Serratia liquefaciens and Bacillus velezensis for Biocontrol of Quinoa Leaf Spot: Bioactive Metabolites Inducing Hyphal Deformation and Plant Growth Promotion

Abstract

Quinoa leaf spot is an important fungal disease that limits quinoa yield and quality. This study aimed to evaluate the pathogenicity of isolates associated with quinoa leaf spot in Qinghai, China, and to screen multifunctional biocontrol bacteria with potential for further development. Pathogenicity assays conducted on healthy quinoa leaves showed that isolates Alternaria alternata AF15 and A. tenuissima AF18 induced typical leaf spot symptoms. The corresponding fungi were successfully re-isolated from the resulting lesions, confirming the pathogenicity of both isolates. Two highly effective biocontrol bacteria, Serratia liquefaciens CB82 and Bacillus velezensis CB316, were subsequently selected through dual-culture assays. Their maximum inhibition rates against the two fungal pathogens reached 56.00% and 57.00%, respectively. Both biocontrol strains exhibited broad adaptability to different temperatures, pH, and NaCl conditions, produced protease, amylase, and cellulase, and showed phosphate-solubilizing activity. Metabolite extraction and fractionation revealed that the antifungal substances were predominantly enriched in the n-butanol fractions, which caused severe shrinkage, surface roughening, breakage, and deformation of the pathogen hyphae. In addition, both strains exhibited strong biofilm-forming capacity and successfully colonized quinoa leaves. Their culturable populations peaked on day 3 after inoculation, reaching 5.58 × 10⁷ and 6.06 × 10⁷ CFU/mL, respectively. In seed germination pouch assays, the bacterial suspensions promoted quinoa root elongation, whereas the fermentation broths increased seedling biomass accumulation in pot experiments. Overall, this study confirmed the pathogenicity of fungal isolates associated with quinoa leaf spot in Qinghai, China, identified two promising biocontrol bacterial strains, and preliminarily characterized their antifungal substances. These findings provide valuable microbial resources and a research basis for the future development of biological control strategies against quinoa leaf spot.

1. Introduction

Quinoa (Chenopodium quinoa Willd.) has a high protein content, a well-balanced amino acid composition, and abundant minerals and bioactive compounds. In recent years, its cultivation has gradually expanded from the traditional production regions of South America to many other parts of the world [1,2]. With the expansion of quinoa production, the adverse effects of diseases on crop yield and quality have received increasing attention. Previous studies have shown that leaf spot and foliar necrosis in quinoa can be caused by diverse microbial pathogens, including fungi belonging to the genera Alternaria, Passalora, and Cercospora, as well as bacteria such as Pseudomonas syringae [3,4,5,6,7]. In Qinghai, China, although quinoa anthracnose caused by Colletotrichum spinaciae has been reported for the first time [8], the pathogens responsible for quinoa leaf spot have not yet been fully characterized, thereby limiting the targeted management of this disease.

Although chemical fungicides remain important tools for controlling fungal plant diseases, their prolonged or inappropriate use may increase the risks of pathogen resistance, pesticide residues, and adverse environmental effects [9]. Therefore, the use of beneficial plant-associated bacteria and their metabolites for the management of foliar diseases has become an important research direction in sustainable agriculture. Bacterial groups investigated for this purpose include Bacillus, Pseudomonas, Paenibacillus, Lysobacter, and Serratia. Different strains of Bacillus and Pseudomonas can inhibit Botrytis cinerea, Alternaria alternata, and other foliar fungal pathogens and reduce the severity of the corresponding diseases [10]. Paenibacillus polymyxa TP3 inhibited B. cinerea and reduced the severity of gray mold in strawberry [11], whereas Lysobacter enzymogenes showed effective control of tobacco powdery mildew [12]. These findings indicate that diverse plant-associated bacteria may serve as candidate biocontrol resources for the management of foliar diseases.

A variety of bacteria with biocontrol or plant growth-promoting potential have been isolated from quinoa seeds, the rhizosphere, and internal plant tissues. Xie et al. screened strains of Bacillus paralicheniformis, B. tequilensis, and B. velezensis from quinoa tissues and evaluated their antagonistic activity against plant pathogens and their plant growth-promoting traits [13]. The quinoa rhizosphere-derived strains B. licheniformis QA1, Enterobacter asburiae QF11, and B. velezensis QA2 exhibited phosphate-solubilizing activity, promoted plant growth, and alleviated salt stress [14,15]. In addition, an endophytic bacterial consortium comprising B. velezensis, B. subtilis, Pseudomonas taiwanensis, and P. putida reduced bacterial leaf spot in quinoa [16]. The quinoa rhizosphere-derived strains B. pumilus CQ5 and B. licheniformis CQ6 also inhibited A. alternata and enhanced tomato resistance to this pathogen [17]. Collectively, quinoa-associated bacteria possess diverse plant growth-promoting and biocontrol functions; however, their roles in the management of quinoa leaf spot caused by Alternaria spp. remain insufficiently understood.

The antagonistic activity of B. velezensis against Alternaria spp. has been widely documented in other crop pathosystems. B. velezensis SYL-3 inhibited A. alternata, the causal agent of tobacco brown spot, and reduced disease severity [18]. Similarly, B. velezensis YXDHD1-7 and HN-Q-8 inhibited A. solani, the causal agent of tomato and potato early blight, respectively, and reduced disease development in the corresponding crops [19,20]. Previous studies have shown that B. velezensis can produce fungal cell wall-degrading enzymes, including β-glucanase, as well as antimicrobial metabolites such as fengycin, surfactin, and bacilysin [21,22]. Its cultures or extracellular metabolites can also induce abnormal changes in pathogen hyphae, including swelling, surface roughening, deformation, and structural damage [21,23]. Furthermore, some B. velezensis strains produce indole-3-acetic acid and exhibit plant growth-promoting activity [24]. These findings highlight the potential of B. velezensis for the management of foliar diseases caused by Alternaria spp. To date, studies of this bacterium in quinoa have mainly focused on pathogen antagonism, salt stress alleviation, and bacterial leaf spot control [13,15,16], whereas its direct activity against quinoa leaf spot caused by Alternaria spp. remains largely unexplored.

Compared with B. velezensis, Serratia liquefaciens has received relatively limited attention as a biological control agent against fungal plant diseases. Previous studies have shown that this bacterium can inhibit phytopathogenic fungi belonging to the genera Alternaria, Fusarium, and Rhizoctonia [25]. For example, S. liquefaciens inhibited A. solani and B. cinerea and reduced the severity of the associated tomato diseases, while the relevant strains also exhibited cellulase activity [25,26]. In addition, some strains of Serratia spp. can produce serratamolide, also known as serrawettin, indicating that members of this genus are capable of synthesizing diverse bioactive metabolites [27]. Certain S. liquefaciens strains have also been reported to promote plant growth, alleviate abiotic stress, and enhance plant disease resistance [28]. These findings demonstrate the potential of S. liquefaciens for plant disease management and growth promotion. However, reports on its activity against quinoa pathogens remain scarce, and its role in the management of quinoa leaf spot caused by Alternaria spp. is still unclear.

Against this background, fungi were isolated from quinoa tissues exhibiting typical leaf spot symptoms in Qinghai, China, and the pathogenicity of the relevant isolates was verified through pathogenicity assays and pathogen re-isolation. Candidate antagonistic bacteria were subsequently screened from bacteria associated with quinoa tissues, and their biocontrol-related characteristics were evaluated in terms of antifungal activity, environmental adaptability, extracellular functional traits, biofilm formation, population dynamics on quinoa leaves, and plant growth-promoting effects. This study aimed to identify the fungal pathogens associated with quinoa leaf spot in selected areas of Qinghai, screen candidate bacteria with biocontrol potential, and preliminarily evaluate their antifungal and plant growth-promoting characteristics, thereby providing an experimental basis for the sustainable management of quinoa leaf spot and subsequent application-oriented research.

2. Materials and Methods

2.1. Experimental Materials

A total of 44 fungal isolates and 212 bacterial isolates were obtained from diseased quinoa tissues collected in Qinghai, China. Among the fungal isolates, 14 were obtained from stem tissues and 30 from leaf tissues, whereas 10 bacterial isolates were obtained from stem tissues and 202 from leaf tissues. The fungal isolates were initially classified according to colony morphology and the morphological characteristics of hyphae, conidiophores, and conidia. The rDNA internal transcribed spacer (ITS) region was subsequently amplified using the universal primers ITS1 (5′-TCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′). The isolates were identified based on sequence comparisons, phylogenetic analyses, and morphological characteristics [3] and were ultimately assigned to nine groups (Table S1).

The bacterial isolates were initially classified according to colony morphology, color, transparency, margin characteristics, elevation, and Gram-staining reactions. The 16S rRNA gene was subsequently amplified using the universal primers 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-ACGGTTACCTTGTTACGACTT-3′), and their taxonomic positions were confirmed through sequence comparisons and phylogenetic analyses [13,29]. The bacterial isolates were ultimately assigned to 13 groups (Table S2).

Molecular identification of the fungal and bacterial isolates was completed during the strain-screening stage. The procedures used for DNA extraction, PCR amplification, sequencing, and phylogenetic analysis have been described in a previous study [29]. Representative isolates were selected based on these identification results and used in subsequent pathogenicity assays, biocontrol strain screening, and functional characterization. The correspondence between the simplified strain designations used in this study and the original complete strain codes, taxonomic information, and GenBank accession numbers is provided in Tables S1 and S2.

2.2. Pathogenicity Testing of Fungal Isolates from Quinoa

The pathogenicity of the fungal isolates was evaluated by inoculating healthy quinoa leaves with spore suspensions, with reference to a previously described method [3]. Healthy quinoa leaves were surface-disinfected with 75% ethanol (Chengdu Kelong Chemical Co., Ltd., Chengdu, China) for 30 s, rinsed with sterile water, and blotted dry. Three wounds were then made in the abaxial epidermis of each leaf using a sterile toothpick.

Purified fungal isolates were inoculated into potato dextrose broth (PDB) and cultured at 25 °C and 150 rpm in a constant-temperature shaker (HYG-A, Jintan Jinda Instrument Manufacturing Co., Ltd., Changzhou, China) for 7 days. The cultures were filtered through sterile gauze, and the spores were collected by centrifugation at 8000 rpm using a refrigerated centrifuge (3-30K, Sigma Laborzentrifugen GmbH, Osterode am Harz, Germany). Spore concentrations were determined using a hemocytometer and adjusted to 1 × 10⁷ spores/mL with sterile water. The resulting spore suspensions were then evenly sprayed onto the wounded leaf surfaces. Control leaves were sprayed with an equal volume of sterile water using the same procedure. The petioles were wrapped with sterile moistened absorbent cotton to maintain moisture and humidity. Each treatment consisted of three independent replicates, with five leaves per replicate.

The inoculated leaves were maintained in an artificial climate chamber (AHRP-1200, Changzhou Aihua Instrument Manufacturing Co., Ltd., Changzhou, China) at 25 ± 1 °C under a 12 h light/12 h dark photoperiod. Disease development was monitored for 7 days after inoculation, and symptom changes were recorded daily. When typical symptoms appeared, tissues were collected from the margins between diseased and healthy areas, and the fungi were re-isolated. The consistency between the re-isolated fungi and the originally inoculated isolates was confirmed by comparing their colony and microscopic morphological characteristics and by aligning their rDNA ITS sequences [3].

2.3. Dual-Culture Screening of Bacterial Isolates Antagonistic to Quinoa Leaf Spot Pathogens

The antagonistic activity of the bacterial isolates was initially and subsequently screened using a dual-culture assay [13,21]. For preliminary screening, a 7 mm mycelial plug was excised from the margin of a 5-day-old pathogen colony and placed at the center of a potato dextrose agar (PDA) plate. Each bacterial isolate was prepared as a suspension at 1 × 10⁸ CFU/mL and spot-inoculated along the plate diagonal at positions 2 cm from the central fungal plug, with 5 μL added to each inoculation point. Plates inoculated only with the pathogen plug served as negative controls. The plates were inverted and incubated in an incubator (MJ-250F-I, Shanghai Yiheng Scientific Instrument Co., Ltd., Shanghai, China) at 28 °C for 5 days. The formation of inhibition zones between the pathogen and bacterial isolates was examined, and isolates exhibiting antagonistic activity were recorded.

The antagonistic isolates obtained from the preliminary screening were quantitatively re-screened using the same procedure. After incubation, the diameter of each pathogen colony was measured in two perpendicular directions using a vernier caliper with a precision of 0.1 mm, and the mean value was calculated. The inhibition rate of each bacterial isolate against the pathogen was calculated using Equation (1). Each bacterial isolate–pathogen combination was evaluated using three independent plates.

$$\mathrm{Inhibition} \mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e} (\%)={\displaystyle \frac{D_{control}-D_{treatment}}{D_{control}}}\times 100$$

(1)

where D_control and D_treatment represent the colony diameters of the control and treatment groups, respectively.

2.4. Preparation of Active Components from Biocontrol Strains

The fermentation broths, cell-free culture filtrates, and bacterial suspensions of the biocontrol strains were prepared according to a previously described method [21]. Activated bacterial strains were inoculated into LB broth and cultured at 30 °C and 180 rpm until the OD₆₀₀, measured using a UV–visible spectrophotometer (T6 New Century, Beijing Purkinje General Instrument Co., Ltd., Beijing, China), reached 1.0–1.2. The resulting cultures were used as seed cultures. Each seed culture was transferred at an inoculation rate of 1% (v/v) into a 250 mL Erlenmeyer flask containing 100 mL of LB broth and incubated under the same conditions for 48 h. The resulting cultures were designated as the bacterial fermentation broths.

A portion of each fermentation broth was filtered twice through a 0.22 μm membrane filter (Tianjin Jinteng Experimental Equipment Co., Ltd., Tianjin, China) to obtain the cell-free culture filtrate, which was stored at 4 °C until use. Another portion was centrifuged at 8000 rpm for 10 min at 4 °C to collect the bacterial cells. The cell pellets were washed twice with sterile phosphate-buffered saline (PBS), resuspended, and adjusted to a viable cell concentration of 1.0 × 10⁷ CFU/mL to prepare the bacterial suspensions. These preparations were used in the subsequent antifungal activity and plant growth-promotion assays.

2.5. Physiological and Biochemical Tests of Biocontrol Strains

The physiological and biochemical characteristics of the biocontrol strains were determined according to Bergey’s Manual of Determinative Bacteriology and previously described methods [13,30]. The tests included starch hydrolysis, gelatin liquefaction, the Voges–Proskauer (V–P) reaction, catalase activity, and the utilization of glucose, mannitol, sucrose, and maltose. Each test was recorded as positive or negative according to the criteria specified for the corresponding medium or detection reagent. All tests were performed in triplicate, and consistent results among the replicates were used as the final determination.

2.6. Determination of Biological Characteristics of Biocontrol Strains

The growth characteristics and environmental adaptability of the biocontrol strains were evaluated by determining their growth curves and growth responses to different NaCl concentrations, initial pH values, and incubation temperatures [21,31].

For growth-curve determination, the activated strains were inoculated into LB broth at 1% (v/v) and incubated at 30 °C and 180 rpm. Samples were collected at 0, 1, 2, 3, 4, 5, 6, 24, 48, 72, 96 and 120 h after inoculation, and the absorbance at 600 nm (OD₆₀₀) was measured. Growth curves were generated from the OD₆₀₀ values obtained at each time point. Three independent cultures were prepared for each strain and treated as three independent biological replicates.

Environmental adaptability was evaluated using a single-factor experimental design. The NaCl-tolerance assay included 13 final NaCl concentrations ranging from 0% to 12% at 1% intervals. The pH-adaptation assay included eight initial pH values of 3, 4, 5, 6, 7, 8, 9, and 10, which were adjusted using a pH meter (AE8601, AZOVTES, Dongguan Fulanke Technology Co., Ltd., Dongguan, China). The temperature-adaptation assay included nine incubation temperatures of 20, 22, 24, 26, 28, 30, 32, 34, and 36 °C. All treatments were inoculated at 1% (v/v). The NaCl and pH treatments were incubated at 30 °C and 180 rpm, whereas the temperature treatments were incubated at the corresponding temperatures and 180 rpm. All other culture conditions were maintained constant, except for the factor being evaluated. After 24 h of incubation, OD₆₀₀ was measured to compare the relative growth of the strains under different conditions and determine their suitable growth ranges. Each strain–treatment combination was evaluated in triplicate.

2.7. Detection of Extracellular Enzyme Activities and Phosphate-Solubilizing Ability of Biocontrol Strains

Cell-free culture filtrates of CB82 and CB316 were prepared as described in Section 2.4. The extracellular enzyme-producing and phosphate-solubilizing abilities of the two strains were evaluated using halo assays [32,33]. Three wells, each 9 mm in diameter, were made in plates containing media for the detection of protease, amylase, cellulase, or chitinase activity or phosphate solubilization. Each well was filled with 100 μL of cell-free culture filtrate, and the plates were incubated at 30 °C for 3 days.

After incubation, the protease and chitinase assay plates were examined directly. The formation of a clear zone around a well indicated the corresponding enzyme activity. The amylase assay plates were stained with Lugol’s iodine solution, and the formation of a yellow clearing zone around a well indicated amylase activity. The cellulase assay plates were stained with Congo red solution for 15 min and then washed with NaCl solution. The formation of a yellow clearing zone around a well indicated cellulase activity. A clearing zone around a well in the phosphate-solubilization medium indicated phosphate-solubilizing activity [32,33].

The total diameter of each clearing zone was measured in two perpendicular directions, and the mean value was calculated. The diameter of the sample well was then subtracted to obtain the net clearing-zone diameter. Three independently prepared batches of culture filtrate were used for each treatment and assayed on separate plates. The measurements obtained from the three wells within each plate were averaged.

2.8. Antifungal Assay of Culture Filtrates of Biocontrol Strains

The antifungal activity of the cell-free culture filtrates against the pathogens was evaluated using an agar well-diffusion assay [21]. A 7 mm mycelial plug was excised from the margin of a 5-day-old pathogen colony and placed at the center of a PDA plate. Wells were made at positions 2 cm from the center of the fungal plug, and each well was filled with 100 μL of cell-free culture filtrate. Filtrates obtained from uninoculated LB medium subjected to the same incubation and filtration procedures were used as controls. Each treatment was performed in triplicate.

After 5 days of incubation, the pathogen colony diameter was measured, and the inhibition rate was calculated using Equation (1). Hyphae were excised from the margin of colonies inhibited by the culture filtrates, mounted on temporary slides, and examined under an optical microscope (ECLIPSE Ci-L plus, Nikon Corporation, Yokohama, Kanagawa, Japan) [34]. Hyphae from pathogen colonies not treated with the culture filtrates were used as morphological controls.

2.9. Determination of Fermentation Broth Extracts of Biocontrol Strains and Their Effects on Pathogen Hyphal Morphology

Based on the sequential liquid–liquid extraction method described by Hur et al. [35], the 30-day incubation period used by Ma et al. [36], and the results of preliminary experiments, 1-day-old cultures of the biocontrol strains were inoculated into LB broth at 1% (v/v) and incubated at 30 °C and 180 rpm for 30 days to prolong the production and accumulation of secondary metabolites [37].

After incubation, the bacterial cells were removed by filtration through a Büchner funnel. The resulting filtrates were sequentially extracted with equal volumes of petroleum ether, chloroform, ethyl acetate, and n-butanol (all from Tianjin Fuyu Fine Chemical Co., Ltd., Tianjin, China) in order of increasing solvent polarity. Each solvent extraction was repeated three times, and the corresponding organic phases were pooled. The remaining aqueous layer was designated as the aqueous fraction. The organic phases were concentrated to dryness under reduced pressure at 40 °C using a rotary evaporator (N-1300, EYELA, Shanghai Ailang Instrument Co., Ltd., Shanghai, China), whereas the aqueous fractions were concentrated to dryness under reduced pressure at 60 °C. The organic extracts were dissolved in chromatographic-grade methanol (Shandong Yuwang Hetianxia New Material Co., Ltd., Yucheng, China) at 25 mg/mL, whereas the aqueous extracts were reconstituted with sterile water to the original filtrate volume. All extracts were filtered through 0.22 μm membrane filters and stored at 4 °C until use.

The antifungal activities of the different fractions against AF15 and AF18 were evaluated using an agar well-diffusion assay according to previously described methods [38]. A 7 mm mycelial plug excised from the margin of a 5-day-old pathogen colony was placed at the center of a PDA plate. Wells were made at positions 2 cm from the center of the fungal plug, and each well was filled with 100 μL of the corresponding extract. Equal volumes of methanol and sterile water were used as controls for the organic and aqueous extracts, respectively. The plates were incubated at 28 °C for 5 days. Colony diameters were measured in two perpendicular directions, and inhibition rates were calculated using Equation (1). Each treatment consisted of three independent plates.

To observe ultrastructural changes in the hyphae following extract treatment, mycelial blocks of approximately 0.5 cm × 0.5 cm were excised from the margins of the inhibition zones and prepared for scanning electron microscopy according to the methods of Abdel-Nasser et al. and Müller et al. [38,39]. The samples were fixed overnight in 2.5% glutaraldehyde at 4 °C, rinsed with PBS, post-fixed with 1% osmium tetroxide for 1–2 h, and rinsed again. The samples were then dehydrated through a graded ethanol series from 30% to 95%, with each concentration applied for 15 min, followed by treatment with 100% ethanol for 20 min. Subsequently, the samples were treated with ethanol/isoamyl acetate (1:1, v/v) for 30 min and then with pure isoamyl acetate for 1 h. After critical-point drying (Quorum Technologies Ltd., Laughton, East Sussex, UK) and sputter coating (Quorum Technologies Ltd., Laughton, East Sussex, UK), hyphal morphology was examined using a scanning electron microscope (Carl Zeiss Microscopy GmbH, Jena, Germany).

2.10. Determination of Biofilm-Forming Ability of Biocontrol Strains

The biofilm-forming capacities of CB82 and CB316 were evaluated using the crystal violet staining method described by Mathur et al., Shao et al., and Fessia et al., with slight modifications [40,41,42]. Activated strains were inoculated into LB broth and cultured for 1 day. The OD₆₀₀ values of the bacterial suspensions were then adjusted to 0.2, 0.4, 0.6, and 0.8. For each concentration, 5 mL of bacterial suspension was added to a sterile glass tube and statically incubated at 30 °C for 2 days. Uninoculated LB broth was used as a blank control.

After incubation, the culture broth was discarded, and the tubes were gently washed with sterile water to remove non-adherent cells. Each tube was filled with 6 mL of 1% crystal violet solution and stained for 20 min. The staining solution was then discarded, and the tubes were thoroughly rinsed. The formation of a blue-violet ring on the tube wall indicated biofilm formation. Subsequently, 95% ethanol was added to dissolve the bound crystal violet, and the absorbance at 570 nm (OD₅₇₀) was measured using the same UV–visible spectrophotometer. Biofilm biomass was evaluated using the blank-corrected OD₅₇₀ value. Three independent culture tubes were used for each strain and initial bacterial suspension concentration.

2.11. Colonization of Biocontrol Strains on Quinoa Plants

The culturable population dynamics of CB82 and CB316 on quinoa leaves were determined according to the methods of Liu et al. and Fu et al. [18,43]. Healthy, fully developed quinoa seeds provided by the Qinghai Academy of Agriculture and Forestry Sciences (Xining, China) were surface-disinfected with 1% NaClO solution for 2 min, thoroughly rinsed with sterile water, and sown in sterilized nutrient soil (Pindstrup Mosebrug A/S, Ryomgaard, Denmark). Plants were cultivated until the five- to six-true-leaf stage. Healthy seedlings of uniform growth were selected, and their leaf surfaces were gently rinsed with sterile water and allowed to air-dry naturally. The fermentation broths of CB82 and CB316 were separately adjusted to 1 × 10⁷ CFU/mL and uniformly sprayed onto the leaf surfaces. Uninoculated control plants were sprayed with an equal volume of sterile water.

Leaves were collected at 1, 2, 3, 4, 5, and 6 days after inoculation. At each sampling time, 0.3 g of leaf tissue was weighed using an electronic balance (JA2003N, Shanghai Yoke Instrument Co., Ltd., Shanghai, China), ground, mixed thoroughly with 3 mL of sterile water, and serially diluted tenfold to 10⁻⁴. An aliquot of 100 μL from the appropriate dilution was spread onto nutrient agar (NA) medium and incubated at 30 °C for 1 day. Colonies were subsequently counted according to their morphological characteristics, and the results were expressed as the number of culturable bacteria in the leaf homogenate (CFU/mL). Three independent leaf samples were collected for each treatment at each sampling time.

2.12. Determination of Plant Growth-Promoting Ability of Biocontrol Strains on Quinoa

The plant growth-promoting potential of CB82 and CB316 was evaluated using a seed germination pouch assay based on previously described methods, with slight modifications [14]. Quinoa seeds were surface-disinfected as described in Section 2.11, thoroughly rinsed with sterile water, and soaked for 5 h. Fifteen seeds were placed in each germination pouch, after which 10 mL of fermentation broth, cell-free culture filtrate, or bacterial suspension prepared as described in Section 2.4 was added. An equal volume of sterile water was used as the control. Root and shoot lengths were measured after 5 days of incubation. Three independent germination pouches were used for each treatment, and the mean value of the seeds within each pouch was treated as one experimental observation.

A greenhouse pot experiment was subsequently conducted to further evaluate the effects of the biocontrol strains on quinoa seedling growth [17]. Surface-disinfected seeds were sown in sterilized nutrient soil, and eight uniformly growing seedlings were retained in each pot after emergence. Each seedling was treated by root drenching with 4 mL of fermentation broth containing 1.0 × 10⁷ CFU/mL viable cells, and the treatment was repeated every 7 days. Plant height and root length were measured 15 days after the first treatment. The whole seedlings were then harvested, washed, and blotted dry before fresh weight determination. The samples were heated at 105 °C for 30 min and subsequently dried at 75 °C to constant weight in a forced-air drying oven (BPG-9240A, Shanghai Yiheng Scientific Instrument Co., Ltd., Shanghai, China) for dry-weight determination [44]. Three independent pots were used for each treatment, and the mean value of the plants within each pot was treated as one experimental observation.

2.13. Statistical Analysis

Except for qualitative observations, all quantitative experiments included three independent biological replicates (n = 3), and the data are presented as the mean ± standard deviation. Technical replicates or subsamples within the same independent experiment were first averaged, and the resulting mean was treated as one independent observation for statistical analysis.

Statistical analyses were performed using IBM SPSS Statistics, version 26.0 (IBM Corp., Armonk, NY, USA). One-way analysis of variance was conducted according to the corresponding experimental design. Before parametric analyses, the normality of the model residuals was assessed using the Shapiro–Wilk test, and homogeneity of variance was evaluated using Levene’s test. When the assumptions for parametric analysis were satisfied, analysis of variance was performed, followed by Duncan’s new multiple range test when the overall effect was significant. Statistical significance was set at p < 0.05. Figures and tables were prepared using Origin 2024 (OriginLab Corporation, Northampton, MA, USA).

3. Results

3.1. Pathogenicity of Fungal Isolates from Quinoa

In preliminary work, nine representative fungal isolates with distinct characteristics were obtained from diseased quinoa tissues collected in Qinghai, China, and used in subsequent experiments. Pathogenicity assays were conducted by inoculating healthy quinoa leaves separately with spore suspensions of the nine representative isolates. As shown in Figure 1b,c, light-brown lesions developed on leaves inoculated with the spore suspensions of AF15 and AF18 and gradually expanded as the disease progressed. These symptoms closely resembled those observed in the original diseased samples collected from the field.

Fungi were subsequently re-isolated from the symptomatic leaves and purified. Isolates with morphological characteristics consistent with those of the originally inoculated fungi were repeatedly recovered. Based on the pathogenicity assay and morphological identification, AF15 and AF18 were identified as Alternaria alternata and A. tenuissima, respectively, and were confirmed as pathogenic isolates associated with quinoa leaf spot in selected areas of Qinghai.

3.2. Screening of Biocontrol Strains Antagonistic to Quinoa Leaf Spot Pathogens

The antifungal activities of 13 bacterial strains previously isolated from quinoa tissues were evaluated. Preliminary screening identified seven strains with inhibitory activity against AF15 and AF18, namely CB82, CB510, CB234, DB105, AB27, CB316, and DB98.

As shown in

Table 1. Inhibition rate of active strains against quinoa leaf spot pathogens.

Strain Number	AF15		AF18
	Colony Diameter/cm	Inhibition Rate/%	Colony Diameter/cm	Inhibition Rate/%
CB82	2.49 ± 0.20 d	43.31	2.75 ± 0.09 c	56.00
CB510	3.24 ± 0.08 b	26.21	3.14 ± 0.23 b	49.80
CB234	2.95 ± 0.10 c	32.76	2.70 ± 0.09 c	56.80
DB105	3.26 ± 0.14 b	25.64	3.27 ± 0.31 b	47.80
AB27	3.43 ± 0.21 b	21.82	3.25 ± 0.10 b	48.08
CB316	2.30 ± 0.08 d	47.58	2.69 ± 0.10 c	57.00
DB98	2.33 ± 0.05 d	47.00	2.84 ± 0.08 c	54.60
CK	4.39 ± 0.30 a	0.00	6.25 ± 0.07 a	0.00

Note: The values in the table are mean ± standard deviation, and the letters represent the significant difference at the p < 0.05 level by Duncan’s new multiple range test.

and Figure 2, further screening using the dual-culture assay showed that CB510, DB105, and AB27 exhibited relatively low overall antifungal activity. Their inhibition rates ranged from 21.82% to 26.21% against AF15 and from 47.80% to 49.80% against AF18. The inhibition rates of the remaining active strains against the two pathogens ranged from 32.76% to 57.00%. Among all tested strains, CB82 and CB316 exhibited the strongest antifungal activity. Their inhibition rates against AF15 were 43.31% and 47.58%, respectively, whereas those against AF18 were 56.00% and 57.00%, respectively.

Based on these results, CB82 and CB316 were selected as candidate biocontrol strains for subsequent experiments.

3.3. Physiological and Biochemical Characteristics of Biocontrol Strains

To characterize the physiological and metabolic properties of the biocontrol strains CB82 and CB316, their physiological and biochemical traits were examined, and the results are presented in

Table 2. Physiological and biochemical characteristics of strain CB82 and strain CB316.

Test Item	CB82	CB316
Starch hydrolysis	+	+
Gelatin liquefaction	+	+
Voges–Proskauer test (V–P test)	+	+
Glucose utilization	+	+
Mannitol utilization	+	+
Sucrose utilization	+	+
Maltose utilization	+	+
Catalase test	+	+

Note: +, positive.

. Both strains tested positive for gelatin hydrolysis, the Voges–Proskauer reaction, and catalase activity. In addition, both strains utilized glucose, mannitol, sucrose, and maltose and were capable of hydrolyzing starch.

3.4. Biological Characteristics of Biocontrol Strains

To characterize the biological properties of the biocontrol strains CB82 and CB316, their growth curves and responses to different environmental factors were determined. The two strains exhibited similar growth patterns (Figure 3a). Growth was relatively slow during the first 4 h, followed by a logarithmic growth phase from 4 to 24 h. After 24 h, both strains entered the decline phase.

Both strains were able to grow at NaCl concentrations ranging from 0% to 12% (Figure 3b). The optimal NaCl concentration ranges for CB82 and CB316 were 6–8% and 6–7%, respectively. The growth of CB82 was markedly inhibited when the NaCl concentration exceeded 9%, whereas that of CB316 was markedly inhibited at NaCl concentrations above 8%.

Both strains also grew over a pH range of 3–10 (Figure 3c). CB82 showed favorable growth at pH 5–7, whereas its growth was markedly inhibited at pH 3 and under alkaline conditions above pH 7. In contrast, CB316 grew favorably at pH 7–9 but was markedly inhibited at pH 3 and pH 10.

The two strains differed in their responses to temperature (Figure 3d). The optimal growth temperatures of CB82 and CB316 were 34 °C and 28 °C, respectively. The growth of CB316 decreased significantly when the temperature exceeded 28 °C.

3.5. Extracellular Enzyme Activities and Phosphate-Solubilizing Ability of Biocontrol Strains CB82 and CB316

The extracellular hydrolytic enzyme activities and phosphate-solubilizing abilities of CB82 and CB316 were evaluated. Both strains produced protease, amylase, and cellulase and exhibited phosphate-solubilizing activity, whereas no detectable chitinase activity was observed (Figure 4).

As shown in

Table 3. Determination of enzyme production and phosphate solubilization ability of strain CB82 and CB316.

Strain	Net Clearing-Zone Diameter (cm)
	Chitinase	Protease	Amylase	Cellulase	Phosphate-Solubilizing Ability
CB82	-	0.74 ± 0.08	2.18 ± 0.12	0.70 ± 0.06	2.10 ± 0.09
CB316	-	2.50 ± 0.01	2.34 ± 0.09	2.30 ± 0.12	2.16 ± 0.11

Note: Values are presented as mean ± standard deviation (n = 3). -, no detectable activity.

, CB82 exhibited larger net halo diameters in the amylase and phosphate-solubilization assays than in the protease and cellulase assays, with values of 2.18 and 2.10 cm, respectively. CB316 showed strong protease, amylase, and cellulase activities, as well as phosphate-solubilizing ability, with net halo diameters of 2.50, 2.34, 2.30, and 2.16 cm, respectively.

3.6. Antifungal Activity of Culture Filtrates of Biocontrol Strains Against Pathogens

Preliminary experiments compared the antifungal activities of the fermentation broth, bacterial suspension, and cell-free culture filtrate of each biocontrol strain. Among the three preparations, the cell-free culture filtrates exhibited more consistent and pronounced inhibitory activity. Therefore, they were selected as the primary test materials for the subsequent evaluation of antifungal activity.

The inhibitory effects of the cell-free culture filtrates of CB82 and CB316 against AF15 and AF18 are shown in

Table 4. Inhibitory effect of culture filtrate of strain CB82 and strain CB316 on pathogen AF15 and AF18.

Activity Strains	Inhibition Rate/%
	AF15	AF18
CB82	29.71 ± 0.02 b	42.27 ± 0.00 a
CB316	34.71 ± 0.01 a	36.46 ± 0.00 b

Note: The values in the table are mean ± standard deviation, and the letters represent the significant difference at the p < 0.05 level by Duncan’s new multiple range test.

and Figure 5. The inhibition rates of the CB82 and CB316 culture filtrates against AF15 were 29.71% and 34.71%, respectively (Figure 5a2,a3). Their corresponding inhibition rates against AF18 were 42.27% and 36.46%, respectively (Figure 5b2,b3).

3.7. Effects of Culture Filtrates of Active Strains on Pathogen Hyphal Morphology

Hyphae from the colony margins of the pathogens after inhibition treatment were collected and observed microscopically, and the results are shown in Figure 6. For Alternaria alternata AF15, the hyphae in the control group were uniform in thickness, structurally intact, transparent, and full (Figure 6a1). After treatment with the culture filtrate of CB82, the hyphae became swollen and widened, with lighter coloration, accompanied by the formation of a small number of sclerotium-like and vesicle-like structures (Figure 6a2). After treatment with the culture filtrate of CB316, the hyphae became markedly darker, and a large number of sclerotium-like and vesicle-like structures were formed. In addition, hyphal transparency was severely reduced, and morphological deformation was more pronounced (Figure 6a3).

For A. tenuissima AF18, the hyphae in the control group were slender, septate, uniformly colored, and structurally intact (Figure 6b1). After treatment with the culture filtrate of CB82, the hyphae became darker and showed bending and twisting, accompanied by the formation of a small number of sclerotium-like and vesicle-like structures (Figure 6b2). After treatment with the culture filtrate of CB316, the hyphae became further darkened, hyphal transparency decreased significantly, and obvious morphological deformation was observed (Figure 6b3).

3.8. Antifungal Activity of Fermentation Broth Extracts from Biocontrol Strains CB82 and CB316

The above results showed that the cell-free culture filtrates of the two biocontrol strains exhibited significant antagonistic activity against the pathogens causing quinoa leaf spot disease and induced morphological deformation of pathogen hyphae, suggesting the presence of antifungal active substances in their metabolites. To further analyze the polarity characteristics of these active substances, enrich the effective components, and provide a basis for subsequent isolation and purification, the fermentation broths of CB82 and CB316 were extracted using organic solvents with different polarities. The antifungal activities of the resulting extract fractions against the two quinoa leaf spot pathogens were then determined.

3.8.1. Antifungal Activity of Extracts Against Quinoa Leaf Spot Pathogens

As shown in

Table 5. Antifungal activity of fermentation broth extracts of CB82 and CB316 against AF15 and AF18.

Strain Number	Extract Solution	Inhibition Rate/%
		AF15	AF18
CB82	Petroleum ether	17.58 ± 0.03 a	9.71 ± 0.11 b
	Chloroform	5.15 ± 0.02 c	15.43 ± 0.08 b
	Ethyl acetate	9.09 ± 0.02 c	20.57 ± 0.16 a
	n-Butanol	20.3 ± 0.03 a	25.71 ± 0.12 a
	Aqueous phase	16.36 ± 0.10 ab	13.9 ± 0.11 b
CB316	Petroleum ether	4.24 ± 0.05 d	12.00 ± 0.02 c
	Chloroform	10.00 ± 0.03 c	14.29 ± 0.02 c
	Ethyl acetate	15.76 ± 0.03 b	19.14 ± 0.03 b
	n-Butanol	21.21 ± 0.02 a	27.43 ± 0.02 a
	Aqueous phase	5.76 ± 0.02 cd	7.43 ± 0.02 d

Note: The values in the table are mean ± standard deviation, and the letters represent the significant difference at the p < 0.05 level by Duncan’s new multiple range test.

, the n-butanol, petroleum ether, and aqueous extracts of biocontrol strain CB82 inhibited the growth of pathogen AF15 by 20.30%, 17.58%, and 16.36%, respectively, showing significant differences from the inhibition rates of the ethyl acetate and chloroform extracts. For pathogen AF18, the inhibition rates of the n-butanol and ethyl acetate extracts of strain CB82 were 25.71% and 20.57%, respectively, which were significantly higher than those of the petroleum ether, chloroform, and aqueous extracts. Among all extracts of strain CB82, the n-butanol extract showed the strongest inhibitory activity against both AF15 and AF18. In addition, the petroleum ether and aqueous extracts of strain CB82 showed stronger inhibitory effects against AF15 than against AF18, whereas the ethyl acetate, chloroform, and n-butanol extracts exhibited stronger inhibitory effects against AF18 than against AF15.

For biocontrol strain CB316, the inhibition rates of the n-butanol and ethyl acetate extracts against pathogen AF15 were 21.21% and 15.76%, respectively. Against pathogen AF18, the inhibition rates of the n-butanol and ethyl acetate extracts were 27.43% and 19.14%, respectively. The n-butanol and ethyl acetate extracts of CB316 showed significantly higher inhibitory activity against both AF15 and AF18 than the petroleum ether, chloroform, and aqueous extracts. Moreover, all organic solvent and aqueous extracts of strain CB316 showed stronger inhibitory effects against AF18 than against AF15.

3.8.2. Observation of the Effects of Extracts on Hyphal Morphology of Quinoa Leaf Spot Pathogens

As shown in Figure 7 and Figure 8, the hyphae of pathogens AF15 and AF18 showed varying degrees of morphological changes after treatment with extracts from active strains CB82 and CB316. These changes included hyphal shrinkage and flattening, uneven surfaces, breakage and damage, and irregular hyphal thickness.

As shown in Figure 7, the hyphae of the control group of pathogen AF15 were evenly distributed, with smooth surfaces and slight depressions (Figure 7a). Compared with the control, hyphae treated with the petroleum ether extract of the fermentation broth of strain CB82 showed obvious surface depressions and damage, while the hyphal thickness remained relatively uniform (Figure 7b1). Hyphae treated with the chloroform extract were uniform in thickness, with small particles on the surface and deeper depressions (Figure 7b2). Hyphae treated with the ethyl acetate extract showed uneven thickness, slight surface protrusions, and partial flattening (Figure 7b3). Hyphae treated with the n-butanol extract were flattened, irregularly distributed, and extremely rough on the surface, with obvious damage (Figure 7b4). Hyphae treated with the aqueous extract showed large differences in thickness, became flattened and widened, and had obvious protrusions on some hyphae (Figure 7b5).

After treatment with the petroleum ether extract of the fermentation broth of strain CB316, the hyphae showed obvious flattening and bending, with fine particles distributed on the surface (Figure 7c1). Hyphae treated with the chloroform extract displayed a granular appearance, and surface protrusions were observed (Figure 7c2). Hyphae treated with the ethyl acetate extract were relatively uniform in thickness but showed obvious shrinkage, with uneven surfaces (Figure 7c3). Hyphae treated with the n-butanol extract were irregularly distributed and broken, with obvious surface damage (Figure 7c4). Hyphae treated with the aqueous extract showed marked differences in thickness, obvious depressions, and relatively rough surfaces (Figure 7c5).

As shown in Figure 8, the hyphae of the control group of pathogen AF18 were full and relatively uniform in thickness (Figure 8a). Compared with the control, hyphae treated with the petroleum ether extract of the fermentation broth of strain CB82 showed uneven thickness and became flattened (Figure 8b1). Hyphae treated with the chloroform extract were mostly flattened and bent, with widening and surface protrusions (Figure 8b2). Hyphae treated with the ethyl acetate extract showed multiple areas of surface damage and slight shrinkage (Figure 8b3). Hyphae treated with the n-butanol extract displayed extremely rough surfaces, obvious widening, and severe damage (Figure 8b4). Hyphae treated with the aqueous extract became widened and showed surface depressions (Figure 8b5).

After treatment with the petroleum ether extract of the fermentation broth of strain CB316, the hyphae exhibited dense granular protrusions on the surface and became flattened and thickened (Figure 8c1). Hyphae treated with the chloroform extract showed rough surfaces; some hyphae became flattened and thickened, whereas others showed swelling and thickening (Figure 8c2). Hyphae treated with the ethyl acetate extract showed uneven thickness and varying degrees of shrinkage and depression (Figure 8c3). Hyphae treated with the n-butanol extract were closely arranged, with rough surfaces, different degrees of damage, and protrusions (Figure 8c4). Hyphae treated with the aqueous extract showed shrinkage and uneven surfaces (Figure 8c5).

Sequential extraction with solvents of different polarities showed that the antifungal active substances of the two biocontrol strains were mainly enriched in the n-butanol fraction, followed by the ethyl acetate fraction. The antifungal activities of these fractions were significantly higher than those of the petroleum ether, chloroform, and aqueous fractions. The n-butanol fraction was the most active fraction for both strains, with inhibition rates of 20.30–25.71% for CB82 and 21.21–27.43% for CB316 against the two pathogens.

In terms of activity differences, all extract fractions of CB316 showed stronger inhibitory effects against AF18 than against AF15. For CB82, the low-polarity fractions and aqueous fraction showed stronger activity against AF15, whereas the medium- to high-polarity fractions exhibited stronger activity against AF18.

3.9. Biofilm-Forming Ability of Biocontrol Strains CB82 and CB316

The biofilm-forming ability of the strains was evaluated using the crystal violet staining method, and the results are shown in Figure 9. No purple ring was observed in the control glass tubes, indicating that no biofilm was formed. In contrast, purple rings were observed in the tubes treated with different initial concentrations of strains CB82 and CB316 (Figure 9a,b), confirming that both strains were capable of forming biofilms.

The quantitative results (Figure 9c) showed that the absorbance values of strain CB82 at 570 nm were 0.57 and 0.44 when the initial OD₆₀₀ values were 0.2 and 0.6, respectively. Under the same initial concentrations, the absorbance values of strain CB316 were both 1.80, indicating that CB316 had a stronger biofilm-forming ability under the tested conditions.

3.10. Colonization of Biocontrol Strains CB82 and CB316 on Quinoa Plants

To monitor the population dynamics of the tested strains on quinoa leaves, samples were collected at regular intervals and analyzed using the serial dilution plating method. Colonies of the target strains were preliminarily identified and counted according to their characteristic colony morphologies. Leaf homogenates from the uninoculated controls contained 2.0 × 10⁴–1.6 × 10⁵ CFU/mL of culturable bacteria with colony morphologies similar to those of the target strains. These populations were approximately two orders of magnitude lower than those detected in the inoculated treatments, indicating that interference from morphologically similar background bacteria was relatively limited.

As shown in Figure 10, the culturable populations of strains CB82 and CB316 peaked on day 3 after inoculation, reaching 5.58 × 10⁷ and 6.06 × 10⁷ CFU/mL, respectively. Thereafter, the culturable populations of both strains gradually declined over time.

3.11. Plant Growth-Promoting Ability of Biocontrol Strains CB82 and CB316 on Quinoa

3.11.1. Effects of Biocontrol Strains CB82 and CB316 on Quinoa Seed Germination

To clarify the key components responsible for the plant growth-promoting activity of the strains, a seed germination pouch assay was conducted to evaluate the effects of different treatments, including bacterial suspension, fermentation broth, and cell-free fermentation filtrate, on quinoa seed germination. As shown in

Table 6. Growth parameters of quinoa seeds treated with strain CB82 and strain CB316.

Treatment	Root Length/cm	Shoot Length/cm
CK (sterile water)	6.33 ± 1.93 b	0.73 ± 0.13 bc
Fermentation broth of strain CB82	1.04 ± 0.41 e	0.72 ± 0.09 bc
Cell-free fermentation broth of strain CB82	2.61 ± 0.97 c	0.81 ± 0.12 a
Bacterial suspension of strain CB82	7.67 ± 2.23 a	0.73 ± 0.14 bc
Fermentation broth of strain CB316	0.80 ± 0.23 e	0.75 ± 0.12 b
Cell-free fermentation broth of strain CB316	1.75 ± 0.62 d	0.71 ± 0.12 bc
Bacterial suspension of strain CB316	7.87 ± 2.18 a	0.70 ± 0.12 bc

Note: Values are means ± standard deviation (n = 3 independent germination pouches). Different lowercase letters within the same column indicate significant differences among treatments at p < 0.05 according to Duncan’s new multiple range test.

, compared with the CK treatment, the bacterial suspensions of strains CB82 and CB316 markedly promoted quinoa root growth. The root lengths of quinoa seedlings after treatment were 7.67 and 7.87 cm, respectively, which were significantly higher than those in the control and other treatment groups.

After treatment with the cell-free fermentation filtrate of strain CB82, the shoot length of quinoa seedlings reached 0.81 cm, which was significantly higher than that in the control and other treatment groups. No significant differences in shoot length were observed among the other treatments.

3.11.2. Effects of Fermentation Broths of Biocontrol Strains CB82 and CB316 on Quinoa Seedling Growth

To further evaluate the practical application potential of the strains at the seedling stage, quinoa seedlings were treated with bacterial fermentation broths under conditions more representative of field application, and their overall plant growth-promoting effects were assessed. Compared with the CK treatment, the fermentation broth of strain CB82 increased all measured growth parameters of quinoa seedlings, with increases of 1.27 cm in plant height, 0.38 cm in root length, 0.58 g in whole-plant fresh weight, and 0.08 g in whole-plant dry weight. Treatment with the fermentation broth of strain CB316 also resulted in higher plant height, whole-plant fresh weight, and whole-plant dry weight than the CK treatment (

Table 7. Growth parameters of quinoa seedlings treated with strain CB82 and strain CB316.

Treatment	Plant Height/cm	Root Length/cm	Whole Plant Fresh Weight/g	Whole Plant Dry Weight/g
CK (sterile water control)	5.30 ± 0.78 b	4.58 ± 1.26 a	0.79 ± 0.06 b	0.09 ± 0.01 c
Fermentation broth of strain CB82	6.57 ± 0.90 a	4.96 ± 1.41 a	1.37 ± 0.11 a	0.17 ± 0.02 a
Fermentation broth of strain CB316	5.77 ± 0.80 b	4.47 ± 0.93 a	1.28 ± 0.10 a	0.13 ± 0.01 b

Note: Values are means ± standard deviation (n = 3 independent pots). Different lowercase letters within the same column indicate significant differences among treatments at p < 0.05 according to Duncan’s new multiple range test.

).

4. Discussion

Two pathogenic fungal isolates, Alternaria alternata AF15 and A. tenuissima AF18, were isolated and identified from quinoa plants exhibiting typical leaf spot symptoms in Qinghai, China. Pathogenicity assays and pathogen re-isolation confirmed that both fungi induced typical leaf spot symptoms on quinoa leaves. These findings are consistent with reports of quinoa leaf spot caused by Alternaria spp. in other production regions [3,4]. For example, A. alternata and A. scrophulariae have been detected in quinoa affected by brown spot disease in Colombia [3]. These results indicate that Alternaria spp. constitute an important group of pathogens associated with foliar diseases of quinoa, although the predominant species may vary among production regions.

Two biocontrol bacterial strains, Serratia liquefaciens CB82 and Bacillus velezensis CB316, were selected against these pathogens. The two strains exhibited distinct adaptive responses to temperature, pH, and salinity. CB82 was better adapted to relatively high temperatures and acidic to neutral conditions and tolerated a comparatively broad salinity range, whereas CB316 grew more favorably under neutral to slightly alkaline conditions. Although these in vitro growth characteristics cannot be directly extrapolated to bacterial performance on quinoa leaves, the observed differences suggest that the two strains may maintain their survival and biocontrol functions under different environmental conditions. Functional differences among members of multi-strain biocontrol systems may contribute to improved environmental adaptability and functional stability [45,46].

In addition to their environmental adaptability, CB82 and CB316 differed in several biocontrol-related functional traits. Both strains inhibited AF15 and AF18 in dual-culture and cell-free culture filtrate assays, indicating that extracellularly secreted products contributed to pathogen suppression. Both strains also produced protease, amylase, and cellulase, with CB316 exhibiting stronger protease and cellulase activities. Proteases and cellulases may contribute to antagonism by affecting pathogen surface structures, extracellular polymers, or nutrient utilization processes [22]. However, halo assays reflect only the enzyme-producing capacity of a strain and cannot determine the actual contribution of individual enzymes to antifungal activity. The phosphate-solubilizing capacity of both strains also suggests potential roles in nutrient transformation and plant growth promotion [15]. Nevertheless, measurements of soil-available phosphorus and plant phosphorus uptake are required to verify this relationship.

To characterize the distribution of extracellular antifungal substances produced by the two strains, cell-free culture filtrates and fermentation broth fractions obtained using solvents of different polarities were further evaluated. Solvent fractionation showed that the antifungal activity of both strains was predominantly enriched in the n-butanol fractions, followed by the ethyl acetate fractions, suggesting that the major recoverable active substances may be moderately polar or amphiphilic compounds. CB82 exhibited a relatively broad distribution of antifungal activity, as its petroleum ether and aqueous fractions also inhibited AF15, suggesting that this strain may produce multiple active substances with different physicochemical properties. Based on previous studies, its n-butanol fraction may contain serrawettin W2-like amphiphilic compounds or other moderately polar antimicrobial metabolites [47]. In contrast, the activity of CB316 was mainly concentrated in the n-butanol and ethyl acetate fractions and may be associated with cyclic lipopeptides, such as fengycin, iturin, and surfactin, or other antifungal peptides [48]. The fractions derived from CB316 generally exhibited stronger inhibitory activity against AF18 than against AF15, further suggesting that different Alternaria pathogens may vary in their sensitivity to bacterial metabolites. It should be emphasized that these candidate compounds were inferred solely from solvent partitioning patterns and previous reports and have not yet been chemically identified. Rafiq et al. used bioassay-guided solvent fractionation to screen highly active antifungal fractions and subsequently analyzed their composition using gas chromatography–mass spectrometry (GC–MS), demonstrating that the integration of activity-guided fractionation with chromatographic and mass spectrometric analyses can facilitate the localization and identification of potential antifungal constituents [49]. Accordingly, future studies should employ activity-guided chromatographic separation and liquid chromatography–tandem mass spectrometry (LC–MS/MS) to characterize nonvolatile active substances, together with GC–MS to detect potential volatile or semivolatile compounds, thereby identifying the major antifungal constituents produced by the two strains and clarifying the differences between them.

Optical microscopy and scanning electron microscopy showed that treatment with the culture filtrates and solvent-extracted fractions caused marked abnormalities in the hyphae of AF15 and AF18, including swelling, bending, shriveling, flattening, surface roughening, and rupture. These changes indicate that extracellular products from the two bacterial strains interfered with normal hyphal development and caused structural damage. Similarly, Baptista et al. observed that cell-free culture supernatants of B. velezensis not only inhibited the growth of phytopathogenic fungi but also induced surface protrusions, deformation, and structural damage in their hyphae, suggesting that hyphal morphological damage can serve as important phenotypic evidence of antifungal activity [21]. When considered together with the activity of the culture filtrates, extracellular enzyme production, and the enrichment of antifungal activity in the n-butanol fractions, these findings suggest that hydrolytic enzymes and lipopeptides or other moderately polar antimicrobial metabolites may jointly contribute to pathogen suppression [50,51]. Hydrolytic enzymes may weaken hyphal structures by acting on fungal cell wall-associated components [52], whereas lipopeptides and related metabolites may interfere with membrane-associated processes and normal cellular metabolism [53]. However, hyphal abnormalities provide only morphological evidence of antifungal activity and cannot be used to identify specific active substances, molecular targets, or precise mechanisms of action. Further studies combining enzyme inhibition assays with measurements of cell membrane permeability and intracellular material leakage are therefore required to clarify the relationship between the active substances and hyphal damage.

In addition to direct antagonism, biofilm formation and the maintenance of bacterial populations on plant surfaces are important factors affecting the persistence of biocontrol activity [54,55]. Both CB82 and CB316 formed biofilms, although CB316 exhibited a stronger biofilm-forming capacity under the tested conditions. Leaf recovery assays showed that the culturable populations of both strains on quinoa leaves peaked on day 3 after inoculation and subsequently declined. Previous studies have demonstrated close relationships among surfactin production, biofilm formation, strain persistence, and subsequent protective effects on plant surfaces [56,57]. Biofilm formation may therefore facilitate bacterial establishment by enhancing attachment to plant surfaces and supporting the maintenance of local populations, thereby providing favorable conditions for sustained biocontrol activity [58,59]. It should be noted that the leaf assay lasted only 6 days and that the recovered colonies were presumptively identified mainly according to colony morphology. Consequently, a small number of naturally occurring background colonies may have led to an overestimation of the absolute populations of the inoculated strains. Nevertheless, the numbers of morphologically similar colonies in the uninoculated controls were substantially lower than those in the inoculated treatments, and the same culture and identification criteria were applied across all treatments and sampling times. Therefore, the present results still provide a reasonable representation of the short-term population dynamics of the two strains on quinoa leaves. Future studies should employ strain-specific quantitative PCR, genetic markers, or molecular fingerprinting techniques for more accurate quantification and long-term monitoring and to further clarify the relationships among biofilm formation, survival on plant surfaces, and sustained disease suppression.

Overall, this study identified two fungal pathogens associated with quinoa leaf spot in Qinghai and systematically evaluated the biocontrol-related characteristics of CB82 and CB316. Both strains inhibited AF15 and AF18, but they differed in their biocontrol traits. CB316 exhibited stronger activities of certain extracellular enzymes and a greater biofilm-forming capacity, whereas CB82 showed distinct environmental adaptation characteristics and a broader distribution of active fractions. These differences suggest potential functional complementarity between the two strains and provide a rationale for further evaluating their combined use in a multi-strain formulation. However, most experiments were conducted under controlled laboratory conditions, and the results cannot fully represent the performance of the strains in complex natural environments or directly predict their efficacy under greenhouse or field conditions. Further studies should identify the active compounds, characterize colonization-related functions, and evaluate functionally complementary multi-strain formulations through greenhouse and field trials, with particular emphasis on disease control efficacy, functional stability, application methods, and environmental safety [60,61]. Such studies would provide a basis for the sustainable management of quinoa leaf spot and the large-scale production and commercial development of quinoa-specific microbial formulations.

5. Conclusions

Pathogenicity tests demonstrated that Alternaria alternata AF15 and A. tenuissima AF18 were confirmed as pathogenic isolates associated with quinoa leaf spot in selected areas of Qinghai. Two biocontrol strains, Serratia liquefaciens CB82 and Bacillus velezensis CB316, were successfully screened. These two strains inhibited pathogen growth, possibly through the secretion of extracellular hydrolytic enzymes and the production of active metabolites enriched in medium-polarity fractions. They also promoted quinoa root elongation in the seed germination pouch assay and enhanced seedling growth in the pot experiment. In addition, both strains showed broad adaptability to salt and pH stress, as well as strong biofilm-forming ability and efficient colonization on quinoa plants. Overall, S. liquefaciens CB82 and B. velezensis CB316 exhibited promising application potential for the green control of quinoa diseases.