The pepper (cultivar “Tianjiao 6”) and fennel (cultivar “Siji”) seeds, and Phytophthora capsici (strain 501) were provided by the State Key Laboratory for Conservation and Utilization of Bio-Resources, Yunnan Agricultural University. P. capsici was cultured on V8 medium (30 mL of V8 juice, 6.0 g of agar, 0.5 g of CaCO₃, 300 mL of ddH₂O, 121°C for 20 min) at 25°C under a 12-h light-dark cycle for 8 days to produce sporangia. Then, 15 mL of sterilized water was added to the plate and placed in a 4°C refrigerator for 30 min before returning to room temperature for 30 min to release zoospores. The zoospore suspension was collected after filtration with gauze (8 m × 0.84 m) and diluted to a concentration of 10⁵ zoospores/mL for further use.

To determine the effect of the pepper and fennel intercropping system on Phytophthora pepper disease suppression, greenhouse experiments were designed. Pepper and fennel seeds were surface sterilized with 3% sodium hypochlorite and then pregerminated in a humid chamber for 3 days at 25°C in the dark. Pepper seeds were sown after the fennel seeds were planted for 7 days in a plastic basin (50 by 30 cm). The greenhouse temperature range was approximately 25–30°C. To confirm the role of fennel plants in the intercropping system, pepper/fennel intercropping was performed (P2/F2), as shown in Figure 1A. In the second experiment, all pepper plants were planted with the same row spacing (P2/F2-) to eliminate the influence of physical distance (Figure 1B). In the third experiment, a pepper monoculture (P2) was designed (Figure 1C). Each treatment was replicated three times. One milliliter of a zoospore suspension (10⁵ mL^–1) of P. capsici was injected into the central pepper seedling in the central pepper row after 30 days (seeding stage). Intercropping borderlines (IB) and monoculture borderlines (MB) indicate the borderlines (indicator line) in the intercropping and monoculture fields, respectively. Intercropping centerlines (IC) and monoculture centerlines (MC) indicate the centerlines (inoculate online) in intercropping and monoculture fields, respectively. Disease incidence in inoculation and indicator rows was investigated from 3 dpi (days postinoculation) until all pepper plants in the inoculation rows were dead.

Disease incidence rate of pepper border row = 100 × [Number of infected plants in border row/Total number of investigated plants in border row]

A field trial was conducted in 2021 at the Xundian County, Kunming City (N25°20′, E102°41′), Yunnan Province, China. To determine the effect of the pepper and fennel intercropping system on pepper Phytophthora blight suppression, we conducted field experiments using a single factor randomized block design with three replicates. And they include pepper monoculture and removing fennel rows from the intercropping system and pepper/fennel intercropping. Each treatment was replicated three times. All the plots were located in the same field and arranged using a randomized block design. When the pepper plants had grown to the fruit period, Phytophthora capsicum cake was used to inoculate the pepper. In brief, the pepper plants were cut with a blade at 0.5 cm below the stem base, and then the cake was put in the wound. The incidence rates of pepper Phytophthora disease in the center and border rows were surveyed to show the ability to spread in and across rows. The incidence was calculated by using the following formula.

Disease incidence rate of pepper in the center row = Number of infected plants in center row/Total number of investigated plants in center row × 100%.

Disease incidence rate of pepper border row = Number of infected plants in border row/Total number of investigated plants in border row × 100%.

A special apparatus was used to monitor the interaction between the spores of P. capsici and fennel roots (Yang et al., 2014). Briefly, a U-shaped chamber was formed by placing a bent capillary tube on a glass slide and covering this with a coverslip. The root cap side of fennel roots was inserted into a zoospore suspension (10⁵ spores mL^–1) in the chamber. The behavior of the zoospores in the root cap was recorded every 5 min for a period of 30 min by taking five photographs of each zone (Olympus BX43F, Japan). A capillary tube replaced the root as the control. The numbers of zoospores, cystospores and ruptured cystospores in the different root zones were counted. The chemotactic ratio (CR) was calculated as “scores of the zoospores and cystospores on the test root” divided by “scores of the zoospores and cystospores in the control” (Halsall, 1976). CR > 1 indicated positive chemotactic activity. Each interaction assay was replicated three times. Then, the inhibition ratio of swimming zoospores and cystospore germination were calculated (Zhang et al., 2020).

Ratio of zoospore rupture (%) = (Number of ruptured zoospores)/(Number of total zoospores) × 100.

Fennel plants were cultured using a previously described method (Yang et al., 2014), and root exudates were collected by the water culture method (Yuan et al., 2015). The fennel seeds were sown in black plastic pots with 40% humus soil and 60% field soil (poor soil outside the planting area) and grown for 45 days. The collected liquids were filtered, extracted twice with ethyl acetate, and concentrated under reduced pressure (Rotavapor R-200, Buchi). Finally, the concentrate was weighed, redissolved in 1 mL of methanol and filtered through a 0.22 μm filter. The concentration of the fennel root exudate stock was 10,000 μg/mL and that of the collected fluids before concentration was 25 μg/mL. The root exudates were prepared for antimicrobial assays.

Headspace solid-phase microextraction (HS-SPME) of fennel root volatiles was conducted. The volatiles were collected from fennel roots that had been growing for 45 days. First, 20 fennel roots were put into the collection flask, and the SPME head was inserted at the top of the collection flask and collected at 25°C for 1 h (Li et al., 2018). Then, the head of extraction for collecting volatiles was directly inserted into the thermal analytical injection port for gas chromatography–mass spectrometry (GC–MS) analysis. The GC–MS fingerprints of the root volatiles were obtained on a GC–MS instrument (Shimadzu QP2010, Japan) (Block, 2010). Then, fennel root volatiles were separated on an SH-Rxi-5Sil MS capillary column (221-75954-30, 30 m × 0.25 mm × 0.25 μm, Shimadzu, Japan). The pressure was maintained at 49.5 kPa, giving a column flow of 1 mL/min. The injection volume was 1 μL in split-less mode, and the injector temperature was 250°C. The initial column temperature was 40°C (hold 2 min) and programmed to increase at a rate of 3°C/min to 80°C and then continue increasing to 270°C at a rate of 5°C/min, where it was then held for 10 min. The ion source temperature was 230°C with an interface temperature of 250°C. Helium (99.999% purity) was used as the carrier gas at a flow rate of 1 mL/min. Mass spectra were obtained in electron impact (EI) ionization mode at 70 eV by monitoring the full-scan range (m/z 35-500). The compounds were identified by matching the obtained mass spectra with those of reference compounds stored in the NIST14 library, except for the compounds that appeared in the control. Then, the standards of terpene compounds were analyzed by GC–MS under similar conditions (Zhang et al., 2020).

The collection of rhizosphere soil was based on the methods of Bai et al. (2015) with slight modifications. The collected soil (with cultured fennel) and control soil (without any plants) were dried in an oven at 60°C for 12 h. At this time, the water content of the collected soil was 43.02%, and the water content of the control soil was 33.4%. Based on this result, the content of terpenes in the soil water could be calculated.

Experiments were designed to obtain the final concentration range of terpene compounds in fennel rhizosphere terpene standards with different concentrations (0, 1, 10, 20, 30, 100 μg/mL) were added to the blank dry soil (B-soil) to reach the same soil water content as the rhizosphere soil of fennel, and the method of collecting B-soil volatiles and determining the GC–MS fingerprint were the same as those described for fennel root exudates. The regression equation between the terpene standards with different concentrations and the GC–MS fingerprints of the fennel rhizosphere soil were obtained. The concentration-peak area standard curve was used to obtain the contents of terpenes in the soil sample to be measured.

The fennel root exudate stock (10,000 μg/mL) was diluted 2, 5, 10, 20, 50, and 100 times, and distilled water containing the same concentration of methanol was used as a control treatment. The chemotaxis of fennel root exudates on zoospores of P. capsica was observed using a previously reported method (Fang et al., 2016). A square groove was made with a capillary of 1 mm in diameter, and then the square groove was placed on a glass slide (length 25 mm × width 25 mm × height 1 mm). Zoospore suspension at a concentration of 10⁵ spores/mL was added to the tank. One end was treated with a capillary tube to which a diluted root exudate was added, and a capillary containing the same concentration of methanol at one end was used as a control. The behavior of zoospores was observed under a microscope with 40-fold magnification. The experiment was performed three times, each time in triplicate.

The zoospore motility and cystospore germination effects of the root exudates on P. capsici were tested as described by Yang et al. (2014). Briefly, 5 μL of root exudates and 20 μL of zoospore suspension (10⁵ zoospores/mL) or cystospore suspension (10⁵ cystospores/mL) were immediately mixed in glass slides. Then, the final concentration of the root exudates was diluted 5, 10, 25, 100, 150, and 250 times, and the concentrations reached 100, 50, 20, 10, 5, and 2 μg/mL in the slide. The slides were placed in Petri dishes containing moist filter paper and incubated at 25°C in the dark.

Photographs of immobilized zoospores and germinated cystospores were taken under a light microscope. Then, the percentage of zoospores encysted into cystospores was recorded every 1 min for a period of 5 min, and the percentage of germinated-cystospores was calculated after incubation for 1.5 h. Three fields were observed in each treatment, with 100–150 spores per field and three replicates per treatment.

Based on actual concentrations of the rhizosphere measured by GC–MS, the antimicrobial effect of the terpene compounds from the fennel exudates on the infection behavior of P. capsici (chemotaxis, zoospore motility, cystospore germination, and hyphal growth) was determined. The terpene compounds were prepared as 1 × and 2 × mixed terpene compounds (MTCs) according to their concentration proportions in the rhizosphere soil of fennel, and each single terpene compound at different concentrations was tested separately using the method with a few modifications described below (Miller et al., 1983).

Six mycelial plugs (7 mm in diameter) of P. capsici were placed in 100 mL flasks containing 60 mL of carrot liquid medium and cultured for 36 h at 28°C and 140 rpm. Then, 600 μL of terpene compound solutions of different concentrations were added to each flask, and 600 μL of methanol solution without terpene compound was added as the control; each treatment was repeated four times. After continuous culturing for 12 h, the liquid medium was removed by filtration; the hyphae were then wrapped in filter paper, dried and weighed to calculate the inhibition rate of the hyphae. The calculation method was as follows: mycelium inhibition rate (%) = (mycelia weight of control 1- treated mycelia weight)/(mycelia weight of control 1- mycelia weight of control 2) × 100. Control 1 was the mycelial weight after treatment with 600 μL of methanol solution without terpene compounds, and control 2 was the mycelial weight after culturing for 36 h at 28°C and 140 rpm.

Intracellular ROS generation was monitored by fluorescence microscopy using a 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCFDA) fluorescent probe. The non-fluorescent compound can be oxidized by cellular oxidants into the highly fluorescent compound 2′,7′-dichlorofluorescin (DCF), which is trapped inside the cells (Murphy et al., 2011). Thus, the fluorescence intensity is proportional to the level of peroxide produced by the cells. Briefly, the mycelium was washed with PBS and then stained with 10 μM H₂DCFDA at 25°C for 30 min after treatment with fennel root exudates and terpene standard compounds for 24 h (Xiao et al., 2013). The fluorescence of DCF was detected by an inverted fluorescence microscope (Leica DM 2000, Germany).

Statistical analysis was performed with SPSS statistics 24.0 (Stanford University, Stanford, CA, United States) using Chi square test, one-way ANOVA and Duncan’s multiple comparisons test, and figures were generated by GraphPad Prism 8 for Windows 10 (Microsoft Corporation, Redmond, WA, United States).

As Figure 1 shows, the intercropping of fennel and pepper (P2/F2) significantly decreased pepper Phytophthora blight in the indicator rows (p < 0.05) (Figures 1A–C) compared with pepper monoculture (P2). Moreover, even though the pepper plants in the monoculture (P2/F2-) were planted with the same row spacing used in intercropping (Figure 1B), the disease incidence in the indicator row of the intercropping group was significantly lower than that in the monoculture, and the incidence of IB (P2/F2), MB (P2/F2-) and MB (P2) was 58%, 75%, 91.67%, respectively (Figure 1D), during the 23-day observation period. Therefore, the intercropping of fennel and pepper could significantly inhibit the occurrence and spread of Phytophthora pepper blight.

The interaction between the fennel roots and P. capsici spores was observed under a microscope. As shown in Figure 2, zoospores of P. capsici showed strong chemotaxis to fennel root (Figure 2A). The dynamic chemotactic process showed that zoospores were significantly attracted by fennel root in 5 min, with chemotaxis ratios (CR) of 4.8 (Figure 2D). After approaching the fennel roots, the zoospores rapidly lost their swimming ability and transformed into cystospores, and the ratio of zoospore motility was only 4.14% (Figures 2B,E). After 30 min of incubation, all zoospores stopped swimming compared with those in the control treatment. After 120 min, fennel roots effectively promoted the germination of cystospores, which was 62.23% higher than that of the control (Figures 2C,E).

The fennel root exudates showed a strong ability to attract zoospores of P. capsici in a 20 μg/mL methanol solution, and the chemotaxis index of P. capsici was positively correlated with the concentration of root exudates (Figure 3A). Fennel root exudates inhibited the motility of zoospores and germination of cystospores in a dose-dependent manner. When the concentration of root exudates reached 20 μg/mL (close to the concentration of the root exudates before concentration), the inhibition rates for zoospore motility and cystospore germination were 40.5% and 13.2%, respectively (Figures 3B,C).

As Figure 4 shows, the GC–MS profiles of fennel root volatiles and rhizosphere soil showed five peaks at the following retention times (RT): 12.25 min (beta-myrcene), 13.80 min (D-limonene), 16.30 min (gamma-terpinene), 20.25 min (estragole), and 23.10 min (anethole) (Figures 4A,B). The characteristic fragment of terpene compounds showed more than 90% similarity in root exudates or rhizosphere soil compared with the standards of terpene compounds. However, none of the five terpenes were detected in the control soil (Figure 4C). The contents of D-limonene, estragole, anethole, gamma-terpinene and beta-myrcene were 67.48, 13.54, 18.30, 66.30 and 6.13 μg/g, respectively, in fennel rhizosphere soil. Quantitative analysis of terpenes in rhizosphere soil showed that the content of D-limonene in the rhizosphere soil of fennel was the highest (67.48 ± 12.99 μg/g), and the content of beta-myrcene was the lowest (6.13 ± 0.70 μg/g) (Table 1).

As shown in Figures 5A–E, the CR of each compound exhibited the ability to attract zoospores of P. capsici to a certain extent. Among the five terpenes, D-limonene (100 μg/mL), estragole (10 μg/mL), and anethole (50 and 100 μg/mL) were associated with positive chemotaxis, and gamma-terpinene and beta-myrcene showed chemotaxis ratios close to 1 at 1, 10, 50, and 100 μg/mL. However, the five terpenes had strong inhibitory effects on the motility of P. capsici zoospores at 100 μg/mL. The inhibition ratios of swimming ability for each of the five terpenes were 50.88, 66.83, 32.16, 55.27, and 79.05% at 100 μg/mL, respectively (Figures 5F–J). In addition, D-limonene and anethole inhibited cystospore germination in P. capsici; the inhibition ratio was the highest at a concentration of 100 μg/mL; and estragole showed strong activity against cystospore germination at concentrations of 1, 10, 50, and 100 μg/mL (Figures 5K–O). These five terpenes significantly inhibited the mycelial growth of P. capsici within the effective concentration range, and the inhibition ratio reached 96.5% at 10 μg/mL anethole. Except for anethole and estragole, the higher the concentration of the other three terpene compounds was, the higher the mycelial inhibition (Figures 5P–T).

To further explore the effect of terpenes on the behavior of P. capsici. The five terpene compounds were mixed according to their concentration proportions in the rhizosphere soil of fennel. As shown in Table 2, the five terpene compounds were not attractive to zoospores, and the terpene mixture was slightly more attractive than its corresponding single compound. In contrast to the ability of the terpene mixtures and D-limonene to attract P. capsica zoospores, the other four terpene compounds could not attract zoospores at twice the rhizosphere concentration. The inhibitory effect of the mixture on zoospore motility was significantly higher than that of any terpene compound alone, and both 1 × and 2 × MTC reached 100% (Table 2). Analysis of the inhibitory effect of the terpene mixtures on cystospore germination also showed that 2 × MTC had a 19.08% stronger effect than 1 × MTC, which was significantly higher than that of any terpene compound alone (Table 2). Regarding the inhibition of hyphal growth, estragole and anethole showed similar activity against hyphal growth and exhibited no significant difference from the terpene mixtures. However, the terpene mixtures of 1 × and 2 × MTC showed the highest ratios of zoospore rupture, 89.43% and 98.85%, respectively, compared to a corresponding single compound.

As determined by the DCHF-DA signal (Figure 6), fennel root exudates and the five terpene compounds (D-limonene, beta-myrcene, anethole, estragole, and gamma-terpene) significantly induced intracellular ROS accumulation, whereas almost no fluorescence was detected in the control group, suggesting that ROS accumulation may play a key role in hyphal growth inhibition or cell death.