Synergistic attraction: Development and field evaluation of a multicomponent lure for Zeugodacus cucurbitae based on host‐plant volatiles and gut bacterium‐derived volatiles

Abstract

BACKGROUND

Zeugodacus cucurbitae is a global quarantine pest. Management of this pest relies heavily on chemical insecticides, which impose substantial environmental burdens. Male‐specific lures [e.g. cue‐lure (CL)] are widely used for monitoring and male‐targeted control, but on their own, they often have a limited impact on population suppression. Therefore, efficient attractants targeting both sexes are urgently needed to advance sustainable and integrated pest management (IPM). To address this need, we screened and optimized multicomponent formulations by combining host‐plant volatiles, volatiles from the culture supernatant of the gut bacterium Morganella sp. M72, and CL.

RESULTS

In behavioral assays, both the optimized host‐plant volatile blend (methyl isovalerate: heptanal: methyl myristate: methyl laurate = 10:10:1:0.1) and the volatile blend from the culture supernatant of the gut bacterium Morganella sp. M72 (indole: benzyl alcohol: 3‐methyl‐1‐butanol: 1‐undecanol = 0.01:10:10:1) significantly attracted both female and male Z. cucurbitae. When CL was incorporated, the resulting ternary formulation produced significantly higher attraction than any individual blend or single‐source formulation. Field trials further supported this improvement, confirming that the ternary blend effectively attracted adults of both sexes under practical conditions.

CONCLUSION

This multicomponent formulation helps overcome the male‐only attraction of conventional lures and showed stable field performance across early‐season, peak‐season and late‐season periods. These results support the development of more effective monitoring and integrated management strategies for Z. cucurbitae. © 2026 The Author(s). Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

The newly developed ternary blend was validated in field trials and proved highly effective in attracting both male and female Zeugodacus cucurbitae.

1. INTRODUCTION

Zeugodacus cucurbitae, a melon fly, is a major invasive pest of the family Tephritidae (Diptera). ¹ Originally from India, this pest has now spread and established in >30 countries and regions worldwide. ² It exhibits a broad host range, infesting 81 plant species, ² with a strong preference for cucurbit crops. This pest is particularly destructive, frequently causing yield losses of 30–100%. ³ In China, the current distribution of this species is primarily limited to regions south of latitude 25 ° N. ⁴ However, owing to the influence of climate change and commercial activities, the potentially suitable habitat is gradually expanding toward the Yangtze River Basin. ⁵ Insects damage crops through two mechanisms: adult oviposition and larval feeding, ⁶ resulting in fruit deformation and rot, and premature drop. Bitter gourds are among the most severely affected hosts, with yield losses of >50%. ⁷ Since its first interception at the Port of Shenzhen in 1985, Z. cucurbitae has been listed as a key quarantine pest for entry into China. ⁸ Therefore, the development of effective control measures remains a critical concern in China.

Plant volatile organic chemicals (VOCs) and insect gut microbial metabolites collectively form an integrated chemical signaling network that regulates key insect behaviors. ⁹ In plant–insect interactions, specific plant VOCs influence the ecological fitness of tephritid flies by modulating key behaviors such as host location, oviposition site selection and mate choice. ¹⁰ The characteristic mango volatile 1‐octen‐3‐ol specifically directs the oviposition behavior of Bactrocera dorsalis, ¹¹ whereas linalool significantly inhibits oviposition in Ceratitis capitata. ¹² Guava (Psidium guajava) VOCs can induce sexual excitement and enhance attractiveness in Anastrepha fraterculus males, ¹³ presenting potential opportunities for the development of innovative plant volatile‐based management approaches. Gut microbial communities are instrumental in modulating insect behavior by releasing specific VOCs. ¹⁴ Klebsiella oxytoca significantly attracted Z. cucurbitae, an effect mediated by specific microbial VOCs, such as 3‐methyl‐1‐butanol and 2‐phenylethanol. ¹⁵ The discovery that cue‐lure (CL), a 4‐phenyl‐2‐butanone derivative, is more effective than anisylacetone at attracting Z. cucurbitae led to the rapid replacement of earlier lures. The CL‐based monitoring approach first reported by Beroza et al. ¹⁶ continues to underpin pest‐monitoring programs worldwide.

Chemical control is currently the primary method for managing Z. cucurbitae. ¹⁷ However, the prolonged and intensive use of these chemicals presents increasing challenges regarding pesticide residues, environmental pollution, and public health risks. ¹⁸ Therefore, developing efficient and sustainable pest management tools is essential. Pest attraction techniques based on behavioral manipulation have emerged as a key research focus owing to their environmental friendliness and sustainability. ¹⁹ Although numerous studies have identified attractants for Z. cucurbitae, most have focused on single compounds, ²⁰ with limited investigations on the use of multicomponent blends. In natural environments, Z. cucurbitae encounters complex olfactory backgrounds comprising multicomponent volatile blends rather than single compounds. ²¹ Recent advances have shown that insect behavioral responses are frequently determined by blend composition and component ratios, and that mixture effects can be nonlinear owing to interactions in peripheral detection and central olfactory processing. ²¹ Moreover, background odors may modulate the behavioral efficacy of lures, highlighting the importance of developing attractants that better reflect ecologically relevant multi‐odor contexts. ²² Single‐component attractants frequently exhibit several limitations, such as a narrow target range, low attractiveness or limited effectiveness to only one sex of the pest. ²³ Multicomponent attractant blends can broaden the target spectrum and enhance attraction efficiency through synergistic interactions among their constituents. ²⁴ Therefore, developing multicomponent attractants based on enhanced blend interactions is a crucial strategy for enhancing pest attraction efficacy. ²⁵ In this study, we present a novel strategy for developing composite attractants.

We introduced a new strategy for formulating a multisource attractant targeting Z. cucurbitae, leveraging the behavioral synergy between host‐plant volatiles, volatiles from the culture supernatant of the gut bacterium Morganella sp. M72, and synthetic lures. First, the active components were screened from the supernatant of the gut symbiont Morganella sp. M72 using electroantennography (EAG). Subsequently, the effective concentrations were determined using behavioral bioassays. Finally, the formulation of the multisource volatile blend was optimized, and its efficacy was evaluated in the field trials. This study provides a foundation for designing efficient, environmentally benign blended attractants, thereby supporting sustainable management of Z. cucurbitae.

2. MATERIALS AND METHODS

2.1. Insects

The initial population of Z. cucurbitae was established using adults collected from infested and rotting fruits in fields near Nanning, Guangxi, China. The colonies were maintained for >20 generations at the Plant Protection Research Institute of Guangxi Academy of Agricultural Sciences, China. The larvae were fed the host plant, zucchini (Cucurbita pepo L.), purchased from Guangbaijia Supermarket. After emergence, adult flies were provided a mixed diet (sucrose: yeast = 2:1) and reared in cages measuring 30 × 30 × 30 cm. All insects were maintained under controlled conditions of 26 ± 1 °C, 70 ± 10% relative humidity (RH) and a 14 h:10 h, light:dark photoperiod. Adults at ≈2 weeks post‐eclosion were used in all subsequent experiments.

2.2. Chemicals

Based on our previous study, we screened 12 behaviorally active single compounds from the volatiles of bitter gourd (Momordica charantia) and sponge gourd (Luffa cylindrica) against adult Z. cucurbitae, ¹² and identified the gut bacterium Morganella sp. M72, whose supernatant attracts both sexes. ²⁶ Although the volatile compounds in this supernatant have been identified, their behavioral effects remain unclear. Therefore, in the present study, we selected 25 volatiles identified in Morganella sp. The M72 supernatant contained dimethyl disulfide, indole, 2‐nonanone and phenethyl alcohol (purity ≥95%; Guangxi Friend Biotechnology Co., Ltd, China). Concurrently, four host‐plant volatiles previously observed to elicit behavioral responses in both female and male flies were selected: methyl isovalerate, heptanal, methyl myristate and methyl laurate (purity ≥98%; Guangxi Friend Biotechnology Co., Ltd). Additionally, a standard CL (Guangxi Friend Biotechnology Co., Ltd) and n‐hexane (chromatographic grade; Tianjin Damao Chemical Reagent Factory, China) were used in the study. All volatiles were commercially sourced and had a purity of ≥95% (Table 1).

Table 1.

Formulation ratio of the M72 supernatant volatile blend

Formula no.	Benzyl alcohol (μg μL−1)	3‐methyl‐1‐butanol (μg μL−1)	Indole (μg μL−1)	1‐Undecanol (μg μL−1)
1	10	10	–	–
2	10	–	0.01	–
3	10	–	–	1
4	–	10	0.01	–
5	–	10	–	1
6	–	–	0.01	1
7	10	10	0.01	–
8	10	10	–	1
9	10	–	0.01	1
10	–	10	0.01	1
11	10	10	0.01	1

2.3. Electrophysiological assays

Each volatile compound identified from the culture supernatant of the gut bacterium Morganella sp. M72 was dissolved in n‐hexane to prepare solutions at five concentrations (0.01, 0.1, 1, 10 and 100 μg μL⁻¹), which were stored at −20°C until use. Pure n‐hexane (all chemicals were purchased from Guangxi Friend Biotechnology Co., Ltd) served as a negative control, and dose–response experiments were performed using an EAG system (IDAC‐2; Ockenfels Syntech GmbH, Kirchzarten, Germany). Actively moving adult melon flies were selected from rearing cages. The head of each insect was swiftly excised using a razor blade, and ≈1 mm of the antennal tip was cut off. The antennal tip was then placed in the recording electrode using fine forceps, and the antennal base was connected to the reference (ground) electrode. Both glass electrodes were filled with a 0.1 mol L⁻¹ potassium chloride (KCl) solution.

For olfactory stimulation, 10 μL of the test solution was applied to a filter paper strip (40 mm × 5 cm). The strip was placed inside a 10‐cm glass cartridge connected to a stimulus controller (CS‐55; Ockenfels Syntech GmbH). During testing, a continuous stream of purified humidified air (400 mL min⁻¹, filtered through activated charcoal) was directed toward the antenna via a stainless steel tube with the outlet positioned ≈1 cm from the antenna. Stimuli were delivered as 0.5‐s puffs of odorized air with an interstimulus interval of 30 s to allow antennal recovery.

The stimulation sequence for each test compound was as follows: n‐hexane (control), test volatile compound and n‐hexane (control). The EAG responses to the two n‐hexane stimulations were averaged and used as the baseline for correcting the responses to the test volatiles.

The relative EAG response was calculated as follows: (CT – CK)/CK, where CT represents the EAG amplitude in response to the test compound and CK is the average EAG amplitude of the control (n‐hexane). At least 15 female and 15 male adults were tested for each compound and concentration, and each antenna was used only once for the tests.

2.4. Two‐choice behavioral assays

The experimental setup consisted of three identical mesh cages (30 × 30 × 30 cm) interconnected via circular openings [5 cm diameter; Fig. 1(A)]. The two‐choice cage assay comprised a central release cage (M) connected to two side cages (L and R). The treatment (single compound at the indicated concentration) and control (n‐hexane) were placed in cages L and R, respectively. For each replicate, 200 μL of the test solution was evenly applied to a filter paper strip (4 cm × 3 mm), which served as the odor source and was placed in the designated side cage. No food or water was provided in cages L or R during the experimental period. Sixty 15‐day‐old adult Z. cucurbitae (starved for 12 h) were released into cage M at 09:00 h, and the numbers of flies that entered cages L (treatment) and R (control) were recorded at 18:00 h. Females and males were tested in separate assays, with 60 individuals per replicate (females, n = 4 replicates; males, n = 4 replicates). For each replicate, the attraction rate (%) was calculated as the proportion of responding flies choosing the treatment: treatment/ (treatment + control) × 100. The sex‐specific attraction responses are shown in Figs 5 and 6.

Figure 1.

Experimental setup for adult attraction assays of Z. cucurbitae. (A) Set‐up of an adult two‐choice trap assay. (B) Schematic of the behavioral free‐choice assay.

Figure 5.

Behavioral responses of adult Z. cucurbitae to 14 volatiles derived from Morganella sp. M72 supernatant at different concentrations. Responses from females and males are shown separately. Data are presented as mean ± SEM (n = 60). Asterisks indicate significant differences between the control and corresponding treatment (*, P < 0.05; **, P < 0.01; ***, P < 0.001), whereas ‘ns’ denotes no significant difference (χ ² test).

Figure 6.

Behavioral responses of adult Z. cucurbitae to different attractant formulations. (A) Attractant formulations composed of volatiles from the gut bacterial supernatant. (B) Attractant formulations composed of host‐plant volatiles. Data are presented as mean ± SEM (n = 60). Asterisks indicate significant differences between the control and each attractant treatment (*, P < 0.05; **, P < 0.01; ***, P < 0.001), whereas ‘ns’ indicates no significant difference (χ ² test). The specific compounds included in the formulations were as follows: A, benzyl alcohol; B, 3‐methyl‐1‐butanol; C, indole; D, 1‐undecanol; a, methyl isovalerate; b, heptanal; c, methyl myristate; and d, methyl laurate.

In order to minimize experimental bias, the following quality control measures were implemented: (i) each treatment was replicated five times biologically; (ii) the positions of the L and R cages were swapped for each replicate to eliminate directional bias; and (iii) all tests were conducted in complete darkness to exclude visual cues. The experiments were performed inside an environmentally controlled growth chamber maintained at 26 ± 1 °C, 70 ± 10% RH, and a 14 h:10 h, light:dark photoperiod. Adult females and males were tested separately in this study.

In order to identify the optimal attraction concentration, we selected the following: (i) four volatiles common to both bitter and sponge gourds: methyl isovalerate, heptanal, methyl myristate and methyl laurate, and (ii) individual volatiles from the supernatant of the gut symbiont Morganella sp. M72 elicited electrophysiological activity in both female and male adults during EAG screening. Each compound was dissolved in n‐hexane to prepare five concentrations (0.01, 0.1, 1, 10 and 100 μg μL⁻¹). The attraction effect of each compound was systematically evaluated using behavioral assays to determine its optimal concentration.

Following the determination of the optimal attractive concentration for each compound, systematic blending optimization was conducted separately for plant and gut bacterial volatiles. A combinatorial experimental design was used to generate various formulation ratios for each category, as presented in Table 1 (bacterial volatiles) and Table 2 (host‐plant volatiles). Therefore, the optimal ratio for each group was determined.

Table 2.

Formulation ratio of host‐plant volatile blend

Formula no.	Methyl isovalerate (μg μL−1)	Heptanal (μg μL−1)	Methyl myristate (μg μL−1)	Methyl laurate (μg μL−1)
12	10	10	–	–
13	10	–	1	–
14	10	–	–	0.1
15	–	10	1	–
16	–	10	–	0.1
17	–	–	1	0.1
18	10	10	1	–
19	10	10	–	0.1
20	10	–	1	0.1
21	–	10	1	0.1
22	10	10	1	0.1

2.5. Free‐choice assays

The optimal blend from each category was initially evaluated against its constituent compounds to confirm enhanced efficacy. Subsequently, a combined host–bacterial complex was formulated using these two optimal blends, and its attractiveness was compared with that of the separate host and bacterial blends. Finally, a ternary attractant system comprising host volatiles, gut bacterial volatiles, and other cues was constructed. Its attraction efficacy was systematically evaluated against both the binary host–bacteria complex and CL alone to analyze the blend effects of the composite system.

Free‐choice assays were conducted within a large white mesh cage measuring 1.9 m (L) × 1.8 m (W) × 2.0 m (H) ²⁷ [Fig. 1(B)]. Before testing, a light meter was used to ensure uniform light intensity across all areas inside the cage, thereby minimizing interference from lighting conditions on the experimental outcomes. Four yellow sticky traps (250 × 200 mm) were suspended in each corner of the cages. A sponge stick impregnated with the test lure and blank control was hung directly beneath each trap. Two hundred sexually mature adult females and males were released into the center of each cage. The assay was conducted between 11:00 and 18:00 h. At the end of each trial, the flies captured in each trap were counted. The cage was ventilated for 2 h to dissipate any residual odor before being set up for subsequent tests. Each treatment was performed in triplicate as biological replicates.

2.6. Field trapping trial

Field trapping trials were conducted from 15 April to 10 November 2025 (excluding rainy days), in a bitter gourd planting area in Sundong Village (109 °E, 21 °N), Beihai City, Guangxi Zhuang Autonomous Region, China. To systematically evaluate the field performance of the attractants, a trial period was designed to cover the initial occurrence, peak abundance, and late seasonal stages of Z. cucurbitae population. The test site employed an open‐field arched tunnel cultivation model with conventional fertilization and irrigation practices. This area has been cultivated with cucurbit crops for years, maintaining a high population density of fruit flies, which provided ideal conditions for the experiment.

Yellow sticky traps (20 × 25 cm; cdbeyond) were used as trapping devices. A sponge lure (Changjiang Horticulture) impregnated with 2 mL of the test attractant was fixed to a metal wire at the center of each trap. The traps were hung 2–2.5 m aboveground, spaced 5 m apart and arranged in a completely randomized manner. Three treatments were evaluated: (i) a formulated attractant blend, (ii) CL control (with an equivalent volume added) and (iii) blank yellow sticky trap control. Twenty‐four traps were installed with four replicates per treatment.

2.7. Statistical analysis

Data analysis was performed using IBM spss statistics v20.0. Electroantennographic response data were analyzed using one‐way ANOVA, followed by Bonferroni's post hoc test for multiple comparisons. Significant differences in responses between females and males were assessed using an independent sample Student's t‐test. Data from the trapping experiments were analyzed using the chi‐square (χ ²) test. The significance level was set at P < 0.05. All figures were generated using Prism 9.0 (GraphPad Inc., San Diego, CA, USA).

3. RESULTS

3.1. EAG responses of Z. cucurbitae to volatiles from the culture supernatant of Morganella sp. M72 at different concentrations

Among the 25 volatiles characterized in the M72 supernatant, 14 were EAG‐active in Z. cucurbitae at different concentrations (Fig. 2). Specifically, 1‐octanol at 100 μg μL⁻¹ elicited the strongest EAG responses [female: 1.17 ± 0.09 mV; male: 1.09 ± 0.12 mV; Fig. 2(A)]. By contrast, indole at the same concentration elicited the weakest responses [female: 0.16 ± 0.01 mV; male: 0.15 ± 0.01 mV; Fig. 2(I)]. Dose‐dependent EAG responses to the 14 compounds were observed in both females and males. Notably, female responses were significantly stronger than male responses for most compounds. The concentration thresholds for eliciting antennal responses varied significantly among the different volatiles. Specifically, 2‐nonanone, phenylacetaldehyde and 1‐nonanol elicited peak EAG responses at 10 μg μL⁻¹ concentrations [Fig. 2(K),(M),(H)]. EAG responses to the various volatiles at identical concentrations were compared between female and male Z. cucurbitae (Fig. 3). The results indicated that the order of functional groups based on their stimulatory effect on the olfactory system was aldehydes > alcohols > ketones. Consequently, all 14 compounds that demonstrated antennal activity in this screening were selected for subsequent behavioral experiments.

Figure 2.

Electroantennography (EAG) responses of adult Z. cucurbitae to 14 volatiles derived from Morganella sp. M72 supernatants at different concentrations. Female and male antennal responses are shown separately in the figure. Data are presented as mean ± SEM (n = 10). Different lowercase letters indicate significant differences among the concentrations of the same volatile within each sex, as determined by one‐way ANOVA followed by Tukey's HSD test (P < 0.05).

Figure 3.

Comparison of EAG responses of Z. cucurbitae to different volatiles at the same concentration. (A) Female and (B) male antennae. The color‐coded heatmaps show the mean EAG response (mV) of Z. cucurbitae to each compound at the tested concentrations. The color bar indicates the magnitude of the EAG responses. Within each concentration row, different lowercase letters denote significant differences among the compounds (P < 0.05; one‐way ANOVA followed by Tukey's HSD test). Sample size: n = 10 for each measurement.

3.2. Optimal attractive concentration of single compounds to Z. cucurbitae

Based on our previous screening, EAG‐active volatiles shared by both bitter gourd and sponge gourd were selected for behavioral assays. The results showed that methyl isovalerate at 10 μg μL⁻¹ (females: χ ² = 7.778, P = 0.005; males: χ ² = 4.728, P = 0.030), heptanal at 10 μg μL⁻¹ (females: χ ² = 9.187, P = 0.002; males: χ ² = 4.913, P = 0.027), methyl myristate at 1 μg μL⁻¹ (females: χ ² = 5.552, P = 0.018; males: χ ² = 5.556, P = 0.018) and methyl laurate at 0.1 μg μL⁻¹ (females: χ ² = 4.728, P = 0.030; males: χ ² = 4.261, P = 0.039) significantly attracted both female and male adults (Fig. 4).

Figure 4.

Behavioral responses of adult Z. cucurbitae to gradient concentrations of four host‐plant volatiles: methyl isovalerate (A), heptanal (B), methyl myristate (C) and methyl laurate (D). Data are presented as mean ± SEM (n = 60). Asterisks indicate significant differences between the control and corresponding treatment (*, P < 0.05; **, P < 0.01), whereas ‘ns’ denotes no significant difference (chi‐square test).

Behavioral assays of individual volatiles from M72 supernatant showed that benzyl alcohol (females: χ ² = 36.461, P < 0.001; males: χ ² = 25.920, P < 0.001) and 3‐methyl‐1‐butanol (females: χ ² = 19.488, P < 0.001; males: χ ² = 24.296, P < 0.001) at 10 μg μL⁻¹ significantly attracted adults of both sexes. In addition, 1‐undecanol at 1 μg μL⁻¹ (females: χ ² = 22.831, P < 0.001; males: χ ² = 10.945, P < 0.001) and indole at 0.01 μg μL⁻¹ (females: χ ² = 10.219, P = 0.001; males: χ ² = 21.429, P < 0.001) significantly attracted both sexes (Fig. 5).

3.3. Behavioral response of Z. cucurbitae to different attractant formulations

Volatile compounds with an attraction rate of >40% were selected as candidates for blend formulation. These candidates were then combined into different mixtures and evaluated in behavioral choice assays. Among the M72 supernatant volatile blends, blend 1 (benzyl alcohol: 3‐methyl‐1‐butanol: indole: 1‐undecanol = 10:10:0.01:1) exhibited the most significant attraction to both female and male adults (females: choice index = 63%, χ ² = 16.835, P < 0.001; males: choice index = 66%, χ ² = 14.617, P < 0.001). Additionally, two other blends showed significant attractant activity: blend 5 (benzyl alcohol: 3‐methyl‐1‐butanol: indole =10:10:0.01; females: 51%, χ ² = 7.435, P = 0.006; males: 54%, χ ² = 8.972, P = 0.003) and blend 11 [benzyl alcohol: 3‐methyl‐1‐butanol =10:10; females: 56%, χ ² = 5.794, P = 0.016; males: 40%, χ ² = 7.574, P = 0.006; Fig. 6(A)].

Among the plant volatile formulations, blend 12 (methyl isovalerate: heptanal: methyl myristate: methyl laurate = 10:10:1:0.1) significantly attracted both female and male adults (females: choice index = 62%, χ ² = 15.705, P < 0.001; males: choice index = 64%, χ ² = 13.525, P < 0.001). Several other blends also showed significant attractant effects, including blend 15 (methyl isovalerate: heptanal: methyl laurate =10:10:0.1; females: 55%, χ ² = 6.522, P = 0.011; males: 52%, χ ² = 3.960, P = 0.047), blend 16 (methyl isovalerate: heptanal: methyl myristate =10:10:1; females: 53%, χ ² = 9.023, P = 0.003; males: 62%, χ ² = 9.684, P = 0.002), blend 20 (methyl isovalerate: methyl laurate = 10:0.1; females: 53%, χ ² = 9.934, P = 0.002; males: 52%, χ ² = 14.208, P < 0.001) and blend 22 [methyl isovalerate: heptanal = 10:10; females: 51%, χ ² = 10.083, P = 0.001; males: 43%, χ ² = 4.253, P = 0.039; Fig. 6(B)].

3.4. Comparison of composite lure efficacy with individual components, and the effect of adding cue‐lure on attraction of Z. cucurbitae

Based on the behavioral screening results, the two most effective formulations (blends 1 and 12, each with a choice rate > 60%) were selected for comparison with their individual components [Fig. 7(A),(B)]. Consistently, the attraction rates of blend 1 for both female and male adults were significantly higher than those of its single compounds. Likewise, blend 12 significantly outperformed the individual components. These findings indicate that scientifically formulated mixtures can trigger stronger attraction responses in Z. cucurbitae than single compounds.

Figure 7.

Attraction efficacy of the optimized formulations and individual components against adult Z. cucurbitae. (A) Attraction efficacy of the gut bacterial (M72) supernatant volatile blend 1 and its constituent compounds. (B) Attraction efficacy of host‐plant volatile blend 12 and its constituent compounds. (C) Attraction efficacy of blends 1, 12 and their combination (blend 1 + 12). (D) Attraction activity of CL alone and in combination with the formulated blends. The female and male attraction rates are shown separately. Data are presented as mean ± SEM. Different lowercase letters above the bars indicate significant differences among treatments within the same sex in each panel (P < 0.05; one‐way ANOVA followed by Tukey's HSD test).

Reformulation assays combining blends 1 and 12 showed that the combination of benzyl alcohol, 3‐methyl‐1‐butanol, indole, 1‐undecanol, methyl isovalerate, heptanal, methyl myristate and methyl laurate in the ratio of 10:10:0.01:1:10:10:1:0.1 attracted significantly more Z. cucurbitae adults than either blend 1 or 12 alone [P < 0.05; Fig. 7(C)]. No significant differences in attraction were observed between blends 1 and 12. The subsequent addition of CL to the combined blend further enhanced attraction compared with the combined blend without CL. Furthermore, the final enhanced formulation significantly outperformed the CL‐only formulation [Fig. 7(D)].

3.5. Field attraction efficacy of optimized attractant formulations for Z. cucurbitae

The field trial results indicated that all attractant formulations exhibited trapping efficacy against both female and male Z. cucurbitae in bitter gourd fields (Fig. 8). A comparison of effectiveness across different periods showed that the overall number of flies trapped was the lowest in April, whereas July, corresponding to the peak occurrence period, exhibited the highest trapping performance. Throughout the initial, peak and late occurrence periods of Z. cucurbitae, the ternary host–gut bacteria–CL composite blend consistently demonstrated significantly higher trapping efficacy than all other formulations. In terms of sex‐specific catches, CL alone was significantly more attractive to males than the formulations that comprised composite blends. By contrast, the host–gut bacteria–CL composite blend significantly increased female captures. Under this treatment, the mean total catch reached 61.5 ± 2.06 flies per trap.

Figure 8.

Comparison of the attractant efficacy of different lure formulations against adult Z. cucurbitae in bitter gourd field. Panels (A–C) show captures recorded in April, July and November, respectively. Female (light gray), male (dark gray) and total (brown) trap catches are presented for each of the treatments. Data are expressed as mean ± SEM. Different lowercase letters above the bars indicate significant differences among treatments within each sampling month (P < 0.05; one‐way ANOVA, followed by Tukey's HSD test). Each treatment was replicated using four traps during each sampling period.

4. DISCUSSION

The attractant effect of odors on pests is a fundamental premise in the development of lures. ²⁸ Consequently, attractants are important for efficient and environmentally friendly pest control strategies within integrated management systems. ²⁹ Z. cucurbitae can detect and respond to a broad spectrum of chemical blends compared to oligophagous or monophagous tephritids. ³⁰ The development of effective lures requires the screening of behaviorally active semiochemicals for Z. cucurbitae and their combinations in defined formulations. Initially, 14 antennal‐active compounds were identified from the 25 compounds tested on Morganella sp. M72 supernatant using EAG. The EAG response amplitudes to these active compounds were significantly higher in females than in males. This suggests that female antennae are aligned to more precisely detect host plant odors to support multiple behavioral processes, such as feeding and oviposition, whereas males primarily utilize host odor detection for feeding and mate‐seeking. ³¹ This is consistent with the general pattern observed in many phytophagous insects, where males exhibit increased sensitivity to female‐emitted sex pheromones than to host‐plant volatiles. At 100 μg μL⁻¹, 1‐octanol elicited the strongest EAG responses (females: 1.17 ± 0.09 mV; males: 1.09 ± 0.12 mV), whereas indole elicited the weakest responses (females: 0.16 ± 0.01 mV; males: 0.15 ± 0.01 mV). EAG is a widely used tool in chemical ecology for assessing insect peripheral olfactory sensitivity to plant volatiles. However, it only indicates whether an odorant activates the antenna and the relative strength of that stimulation, and does not demonstrate behavioral effects. ³² As EAG amplitude reflects the summed receptor potentials of many olfactory receptor neurons, whereas behavioral orientation depends on central processing, odorants that are readily detected may evoke strong EAG signals yet be behaviorally neutral or even repellent. ³³ Therefore, behavioral assays are required to validate the functional roles of EAG‐active compounds. In this study, we used a two‐trap assay to further screen for behaviorally active compounds from the antennally active candidates, followed by behavioral analyses to confirm the functions of the selected volatiles.

Behavioral assays were performed to validate behavioral activity of the active volatiles. The results demonstrated that among the M72 supernatant volatiles, benzyl alcohol, 3‐methyl‐1‐butanol, 1‐undecanol and indole were significantly attractive to both female and male adults at specific concentrations. The activity of these volatiles was consistent with the findings of previous studies. For instance, 3‐methyl‐1‐butanol and 2‐phenylethanol produced by the gut bacterium Klebsiella pneumoniae in Z. cucurbitae enhance female mating behavior. ³⁴ Likewise, in the spotted‐wing drosophila (Drosophila suzukii), 3‐methyl‐1‐butanol, 2‐phenylethanol and indole have been identified as attractants. ³⁵ Additionally, this study established well‐defined effective dose ranges. Specifically, benzyl alcohol and 3‐methyl‐1‐butanol exhibited the strongest attraction at 10 μg μL⁻¹ concentration. The optimal concentration of 1‐undecanol was 1 μg μL⁻¹, whereas indole showed significant attractiveness even at an extremely low concentration of 0.01 μg μL⁻¹. This indicates that the behavioral response of Z. cucurbitae to volatiles has an apparent concentration threshold; effective attraction occurs only when a compound reaches its specific concentration‐dependent with an apparent threshold. ³⁶ This concentration‐dependent response is consistent with findings in other insect systems, such as the dose–response of leafhoppers to the tea plant volatile β‐ionone, which exhibited a threshold with significant attraction achieved at 1 × 10⁻⁴ g mL⁻¹. ³⁷

Attraction of polyphagous insects does not require all host‐plant volatiles. The behavioral response is dependent on the presence of specific key components. ³⁸ Based on our previous screening of common volatiles in bitter and sponge gourds, this study focused on four key compounds: methyl isovalerate, heptanal, methyl myristate and methyl laurate. This focus is supported by the existing literature; for example, heptanal, a volatile identified in peanuts, has been shown to significantly attract female Plodia interpunctella and increase its oviposition rate. ³⁹ Furthermore, we established dose–response relationships and identified the optimal behaviorally active concentrations of each of these four compounds for Z. cucurbitae. Methyl isovalerate and heptanal exhibited the strongest attraction at 10 μg μL⁻¹ concentrations. Methyl myristate was most effective at 1 μg μL⁻¹, whereas the optimal concentration of methyl laurate was as low as 0.1 μg μL⁻¹. The substantial variation in the effective concentrations of the compounds indicates the differential sensitivity of insects to individual host‐plant volatiles. ⁴⁰ The determination of these concentrations provides critical parameters for subsequent formulation experiments.

Formulation experiments revealed that both optimized blends, regardless of whether they were microbe‐ or plant‐derived, were significantly more attractive to Z. cucurbitae than any single compound. Specifically, the microbe‐derived blend 1 (benzyl alcohol: 3‐methyl‐1‐butanol: indole: 1‐undecanol = 10:10:0.01:1) and the plant‐derived blend 12 (methyl isovalerate: heptanal: methyl myristate: methyl laurate = 10:10:1:0.1) were each significantly more attractive to Z. cucurbitae than their corresponding individual components. Evidence from other systems supports this enhanced attraction: a blend of five esters was highly attractive to Rhagoletis pomonella, ⁴¹ and a nine‐component cucumber volatile mix was twice as attractive to female Z. cucurbitae as standard protein baits. ⁴² , ⁴³ Thus, these findings suggest that combining key active compounds into optimized blends can markedly enhance attraction.

Microbes that attract insects can be used as attractants for pest control. ⁴⁴ However, it remains poorly understood whether volatiles emitted by the insect gut microbiota interact with host‐plant volatiles. Here, we combined gut microbial blend 1 with host plant blend 12. The resulting composite demonstrated significantly superior attraction compared to the blends from either single source. This indicates functional complementarity between host plant and gut microbial volatiles in guiding insect behavior. Plant volatiles primarily simulate the host‐feeding habitat, ⁴⁵ whereas gut microbial volatiles may be linked to specific intraspecific communication pathways. ⁴⁶ Thus, their combination suggests a synergistic interaction and provides a more complete ecological context. The addition of CL to the binary blend produced a novel ternary ‘host‐microbe‐CL’ complex, which significantly enhanced the attraction of Z. cucurbitae. Specifically, although CL alone strongly attracted males, its combination with our optimized binary blend notably increased its efficacy against females. Recent studies have shown similar synergistic mechanisms; for example, a study on B. dorsalis reported a synergistic effect between l‐prolinamide, a metabolite of its gut bacterium Enterobacter cloacae, and methyl eugenol. ⁴⁷

The ternary host‐microbe‐CL blend was evaluated in the field trials. Although all tested formulations demonstrated a certain level of attraction to both female and male Z. cucurbitae, the composite host‐microbe‐CL blend significantly outperformed all other formulations in terms of trapping efficacy. Analysis of the trapping efficacy across different seasonal periods revealed that the ternary blend demonstrated consistent superiority over the other formulations during the initial occurrence, peak abundance and late‐season stages of Z. cucurbitae, indicating its potential for field applications throughout the pest activity period. Although CL remains highly attractive to males, the combination of host‐plant and gut microbial volatiles significantly increased overall attraction to both sexes. ⁴⁸ These findings have practical implications for improving field control strategies against Z. cucurbitae.

5. CONCLUSION

This study integrated multidisciplinary approaches, including insect olfactory physiology, behavioral ecology and field ecology. Using these approaches, we systematically evaluated the control potential of a composite attractant derived from three odor sources: volatiles from the culture supernatant of the gut bacterium Morganella sp. M72, host‐plant volatiles, and CL. Field trials demonstrated that the optimized blend (benzyl alcohol: 3‐methyl‐1‐butanol: 1‐undecanol: indole: methyl isovalerate: heptanal: methyl myristate: methyl laurate: CL = 10:10:1:0.01:10:10:1:0.1:10) overcame the limitation of conventional tephritid lures that primarily attract males. The blend also showed consistent and effective field performance across different seasonal periods for Z. cucurbitae. These findings support the development of sustainable, environmentally friendly control technologies for Z. cucurbitae.

CONFLICT OF INTEREST

The authors declare no conflicts of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.