁻¹ to 150 °C and subsequently held at this temperature for 1 min, and then programmed at 10 °C min⁻¹ to 230 °C, followed by a 10 min hold. The carrier gas was ultra-high purity helium (1.2 mL-min⁻¹).

The electrophysiologically active gas chromatography peaks were analyzed using an Agilent 7200 Q-TOF mass spectrometer (EI, 70 eV, source temperature 250 °C, m/z 50–650). Tentative identification of the C. cunea EAD-active compounds was based on (1) Kovats indices, (2) electron ionization mass spectra, and (3) comparison with authentic chemicals.

2.5. Gas Chromatography-Electroantennography

The electrophysiological preparation was initiated by excising a single antenna from a 1-day-old mated C. cunea female using microsurgical scissors under a magnifying lens. A glass capillary (0.5 mm inner diameter) filled with 0.1 M KCl solution was used as an electrode. The reference electrode was attached to the base of the isolated antenna, and the recording electrode was connected to the cut tip of the antenna. Chlorinated silver–silver chloride junctions were used to ensure electrical contact between the electrodes and input of the preamplifier.

The analog signal captured by a probe (INR-II; Syntech, Kirchzarten, Germany) was processed using a data acquisition controller (IDAC-232; Syntech) and was analyzed subsequently using customized software (GC-EAD 1.2.5, Syntech). For GC-EAD coupling, the GC column effluent was simultaneously split to the antennal preparation and flame ionization detector, FID. The collected volatiles were separated by GC (6820, Agilent Technologies, Stockport, UK) using a 30 m × 0.32 mm × 0.25 μm ID HP-5MS column (Agilent). Carrier gas flow (helium) and temperature programming followed the previously described GC-MS method. Both EAD amplifier output and the FID were monitored simultaneously and analyzed using Syntech software. A peak was deemed electrophysiologically active if it elicited responses on three or more independent antennal replicates.

2.6. Chemicals

Synthetic standards necessary for the confirmation of identity and behavioral bioassays were obtained from J&K Chemical Co. (Beijing, China) (nonadecane 99.5%, CAS 629-92-5; hexadecane 99%, CAS 544-76-3) and Aladdin Co. (Shanghai, China) (undecane ≥ 99%, CAS 1120-21-4; heneicosane > 99%, CAS 629-94-7; tridecane 99%, CAS 629-50-5; tetradecane 99%, CAS 629-59-4).

2.7. Behavioral Assays

2.7.1. Responses of Mated Female H. cunea to Different Plants

Due to the potential impact of plant leaf consumption on H. cunea behavior, H. cunea specimens were provided with an artificial diet. Individual leaves (not previously infested by H. cunea) of approximately equivalent mass (5 g) from eight plant species were cleansed with water and allowed to air-dry naturally. Two leaves were affixed to each side of a cage (30 × 30 × 30 cm) using adhesive tape. The experiments were performed in a darkened room illuminated by red light. Fifteen mated female moths were placed in three Petri dishes, which were positioned in the center of the cage. The Petri dishes were left uncovered, and moth behavior was observed for a duration of 30 min. For each of the eight replications, fresh plant leaves and new H. cunea females were utilized.

2.7.2. Responses of 1-Day-Old Mated C. cunea Females to Various Odor Sources

Both Y-tube (binary choice) and four-arm olfactometers were employed to assess the behavioral preferences of 1-day-old mated C. cunea females to various chemical stimuli. The Y-tube olfactometer consisted of a common glass tube (11.5 cm long × 2.2 cm diameter) and two lateral glass arms (7.5 cm long × 2.2 cm diameter). A purified air stream (200 mL-min⁻¹) was generated by an air pump through activated charcoal and doubly distilled deionized water. The experiment was conducted in a dark room. A lamp (25 W, 250 lx) was positioned 55 cm above the olfactometer for illumination.

The four-arm olfactometer (QT-WII01, Channel Technology Co., Beijing, China) was constructed with an acrylic chamber (10 × 10 cm) partitioned into four zones. The top of the chamber was covered with a glass plate to prevent insect escape. The experiment was conducted in a dark room. A light bulb was positioned on top of the arena, such that the amount of illumination (∼250 lx) was consistent for all arms.

The experiments involved the following comparisons, all examining the attraction of C. cunea (i) to different plant leaves before and after they were consumed by H. cunea larvae; (ii) to the eight plant leaves after they were consumed by H. cunea larvae; (iii) to single compounds at different doses versus hexane (carrier solvent); and (iv) to different synthetic blends (see detailed methods below).

(i)

In total, 5 g of uncontaminated plant leaves were positioned in one arm of the Y-tube olfactometer, while 5 g of leaves from the same plant, consumed by H. cunea, were placed in the other arm. The leaves were separated using 70 mesh absorbent cotton gauze to prevent physical contact. Forty mated female C. cunea (1-day-old) were released at the base of the common arm of the Y-tube and monitored for a maximum of 5 min. Females failing to exhibit orientation behavior within 5 min were excluded from the analysis. Females that traversed 1 cm beyond the Y-junction and remained there for at least 10 s were recorded to have made a selection. Following each assay, the tube was cleansed with soap, water, hexane, and air dried, and its orientation was reversed. Five independent biological tests were performed for each odor source. Each parasitoid was utilized only once. All bioassays were conducted at 25 °C and 60% relative humidity (RH). The female response percentage was calculated as follows: Rs = 100% × (parasitoids that selected the treatment arm)/(parasitoids that selected the treatment arm + parasitoids that selected the control arm).

(ii)

The responses of C. cunea to the eight different plant leaves after they were consumed by H. cunea were investigated using a four-arm olfactometer. The experiment was conducted as follows: group 1: the attraction of four plants to the wasps was compared; group 2: attraction to the other four plants was compared; group 3: the two most attractive plants in group 1 were compared with the two most attractive plants in group 2; and group 4: the two least attractive plants in group 1 were compared with group 2 again. Leaves (5 g) consumed by H. cunea were placed inside the chambers of four randomly selected arms of the olfactometer. Forty 1-day-old mated female C. cunea were introduced in the center of the arena, and their selection of the four arms was recorded. A selection was defined as entering one of the four arms and remaining there for at least 10 s. The odor sources in the olfactometer were replaced, and their positions were altered at every assay. Additionally, the entire arena and odor chambers were cleansed with soapy water followed by hexane and then air-dried. Each parasitoid was utilized only once. Nine independent biological tests were conducted, and the bioassays were performed at 25 °C and 60% RH.

(iii)

For chemical standard testing, test compounds at various concentrations (10, 100, 1000, and 10,000 ng/μL, 10 μL) were applied to filter papers. After a 20 s period to allow for solvent evaporation, the filter papers were positioned in one arm of the Y-tube olfactometer. A corresponding filter paper with 10 μL of hexane was placed in the opposite arm (solvent control). The bioassay procedures with C. cunea were conducted as previously described.

(iv)

The responses of C. cunea to 10 μL of synthetic blends of EAD-active compounds were examined using a four-arm olfactometer. Filter papers containing synthetic blends of EAD-active compounds (10 ng/uL for each component), synthetic blends (excluding tridecane), and hexane were positioned within the chambers of three randomly selected arms of the olfactometer. The fourth arm received 10 μL hexane and functioned as a solvent control. The bioassay procedures with C. cunea were conducted as previously described.

2.7.3. Responses of Female Adult H. cunea to Different Compounds

The olfactory responses of mated H. cunea females to various compounds were examined utilizing a Y-tube olfactometer. In group 1, a filter paper containing 10 μL tridecane (10 ng/μL) was positioned in one arm of the Y-tube olfactometer, while hexane (solvent control) was placed in the other arm. For group 2, a filter paper with a synthetic blend of EAD-active compounds excluding tridecane was situated in one arm of the Y-tube olfactometer, and 10 μL of the identical synthetic blend including tridecane (10 ng/μL for each component) was positioned in the other arm. The bioassay procedures for C. cunea were conducted in accordance with the aforementioned protocol.

2.8. Statistical Analysis

One-way analysis of variance followed by a Tukey post hoc multiple comparison test was used to compare (i) the performance of H. cunea larvae (body length, body weight, and pupation rates) fed on different plant species leaves (Section 2.2); (ii) the preferences of mated female H. cunea to different host plants (Section 2.7.1); (iii) the preferences of C. cunea to different plant leaves that had been fed upon by H. cunea in a four-arm olfactometer (Section 2.7.2, ii); and (iv) the responses of C. cunea to different synthetic blends of EAD-active compounds in a four-arm olfactometer (Section 2.7.2, iv). All behavioral responses of female C. cunea in the Y-tube olfactometer were analyzed using the Wilcoxon rank sum test, including (i) the responses of C. cunea to the leaves of eight plant species before and after H. cunea feeding (Section 2.7.2, i); (ii) the responses of C. cunea females to different concentrations of tridecane (Section 2.7.2, iii); and (iii) the responses of H. cunea to hexane control and tridecane (Section 2.7.3). All analyses were performed in R 4.3.0 (www.r-project.org).

3. Results

3.1. Growth of H. cunea Fed Different Plant Leaves

Initially, no significant differences were observed in the length and weight of larvae feeding on the leaves of different plants (p > 0.05, 12th day after hatching). Subsequently, H. cunea specimens that consumed the leaves of A. altissima, F. chinensis, and P. tomentosa exhibited greater length and weight compared to those feeding on M. alba, P. davidiana, P. orientalis, A. persica, and S. babylonica (p < 0.05) (Figure 1). For example, on the 18th day after hatching, H. cunea larvae feeding on A. altissima exhibited optimal growth performance, with a mean body length of 25.47 ± 1.19 mm and body weight of 0.20 ± 0.00 g. Those consuming M. alba leaves measured 23.77 ± 0.72 mm in body length and 0.18 ± 0.01 g in weight. Larvae reared on A. persica showed the smallest dimensions, with a mean body length of 20.97 ± 0.76 mm and weight of 0.13 ± 0.01 g. Pupation rate demonstrated a similar trend, with significantly higher pupation rates observed in H. cunea specimens that consumed the leaves of A. altissima, F. chinensis, and P. tomentosa compared to those that fed on other plant species (p < 0.05) (Figure 2). These findings suggest that A. altissima, F. chinensis, and P. tomentosa provided superior nutritional content for the growth and development of H. cunea compared to M. alba and the other plant species examined.

3.2. H. cunea Preference of Eight Plant Species

Mated female H. cunea selected plants inconsistent with larval growth parameters (Figure 1 and Figure 2). The preference hierarchy was as follows: 1, M. alba; 2, P. tomentosa, A. altissima, and F. chinensis; 3, P. orientalis and P. davidiana; and 4, S. babylonica and A. persica (p < 0.05) (Figure 3). The preferred host plant of female H. cunea (M. alba) is not optimal for larval growth and development.

3.3. Responses of C. cunea to Leaves Before and After Feeding by H. cunea

The leaves of all eight plant species exhibited significantly higher attractiveness to C. cunea after H. cunea feeding compared to before feeding (F. chinensis, p = 0.009; P. orientalis, p = 0.008; A. altissima, p = 0.009; P. davidiana, p = 0.009; M. alba, p = 0.009; P. tomentosa, p = 0.009; A. persica, p = 0.009; S. babylonica, p = 0.009) (Figure 4).

3.4. Comparison of Responses of C. cunea to Different Plant Leaves Fed upon by H. cunea

The four-arm olfactometer was utilized to assess C. cunea preferences for different plant leaves that had been subjected to H. cunea herbivory. Four experimental groups were established to compare the attraction of C. cunea to various plant species (Figure 5):

Group I: F. chinensis (34.21%) and A. altissima (32.29%) elicited greater attraction of C. cunea compared to P. tomentosa (21.45%) and M. alba (12.05%).

Group II: P. orientalis (32.67%) and P. davidiana (31.75%) demonstrated higher attraction of C. cunea in comparison to A. persica (16.99%) and S. babylonica (18.60%).

Group III: The four most attractive plants in groups I and II were subjected to comparative analysis. No statistically significant differences were observed in the attractiveness of F. chinensis, A. altissima, P. orientalis, and P. davidiana (p = 0.395).

Group IV: The four least attractive plants in groups I and II were compared. P. tomentosa (38.26%) exhibited significantly higher attraction of C. cunea compared to A. persica (24.63%) and S. babylonica (26.46%), while M. alba attracted the fewest C. cunea (10.65%).

Consequently, the attractiveness to C. cunea of host plants subjected to H. cunea herbivory can be categorized into four levels. F. chinensis, A. altissima, P. orientalis, and P. davidiana exhibited the highest attractiveness to C. cunea. P. tomentosa, A. persica, and S. babylonica demonstrated significantly lower attractiveness compared to the aforementioned five plant species. M. alba, which is the most preferred host of H. cunea, exhibited the least attractiveness to C. cunea.

3.5. Chemical Analysis

The leaf volatiles of eight different plant species before and after H. cunea herbivory were analyzed using GC-EAD and GC-MS techniques. Six compounds (undecane, tridecane, tetradecane, hexadecane, nonadecane, and heneicosane) present in the plant leaf volatiles elicited EAD responses in the antennae of C. cunea females. Among the eight plants examined, tridecane was not detected solely in M. alba (

Table 1. Compounds were identified by GC peak enhancement using authentic standards on an Agilent DB-5 column.

	Compound	Retention Time	Emission Rate (ng/min from 1 g Leaves)
			M. alba	P. davidiana	P. orientalis	A. altissima	S. babylonica	A. persica	P. tomentosa	F. chinensis
1	Undecane	8.85	31.48	9.02	17.87	44.36	16.52	15.70	17.51	19.07
2	Tridecane	11.43	N	30.76	30.26	30.40	4.98	9.12	21.10	27.72
3	Tetradecane	12.60	22.85	15.11	11.41	19.25	13.70	19.91	14.11	35.97
4	Hexadecane	14.86	20.63	15.98	23.20	42.31	13.50	21.01	17.21	13.36
5	Nonadecane	17.85	17.23	8.91	21.80	24.60	13.68	15.68	13.62	14.68
6	Heneicosane	24.81	17.17	11.47	18.57	20.24	13.42	13.95	13.56	22.73

).

3.6. Olfactometer Bioassays

3.6.1. Tridecane vs. Hexane

As tridecane, a HPIV in seven of the eight host plants, was not detected solely in M. alba leaves, the effect of tridecane on the wasp was evaluated utilizing a two-choice Y-tube olfactometer. In comparison with hexane, none of the concentrations of tridecane elicited significant attraction or repellency in the wasp (10 ng/µL: p = 0.288; 100 ng/µL: p = 0.307; 1000 ng/µL: p = 0.353; 10,000 ng/µL: p = 0.964) (Figure 6).

3.6.2. Responses of C. cunea to Synthetic Blends of EAD-Active Compounds

Employing a four-arm olfactometer, the blank arm and the hexane control attracted minimal wasps, whereas the synthetic blend of hydrocarbons demonstrated high attractiveness. However, the removal of tridecane from the blend resulted in a significant reduction in attraction to the blend (p < 0.05) (Figure 7).

3.6.3. Responses of H. cunea Females to Different Compounds

Similarly to C. cunea females, H. cunea females exhibited no significant preference for tridecane over the hexane control (p = 0.965) (Figure 8A). However, while C. cunea females demonstrated a preference for the hydrocarbon blend with tridecane, H. cunea females significantly preferred the synthetic blend without tridecane (p < 0.0001) (Figure 8B).

4. Discussion

Based on the ‘enemy-free space’ hypothesis, insects may exhibit a preference for host plants that are nutritionally suboptimal for growth and development but are less frequented by their natural enemies, thereby reducing exposure to predators and parasitoids. In this study, H. cunea larvae were reared on leaves of eight host plant species, including M. alba, A. persica, P. tomentosa, A. altissima, S. babylonica, P. davidiana, F. chinensis, and P. orientalis. Previous findings indicated that M. alba, P. davidiana, P. tomentosa, A. altissima, F. chinensis, and P. orientalis were the preferred host plants of H. cunea [35], whereas S. babylonica and A. persica were less preferred host plants [36].

Our study results demonstrated that H. cunea larvae achieved greater body length, body mass, and pupation rate when feeding on the leaves of A. altissima, F. chinensis, and P. tomentosa compared to the leaves of M. alba, followed by P. davidiana, P. orientalis, A. persica, and S. babylonica. Thus, our assays support the hypothesis that A. altissima, F. chinensis, and P. tomentosa provided superior nutrition for the growth and development of H. cunea compared to M. alba and the other plant species.

In contrast to their growth performance on the various plant species, H. cunea females exhibited the highest attraction to M. alba, followed by P. tomentosa, A. altissima, F. chinensis, P. orientalis, P. davidiana, S. babylonica, and A. persica. Yu (2016) [37] observed that H. cunea oviposited more eggs on M. alba than on 11 other host plant species, which is consistent with our preference results.

The preferred host plant of female H. cunea (M. alba) is not its optimal nutritional host for growth and development. Conversely, the optimal nutritional hosts (A. altissima, F. chinensis, and P. Tomentosa) are not the most preferred host plants. A plausible explanation for this discrepancy between behavioral preference and physiological performance is the influence of the third trophic level—natural enemies that exert selection pressure on behavioral preferences.

Herbivore-induced plant volatiles (HIPVs) play crucial roles in tri-trophic interactions [3,38]. Therefore, we initially investigated the HIPVs released by plants in response to H. cunea feeding that attract the parasitoid C. cunea. C. cunea exhibited significantly higher attraction to all eight plant species after feeding by H. cunea compared to before feeding. Thus, it is evident that herbivory by H. cunea induces the plants to release HIPVs that subsequently attract the natural enemy C. cunea.

Subsequently, we compared the attraction capabilities of C. cunea to various HIPVs. C. cunea exhibited the highest attraction to H. cunea-fed F. chinensis, A. altissima, P. orientalis, and P. davidiana, followed by P. tomentosa, A. persica, and S. babylonica. M. alba demonstrated the least attraction for C. cunea based on its HIPVs. Consequently, M. alba, which is the most preferred host of H. cunea, is the least attractive to C. cunea. This phenomenon may elucidate why M. alba is the most preferred host of H. cunea, despite not being its optimal nutritional host for growth and development. A. persica and S. babylonica exhibited weak attraction to C. cunea, yet H. cunea still did not prefer these plant species, potentially due to their poor nutritional value for growth and development. Therefore, it appears that H. cunea engages in a trade-off between performance and survival, preferring a host plant with suboptimal nutrition in favor of mitigating parasitism. A similar correlation between plant quality and larval parasitism was observed in field investigations. H. cunea larvae reared on higher-quality host plants also showed a greater proportion of mortality attributable to parasitism [39].

To identify the biologically active HIPVs, we analyzed the leaf volatiles of eight different plant species before and after H. cunea feeding. GC-EAD, followed by GC-MS analysis, identified six electrophysiologically active saturated hydrocarbons (undecane, tridecane, tetradecane, hexadecane, nonadecane, and heneicosane) [40].

HIPVs comprise a variety of divergent VOCs, including alkanes, alkenes, aldehydes, alcohols, ketones, ethers, esters, and carboxylic acids [40]. The HIPVs identified in this study are all alkanes. We hypothesize that in addition to these alkanes, there might be a variety of other compounds present. In subsequent research, we will continue to investigate other active compounds that guide host preference. The alkanes identified in this study have also been observed in other studies as active components in HIPVs. Undecane has been demonstrated to be a HIPV from immature fruits of Java plum Syzygium cuminii [41]. Tetradecane, as a major bioactive component of HIPVs produced by fruits, has been shown to regulate the behavioral response of the egg–pupal endoparasitoid Fopius arisanus [42]. The parasitoid Microplitis similis selection of cabbage was positively correlated with the release of hexadecane, as one of the HIPV components to defend against Spodoptera litura [43]. Cucurbita maxima damaged by Aulacophora foveicollis released nonadecane [44]. Nicotiana tabacum, infested by sap-sucking aphid Myzus persicae Sulzer, produced heneicosane as a HIPV [45].

Of the six compounds, tridecane was not detected in only one host plant, M. alba. In several studies, tridecane has been demonstrated to be a herbivore-induced plant volatile (HIPV) that attracts natural enemies. For instance, tobacco plants (Nicotiana tabacum L.), infested by sap-sucking aphids (Myzus persicae Sulzer) produce tridecane and other compounds as HIPVs [45]. The Asian egg parasitoid Trissolcus japonicas utilizes tridecane as the kairomone of the brown marmorated stink bug, Halyomorpha haly, for short-range host location [46]. Additionally, the predaceous minute pirate bug, Orius insidiosus (Say), employs tridecane as a kairomone to locate the brown marmorated stink bug [47]. Tridecane is also a component of the alarm volatiles of stinkbugs (Tessaratoma papillosa), released when stinkbugs are disturbed or irritated [48]. Furthermore, tridecane is a component of the pygidial gland of larval secretions, playing a role in chemical defense against predators such as ants [49].

Our investigation revealed opposite responses to HPIVs by the herbivore (H. cunea) and its parasitoid (C. cunea). In isolation, tridecane, a prominent HPIV in seven of the eight host plants, did not elicit any behavioral responses, either attraction or repellency, in H. cunea or C. cunea. However, when combined with other HPIV compounds, the mixture that included tridecane significantly attracted C. cunea and repelled H. cunea. M. alba, the only host plant that did not emit tridecane in its HPIVs, did not attract the wasps and was preferred by the moth over the other seven host plants. Most insect olfactory receptors require cooperative binding of multiple volatile molecules to trigger responses. A single chemical is typically insufficient to activate the olfactory receptor channel [50]. Thus, when acting alone, tridecane exhibits insufficient binding affinity, whereas other compounds in the mixture may provide auxiliary binding sites that collectively reach the receptor activation threshold.

It is generally accepted that natural enemies rely on HIPVs for long-range localization of an appropriate habitat [51] and the host insect’s volatiles for the short-range location of the host [52,53]. In this investigation, we identified tridecane, hexadecane, nonadecane, and heneicosane as HIPVs that attracted C. cunea wasps. In a previous study, we identified six compounds (1-dodecene, hexadecane, heptadecane, nonadecane, heneicosane, and heptacosane) emitted from the pupae of three lepidopteran species, including H. cunea, that serve as hosts for C. cunea [54]. It is particularly noteworthy that both behaviorally active blends (from plant hosts and insect hosts) comprise hydrocarbons with similar characteristics. We therefore postulate that the semiochemicals released by plants and those released by host insects may be chemically related and thus perhaps represent the convergence of tritrophic systems on suites of related compounds. Identifying common semiochemicals may be a more efficient strategy for C. cunea, corresponding to the parasitoid wasp’s limited neural capacity [55]. The responses of the mated female H. cunea to different plant leaves pre-infested by H. cunea will be examined in subsequent research, which is crucial for developing practical applications in the field ecosystem. It is also of interest whether the leaf volatiles (HIPVs) before and after feeding by other herbivores consist of the same emissions, or if H. cunea elicits specific HIPVs.

5. Conclusions

The main findings of this study are summarized as follows: The preferred host plant of female H. cunea was M. alba. However, compared with other host plants, M. alba was not the optimal nutritional host for H. cunea growth and development. Compared with other host plants, M. alba was the least attractive to the parasitoid C. cunea, a natural enemy of H. cunea. GC-EAD and GC-MS analyses identified six volatiles as HIPVs of different host plants fed upon by H. cunea. These HIPVs elicited EAD responses in the antennae of female C. cunea. Tridecane was emitted by seven of eight host plant species, but not by M. alba. Tridecane alone did not attract or repel either C. cunea or H. cunea, but when combined with other HPIVs, the mixture significantly attracted C. cunea and repelled H. cunea. M. alba represents “enemy-less space” for H. cunea. By preferring M. alba, which does not emit tridecane, H. cunea effectively avoids parasitism by C. cunea, which is highly attracted to HIPVs that contain tridecane.