Prevalence of Quinone Outside Inhibitor Fungicide Resistance and Dynamics of Cytochrome b Genotypes Under Different Fungicide Selection Pressure in Botrytis cinerea Isolates from Korean Strawberry Farms

Abstract

Botrytis cinerea, the causal agent of gray mold, is a major pathogen in strawberry production. Quinone outside inhibitor (QoI) fungicides have been widely used for disease control, but resistance driven by the G143A mutation in cytochrome b (cytb) has limited their field efficacy. This study evaluated the sensitivity of 229 B. cinerea isolates collected from farms in major Korean strawberry-growing regions during 2022–2024 to the QoI fungicide pyraclostrobin. The EC₅₀ values for pyraclostrobin ranged from 0.10 to 329.21 μg mL⁻¹. In total, 86.46% of the population showed resistance to pyraclostrobin, harboring the G143A mutation in cytb. In planta and greenhouse assays confirmed that pyraclostrobin failed to control the QoI-resistant isolates. Furthermore, strong cross-resistance to pyraclostrobin, mandestrobin, and pyribencarb indicates that QoI-resistant populations substantially limit the field efficacy of QoI fungicides. To assess how the QoI resistance mutation affects selection under non-QoI fungicide pressure, mixed QoI-sensitive and -resistant isolates were treated with alternative fungicides. Allele-specific qPCR revealed that pydiflumetofen, fludioxonil, and fluxapyroxad maintained a higher proportion of the QoI-sensitive genotype at all concentrations, whereas pyraclostrobin consistently favored the resistant genotype. These findings indicated that QoI fungicides impose strong selection pressure promoting resistant B. cinerea populations, while non-QoI fungicides help maintain sensitive individuals. Our study suggests the need for fungicide class rotation and the targeted use of non-QoI fungicides to delay QoI resistance development and preserve the efficacy of gray mold management in strawberry production.

Botrytis cinerea, the causal agent of gray mold, is considered one of the top 10 most important plant-pathogenic fungi worldwide (Dean et al., 2012). This necrotrophic fungus infects a broad range of hosts, including fruits, vegetables, and ornamentals, and causes major losses before and after harvest (Nakielska et al., 2024; Williamson et al., 2007). In Korea, strawberry is a major fruit crop, and its industry has expanded over the past decade (Jee et al., 2024). By 2020, the cultivation area for strawberries reached 5,683 ha, and the production value was KRW 1.227 trillion, accounting for 10.9% of the total vegetable production value (Korea Agro-Fisheries & Food Trade Information, 2023). The cultivar 'Seolhyang' represents approximately 50% of national production, and its gray mold infection rate has been reported to exceed 50% (Kim et al., 2016; Lee and Chae, 2012). Gray mold also reduces postharvest fruit quality and causes substantial economic losses throughout the supply chain (Droby and Lichter, 2007; Nakielska et al., 2024).

While chemical control using multiple fungicide classes remains a key component of managing B. cinerea (Leroux et al., 2002), the efficacy of chemical control is increasingly threatened by the widespread emergence of fungicide resistance. Continuous exposure to active ingredients has not only led to documented resistance against benzimidazoles, succinate dehydrogenase inhibitor (SDHI), and quinone-outside inhibitor (QoI) fungicides across many regions, but also selected for multidrug-resistant phenotypes that severely complicate disease management (Kim et al., 2014, 2023; Kretschmer et al., 2009; Liu et al., 2024). In particular, QoI fungicides, such as azoxystrobin, mandestrobin, orysastrobin, pyribencarb, and pyraclostrobin, are widely used and are highly effective against a range of plant diseases (Bartlett et al., 2002; Lee et al., 2024; Lim et al., 2025; Liu et al., 2025; Soe et al., 2024). These compounds inhibit mitochondrial respiration by binding to the Qo site of the cytochrome bc1 complex and blocking electron transfer, thereby preventing ATP synthesis (Von Jagow et al., 1986). However, their single-site mode of action facilitates the rapid emergence of resistance and Fungicide Resistance Action Committee (FRAC) classified QoIs as a high-risk group (FRAC, 2025). In B. cinerea, target-site resistance is predominantly driven by a point mutation in the cytochrome b (cytb) gene resulting in the G143A amino acid substitution (Banno et al., 2009; De Miccolis Angelini et al., 2012; Fernández-Ortuño et al., 2014). Isolates harboring this mutation exhibit high levels of resistance and often lead to field control failure (Samuel et al., 2011; Yin et al., 2018). In Korea, azoxystrobin resistance has been reported in isolates collected from several host crops, with strawberry isolates exhibiting the highest resistance rates among those surveyed (Kim et al., 2014). Because this single point mutation confers such strong resistance, incorporating fungicides with different modes of action into gray mold management programs has become increasingly important (Bartlett et al., 2002).

The emergence, spread, and persistence of fungicide resistance in fungal populations are influenced by multiple factors (Jeger et al., 2008). Selection pressure from fungicide applications strongly selects for target-site mutations, and resistant phenotypes may persist or decline depending on associated fitness costs (Jeger et al., 2008; Leroux et al., 2002). Fitness costs associated with resistance mutations typically reduce competitive ability through impaired growth, reproduction, or stress tolerance, potentially limiting the persistence of resistant populations in the absence of selection pressure. Conversely, when resistance mutations impose minimal fitness penalties, resistant genotypes can maintain stable frequencies in field populations regardless of fungicide use (Leroux et al., 2002). In B. cinerea, QoI-resistant isolates carrying the G143A mutation have been reported to show no fitness cost (Chen et al., 2016), however it remains unclear whether this mutation affects population structure under non-QoI fungicide pressure. Non-QoI fungicides such as SDHIs and phenylpyrroles, which target different cellular pathways, may impose different selection pressures on QoI resistant populations.

Monitoring QoI resistance in B. cinerea field populations and evaluating QoI efficacy against sensitive and resistant isolates are essential for integrated gray mold management. Accordingly, this study aimed to (1) monitor QoI resistance in B. cinerea isolates collected from major Korean strawberry-growing regions during 2022–2024, (2) assess cross-resistance among QoI fungicides, and (3) compare the fitness of isolates with and without the G143A mutation under selection pressure from fungicides with different modes of action. These findings provide insight into resistance dynamics and support sustainable management strategies beyond reliance on the introduction of new fungicides.

Materials and Methods

Sample collection and fungal isolation

A total of 229 B. cinerea isolates were collected from strawberry fields in 91 cities across Korea between 2022 and 2024. Some isolates from 2022 and 2023 were used in the previous studies (Choi et al., 2026; Kim et al., 2023) (Supplementary Table 1). Single-conidium isolation was performed for purification, and isolates were cultured on potato dextrose agar (PDA; BD Difco, Sparks, MD, USA) at 25°C in the dark for 3 days. These isolates were stored at −80°C in 15% glycerol stock solution in the Molecular Microbiology Laboratory, Dept. of Integrative Food, Bioscience and Biotechnology, Chonnam National University, Gwangju, Republic of Korea.

Molecular identification

Genomic DNA was extracted from isolates cultured on PDA using the CTAB method (Cubero et al., 1999). For species identification, the internal transcribed spacer (ITS) region and the necrosis and ethylene-inducing protein 2 (NEP2) gene were amplified using the primer pairs ITS5/ITS4 and NEP2forD/NEP2revD, respectively (Staats et al., 2007). PCR was performed in 10 μL reaction volumes containing 0.5 μL of template DNA (50 ng μL⁻¹), 0.5 μL of each primer (10 pmol μL⁻¹), 0.2 μL of dNTPs (10 mM), 2 μL of 5× Phusion GC buffer, and 0.1 μL of Phusion DNA polymerase (0.02 U μL⁻¹; New England Biolabs, Ipswich, MA, USA). Thermocycling consisted of an initial denaturation at 98°C for 30 s; followed by 35 cycles of 98°C for 10 s, 58°C (ITS) or 52°C (NEP2) for 30 s, and 72°C for 30 s; and a final extension at 72°C for 10 min. Amplicons were purified using ExoSAP-IT (Thermo Fisher Scientific, Waltham, MA, USA) and bidirectionally sequenced by Macrogen Inc. (Seoul, Korea). Raw sequences were assembled and aligned with closely related reference sequences from GenBank using MUSCLE in MEGA X (Kumar et al., 2018). Phylogenetic trees were constructed using the Maximum Likelihood (ML) method based on the Tamura-Nei model, with 1,000 bootstrap replicates to evaluate branch support.

Fungicides

Pyraclostrobin (Standby; BASF, Ludwigshafen, Germany) was used for EC₅₀ determination and in planta assays. For cross-resistance testing, two additional QoI fungicides, pyribencarb (Seondujuja; Kumiai Chemical Industry, Tokyo, Japan) and mandestrobin (Mandes; Sumitomo Chemical, Tokyo, Japan), were included. Fluxapyroxad (Kadis, BASF, Ludwigshafen, Germany), fludioxonil (Sapphire; Syngenta, Basel, Switzerland), and pydiflumetofen (Miraebit, Syngenta, Basel, Switzerland) were used for sensitivity assays against alternative fungicides.

In vitro sensitivity assays of Botrytis cinerea isolates to QoI fungicides

Sensitivity of 229 B. cinerea isolates to pyraclostrobin was evaluated by determining the EC₅₀ values for mycelial growth inhibition. Isolates were cultured on PDA at 25°C in the dark for 3 days. Mycelial plugs (5 mm diameter) were excised from the actively growing colony margins and transferred onto PDA plates amended with pyraclostrobin at 0, 0.1, 1, 10, 100, and 1,000 μg mL⁻¹. Salicylhydroxamic acid (SHAM; Sigma-Aldrich, St. Louis, MO, USA) was added to all plates at 100 μg mL⁻¹ to inhibit alternative respiration. Plates were incubated at 25°C in the dark for 3 days, and colony diameters were measured using a digital caliper (Mitutoyo, Japan). EC₅₀ values were calculated with the R package drc (v4.2.3) (Ritz et al., 2015). A representative subset of QoI-sensitive (CMML23-BC12, CMML23-BC13, CMML23-BC16, CMML23-BC39, CMML23-BC73, CMML23-BC95, CMML23-BC96, and CMML23-BC97) and QoI-resistant (CMML23-BC10, CMML23-BC18, CMML23-BC20, CMML23-BC26, CMML23-BC30, CMML23-BC38, CMML23-BC59, CMML23-BC69, CMML23-BC80, and CMML23-BC82) isolates was additionally tested for sensitivity to pyribencarb and mandestrobin under the same conditions. Two replicate plates were used per treatment, and the experiment was repeated twice.

Cytochrome b sequence analysis

The cytb gene was amplified using primers Qo13ext (5′-GGT ATA ACC CGA CGG GGT TAT AGA ATA G-3′) and Qo14ext (5′-AAC CAT CTC CAT CCA CCA TAC CTA CAA A-3′) (Leroux et al., 2010). PCR reactions (20 μL) contained 2 μL 10× nPfu-Forte buffer, 2 μL of 2 mM dNTPs, 1 μL genomic DNA (100 ng), 0.5 U nPfu polymerase, and 13.8 μL nuclease-free water. PCR cycling conditions were as follows: 95°C for 2 min; 35 cycles at 95°C for 30 s, 48°C for 30 s, and 72°C for 1 min; and a final extension at 72°C for 5 min. PCR products were purified with ExoSAP-IT (Applied Biosystems, Foster City, CA, USA) and Sanger sequenced by Macrogen Inc. (Seoul, Korea). Resulting sequences were aligned using MEGAX (Kumar et al., 2018) and examined for the presence of the G143A mutation in the cytb gene.

Sensitivity assays of Botrytis cinerea isolates to pyraclostrobin on strawberry leaves

Sensitivity of B. cinerea isolates to pyraclostrobin was evaluated using detached leaves of the strawberry cultivar ‘Seolhyang’. For this assay, two QoI-sensitive (CMML23-BC13 and CMML23-BC96) and two resistant isolates (CMML23-BC74 and CMML23-BC79) were selected. Conidia of the isolates were harvested from 7-day-old colonies grown under near-UV light (16 h light and 8 h dark at 25°C), and suspensions were adjusted to 1 × 10⁵ conidia mL⁻¹. Leaves were surface-sterilized with 1% sodium hypochlorite for 3 min, rinsed three times with distilled water, and allowed to air-dry for 20 min. Pyraclostrobin at 28.625 μg mL⁻¹ (0.5×), 57.25 μg mL⁻¹ (1×), or 114.5 μg mL⁻¹ (2×) was sprayed on the leaves, while sterile water served as the control. After drying for 1 h, a 10 μL droplet of the conidia suspensions was placed on each leaf. Inoculated leaves were incubated at 25°C in the dark at 80% relative humidity for 3 days. Lesion sizes were measured using a digital caliper. Three leaves were used per treatment, and the experiment was repeated twice.

Fruit assay of QoI-sensitive and -resistant Botrytis cinerea isolates to different QoI fungicides

Three QoI fungicides, pyraclostrobin, mandestrobin, and pyribencarb, were evaluated using the methods described above. Cross-resistance among these fungicides was further evaluated in vivo using strawberry cultivar ‘Seolhyang’ fruits. Four isolates differing in cytb genotype were used: two QoI-sensitive isolates (CMML23-BC15 and CMML23-BC21) without the G143A mutation in cytb, and two QoI-resistant isolates (CMML23-BC26 and CMML23-BC69) carrying the G143A substitution in cytb. Fruits were surface-sterilized with 1% sodium hypochlorite for 3 min, rinsed three times, and dried for 20 min. Fruits were sprayed with pyraclostrobin (57.25 μg mL⁻¹), mandestrobin (200 μg mL⁻¹), or pyribencarb (100 μg mL⁻¹). Sterile water served as a control. Conidial suspensions were adjusted to 1×10⁵ conidia mL⁻¹, and 10 μL droplets were applied to the fruit surface. After drying for 1 h, the inoculated fruits were incubated at 25°C in the dark with 80% relative humidity for 5 days. Lesion sizes were measured using a digital caliper. Two fruits were used per treatment, and the experiment was repeated twice.

Control efficacy assay of QoI fungicides against Botrytis cinerea isolates in a strawberry greenhouse

In March 2025, a greenhouse trial was conducted in Gwangju, South Korea, to evaluate the control efficacy of QoI fungicides against strawberry gray mold. Plants with no prior fungicide treatment were sprayed with 100 mL of pyraclostrobin (57.25 μg mL⁻¹), mandestrobin (200 μg mL⁻¹), or pyribencarb (100 μg mL⁻¹). Control treatments included fludioxonil (100 μg mL⁻¹) and an inoculated control treated with sterile water. Fungicides were applied at weekly intervals for three consecutive weeks. Once the first application had dried, the plants were inoculated with 10 mL of a conidial suspension (1 × 10⁵ spores mL⁻¹) from QoI-sensitive (CMML23-BC13) and resistant (CMML23-BC81) B. cinerea isolates using a hand sprayer. Disease severity was evaluated at 14 and 21 days after the first fungicide application. Each plot contained three plants, with three replicate plots per treatment. All ripened fruits were rated on a 0–10 scale (0 = no disease, 10 = severe disease). The disease severity index (DSI) was calculated using the following formula:

$$\text{DSI}(\%)=\frac{\mathrm{\Sigma }(\text{rating}\times \text{number\hspace{0.17em}of\hspace{0.17em}fruits\hspace{0.17em}at\hspace{0.17em}that\hspace{0.17em}rating})}{10\times \text{total\hspace{0.17em}number\hspace{0.17em}of\hspace{0.17em}fruits}}\times 100$$

Fitness evaluation of QoI-resistant and -sensitive isolates

Six representative isolates were selected for fitness evaluation based on their cytb genotype: three QoI-sensitive isolates (CMML23-BC12, CMML23-BC13, and CMML23-BC95) lacking the cytb-G143A mutation and three QoI-resistant isolates (CMML23-BC14, CMML23-BC64, and CMML23-BC81) carrying the G143A substitution in cytb. For mycelial growth assessment, mycelial plugs (5 mm diameter) were excised from colony margins and transferred to PDA plates. Plates were incubated at 25°C in the dark, and colony diameters were measured at 48 and 72 hours using the digital caliper. For sporulation evaluation, isolates were cultured on PDA for 2 days at 25°C in the dark and subsequently incubated at 20°C under near-UV light for 5 days to induce sporulation. Conidia were harvested by flooding plates with 3 mL of sterilized distilled water supplemented with 0.05% Tween 20, filtering the suspension through a 40 μm cell strainer, centrifuging for 2 min (14,000 rpm), and washing the conidial pellet with sterile water. Conidial concentrations were adjusted to 1 × 10⁵ conidia mL⁻¹ using a hemocytometer, and 30 μL aliquots were inoculated onto water agar plates and incubated at 25°C for 24 h. Germination rates were determined by microscopic examination, counting germinated and non-germinated conidia in six random fields per plate. Stress tolerance was evaluated using PDA plates amended with NaCl (0.4 M), CaCl₂ (0.4 M), sorbitol (0.8 M), and H₂O₂ (20 mM). Colony diameters were measured after 3 days of incubation at 25°C in the dark using the digital caliper. Two replicate plates were used per treatment, and the experiment was repeated twice.

Comparative analysis of QoI-resistant and -sensitive isolates to fungicides with different modes of action

To investigate proportion changes between QoI-sensitive (non-mutation in cytb) and -resistant (G143A in cytb) isolates under selection pressure from fungicides with different modes of action, comparative assays using two isolates were performed. The QoI-sensitive isolate CMML23-BC13 and the QoI-resistant isolate CMML23-BC81 were cultured on 50 mm PDA plates for 2 days and subsequently incubated at 20°C under near-UV light for 3 days to induce conidiation. Conidia were harvested by flooding plates with 3 mL of sterilized distilled water supplemented with 0.05% Tween 20, filtering the suspension through a 40 μm cell strainer, centrifuging for 2 min, and washing the conidial pellet with sterile water. Conidial concentrations were adjusted to 1 × 10⁶ mL⁻¹. Equal volumes of the two suspensions were mixed at a 1:1 (S:R) ratio, and 20 μL aliquots were inoculated onto PDA plates amended with each fungicide. EC₅₀ values were determined for the QoI-sensitive and QoI-resistant isolates (Supplementary Table 2). The geometric mean EC₅₀ values were calculated as follows: pyraclostrobin, 3.96 μg mL⁻¹; fluxapyroxad, 0.62 μg mL⁻¹; pydiflumetofen, 0.019 μg mL⁻¹; and fludioxonil, 0.021 μg mL⁻¹. Experimental concentrations were established at 0.5×, 1×, and 2× the respective mean EC₅₀ values for each fungicide (Supplementary Table 3). After 2 days of incubation, mycelial tissue was harvested for genomic DNA extraction by the CTAB method.

The proportions of resistant and sensitive alleles were determined using a TaqMan allele-specific qPCR assay targeting the cytb gene (Samuel et al., 2011). The assay employed four oligonucleotides, as presented in Supplementary Table 4. The probes differed by a single nucleotide substitution at the cytb locus. The resistant allele-specific probe was labeled with FAM, and the sensitive allele-specific probe with Cy5. Ct values obtained from resistant-specific (Ct_R) and sensitive-specific (Ct_S) probes were used to calculate the proportion of the resistant allele according to the following equation:

$$\text{R}(\%)=\frac{2^{-C{t}_R}}{2^{-C{t}_R}+2^{-C{t}_S}}\times 100$$

Higher R (%) values indicate a greater proportion of resistant alleles. Each treatment consisted of three samples, each analyzed four times.

Results

Identification of Botrytis cinerea isolates from strawberry farms

PCR amplicons of ITS region and NEP2 gene were approximately 433 bp and 631 bp in size, respectively. The ITS phylogenetic analysis included 28 taxa, utilizing Sclerotinia sclerotiorum (F113458) as an outgroup. While the ITS phylogeny successfully classified the isolates within the Botrytis genus, it lacked sufficient resolution to differentiate them from other closely related Botrytis species (Fig. 1A). To achieve precise species-level identification, a subsequent phylogenetic analysis was performed using the NEP2 gene, incorporating 25 reference taxa with Botrytis prunorum (Bpru-21) as the outgroup. The NEP2 phylogeny provided significantly higher resolution. All tested isolates clustered into a single, well-supported clade (80% bootstrap value) alongside reference strains of B. cinerea (Fig. 1B). The ITS and NEP2 sequences were deposited in GenBank (Acc. nos. PZ707233-PZ707346, PZ715747-PZ715769 for ITS, PZ716636-PZ716762 for NEP2).

Fig. 1.

Maximum likelihood phylogenetic trees of Botrytis isolates. The trees were generated based on sequences of (A) the internal transcribed spacer (ITS) region and (B) the necrosis and ethylene-inducing protein 2 gene. The analysis includes B. cinerea field isolates from this study alongside reference sequences retrieved from GenBank. Neotype strains are indicated by “T”. Bootstrap values greater than 60% (based on 1,000 replicates) are shown at the nodes.

In vitro sensitivity of Botrytis cinerea isolates to pyraclostrobin

Across the isolates collected during 2022–2024 (n = 229), pyraclostrobin EC₅₀ values ranged from 0.10 to 329.21 μg mL⁻¹ (Fig. 2). Year-specific means ± SD were 73.75 ± 62.33 (2022, n = 62), 62.93 ± 40.34 (2023, n = 90), and 82.52 ± 58.05 (2024, n = 77). Using an EC₅₀ threshold of 10 μg mL⁻¹, below which target-site mutations were absent, the proportions of resistant isolates were 88.7%, 83.3%, and 88.3% for 2022, 2023, and 2024, respectively. The relative mycelial growth (RMG) at 10 μg mL⁻¹ showed the strongest correlation with log₁₀(EC₅₀) among the tested concentrations (R² = 0.888, P < 0.001), supporting its use as the discriminatory dose (Supplementary Fig. 1). A high-resistance subset (EC₅₀ ≥ 100 μg mL⁻¹) accounted for 24.2%, 15.6%, and 29.9% in each respective year. Overall, resistance frequencies remained high and stable across years, with no evidence of decline. Sequencing of the cytb gene confirmed that all resistant isolates carried the G143A mutation in cytb. Geographically, resistance frequencies varied considerably across provinces. Jeolla Province exhibited the highest frequency (93%, 41 isolates), followed by Gangwon (88%, 40 isolates), Chungcheong (87%, 52 isolates), Gyeongsang (85%, 52 isolates), and Gyeonggi (82%, 44 isolates) Provinces (Fig. 3). EC₅₀ distributions and G143A mutation status are shown in Supplementary Fig. 2.

Fig. 2.

Distribution of EC₅₀ values for pyraclostrobin against Botrytis cinerea isolates (2022–2024). Bars show the number of isolates per EC₅₀ bin (μg mL⁻¹; log₁₀ scale). Colors denote the collection year; dashed lines indicate the year-specific geometric mean EC₅₀. Sensitive isolates had EC₅₀ <10 μg mL⁻¹; resistant isolates had EC₅₀ ≥10 μg mL⁻¹. The LL.3 model of the R package “drc” was used to calculate EC₅₀.

Fig. 3.

Geographic distribution of pyraclostrobin resistance in Botrytis cinerea isolates collected from strawberry greenhouses across five Korean provinces from 2022 to 2024. Each province included ≥40 isolates. Red indicates resistant isolates; blue indicates sensitive isolates.

In planta sensitivity of Botrytis cinerea isolates to pyraclostrobin on strawberry leaves

The sensitivity of B. cinerea isolates to pyraclostrobin on detached strawberry leaves was evaluated. The pyraclostrobin-sensitive isolates (CMML23-BC13 and CMML23-BC96) were effectively controlled even at half the recommended field concentration (28.625 μg mL⁻¹) (Fig. 4A). Disease development was largely suppressed in CMML23-BC13 (4 of 6 replicates showing no symptoms; mean ± SD: 32.7 ± 50.8%; n = 6) and reduced to 33–96% of the untreated control in CMML23-BC96 (mean ± SD: 60.9 ± 24.7%; n = 6), even at half the standard application rate. In contrast, pyraclostrobin-resistant isolates CMML23-BC74 and CMML23-BC79 showed poor control efficacy across all concentrations, including the 2× rate (114.5 μg mL⁻¹). At the 2× concentration, relative disease severity in CMML23-BC74 and CMML23-BC79 (mean ± SD: 111 ± 13.5% and 101 ± 21.7%, respectively; n = 6) was similar to the untreated control, indicating no disease suppression (Fig. 4B).

Fig. 4.

In planta sensitivity of QoI sensitive and resistant Botrytis cinerea isolates to pyraclostrobin on strawberry leaves. (A) Representative photographs of symptoms on strawberry leaves at the indicated pyraclostrobin concentrations; 1× indicates the label-recommended concentration, and 1/2× and 2× indicate half and twice that concentration, respectively. (B) Relative lesion growth (%) for each isolate across concentrations (mean ± SD). Different letters indicate significant differences among treatments (Tukey’s HSD, α = 0.05). QoI, quinone outside inhibitor.

Cross-resistance of B. cinerea isolates against different QoI fungicides

Cross-resistance among B. cinerea isolates to QoI fungicides was assessed in vitro on PDA amended with pyraclostrobin, mandestrobin, or pyribencarb (Fig. 5). EC₅₀ values of each isolate against pyraclostrobin, mandestrobin, and pyribencarb are presented as a heatmap (Fig. 5A, Supplementary Table 5). In the sensitive isolates, EC₅₀ values for all three QoI fungicides remained below 10 μg mL⁻¹, whereas all resistant isolates exhibited EC₅₀ values exceeding 10 μg mL⁻¹ across all fungicides, with fungicide-dependent variation in resistance levels. Pairwise correlation analysis revealed R² values exceeding 0.8 for all fungicide pairs (Fig. 5B), indicating strong cross-resistance among QoI fungicides.

Fig. 5.

Cross-resistance among QoI fungicides in Botrytis cinerea isolates. (A) Heatmap of EC₅₀ values (μg mL⁻¹) for pyraclostrobin, mandestrobin, and pyribencarb. Darker shading indicates higher EC₅₀ values. (B) Pairwise correlations of EC₅₀ values between QoI fungicide pairs with linear regression lines (red). R² values are indicated for each pair. QoI, quinone outside inhibitor.

This cross-resistance was further evaluated in planta using detached strawberry fruits inoculated with two QoI-sensitive (CMML23-BC15 and CMML23-BC21) and two QoI-resistant (CMML23-BC26 and CMML23-BC69) isolates (Fig. 6A). In the sensitive isolates, all three QoI fungicides effectively reduced lesion sizes to 7.46 ± 3.16 mm for pyraclostrobin, 4.98 ± 1.20 mm for mandestrobin, and 7.83 ± 0.86 mm for pyribencarb, compared to the water control (19.51 ± 1.81 mm). For the resistant isolates, QoI fungicides demonstrated limited efficacy in reducing lesion size: pyraclostrobin (17.76 ± 2.30 mm), mandestrobin (15.97 ± 2.10 mm), and pyribencarb (15.16 ± 4.09 mm), compared to water control (18.35 ± 2.33 mm). Inhibition rates in sensitive isolates ranged from 59.84% to 74.45%, while resistant isolates exhibited considerably lower inhibition rates of 3.24% to 17.38% (Fig. 6B). These findings confirmed the cross-resistance to multiple QoI fungicides in the resistant B. cinerea isolates.

Fig. 6.

Cross-resistance among QoI fungicides against QoI-sensitive (CMML23-BC15 and CMML23-BC21) and -resistant (CMML23-BC26 and CMML23-BC69) B. cinerea isolates on strawberry. (A) Representative images of symptoms on strawberry fruit in each treatment; three QoI fungicides were applied at the label recommended concentration. (B) Bar graph of mycelial growth (mm) for each isolate (mean ± SD). Different letters indicate significant differences among treatments within an isolate (Tukey’s HSD, α = 0.05). QoI, quinone outside inhibitor.

Control efficacy assay of QoI fungicides against B. cinerea isolates in a strawberry greenhouse

No visible gray mold symptoms were observed in any treatments at 7 days post inoculation (dpi). In plots inoculated with the sensitive isolate, disease severity in water treatments was 24% at 14 dpi. Fludioxonil demonstrated superior control efficacy (92.3%), while QoI fungicides showed limited efficacy: pyribencarb (34.4%), mandestrobin (18.1%), and pyraclostrobin (13.3%). Differences between QoI fungicide treatments were not statistically significant (Fig. 7A). By 21 dpi, disease severity in water treatments increased to 50%. Fludioxonil maintained high control efficacy (86.3%), whereas QoI fungicides remained relatively ineffective: pyribencarb (22.5%), mandestrobin (15.8%), and pyraclostrobin (13.6%) (Fig. 7B). In plots inoculated with the resistant isolate, disease severity in water treatments was 24.6% at 14 dpi. Fludioxonil showed moderate efficacy (73.1%), whereas QoI fungicides provided no measurable control: pyraclostrobin (0%), mandestrobin (0%), and pyribencarb (0%). Disease severity in QoI fungicide-treated plots was comparable to water-treated plots (Fig. 7C). By 21 dpi, disease severity in water treatments increased to 52.4%. Fludioxonil demonstrated high efficacy (98.0%), while QoI fungicides remained largely ineffective: pyraclostrobin (0%), pyribencarb (0%), and mandestrobin (15.0%) (Fig. 7D).

Fig. 7.

Disease severity and control efficacy of three QoI fungicides and fludioxonil in strawberries inoculated with a QoI-sensitive Botrytis cinerea isolate at 14 dpi (A) and 21 dpi (B), or a QoI-resistant B. cinerea isolate at 14 dpi (C) and 21 dpi (D). Fungicides (fludioxonil, mandestrobin, pyraclostrobin, and pyribencarb) or water were applied at 7-day intervals at the label-recommended concentration (1×). Control efficacy was calculated relative to the disease severity (%) in water-treated plots inoculated with the B. cinerea isolate. Different letters indicate significant differences among treatments (Tukey’s HSD, α = 0.05). QoI, quinone outside inhibitor.

Fitness evaluation of QoI-resistant and QoI-sensitive isolates

Fitness parameters were compared between QoI-sensitive (CMML23-BC12, CMML23-BC13, and CMML23-BC95) and QoI-resistant (CMML23-BC14, CMML23-BC64, and CMML23-BC81) isolates (Table 1). At 48 hours, colony diameters ranged from 46.73 to 54.78 mm, with no consistent differences between sensitive and resistant isolates. At 72 hours, colony diameters ranged from 67.35 to 77.91 mm across all isolates, with no consistent differences between sensitive and resistant isolates. Conidial germination rates were 7.80%, 5.71%, and 7.50% for sensitive isolates, and 6.56%, 5.34%, and 5.45% for resistant isolates, respectively. Under abiotic stress conditions, H₂O₂ (20 mM) produced inhibition rates of 44.9–52.7% for sensitive isolates and 40.5–55.8% for resistant isolates. Sorbitol (0.8 M), CaCl₂ (0.4 M), and NaCl (0.4 M) stress conditions showed inhibition rates of 0–7.0%, 18.3–20.4%, 0–1.6% for sensitive isolates, and 0%, 0–25.0%, and 0–4.0% for resistant isolates, respectively. No consistent fitness differences were observed between QoI-sensitive and resistant isolates across all evaluated parameters.

Table 1.

Fitness of Botrytis cinerea isolates under stress conditions

Parameter		CMML23-BC12 (WT)	CMML23-BC13 (WT)	CMML23-BC14 (G143A)	CMML23-BC64 (G143A)	CMML23-BC81 (G143A)	CMML23-BC95 (WT)
Colony diameter 48 h (mm)		54.78 ± 2.86	49.46 ± 1.24	46.73 ± 1.56	51.20 ± 5.34	50.73 ± 1.43	47.84 ± 4.88
Colony diameter 72 h (mm)		77.91 ± 2.62	67.35 ± 4.46	68.50 ± 3.60	72.87 ± 3.24	73.00 ± 5.31	71.34 ± 6.18
Germination rate (%)		7.80 ± 2.14	5.71 ± 1.77	6.56 ± 1.85	5.34 ± 4.20	5.45 ± 0.98	7.50 ± 5.40
Inhibition rate (%)	H2O2 20 mM	44.9 ± 5.8	52.7 ± 4.4	40.5 ± 5.8	55.8 ± 4.2	49.8 ± 2.4	44.9 ± 5.1
	Sorbitol 0.8 M	-	-	-	-	-	7.0 ± 6.8
	CaCl2 0.4 M	19.5 ± 4.2	20.4 ± 2.2	-	21.5 ± 7.0	25.0 ± 3.3	18.3 ± 3.7
	NaCl 0.4 M	-	-	-	-	4.0 ± 3.1	1.6 ± 2.3

Values are presented as mean ± standard deviation.

Comparative responses of QoI-resistant and QoI-sensitive isolates to fungicides with different modes of action

The TaqMan allele-specific qPCR successfully distinguished between B. cinerea isolates with the G143A substitution in cytb (QoI-resistant) and those without the substitution (QoI-sensitive). Clear differences in Ct values were observed among the resistant-only, 1:1 mixed, and sensitive-only samples, indicating that the assay reliably detected variations in allele proportions. Furthermore, the relative amplification signals of the FAM (resistant allele) and Cy5 (sensitive allele) probes consistently reflected the expected allele ratios, demonstrating that the quantitative output of the assay can be used to infer the proportion of resistant individuals within a mixed population. Resistant-only samples yielded markedly lower Ct values in the FAM channel and minimal amplification in the Cy5 channel, whereas sensitive-only samples showed the opposite pattern. Mixed samples produced intermediate Ct values between these two extremes, confirming that the assay accurately resolved partial resistance frequencies. When genomic DNA mixtures of known ratios were tested, allele-specific qPCR using both R- and S-allele signals yielded R-allele proportions of 3.78% in the sensitive-only sample, 46.88% in the 1:1 (S:R) mixture, and 99.87% in the resistant-only sample (Supplementary Fig. 3). To further validate the assay across a broader range of allele ratios, defined mixtures containing 0, 25, 50, 75, and 100% resistant-allele DNA were quantified, and the measured resistant-allele proportions showed strong linear agreement with the expected values (R² = 0.95) (Supplementary Fig. 4). These results verified that the method reliably detected shifts in genotype frequencies across the full range of allele ratios.

Experimental concentrations of fungicides were determined using the geometric mean of their EC₅₀ values to ensure comparable inhibition levels (approximately 20–80%) between the two isolates CMML23-BC13 (QoI-sensitive) and CMML23-BC81 (QoI-resistant) (Supplementary Tables 2, 3). Pyraclostrobin-treated samples resulted in consistently higher proportions of the resistant genotype across all application rates (0.5×, 1×, and 2×) (Fig. 8). The samples treated with non-QoI fungicides (pydiflumetofen, fludioxonil, and fluxapyroxad) maintained higher proportions of the QoI sensitive genotype. Pydiflumetofen-treated samples showed the highest proportion of the sensitive genotype across all concentrations (80.11–92.23%), followed by fludioxonil-treated samples (81.82–91.12%). Fluxapyroxad-treated samples exhibited comparatively lower proportion of the sensitive genotype (60.08–80.98%) than other non-QoI fungicides, but the sensitive genotype remained predominant under all treatment conditions. These results indicate that non-QoI fungicides impose relatively weak selection pressure favoring the resistant genotype, in contrast to pyraclostrobin, which strongly promotes the dominance of the resistant allele.

Fig. 8.

Allele-specific qPCR quantification of Botrytis cinerea genotype proportions under fungicide treatments representing different modes of action. Field isolates CMML23-BC13 (QoI-sensitive) and CMML23-BC81 (QoI-resistant) were mixed and exposed to different fungicides (pyraclostrobin, fludioxonil, pydiflumetofen, and fluxapyroxad) at 0.5×, 1×, and 2× rates. The proportions of resistant (R) and sensitive (S) genotypes were quantified using allele-specific qPCR targeting each allele. Gray indicates resistant isolates; Bright green indicates sensitive isolates. QoI, quinone outside inhibitor.

Discussion

This study highlights that QoI fungicides may impose substantial selection pressure on B. cinerea populations, contributing to the persistence of resistance rather than merely initiating its development (Jeger et al., 2008). From 2022 to 2024, the proportion of QoI-resistant isolates in Korean strawberry farms remained consistently high, suggesting that repeated QoI applications may have contributed to maintaining and increasing the proportion of resistant isolates in the field. This pattern reflects the extensive use of QoI fungicides for controlling multiple fungal diseases in strawberry production, such as gray mold, anthracnose, and powdery mildew (Forcelini et al., 2018; Palmer and Holmes, 2021), which continuously exerts selection pressure on resistant populations. Furthermore, resistant populations were geographically widespread across all surveyed regions in Korea, indicating the widespread establishment of resistance alleles in field populations. The consistent presence of the cytb G143A mutation in resistant isolates suggests that this specific mutation has become established and is maintained through continuous selection pressure in agricultural environments (Banno et al., 2009; Leroux et al., 2010). The lack of detectable fitness costs associated with the G143A mutation in this study explains why QoI-resistant populations remain stable across regions and years. This finding indicates that once established, resistant populations are unlikely to decline through natural selection, underscoring the importance of proactive resistance management strategies.

The practical limitations of QoI fungicides are evident once resistant populations establish in production environments (Leroux et al., 2010). QoI fungicides provided poor control when resistant isolates were present, with no measurable efficacy observed across all tested compounds in this study. Although QoI applications successfully controlled sensitive B. cinerea isolates in detached leaf and fruit assays, field experiments resulted in poor disease control due to the presence of a native QoI-resistant population. Isolates recovered from diseased fruits in the QoI-treated plots inoculated with QoI-sensitive isolate consistently carried the cytb G143A mutation, although the origin of this population remains unknown. This native QoI-resistant population therefore acts as a confounding factor, and the limited control in the sensitive-isolate treatment may reflect not only the reduced efficacy of QoI fungicides against the inoculated isolate but also the contribution of this pre-existing resistant population. Furthermore, the consistently low inhibition across pyraclostrobin, mandestrobin, and pyribencarb demonstrates strong cross-resistance within the QoI class in this study due to their shared target-site mutation (Ishii et al., 2009).

The treatment of pyraclostrobin appeared to maintain high frequencies of QoI-resistant isolates in TaqMan qPCR analysis. Previous work on B. cinerea has shown that resistant phenotypes can be less competitive when fungicide pressure is absent but can become dominant under fungicide selection (Chen et al., 2016). In contrast, the non-QoI fungicides, pydiflumetofen, fludioxonil, and fluxapyroxad, which were evaluated, tended to preserve a higher proportion of the QoI-sensitive genotype in this study, suggesting that these fungicides exert weaker selection pressure toward QoI-resistant isolates. The results indicate that these fungicides could help maintain sensitivity within mixed populations and represent viable alternatives for resistance management. However, this assay used a single QoI-sensitive (CMML23-BC13) and a single QoI-resistant (CMML23-BC81) isolate at a 1:1 ratio, so we cannot exclude an effect of the specific isolates used. The individual EC₅₀ values of the two isolates against the four fungicides were therefore compared (Supplementary Table 2). For the three non-QoI fungicides, the resistant isolate (CMML23-BC81) showed EC₅₀ values comparable to or higher than those of the sensitive isolate (CMML23-BC13), suggesting that the maintenance of the sensitive genotype under non-QoI selection is unlikely to result from greater sensitivity of the resistant isolate. Further validation with additional isolate combinations would be needed to generalize these findings.

Regarding population structure, QoI resistance is fundamentally caused by cytb mutations. However, isolates carrying identical G143A mutations exhibited markedly different EC₅₀ values, suggesting the association of both qualitative and quantitative resistance mechanisms. While the cytb mutation confers qualitative resistance through target-site modification, quantitative resistance may arise from non-target-site mechanisms that reduce intracellular fungicide accumulation or enhance detoxification. In B. cinerea, ATP-binding cassette (ABC) transporters such as BcatrB and BcatrD, and major facilitator superfamily transporters such as BcmfsM2, have been implicated in multidrug efflux and are associated with multidrug-resistant phenotypes (Kretschmer et al., 2009). Additionally, xenobiotic detoxification systems, including cytochrome P450 monooxygenases and ABC transporters, may contribute to the metabolic degradation of fungicide active ingredients and to efflux pump activity, further modulating resistance levels independently of target-site mutations (Sang et al., 2018). The interplay between these qualitative and quantitative mechanisms likely accounts for the wide range of EC₅₀ values observed among G143A-carrying isolates in this study. Further investigation into the expression and regulation of these non-target-site resistance determinants is needed to fully explain the phenotypic variation within cytb mutant populations and to inform more precise resistance risk assessments.

In conclusion, continued use of QoI fungicides appears to intensify selection pressure and promote the persistence of resistant and cross-resistant B. cinerea populations. The comparative assay of QoI-sensitive and -resistant genotypes suggests that incorporating non-QoI fungicides such as pydiflumetofen, fludioxonil, and fluxapyroxad may help reduce selection for QoI resistance. If these results are validated under natural field conditions with varying population structures, these fungicides could enhance the durability of disease management programs by reducing reliance on QoI fungicides, thereby delaying resistance development. In addition, incorporating fungicides with novel modes of action, such as ipflufenoquin, represents a critical component of fungicide resistance management strategies against QoI-resistant populations (Choi et al., 2026). These findings provide important guidance for developing fungicide rotation strategies aimed at minimizing selection pressure and preserving the long-term efficacy of available fungicide options in gray mold management.

Electronic Supplementary Material

Supplementary materials are available at The Plant Pathology Journal website (http://www.ppjonline.org/).