Hyphantria cunea (Drury) Showed a Stronger Oviposition Preference for Native Plants after Invading the Subtropical Region of China

Abstract

Hyphantria cunea (Drury) (Lepidoptera: Erebidae) is an invasive alien species that is widely distributed in northern China. H. cunea now occurs for the first time in the subtropical areas of China. Despite the importance of identifying oviposition host plants to control the spread of H. cunea, it is not clear whether H. cunea has a new oviposition preference for plant hosts in the new habitat after invading the subtropical region. At the same time, whether the presence of new plant volatiles of new hosts in new habitats influences the oviposition host preference of H. cunea has not been studied. In this study, we investigated the oviposition host plant preferences of H. cunea in the subtropical region of China. In the presence of multiple potential host plants, we found, for the first time, that H. cunea preferred Carya illinoinensis, Morus alba, and Carya cathayensis for oviposition. Through the effects on plant volatiles and on the oviposition, ten volatile compounds with significant differences in relative abundance between five plants with different fitness levels were significantly correlated with the oviposition preference of females. Cis-Hex-3-en-1-ol, (E)-3-Hexen-1-ol, cis-3-Hexenyl acetate 1, and 3-Pentadiene,4-methyl, had a significant positive correlation with the adult oviposition preference. Our results provide an important research basis for the study of green prevention and control strategies of subtropical H. cunea in China.

1. Introduction

Oviposition is an important factor in the reproduction and population growth of herbivorous insects. To ensure that larvae have sufficient food for growth, females often select ideal host plants for oviposition [1]. Some recent studies on invasive species have shown that understanding their host preferences in newly invaded areas is critical for developing pest management strategies [2,3]. For females, host plant choice reflects a co-evolved adaptive relationship between herbivorous insects and plants. It also determines the exploitation strategy of herbivorous insects towards plants and can influence the population growth of the next generation of larvae [4]. In addition, the larval stage is the key stage that damages plants and the key time for population management [5]. Thus, studying how females choose their hosts can help us better understand how insects and plants interact, and can provide a new vision for green prevention and control.

The fall webworm, Hyphantria cunea (Drury) (Lepidoptera: Erebidae), is an important quarantine pest [6,7,8,9]. It is highly polyphagous and has a high reproductive capacity [9,10]. The subtropical region of China is rich in forest resources. It has a humid subtropical monsoon climate, which is suitable for the survival of H. cunea [9]. In July 2020, adult H. cunea was detected for the first time in Zhejiang Province, a subtropical region of China [11]. Due to the lack of tailored prevention and control strategies, H. cunea now appears to be spreading to the southern part of Zhejiang province and will threaten the local environment and economy [12]. To occupy a niche in a new habitat, invasive alien insects should adapt to local plant hosts to improve the population base [13]. Therefore, it is important to assess the oviposition preference of H. cunea in the newly invaded area. This information can help us understand its host selection mechanism and provide a research basis for population management of H. cunea in the subtropical region of China.

In addition, the fall webworm can feed on >600 species of trees and shrubs, and its preferences vary among the different host plants [9,14,15]. In Shanghai (a city located in a subtropical region of China), H. cunea can feed on conifer species, such as Metasequoia glyptostroboides, Taxodium distichum, and Taxodium ascendens, which have been caused severe damage [16]. The area of first invasion in Zhejiang Province is adjacent to Jinshan District, Shanghai, and has similar climatic characteristics. In this area, most street trees are conifers, such as M. glyptostroboides [16]. What is more, large areas of Carya cathayensis and Carya illinoinensis are cultivated in the mountainous areas northwest of Zhejiang [12]. Based on the host plant distribution in the subtropical region of China and the different host preferences of H. cunea, we selected five different host plants (Morus alba, Populus simonii, C. cathayensis, C. illinoinensis, and M. glyptostroboides) to study oviposition preference. According to the results of previous studies, females lay eggs on suitable hosts to provide food for their larvae, which ultimately sustains population expansion [17,18]. At the same time, there are many factors that influence oviposition preference in phytophagous insects, and plant volatiles play a dominant role in the selection of oviposition host selection [19]. Therefore, what role do the volatiles of the new host, planted in the subtropical climate of China, play in the oviposition preference of H. cunea invading new habitats? Answering these questions is also important for our understanding of the host adaptation mechanism of H. cunea.

2. Materials and Methods

2.1. Insect Specimens

H. cunea larvae were obtained from Jiashan County, Zhejiang Province. The rearing area was maintained at 25 ± 1 °C, 16:8 h (L:D) photoperiod, and 70–80% relative humidity in nylon cages, and the larvae were fed artificial diets [20]. The diets were obtained from the Chinese Academy of Forestry (Beijing, China). Mature larvae were randomly selected and allowed to pupate. In addition, adult females are more active before mating than after mating [21], indicating that pre-mating is a good time to search for host plants. We used unmated H. cunea for the study of oviposition trends indoors.

2.2. Plant Material and Host Selection

Two-year-old seedlings of five host plants, including P. simonii (PL), M. alba (MA), C. illinoinensis (CI), C. cathayensis (CC), and M. glyptostroboides (MG), were used for oviposition tests and extraction of plant volatiles. The five plant hosts were selected based on the findings of HC damage during the ongoing 2020–2022 survey. These seedlings were planted into a 3:1 ratio of field soil and nursery substrate and placed in the plastic greenhouse at Zhejiang A&F University.

2.3. Determination of Tropism of Y Tube in the Laboratory

The Y-tube olfactometer was constructed according to the method of Guo et al. (2020) [22]. It consisted of two conical flares, two conical funnels, a rectangular container, and several bidirectional glass tubes (Figure 1). A total of ten adult females were simultaneously placed in the rectangular glass containers, and leaves from different host plants were placed in the two conical flasks. The volatiles released by the plant were continuously moved by an extraction pump. The experiment lasted from 9:00 a.m. to 9:00 p.m. and the locations of H. cunea were recorded every two hours for a total of seven measurements. For the five selected host plants, we divided them into ten groups based on different combinations. Each group contained two different host plants, and then the homogenates from the leaves of two plants were placed in the conical flasks. A total of ten groups of preference tests were performed, including a (CC vs. CI), b (CC vs. MA), c (CC vs. MG), d (CC vs. PL), e (CI vs. MA), f (CI vs. MG), g (CI vs. PL), h (MA vs. MG), i (MA vs. PL), j (MG vs. PL).

2.4. Oviposition Host Selection in the Field

Five host plants (one per species) were placed evenly in different directions of the circular net cage two meter in diameter, with each host plant separated by about 30 cm. Fifteen post-mating adult females were selected and placed in the center of the cage. Observations were recorded at 10:00 a.m. on the second day. We recorded the numbers of females resting on each host plant every 12 h, for a total of seven consecutive recordings. After each count, the position of these plants was moved clockwise. Finally, the number of egg masses on different host plants was counted. Adults were collected from each generation throughout the annual life history (from early June to late September), and this experiment was repeated three times.

2.5. Extraction of Volatiles from Host Plants

To collect plant volatiles in the wild, plastic collection bags were placed on appropriate plant leaves. The PDMS (polydimethylsiloxane, 50 μm) extractor was then activated at 250 °C for 5 min. After the volatiles released from the host plants reached a dynamic equilibrium, the activated extraction head was quickly inserted into the collection bag for extraction for one hour. After that, the extraction head was inserted into the inlet end of the gas chromatography–mass spectrometry device (Agilent 6890N-5973NGC/MSD), and then gas chromatography-mass spectrometry (GC-MS) analysis was performed. The volatiles from five host plants were analyzed by GC-MS. The carrier gas was high-purity helium gas and the gas flow was 1 mL·min⁻¹. The chromatographic column was an HP-5MS capillary column (30 mm (long) × 0.25 mm (ID) × 0.25 μm (film)). The column temperature heating procedure was performed as described below: 40 °C was held constant for 1 min, heated at a rate of 4 °C·min⁻¹ to 220 °C, and then the temperature was held for 3 min, and the total running time was 49 min. The temperature of the detector was set as 280 °C and the temperature of the forward sample was set as 220 °C. The temperature of the four-stage bar was set at 150 °C, and the temperature of the ion source was set at 230 °C. The ionization mode was set to EI and the mass spectrometry bombardment voltage was set to 70 eV. The scanning mass range was set as 35–450 m·z⁻¹. For qualitative analysis, the mass spectrometry data of these components were automatically analyzed by GC and the NIST08 spectral database. All search results were checked and supplemented by the standard chromatograms. The relative content of each component was determined by normalization of the peak area.

2.6. Statistical Analysis

Based on the position and residence time of the females on the mesh surface of the Y-tube, the scatter density plot was generated using the MASS package and the ggplot2 package of R 4.2.1. Correlation analysis between the number of egg masses and the number of adults was performed using Scatterstats 0.1.1. For quantitative analysis of GC-MS, peak area normalization was performed to determine the relative content of each component and edgeR package and DESeq2 package of R 4.2.1 were used to analyze these differential compounds. The ggplot2 package was used to draw the volcano map to show these differential compounds, and the significance of the residence time fitting for different compounds from different sources was shown using the Sankey diagram. The fitting curves of adult number and volatile peak area were performed using R 4.2.1.

3. Results

3.1. Oviposition Trend Study

Using a Y-tube olfactometer to evaluate the preferences of moths based on the location chosen by the adult females in each area, we found that six groups, c (CC vs. MG) e (CI vs. MA) f (CI vs. MG) g (CI vs. PL) I (MA vs. PL), and j (MG vs. PL), showed significant differences (p < 0.05), and the differences of the other four groups were not significant (p > 0.05) (Figure 2). It is worth noting that the higher the density of blue dots, the stronger the female preference for the host plant. Comparing M. glyptostroboides (MG) with the other four plants (Figure 2c,f,h,j), the dots of MG were more dispersed. However, this kind of dispersion of dots was not found when the pairwise comparisons were performed for the other four plant species, except MG. The results confirmed that M. glyptostroboides was not a preferred host plant of H. cunea adult females.

We further evaluated the oviposition preference of adults with the number of females on different hosts and egg masses in the wild. Compared to PL and MG, adult females showed a significant preference for remaining in the zone covered by MA, CC, and CI (Figure 3a). For the number of egg masses, adults also showed a significant attraction to MA, CC, and CI compared to MG (Figure 3b). In addition, the number of females on different hosts and egg mass showed a significant positive correlation (Figure 3c). Thus, differences in the attractiveness of different host plants to adult females was one of the key factors determining the number of larvae surviving on these host plants. MG was a less attractive host for the fall webworm, both in terms of number of females and egg mass.

3.2. Components and Relative Contents of Volatiles in Five Host Plants

The components of volatiles extracted from M. alba, p. simonii, M. glyptostroboides, C. illinoinensis, and C. cathayensis were tested using GC-MS system. The chromatograms of the five plants are shown in Figure 4. After excluding the impurity peaks, the chromatogram represents the volatile components from the leaves of the five host plants.

3.3. Identification and Analysis of Volatile Components among Different Host Plants

We performed compound differential analysis to determine the difference in volatiles in different plants and found that a total of 112 specific enrichment compounds were associated with oviposition preference of H. cunea (Figure 5, CAS numbers in order of difference significance). The different color of the dots (red or blue) was used to indicate the difference in odorant volatiles between the different host plants. For group b (CI vs. MA), the number of specific enriched compounds was the lowest (five compounds). The highest number (24 compounds) was found in group h (MA vs. MG). In addition, we found that the species of group a (CI vs. CC) both belonged to the genus Carya, but the results showed that CI had 11 pecific enrichment compounds than CC. Furthermore, the results of comparisons among CC, CI, and MA, including group a (CI vs. CA), group b (CI vs. MA), and group e (MA vs. CC), showed a low number of different compounds, all of which were less than 11 (Figure 5). However, when MG and PL were compared with other host plants, the number of different compounds ranged from 15 to 24, showing a higher number (Figure 5).

3.4. Fitting Analysis of Molecular Weight of Volatiles of Host Plants and Residence Number of H. cunea

Based on the results of the volcanic map analysis, 20 odorant volatiles associated with H. cunea oviposition and significantly enriched from different host plants were selected (Figure 6). (Z)-13,7-dimethyl-3,6-octatriene (CAS ID: 003338-55-4), trans-3-hexenyl acetate (CAS ID: 003681-82-1), trans-3-hexene-1-ol (CAS ID: 000544-12-7), beta-elemene (CAS ID: 000515-13-9), trans-2-hexenal (CAS ID: 006728-26-3), cis-3-hexenyl acetate (CAS ID: 003681-71-8), 4-methyl-1,3-pentadiene (CAS ID: 000926-56-7), cis-3-hexenol (CAS ID: 000928-96-1), trans-3-hexenol (CAS ID: 000928-97-2), and l-caryophyllene (CAS ID: 000087-44-5) were found in two or more plant species. Bornyl acetate (CAS ID: 000076-49-3), salicylaldehyde (CAS ID: 000090-02-8), 3-hexenal (CAS ID: 004440-65-7), 1-Propene,2-methyl-3-(2-propenyloxy)-(9CI) (CAS ID: 014289-96-4), 3-carene (CAS ID: 013466-78-9), l-alpha-pinene (CAS ID: 007785-26-4), g-terpinene (CAS ID: 000099-84-3), myrcene (CAS ID: 000123-35-3), m-cymene (000535-77-3), and 2-carene (CAS ID: 000554-61-0) were found in only one plant species.

To further determine the relationship between these compounds and oviposition preference, we performed a fitting analysis to H. cunea residence number and molecular weights of these 20 volatiles (Figure 6). We found 10 compounds with a significant difference (p < 0.05), and what is interesting is that these 10 compounds were mainly extracted from C. illinoinensis and M. glyptostroboides. Five volatile compounds, cis-3-hexenol (000928-96-1), trans-3-hexenol (000928-97-2), cis-3-hexenol acetate (003681-71-8), 4-methyl-1, 3-pentadiene (000926-56-7), and l-alpha-pinene (007785-26-4), showed obvious attraction (p < 0.01), and l-alpha-pinene from M. glyptostroboides was negatively correlated with residence number. In addition, the other five compounds, 1-caryophyllene (000087-44-5), g-terpinene (000099-84-3), myrcene (000123-35-3), m-cymene (000535-77-3), and 2-carene (000554-61-0), all showed a negative correlation (p < 0.05), suggesting that these compounds may be repellent to H. cunea.

4. Discussion

The oviposition behavior of herbivorous insects is the result of long-term co-evolution. Host plant volatiles, a factor of insect–plant interaction, play a key role in host recognition [23,24,25]. Some notorious invasive pests of Lepidoptera, such as H. cunea, Spodoptera frugiperda, and Ostrinia furnacalis, have been the subject of many studies on oviposition preferences, and these Lepidoptera species were inclined to choose suitable oviposition sites that were suitable for larvae and avoid plants that were unfavorable for larval growth and development [26,27,28,29,30,31]. Since 1979, H. cunea has invaded China, its range has continued to expand, and its invasion is now widespread. In northern China, H. cunea prefers M. alba, Acer saccharum, P. simonii, Ailanthus altissima, and many other deciduous tree species [32]. In this study, we found that H. cunea females from Zhejiang, the newly invaded population, showed no obvious preference for M. glyptostroboides, and the new population mainly preferred C. cathayensis, M. alba, and C. illinoinensis (Figure 2 and Figure 3). These results show that there may be a difference between feeding preference and oviposition preference, and the host range of feeding preference may be even wider. This would be consistent with previous research that found the expansion of GRs (gustatory receptor genes) of H. cunea, and this expansion may help larvae to feed on a wider variety of plants [9]. In addition, the volatiles with significant preference were mainly isolated from C. cathayensis, M. alba, and C. illinoinensis, and only a few were isolated from M. glyptostroboides and P. simonii (Figure 6). Therefore, M. glyptostroboides may not be a preferred species for oviposition of adult females.

Currently, H. cunea occurs in the northernmost part of Zhejiang Province, but it may eventually invade other cities in the subtropical region of China. The freshwater-rich southern cities are more susceptible to pesticide pollution, which increases the risk of injury to humans, birds, and silkworms [33]. Therefore, the use of chemical controls should be avoided in these cities. In addition, plant-based attractants, which are not harmful to the environment or organisms, should be further developed and used [34,35]. Herbivorous insects can recognize and locate host plants by the volatile odors released by the plants [36]. Studies on coleopteran insects have shown that 3-carene and nonanal are plant-derived pheromones that stimulate oviposition in Anoplophora glabripennis [37]. Host plant volatiles are closely related to insect oviposition. Many insect species use the scent volatiles released by host plants to find suitable oviposition sites. For example, cotton (Gossypium spp.) and tomato (Lycopersicon esculentum), hosts of Helicoverpa zea, emit volatiles that induce host-oriented behavior in mated females and stimulate females to oviposit. Host odors in undamaged leaves of carrot (Daucus carota) can induce Papilio polyxenes to oviposit [38,39].

Regarding the volatiles found in this study, a total of 112 specific enrichment compounds were detected in five host plants. It is worth noting that the least number of enriched compounds were detected between C. illinoinensis (CI) and M. alba (MA), and many female adults and eggs were found in the two host plants. However, there was a preference difference between CI and MA for larvae [40]. A larval preference study found that black-headed larvae preferred MA, while red-headed larvae preferred CI, and only black-headed larvae are now found in China [40]. The reason for this difference may be that larvae can select plants by taste, but adults can only select hosts by plant volatiles.

Our present results suggest that four compounds (cis-3-hexenol (000928-96-1), trans-3-hexenol (000928-97-2), cis-3-hexenol acetate (003681-71-8), and 4-methyl-1, 3-pentadiene (000926-56-7)) are attractive to H. Cunea female adults, and six compounds (l-alpha-pinene (007785-26-4), 1-caryophyllene (000087-44-5), g-terpinene (000099-84-3), myrcene (000123-35-3), m-cymene (000535-77-3), and 2-carene (000554-61-0)) may be repellent to these adults. These volatiles have good research potential as attractants or repellents. Many volatile odors stimulate insect oviposition, and some compounds may inhibit oviposition [41,42]. In addition, studies have shown that the concentration of plant volatiles is important in regulating insect behavior [43,44,45,46]. For example, limonene at a high concentration is repellent to Neodiprion sertifer, but it is attractive at lower concentrations [47]. Thus, the optimal combination system and concentration of these adult attractants or repellents needs further investigation.

5. Conclusions

For the first time, we investigated the oviposition preference of H. cunea invading Zhejiang Province, the subtropical region of China. In addition, we analyzed the correlation between plant volatile composition and preference. Among the newly invaded H. cunea in Zhejiang Province, there was a greater preference for C. illinoinensis and M. alba compared with P. simonii and M. glyptostroboides. Furthermore, we found that four volatiles, cis-3-hexenol, trans-3-hexenol, cis-3-hexenol acetate, and 4-methyl-1,3-pentadiene, were positively correlated with female preference. These results may provide a new perspective for the development of new attractants and management of H. cunea in the subtropical region of China and provide a basis for screening single compounds with strong attractant effects and mixtures of compounds.