Synergistic Effects of Clonostachys rosea Isolates and Succinate Dehydrogenase Inhibitors Fungicides against Gray Mold on Tomato

Gray mold caused by Botrytis cinerea is a devastating disease in tomatoes. Site-specific fungicide application is still key to disease management; however, chemical control has many drawbacks. Here, the combined application of a biological agent, Clonostachys rosea, with newly developed succinate dehydrogenase inhibitors (SDHI) fungicides showed stronger synergistic effects than the application of SDHI fungicides alone on tomato gray mold control. C. rosea 67-1 has been reported as an efficient biological control agent (BCA) for B. cinerea. Little information is currently available about the combination of C. rosea and fungicides in the control of gray mold. By testing the sensitivity to fungicides with different action mechanisms, C. rosea isolates showed high tolerance to SDHI fungicides (1000 μg mL−1) on PDA, and the conidial germination rate was almost not affected under 120 μg mL−1 of fluxapyroxad and fluopyram. In greenhouse experiments, the control effect of the combination of C. rosea and fluxapyroxad or fluopyram against tomato gray mold was significantly increased than the application of BCA or SDHI fungicides alone, and the combination allows a two-fold reduction of both the fungicide and BCA dose. Further, the biomass of B. cinerea and C. rosea on tomato plants was determined by qPCR. For B. cinerea, the trend of detection level for different treatments was consistent with that of the pot experiments, and the lowest biomass of B. cinerea was found when treated with C. rosea combined with fluxapyroxad and fluopyram, respectively. For C. rosea, qPCR assay confirmed its colonization on tomato plants when mixed with fluopyram and fluxapyroxad. These results indicated that combining C. rosea 67-1 with the SDHI fungicides could synergistically increase control efficacy against tomato gray mold.


Introduction
Gray mold caused by Botrytis cinerea can be a devastating disease in tomatoes worldwide. It is also common with numerous other fruit, vegetables, and ornamental crops [1], which makes it difficult to control. Although cultural methods such as appropriate plant spacing, rational fertilization, and breeding disease-resistant varieties can reduce disease incidence, site-specific fungicide application is still crucial to disease management [2]. However, the polycyclic nature of the disease, abundant sporulation, high genetic variability, and short generation time of the pathogen contribute to a high risk for the development of resistance to site-specific fungicides used for control [3]. Several of the most serious issues of fungicide resistance have been reported in B. cinerea, including resistance to methyl benzimidazole carbamates, dicarboximides, succinate dehydrogenase inhibitors (SDHI), roxad (BASF Corp., Research Triangle Park, NC, USA), 98% fluopimomide (Shandong Zhongnong United Biotechnology Co., Ltd., Jinan, China), and 98% fluopyram (ACMEC, Shanghai, China) were used in this study. Stock solutions were made by dissolving each fungicide in DMSO at the concentration of 10 5 µg a.i. mL −1 . The stock solutions were stored at 4 • C in darkness. Salicylhydroxamic acid (SHAM, 99% a.i.; Syngenta Biotechnology Co. Ltd., Shanghai, China) was added to pyraclostrobin-amended PDA at 100 µg mL −1 to suppress the alternative oxidase pathway [32]. Corresponding control dishes contained SHAM.

Fungicides Sensitivity Assessments of C. rosea and B. cinerea In Vitro
Sensitivity to carbendazim, tebuconazole, boscalid, and pyraclostrobin was assessed on fungicide-amended PDA at 0, 0.1, 0.3, 1, 3, 10, and 30 µg a.i. mL −1 . Furthermore, sensitivity to boscalid, fluxapyroxad, fluopimomide, and fluopyram was assessed on fungicide-amended PDA at 0, 0.1, 0.3, 1, 3, 10, 30, 100, 300, 1000, and 3000 µg a.i. mL −1 . To inoculate test plates, mycelial plugs were removed with a 5-mm cork borer from the margins of 5-day-old colonies and placed upside down on the centers of 9-cm plastic Petri dishes containing fungicide-amended or unamended media. Each isolate was tested in triplicate, and plates were incubated until the diameter reached 60 mm (around five days for B. cinerea and nine days for C. rosea). Fungicide sensitivity, as measured by the 50% effective concentration (EC 50 ) value, was calculated as described by Wong and Wilcox (2002) [33]. Briefly, the percent relative growth (RG) was calculated as (radial growth at fungicide concentration/radial growth on the non-amended control plate) × 100. The EC 50 value was estimated by linear regression of the probit-transformed relative inhibition (RI) value (RI = 1 -RG) on log10 transformed-fungicide concentration. The EC 50 value for each isolate was calculated as the mean of the three replicates.

Effect of SDHI Fungicides to C. rosea Conidia Germination
To determine the inhibition effect of SDHI fungicides boscalid, fluxapyroxad, fluopimomide, and fluopyram on C. rosea and B. cinerea, a spore germination rate test was conducted as described. To stimulate sporulation, C. rosea isolate 67-1 was inoculated in Czapek Dox Liquid Medium (Sigma-Aldrich, St.Louis, MO, USA) [34]. B. cinerea isolate YN80 was inoculated in a PDA medium. Conidia were harvested by flooding 1-2-week-old C. rosea and B. cinerea cultures with a sterile scraper and suspending them in sterile distilled water. The conidial concentration of C. rosea and B. cinerea was then quantified microscopically using a hemocytometer and diluted to a concentration of 1.0 × 10 6 conidia mL −1 . An aliquot of 200 µL of conidia suspension was plated on the YBA medium (10 g L −1 bacto-peptone (Sinopharm, Beijing, China), and 20 g L −1 sodium acetate (Sinopharm, Beijing, China), 10 g L −1 yeast extract (Sinopharm, Beijing, China), and 15 g L −1 agar (Sinopharm, Beijing, China)), then mixed with fungicide using a sterile glass spreader at the final concentrations of 0, 7.5, 15, 30, 60, and 120 µg mL −1 . After 18-24 h incubation at 25 • C in the dark, the number germinated per 100 conidia was counted, and the germination rate of conidia was calculated. The experiment was performed twice.
After 24 h, all of the above tomato seedlings treatments were inoculated with 5-mmagar plugs of B. cinerea isolate YN80 on the leaves referred to Myresiotis et al. [32], except the blank control treatment. Each plant was inoculated with ten agar plugs, one plug for each leaf. Six pots were prepared for each treatment. After inoculation, tomato plants were immediately returned to the chamber to maintain a high relative humidity and an appropriate temperature. Seven days after inoculation, lesion diameters were measured at two perpendicular directions using a caliper, and the control efficacy of each treatment was calculated. The experiments were performed three times.

qPCR for Specific Quantification of C. rosea and B. cinerea
To measure the concentration of the DNA, standard plasmids were constructed. The DNA sequence for B. cinerea was amplified using the primers P1 (5 -GCTGTAATTTCAATGT GCAGAATCC-3 ) and P2 (5 -GGAGCAACAATTAATCGCATTTC-3 ) targeting the Bcos5 gene as reported by Duan et al. [35]. As for C. rosea, primers targeting β-tubulin-encoding genes were retrieved from Genbank (Accession number AF435066). Primers CLO-QF/CLO-QR (CAACAACAACGAGTGGGGAG/ATAAAAGACGGAGCGAAGAC) were designed and used in this study. PCR reactions were performed as follows: 95 • C for 5 min, and then 35 cycles of denaturation at 95 • C for 30 s, annealing at 60 • C for 30 s, extension at 72 • C for 30 s, with a final extension at 72 • C for 10 min. Then, purified PCR products were inserted into the cloning vector pClone007 Vector Kit (Tsingke Biotechnology, Beijing, China), and transformed into an E. coli DH5α competent cell. The transformed competent cells were coated in the LB medium (Luria-Bertani: tryptone (Sinopharm, Beijing) 10 g L -1 , yeast extract (Sinopharm, Beijing) 5 g L −1 , NaCl (Sinopharm, Beijing) 10 g L −1 , agar (Sinopharm, Beijing) 15 g L −1 ) containing 200 µg mL −1 of ampicillin, and incubated at 37 • C to obtain the target cell after 12-16 h. The plasmid DNA was extracted from the target cell using a plasmid mini kit (Tsingke Biotechnology, Beijing, China). The plasmid DNA was used for preparing 10-fold dilution series of eight concentration points starting with about 10 ng/µL, as a "fungal DNA series". The initial stock solution contained around 3 × 10 8 target copies/µL, which was calculated by converting the stock concentration and the mass of the fragment into copy numbers. The concentration of plasmid DNA was quantified by spectrophotometry. The standard curve was prepared in fungal DNA series and amplified to obtain standard curves. Each standard curve was measured in three technical replicates. Standard curves were generated by plotting the logarithmic values of target copies versus the corresponding cycle threshold (Ct) values and fitted into a linear regression model. It was always checked that the R 2 of standard curves ranged from 0.99 to 1. Only Ct values inferior to 40 for B. cinerea and 35 for C. rosea were considered to avoid false positives, and each standard was measured in three technical replicates.
Following the method in Section 2.4, fifteen leaves (five for each plant) were collected from treatment "B. cinerea treatment", "C. rosea treatment", "fluxapyroxad", "fluopyram", "67-1 combined with fluxapyroxad treatment", "67-1 combined with fluopyram treatment" and then ground into a fine powder under liquid nitrogen. For each sample, 150 ± 2 mg was used for DNA extraction to detect fungal content by qPCR. The genomic DNA was subsequently extracted using the Plant Genomic DNA Kit (TIANMO BIOTECH, Beijing, China) according to the manufacturer's instructions.
As for the qPCR detection of B. cinerea and C. rosea, primers P1/ P2 and CLO-QF/CLO-QR for the construction of standard plasmid were used. All qPCR reactions were performed on QuantStudio™ 6 Flex Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) in transparent Multiwell 96-well plates and sealed with adhesive foil. Twenty microliter reaction volume contained 10 µL TSINGKE TSE201 2×TSINGKE ® Master qPCR Mix (SYBR Green I) (Tsingke Biotechnology Co., Ltd., China), 0.8 µL of each primer, 0.4 µL 50×ROX Reference Dye II (Tsingke Biotechnology Co., Ltd., China), 7 µL of DNAse-free water, and 1 µL of DNA sample (unless otherwise stated). The detection wavelength was 520 nm ±10 nm. The following thermal program was applied: an initial denaturation step of 94 • C for 5 min, followed by 40 amplification cycles of 15 s denaturing step (94 • C) and 60 s annealing-extension step (60 • C). All of the experiments were repeated independently twice. Three replications per sample were included in all of the experiments.

Statistical Analysis
Control efficacy = [(lesion diameter of the control − lesion diameter of the treatment)/lesion diameter of the control] × 100%. Results were represented as the mean values ± standard deviation. One-way analysis of variance (ANOVA) with a least significant difference (LSD) test in SPSS software (version 21.0; IBM SPSS Inc. Chicago, IL, USA) was used to evaluate the significant differences between treatments.

In Vitro Mycelial Growth Inhibition of C. rosea and B. cinerea by Differernt Fungicides
To test the compatibility of C. rosea and fungicides, carbendazim, tebuconazole, pyraclostrobin, and boscalid were selected as representative fungicides for Methyl Benzimidazole Carbamates (MBCs), sterol demethylation inhibitors (DMIs), quinone outside inhibitors (QoIs), and SDHIs fungicides, respectively. The sensitivity of C. rosea isolates to those fungicides was tested ( Figure 1). Overall, C. rosea isolates displayed the strongest tolerance to SDHI fungicide boscalid. Boscalid at 10 µg mL −1 or 30 µg mL −1 showed no suppressive activity against mycelium growth of C. rosea on PDA medium. In contrast, C. rosea isolates were quite sensitive to cabendazim and pyraclostrobin, with EC 50 values of 0.34 µg mL −1 -1.66 µg mL −1 and 0.52 µg mL −1 -11.17 mL −1 , respectively. Tebuconazole also had an inhibitory effect on C. rosea mycelia for most of the isolates tested (except for isolate NHH−48-2), with EC 50 values of 0.02 µg mL −1 -21.11 µg mL −1 . C. rosea isolate NHH-48-2 was tolerant to tebuconazole, with EC 50 values of 102.86 µg mL −1 (Table 1).  To further explore the compatibility of SDHI fungicides with C. rosea isolates, mor fungicides from the same categories were tested for their effects on C. rosea isolates. more comprehensive range of concentration was tested for SDHI fungicides boscalid fluxapyroxad, fluopimomide, and fluopyram from 0.1 μg mL −1 to 3000 μg mL −1 . All teste C. rosea isolates displayed strong tolerance to all SDHI fungicides tested. When treate with 100 μg mL −1 of SDHIs, the growth of mycelium was only suppressed by 9.11% t 28.20% (Figure 2). Even when treated with 3000 μg mL −1 of SDHIs, the mycelium coul grow by 53.73% to 77.96% compared to the unamended control. In contrast, the B. cinere isolates YN80 and YN81 were sensitive to all the SDHI fungicides tested, with EC50 les than 15.46 μg mL −1 (Table 1). Based on the EC50 value, fluxapyroxad and fluopyram wer most effective against the B. cinerea isolates used in this study. Thus, those two fungicide were selected for the following experiments. To further explore the compatibility of SDHI fungicides with C. rosea isolates, more fungicides from the same categories were tested for their effects on C. rosea isolates. A more comprehensive range of concentration was tested for SDHI fungicides boscalid, fluxapyroxad, fluopimomide, and fluopyram from 0.1 µg mL −1 to 3000 µg mL −1 . All tested C. rosea isolates displayed strong tolerance to all SDHI fungicides tested. When treated with 100 µg mL −1 of SDHIs, the growth of mycelium was only suppressed by 9.11% to 28.20% (Figure 2). Even when treated with 3000 µg mL −1 of SDHIs, the mycelium could grow by 53.73% to 77.96% compared to the unamended control. In contrast, the B. cinerea isolates YN80 and YN81 were sensitive to all the SDHI fungicides tested, with EC 50 less than 15.46 µg mL −1 (Table 1). Based on the EC 50 value, fluxapyroxad and fluopyram were most effective against the B. cinerea isolates used in this study. Thus, those two fungicides were selected for the following experiments.

Inhibition Effect of Fungicides on the Germination Rate of C. rosea Conidium
The germination inhibition assays of SDHI fungicides were also conducted in our study. The SDHI fungicides had strong inhibitory activity on the spore germination of B. cinerea. The germination rate of YN80 was less than 10% when treated with 15 μg mL -1 of fluxapyroxad and fluopyram ( Table 2). In contrast, the inhibitory activity of fluxapyroxad and fluopyram against C. rosea was very weak. A strong residual growth (with a germination rate above 95%) was observed for C. rosea isolate 67-1 when treated with 120 μg

Inhibition Effect of Fungicides on the Germination Rate of C. rosea Conidium
The germination inhibition assays of SDHI fungicides were also conducted in our study. The SDHI fungicides had strong inhibitory activity on the spore germination of B. cinerea. The germination rate of YN80 was less than 10% when treated with 15 µg mL -1 of fluxapyroxad and fluopyram (Table 2). In contrast, the inhibitory activity of fluxapyroxad and fluopyram against C. rosea was very weak. A strong residual growth (with a germination rate above 95%) was observed for C. rosea isolate 67-1 when treated with 120 µg mL −1 of fluxapyroxad and fluopyram. Thus, good compatibility was observed for SDHI fungicides and C. rosea in vitro.

Synergistic Effects of C. rosea Isolate 67-1 and SDHI Fungicides against Tomato Gray Mold in the Greenhouse
The data regarding the combined effects of C. rosea isolate 67-1 and SDHI fungicides against tomato gray mold in the greenhouse are presented in Figure 3 and Table 3. The average disease diameter in the control group was 2.67 cm in the greenhouse, indicating that B. cinerea was successfully inoculated and well developed (Figure 3). Overall, the combined application of C. rosea and SDHI fungicides, either in a mixture or in a rotation, significantly reduced the disease incidence and severity of tomato gray mold. The highest control efficacy of 77.07% was obtained with pretreatment of isolate 67-1 at 5 × 10 6 conidia mL −1 and then fluopyram at 15 µg mL −1 . The control efficacy of the combined application of isolate 67-1 with fluxapyroxad and fluopyram reached 70.91% and 71.94%, respectively. Sole treatment of fluxapyroxad and fluopyram at 30 µg mL −1 produced a significantly lower control efficacy of 52.28% and 58.31%, respectively, while C. rosea treatment 10 7 conidia mL −1 yielded a control efficacy of 46.42% (Table 3).

qPCR for Specific Quantification of C. rosea and B. cinerea
The evaluation of the Ct values from the standard curve amplification for both cinerea and C. rosea revealed a linear dynamic range from 10 2 to 10 6 target copies, cor sponding to a Ct range of 39~14 for B. cinerea and 32~14 for C. rosea (Figure 4). The low limit of detection of C. rosea was determined around one target copy per reaction as cycles were set to be the cutoff value for the method. Similarly, 40 cycles were set to be cutoff value for B. cinerea. Linear regressions between the log-transformed number of t Figure 3. The disease lesion diameter of tomato gray mold when treated by Clonostachys rosea 67-1, fluxapyroxad, fluopimomide alone, in combination, or in rotation. Data are presented as the mean ± standard deviation (SD). The different lowercase letters indicate significant differences between different treatments in each repeat at the 5% level of probability. "YN80", only inoculated with mycelial plugs of B. cinerea isolate YN80; "67-1", sprayed with conidia suspension of C. rosea isolate 67-1; "Fluxapyroxad", sprayed with fluxapyroxad; "Fluopyram", sprayed with fluopyram; "67-1+Flux", sprayed with the mixture of 67-1 conidia suspension and fluxapyroxad; "67-1+Fluo" sprayed with the mixture of 67-1 conidia suspension and fluopyram; "67-1_Flux", sprayed the 67-1 conidia suspension first and fluxapyroxad 24 h later; "67-1_Fluo" sprayed the 67-1 conidia suspension first and fluxapyroxad 24 h later. 77.07% ± 2.26% a y Treatment "67-1", sprayed with conidia suspension of Clonostachys rosea isolate 67-1; "Fluxapyroxad", sprayed with fluxapyroxad; "Fluopyram", sprayed with fluopyram; "67-1+Flux", sprayed with the mixture of 67-1 conidia suspension and fluxapyroxad; "67-1+Fluo" sprayed with the mixture of 67-1 conidia suspension and fluopyram; "67-1_Flux", sprayed the conidia suspension of C. rosea isolate 67-1 first and fluxapyroxad 24 h later; "67-1_Fluo" sprayed the conidia suspension of C. rosea isolate 67-1 first and fluxapyroxad 24 h later. After 24 h, all of the above tomato seedlings treatments were inoculated with 5-mm-agar plugs of Botrytis cinerea isolate YN80 on the leaves. z Data are presented as the mean ± standard deviation (SD). The different lowercase letters indicate significant differences between different treatments in each repeat at the 5% level of probability. One-way analysis of variance (ANOVA) with a least significant difference (LSD) test in SPSS software (version 21.0; SPSS Inc.) was used to evaluate the significant differences between treatments.

qPCR for Specific Quantification of C. rosea and B. cinerea
The evaluation of the Ct values from the standard curve amplification for both B. cinerea and C. rosea revealed a linear dynamic range from 10 2 to 10 6 target copies, corresponding to a Ct range of 39~14 for B. cinerea and 32~14 for C. rosea (Figure 4). The lower limit of detection of C. rosea was determined around one target copy per reaction as 35 cycles were set to be the cutoff value for the method. Similarly, 40 cycles were set to be the cutoff value for B. cinerea. Linear regressions between the log-transformed number of target copies and the corresponding Ct values revealed R 2 values > 0.99 for both B. cinerea and C. rosea reactions. No PCR inhibition was observed when different amounts of plant DNA isolated from tomato plants were added to the qPCR, increasing concentrations from 1, 10, 25, 50, to 100 ng (data not shown).

qPCR for Specific Quantification of C. rosea and B. cinerea
The evaluation of the Ct values from the standard curve amplification for both B. cinerea and C. rosea revealed a linear dynamic range from 10 2 to 10 6 target copies, corresponding to a Ct range of 39~14 for B. cinerea and 32~14 for C. rosea (Figure 4). The lower limit of detection of C. rosea was determined around one target copy per reaction as 35 cycles were set to be the cutoff value for the method. Similarly, 40 cycles were set to be the cutoff value for B. cinerea. Linear regressions between the log-transformed number of target copies and the corresponding Ct values revealed R 2 values > 0.99 for both B. cinerea and C. rosea reactions. No PCR inhibition was observed when different amounts of plant DNA isolated from tomato plants were added to the qPCR, increasing concentrations from 1, 10, 25, 50, to 100 ng (data not shown). . qPCR standard regression was obtained from the log of the copy number of B. cinerea (a) and C. rosea (b) against the corresponding cycle threshold (Ct) values. Target range was from 5.25 × 10 2 to 5.25 × 10 6 copies per reaction for B. cinerea, and 2.91 × 10 to 2.91 × 10 5 per reaction for C. rosea. The number of target copies on a log-scaled X-axis were plotted against Ct values from 14 to 40 for B. cinerea isolate YN80 and 14 to 32 for C. rosea isolate 67-1 on the Y-axis. Linear regression equation of the B. cinerea standard curve was Y = -6.01x + 55.41 at R 2 = 0.99. Linear regression equation of the C. rosea standard curve was Y = -3.56x + 37.77 at R 2 = 0.99. Thus, the qPCR method was applied to determine the survival of B. cinerea and C. rosea on tomato plants. In sample sets, B. cinerea and C. rosea were always detected when applied and not in the negative control samples. For B. cinerea, the trend of detection level for different treatments was inconsistent with those in the pot experiments. Take repeat 1, for example: two of the lowest levels of detection, with 1.66 × 10 4 copies and 1.08 × 10 4 copies, reflecting the lowest survival of B. cinerea, were found when treated with C. rosea combined with fluxapyroxad and fluopyram, respectively. When treated with C. rosea, the detection levels (with copies of 6.37 × 10 4 ) were higher than those that were treated with fluopyram or fluxapyroxad (with copies of 3.72 × 10 4 and 3.69 × 10 4 , respectively) but lower than those that were treated with distilled water (with copies of 8.55 × 10 4 ). For C. rosea, the qPCR results showed that C. rosea could still be detected on tomato plants when mixed with fluopyram and fluxapyroxad ( Figure 5).

Discussion
A combination of synthetic fungicides with BCA or a combination of different BCAs has been reported to reduce chemical application rates. Several combinations of BCA with fungicides have shown greater efficacy than the individual treatments. For example, combining B. amyloliquefaciens SDTB009 with difenoconazole is an effective strategy for tomato Fusarium wilt management [8]. Synergistic effects have been observed in the combined application of Bacillus subtilis H158 and strobilurins for rice sheath blight control [36]. The

Discussion
A combination of synthetic fungicides with BCA or a combination of different BCAs has been reported to reduce chemical application rates. Several combinations of BCA with fungicides have shown greater efficacy than the individual treatments. For example, combining B. amyloliquefaciens SDTB009 with difenoconazole is an effective strategy for tomato Fusarium wilt management [8]. Synergistic effects have been observed in the combined application of Bacillus subtilis H158 and strobilurins for rice sheath blight control [36]. The combination of Trichoderma and hymexazol enhanced antagonistic effects towards F. oxysporum [37]. Besides, the combination of Metarhizium robertsii and Trichoderma asperellum reduced the malathion doses in controlling ambrosia beetles [38]. However, few studies showed the combination of C. rosea and fungicides or other BCAs in the control of plant disease. In this study, the compatibility of SDHIs fungicides was evaluated and the synergistic effect of the combined use of C. rosea and SDHI fungicides against tomato gray mold was investigated.
The action targets of fungicides against pathogenic fungi include cell membrane integrity, cell mitosis, nucleic acid metabolism, respiration, signal transduction, and protein synthesis [24]. However, some active ingredients of fungicides also act on non-target or beneficial microorganisms such as BCAs, which reduce the growth and population size of BCAs and limit the biocontrol effect [39]. Therefore, knowledge of the compatibility of BCAs and fungicides is essential to allow combined applications. Generally, fungal BCAs resistant to specific fungicides or bacterial BCAs have good compatibility. Compared with the biocontrol fungus, biocontrol bacteria, such as B. amyloliquefaciens and B. subtillis have been reported to tolerate many fungicides and exhibit synergistic effects when applied in combination [40][41][42][43]. The combination of hymexazol-resistant Trichoderma isolate with hymexazol also showed good compatibility and enhanced antagonistic potential [37]. Potential additive or synergistic effects of C. rosea and fungicides depend first on the biological compatibility between the biocontrol agent and the synthetic chemical. In this study, we screened several different categories of fungicides to identify their compatibility with C. rosea. Four FRAC code fungicides that are frequently used for the control of gray mold have been selected. C. rosea isolates were quite sensitive to carbendazim, pyraclostrobin, and tebuconazole in vitro. Fortunately, we found that C. rosea could tolerate SDHI fungicides, including boscalid, fluxapyroxad, fluopimomide, and fluopyram. Even when treated with 3000 µg mL −1 of SDHIs, the mycelium could grow quite well. The natural resistance of fungus to SDHI fungicides are not uncommon. The insensitivity of plant pathogens Colletotrichum species to boscalid, fluxapyroxad, and fluopyram have been confirmed on media and on plants [44]. Penflupen, a novel SDHI fungicide, exhibited good bioactivity against F. fujikuroi, but weak activity against other Fusarium spp. [45]. So far, the inherent resistance mechanisms in the above plant pathogens have remained unknown. As for C. rosea, the natural resistance to SDHIs allows them to be mixed with fungicides.
C. rosea 67-1 isolate has been reported to be a highly efficient biocontrol fungus targeting many plant pathogenic fungi, including B. cinerea [30,31]. Therefore, isolate 67-1 was selected for the following pot experiment. According to our data, the control effect of C. rosea alone was only slightly lower than the application of fungicides, which further proved that C. rosea 67-1 isolate is a promising BCA against B. cinerea. As C. rosea acts by competing for space and nutrients in wounded tissues [46], its efficacy in colonizing the host may depend on the amount of conidia applied. According to Borges et al., who compared the conidial concentration and disease control, the best results for control were obtained at a concentration above 10 6 conidia mL −1 one day before or simultaneously with the pathogens on tomato plants [15]. Thus, we applied C. rosea at 10 7 conidia mL −1 concentration for the control of B. cinerea in our pot experiments and halved the concentration of C. rosea to 5 × 10 6 conidia mL −1 when combined with the fungicides. Based on Chatterton and Punja's research, environmental factors such as temperature and pH were major factors that influenced population levels of C. rosea [14,47]. The optimum temperature for leaf colonization was 20-25 • C, and maximum population densities on the leaves required at least 12 h of continuous leaf wetness [14]. Hence, greenhouse environmental conditions were maintained at 90% relative humidity and 25 • C room temperature for the pot experiment to obtain a stable and efficient control effect.
Our study showed a significant synergistic effect of C. rosea with SDHIs. The control effect of the combination of C. rosea with fluxapyroxad or fluopyram against tomato gray mold was significantly increased compared to that of BCA or SDHI fungicide alone in combination treatment and rotation treatment; the combination allows a two-fold reduction of both the fungicide and BCA dose. Several possible mechanisms for the synergistic effects were observed upon the combined application of C. rosea and SDHIs. Firstly, as the primary biological control mechanism, C. rosea could secrete cell-wall-degrading enzymes (CWDEs) to degrade the cell wall of the host fungus [48][49][50]. Thus, with the lack of an essential barrier for cell protection, the gray mold might become more vulnerable to the fungicides treated. Second, C. rosea produced secondary metabolites such as antibiotics and toxins [51,52], and the combined application of these antibiotics or toxins with SDHIs may show the same synergistic effects as the synergistic effect shown in a combination of fungicides with one another. Third, treating B. cinerea infection with C. rosea has been reported to induce several defense mechanisms in tomatoes, including fortifying the plant cell wall and stimulating the expression of several signaling molecules [16,19,53]. In this way, the resistance of tomato plants to gray mold is enhanced when inoculated with C. rosea. After the fungicide treatment, the plants are less susceptible to gray mold, showing a synergistic effect.
Whether the BCAs survive on plants or colonize the plants successfully after the application is a crucial step for the biological control activity of many BCAs. Rapid activity loss is thought to be the main reason some BCAs are not successful in the field but show excellent performance in the lab [54]. It is reported that B. subtilis was rapidly lost 3 days after application on rice by using real-time qPCR detection [36]. This result is in accordance with a study of B. subtilis on a strawberry based on next-generation sequencing [55]. In terms of C. rosea, it was confirmed that C. rosea could successfully colonize the foliage of geraniums and the roots of cucumbers by using a GUS-transformed isolate, demonstrating the endophytic ability of C. rosea in foliar and root tissues [14,47]. In this study, DNA of C. rosea was directly extracted from tomato plants, and the fungal dynamics were analyzed by real-time qPCR to quantify C. rosea DNA. Although DNA extraction included dead and inactive fungi and may result in a higher gene expression level, it was believed to be the most available method because of its convenience and accuracy [36]. In the qPCR assays, though there were variations between the replicates, the replicates showed a similar trend ( Figure 5). Because the absolute quantifications of B. cinerea and C. rosea were tested, it was very hard to repeat the absolute copy number from the two independent experiments. The environment and the status of the microorganisms can be slightly different from the two replicates, which ultimately influence the colonization. The qPCR test of C. rosea demonstrated that C. rosea could still be detected on tomatoes when used alone and mixed with fungicides. The qPCR test of B. cinerea showed that C. rosea and SDHI fungicide significantly reduced the biomass of B. cinerea. Compared to the control, the biomass of B. cinerea was the lowest in the combination treatment of C. rosea and SDHI fungicide, which is consistent with the control efficacy in the greenhouse.
In conclusion, our study showed that C. rosea isolates could tolerate high concentrations of SDHIs with no adverse growth effects, suggesting that they were fully compatible with these fungicides. Pot experiment and qPCR assays showed a significant synergistic effect of C. rosea with SDHIs in controlling tomato gray mold. These results showed that combining BCA with SDHIs may meet the demands of the Chinese government's "low fertilizer and low pesticides" campaign. Additional field trials and investigations to monitor the behavior of C. rosea in the field can help to determine the optimal timing and the method of this BCA application to control gray mold in tomato production.