A colony of YPMs was established and had been maintained for about 25 generations on maize in climate incubators (RTOP-B, Zhejiang Top Instrument Co., Ltd.) at 23 ± 1°C, RH 75 ± 2%, 16L/8D photoperiod, and 3, 500 lux light intensity (Guo et al., 2021). Adult moths were provided with 5–8% honey solution after emergence. Apples covered with gauze pieces were provided for the oviposition of mated YPM females in the cage.

The P. digitatum isolated from orange fruits were purified using the traditional tissue separation method (Shi et al., 2019). The symptomatic–asymptomatic junction tissue was cut into segments (about 0.5 × 0.5 cm²), which was immersed completely into 2% sodium hypochlorite for 3 min and flushed with sterilized water three times, and then were immersed completely into 75% ethanol for 1 min and flushed with sterilized water for three times. Finally, the segments were incubated onto potato dextrose agar (containing (g/L): potato 200; dextrose 20; agar 18) medium in Petri dishes and placed in a constant temperature incubator at 28°C. After repeated purification for four times, the pure culture of P. digitatum was obtained. Potato dextrose agar medium (7 mm diameter) with fully grown P. digitatum (PPD) was also prepared and incubated at 28°C for 6 d before behavioral assays.

The P. digitatum was cultured on potato dextrose agar at 25°C to prepare the conidial suspension. The potato dextrose agar culture of P. digitatum bearing 7-d-old conidia was gently rinsed in 1.5 mL sterilized distilled water and the density of conidial suspension was adjusted to 8 × 10⁷ conidia/mL.

To test the effects of P. digitatum on the oviposition behavior of mated YPM females, four treatments (NIA, MDA, PDA, and PPD) were simultaneously offered in a wood-frame cage (35 × 27 × 25 cm) with plastic gauzes on side walls to allow the oviposition of mated YPM females. The experimental materials (NIA, MDA, PDA, and PPD) were individually put into two opposite plastic bowls that were punched holes in advance. The materials were renewed every day and their positions were randomly changed. Ten newly emerged naive females and 15 males (no exposure to natural or synthetic sources of apples or fungi-infected apples) had been released into a cage to copulate for 3–4 days before the oviposition selection experiments. Each experiment was replicated 20 times with a total number of 200 females. The egg numbers on each sheet were counted separately and the data were statistically treated on the basis of average number of eggs by 10 females.

Four-arm olfactometer was used to test two bioassay experiments about the behavioral responses of mated YPM females, one is to the odors from NIA, MDA, PDA, and PPD, the other is to different concentrations (10⁻¹, 10⁻², 10⁻³, and 10⁻⁴ (v/V)) of the same specific VOCs from P. digitatum-infected apples. According to our previous study, the four-arm olfactometer consists of four glass chambers (2.5 cm diameter, 10.0 cm long), each with terminal end coupling to a cap (3.0-cm-long arms, 2.0 cm diameter) and central converging into a 10-cm-long common arm (2.5 cm diameter) (Shi et al., 2019). Moistened and charcoal-filtered air was pumped through each odor and then respectively went into one of the caps at a rate of 250 ml/min controlled by flow meters (LZYIA Instrument Co. Ltd, China). One moth was introduced into the entrance of the common arm of the olfactometer using a glass vial, and its behavioral response was observed under a 25-W red light lamp. The test for each moth lasted 3 min, and the behavioral response was classified as a choice if the moth passed over 1/3 length of the arm associated with one of the four odors and stayed there for more than 30 s. Conversely, no-choice was assigned if the test moth remained in the common arm for 3 min. The position of the lateral chambers along with the olfactometer was systematically exchanged after testing 2 moths to avoid positional bias. The olfactometer was flushed following Du et al. (2016). The moths used for test were allowed to acclimatize to the test conditions for 2 h before the start of the test. 3- to 4-day-old mated females were used and each individual moth was used only once. The selection rate in the four-arm olfactometer experiment was defined as the number of females that made a selection for the odor divided by the total number of females that made a selection for any odors offered simultaneously.

The experiment 1: the behavioral responses of mated YPM females to the odors from NIA, MDA, PDA, and PPD. Three apples of each treatment (NIA, MDA, PDA) and 4 pieces of PPD (diameter 7 mm) were placed into an oven bag (Reynolds Kitchens, Richmond, VA, USA) respectively. Each oven bag was tightened and connected with a Teflon tube to one of the caps of the four-arm olfactometer for behavioral tests. A total of 211 individual mated YPM females were tested.

The experiment 2: the behavioral responses of mated YPM females to different concentrations of the five specific VOCs from P. digitatum-infeced apples. For each VOC, four concentrations (10⁻¹, 10⁻², 10⁻³, and 10⁻⁴ (v/V)) were prepared with mineral oil as solvent. And then, an aliquot of 10 μl test solution for each of the four concentrations of the same compound was applied onto a 1 × 5 cm filter paper, which was thereafter placed into one chamber of the four-arm olfactometer for behavioral tests. Totally, 80 mated YPM females (styrene), 86 mated YPM females (methyl 2-methylbutyrate), 80 mated YPM females (methyl caproate), 85 mated YPM females (butyl caprylate), and 80 mated YPM females (n-tetradecane) were tested.

All chemicals (purities ≥ 95%) were purchased from commercial companies, which methyl 2-methyl butyrate and styrene were from J & K Chemical Ltd. (Shanghai, China), methyl caproate, butyl caprylate, and n-tetradecane from TCI Development Co., Ltd. (Shanghai, China).

Five apples of each treatment (NIA, MDA, PDA) and correspondingly similar size of PPD that were placed into a 48.2 × 59.6 cm oven bag respectively were used to collected VOCs according to dynamic headspace collection method reported by Guo et al. (2021).

The bag mouth was tightened with a twist tie around a glass tube (6 mm diameter, 10 cm long) filled with 50 mg of Porapark Q adsorbent (80–100 mesh, Waters Corporation). The humidified and purified air was pushed at a rate of 450 ml/min by a QC-1S pump (Labor Protection Science Research Institute of Beijing) into the bag through the Teflon tube and was pulled out through the glass tube filled with Porapark Q adsorbent. VOCs were trapped by the Porapark Q adsorbent when they passed through the glass tube.

After collection, the trapped VOCs were eluted using chromatography-grade n-hexane (99.9%) and then were analyzed using an Agilent 6,890 gas chromatograph (GC) coupled to an Agilent 5,975 Mass Spectrometer (MS). The procedures for the GC-MS analysis were the same as described in Du et al. (2016) with the exceptions that the GC was equipped with a DB-5MS column (60 m × 0.25 mm × 0.15 μm, Agilent, USA) rather than a HP-5MS column (30 m × 0.25 mm × 0.25 μm) and the injector temperature was 250°C other than 210°C. Following injection, the column temperature was maintained at 37°C for 6 min, followed by an increase in temperature of 2°C/min to 70°C for 5 min, and then an increase of 5°C/min up to 200°C, at last maintained at 200°C for 5 min. Compounds were tentatively identified by comparing mass spectra with NIST Standard Reference Database 98 (Agilent Technologies, Palo Alto, CA, USA). Compounds were quantified by their total ion abundance relative to that of the internal standard (n-nonyl acetate).

Five specific VOCs from PDA, including methyl 2-methyl butyrate, styrene, methyl caproate, butyl caprylate, and n-tetradecane, were chosen for electroantennogram (EAG) measurements. All chemicals (purities ≥ 95%) were purchased from commercial companies, which methyl 2-methyl butyrate and styrene were from J & K Chemical Ltd. (Shanghai, China), methyl caproate, butyl caprylate, and n-tetradecane from TCI Development Co., Ltd. (Shanghai, China). Four concentrations [10⁻¹, 10⁻², 10⁻³, and 10⁻⁴ (v/V)] of five individual compounds were prepared as the four-arm olfactometer experiment 2. The test solutions were stored at −20°C for further EAG analyses.

EAG recordings were performed on 3- to 4-day-old mated YPM females that the moths at this stage were eggs-loading and sensitive to signals used for oviposition location (Belmain et al., 2002). The method of EAG recordings was the same as that described by Du et al. (2016). Stimulus was delivered and tested in increasing doses on the antennae of mated YPM females with mineral oil and n-hexanol being used as control and standard stimuli, respectively. EAG test was run for a variable number of replicates per day, and each compound at each concentration was tested on 15 antennae. In each test, the control and standard stimuli were applied subsequently after four successive stimulations. Normalization was achieved by dividing the peak EAG amplitude of the test puff with the average EAG amplitude of the two nearest standard stimulations after subtracting the amplitude recorded in response to the mineral oil.

The preference of mated YPM females to apples with or without five specific VOCs (methyl 2-methylbutyrate, styrene, methyl caproate, butyl caprylate, and n-tetradecane) from PDA was tested using Y-tube olfactometer. For Y-tube olfactometer assays, apples with and without five specific VOCs from PDA [the amount was referred to the concentrations in fungi-infected apple fruits (Supplementary Table S1)] were separately placed into the chambers of the Y-tube. The test procedure was similar to that in our previous study (Guo et al., 2022). Each individual moth was used only once, and totally 80 mated female moths were tested for each treatment. The selection rate in the Y-tube olfactometer experiment was defined as the number of females that made a selection for apples with exogenous compounds divided by the total number of females that made a selection between apples with and without exogenous compounds.

Data obtained from oviposition selection experiments, behavioral assay in four-arm olfactometers, EAG tests, and the absolute content of host plant VOCs were subjected to analysis of variance (ANOVA) using Tukey-HSD test (P < 0.05). The data of Y-tube olfactometer experiments were analyzed using non-parametric Chi-square analysis (Females with no choice were excluded from statistical analyses, P < 0.05). The quantification of VOCs measured as the absolute content of each compound was analyzed using principal component analysis (PCA) by the software program SIMCA P+ 11.0 (Umetrics AB, Umeå, Sweden) (P < 0.05). All statistics except the PCA analysis were performed using the SPSS16.0 statistical software. Graphs were generated in the program of Graphpad Prism 9.0.

In order to assess the effects of P. digitatum on the oviposition selection of mated YPM females, the apples infected with P. digitatum for 2, 4, 6, and 8 d were simultaneously provided to allow the oviposition selection of mated YPM females. The results showed that the number of eggs on the 8d-PDA was significantly higher than those on the 2d- and 4d-PDA, and was larger (but not significant) than that on the 6d-PDA (F_{3, 79} = 11.699, P < 0.01; Figure 1A). When NIA, MDA, PDA, and PPD were simultaneously offered in a cage to allow for oviposition, the average number of eggs laid by 10 mated YPM females (173.2) was significantly higher on PDA than those on MDA (22.3), NIA (22.2), and PPD (5.7) (F_{3, 79} = 111.413, P < 0.01; Figure 1B), respectively.

The behavioral responses of mated YPM females were tested in a four-arm olfactometer. The selection rate of mated YPM females to the odor of PDA (42.03%) was significantly higher than those to the odor of NIA (17.91%), the odor of MDA (24.17%), and the odor of PPD (15.89%) (F _{3, 19} = 18.948, P < 0.01; Figure 1C), suggesting that mated YPM females preferred to the odor of PDA.

In total, 38 volatile compounds were detected from the emissions of NIA, MDA, PDA, and PPD, including 24 compounds in NIA, 22 compounds in MDA, 24 compounds in PDA, and 1 compound in PPD (Supplementary Figure S1; Table 1). The results showed that α-farnesene was the most abundant compound in NIA, MDA, and PDA. Compared with MDA and NIA, five compounds, including methyl 2-methylbutyrate, styrene, methyl caproate, butyl caprylate, and n-tetradecane, were specifically detected in PDA. Only one compound, ethyl butyrate, was found in PPD (Table 1).

A principal component analysis (PCA) based on the absolute content of the 38 compounds was then performed to determine the major sources of variations in the four treatments (NIA, MDA, PDA, and PPD). The biplot depicted by graphical PCA expounded the first two principal components (PCs) with an explication of 70.61% of the total variance in the four treatments. The first component (PC1), which explained 50.68% of the total variance, was clearly isolated NIA. The second component (PC2) accounted for 19.93% of total variance (Figure 2A). Further analyses indicated that the major loadings of PC1 were for n-butyl butyrate (-0.995), isoamyl 2-methylbutyrate (−0.995), hexyl hexanoate (-0.989), amyl hexanoate (-0.989) and major loadings of PC2 were for styrene (0.887), ethyl octanoate (0.887), α-farnesene (0.868), ethyl caproate (0.865). And then, PC scores were subjected to one-way ANOVA. The results showed that NIA was significantly higher than PDA and MDA in PC1 scores, and PDA had the highest PC2 scores among four treatments (Figures 2B, C).

Five specific VOCs emitted from PDA, including methyl 2-methylbutyrate, styrene, methyl caproate, butyl caprylate, and n-tetradecane, were selected for EAG and behavioral tests. Each of the five synthetic compounds could trigger significant EAG responses of mated YPM females, with dose-dependent decreasing from 10⁻¹ to 10⁻⁴ (v/V) (Figure 3). Of which, the strongest EAG response was elicited by methyl caproate at 10⁻¹ (v/V) (EAG value = 1.314 μA) (Figure 3C), followed by methyl 2-methylbutyrate at 10⁻¹ (v/V) (EAG value = 0.995 μA) (Figure 3B) and butyl caprylate at 10⁻¹ (v/V) (EAG value = 0.790 μA) (Figure 3D). The compound methyl 2-methylbutyrate at concentration of 10⁻² (v/V) (EAG value = 0.620 μA) (Figure 3B) could trigger stronger EAG responses than styrene at 10⁻¹ (v/V) (EAG value = 0.602 μA) (Figure 3A) and, butyl caprylate at 10⁻² (v/V) (EAG value = 0.398 μA) (Figure 3D) and styrene at 10⁻² (v/V) (EAG value = 0.374 μA) (Figure 3A) both trigger stronger EAG responses than n-tetradecane at 10⁻¹(v/V) (EAG value = 0.347 μA) (Figure 3E).

To further predict the optimal concentration for attraction, the four-arm olfactometer was further used to determine the simultaneous selection of C. punctiferalis mated females among four concentrations [10⁻¹, 10⁻², 10⁻³, and 10⁻⁴ (v/V)] of each same compound. The behavioral responses indicated that the five individual compounds all showed the strongest attractiveness to the mated YPM females at 10⁻² (v/V) in the four-arm olfactometer experiments (Figure 4). For styrene, the selection rates of mated YPM females were 35.89% at 10⁻³ (v/V) and 39.56% at 10⁻²(v/V), which was significantly higher than those of 14.19% at 10⁻⁴ (v/V) and 10.35% at 10⁻¹ (v/V) (Figure 4A). The selection rate of mated YPM females was highest at 10⁻² (v/V) (41.82%) among four concentrations of methy1 2-methylbutyrate, followed by those at 10⁻¹ (v/V) (25.58%) and 10⁻⁴ (v/V) (17.48%), and the lowest was at 10⁻³ (v/V) (15.13%) (Figure 4B). The selection rate of mated YPM females was significantly higher at 10⁻² (v/V) (34.24% of methyl caproate and 36.97% of n-tetradecane) than those at 10⁻¹ (v/V) (13.64% of methyl caproate and 17.79% of n-tetradecane), while it was intermediate at 10⁻³ (v/V) (24.72% of methyl caproate and 24.55% of n-tetradecane) and 10⁻⁴ (v/V) (27.39% of methyl caproate and 20.69% of n-tetradecane) (Figures 4C, E). For butyl caprylate, the selection rate of mated YPM females at 10⁻² (v/V) (37.88%) was the highest, followed by those at 10⁻³ (v/V) (24.24%), 10⁻⁴ (v/V) (19.70%), and 10⁻¹ (v/V) (16.67%) (Figure 4D).

To address the critical synergist in attracting mated YPM females to fungi-infected apples, the individual or mixed blends of five specific VOCs were added onto the mechanically damaged apples (MDA) respectively to test the behavioral responses, with MDA as control (Figure 5; Supplementary Table S1). For individual compounds, mated YPM females preferred to apples with methyl 2-methylbutyrate (the selection rate of 66.25%, χ²= 8.45, P = 0.004) (Figure 5B), butyl caprylate (the selection rate of 61.25%, χ²= 4.05, P = 0.044) (Figure 5D), and n-tetradecane (the selection rate of 66.25%, χ²= 8.45, P = 0.004) (Figure 5E). However, mated YPM females had significant repellence to apples with styrene (the selection rate of 37.5%, χ²= 5.00, P = 0.025) (Figure 5A). For mixed compounds, mated YPM females had significant preference to apples with methyl 2-methylbutyrate and butyl caprylate (the selection rate of 62.5%, χ² = 5.00, P = 0.025) (Figure 5K), methyl 2-methylbutyrate and n-tetradecane (the selection rate of 63.75%, χ² = 6.05, P = 0.014) (Figure 5L), butyl caprylate and n-tetradecane (the selection rate of 70%, χ² = 12.8, P < 0.001) (Figure 5O), styrene, methyl 2-methylbutyrate, and n-tetradecane (the selection rate of 62.5%, χ² = 5.00, P = 0.025) (Figure 5R), methyl 2-methylbutyrate, methyl caproate, and butyl caprylate (the selection rate of 67.5%, χ² = 9.80, P = 0.002) (Figure 5V), styrene, methyl 2-methylbutyrate, methyl caproate, and butyl caprylate (the selection rate of 61.25%, χ² = 4.05, P = 0.044) (Figure 5Z), methyl 2-methylbutyrate, methyl caproate, butyl caprylate, and n-tetradecane (the selection rate of 61.25%, χ² = 4.05, P = 0.044) (Figure 5AD), styrene, methyl 2-methylbutyrate, methyl caproate, butyl caprylate, and n-tetradecane (the selection rate of 63.75%, χ² = 6.05, P = 0.014) (Figure 5AE).