The C. rosea strain CMAA1284 (GenBank accession MG489966) used in these studies was isolated from rose crops in Viçosa, Minas Gerais State, Brazil; the S. sclerotiorum strain CMAA1105 (GenBank accession OM348513) used was isolated from soybean. All fungi used in this research were deposited at Embrapa Environment Collection of Microorganisms of Agricultural and Environmental Importance (CMAA). The fungi were preserved by means of fragments, 5 mm in diameter, of fully-grown colonies, added in cryovials filled with 1.5 ml of sterile 20% glycerol solution (v/v), and stored at −40°C. From the cryotubes, the microorganisms were routinely transferred to Petri dishes (90 × 15 mm, Pleion^®) that contained 20 ml of potato–dextrose–agar (PDA, Difco^®) medium and incubated in growth chamber for 14 days at 25 ± 1°C and with 12:12 h photoperiod. The fungal strains have been registered under the Brazilian genetic heritage – Sisgen – protocol A00AFAF.

To study the biphasic fermentation of C. rosea, 300 g of parboiled rice were moistened with 600 ml of distilled water for 1 h. Then, the excess water was removed and 15 g of the wet rice (initial moisture content of 39% w/w) were distributed in each Erlenmeyer flask (125 ml, baffled type, Exom^®, São Paulo, SP, Brazil), sealed with hydrophobic cotton plugs and then covered by aluminum foil for autoclaving for 20 min at 121°C. After autoclaving, the aluminum foil was removed to allow the gas exchange during the fermentation process. This trial included two treatments, whose objective was to investigate the influence of two types of inoculum on the conidial production by a solid-state fermentation using the parboiled rice grains. Treatment 1 consisted of the suspension containing submerged conidia produced by liquid culture (hereafter referred to as liquid pre-culture), and treatment 2 was assigned to the suspension of aerial conidia obtained from sporulated cultures grown on PDA for 14 days and then suspended in 10 ml of sterile 0.04% polysorbate solution (Tween 80^®), in order to obtain a inoculum adjusted to 1 × 10⁷ conidia/ml for further inoculation of 90 ml of liquid medium with 10 ml (10% v/v) of this spore suspension. Briefly, the pre-culture of C. rosea was obtained with a liquid medium consisting of dextrose monohydrate and yeast extract with a 50:1 C:N ratio and 36 g/L of carbon content (total volume of 100 ml in 250-ml baffled Erlenmeyer flask) grown for 4 days (the pre-culture is described in Table 1). These pre-cultures were quantified based solely on submerged conidia with barely hyphal fragments in the final inoculum concentration. The flasks that contained 15 g of sterile moistened rice mass were inoculated with 1.5 ml (10% inoculum volume) of each suspension of C. rosea adjusted to 5 × 10⁶ aerial or submerged conidia/ml to deliver a final concentration of 5 × 10⁵ aerial or submerged conidia/g rice (on wet weight basis). The flasks were vigorously hand-shaken for 1 min to homogenize the inoculum throughout the rice mass. The inoculated rice cultures were kept static (without agitation) in a growth chamber set to 25 ± 1°C with a 12:12 h photoperiod for 7 days. After this period, fully sporulated fungus-colonized rice mass was suspended in 50 ml of surfactant solution (0.05% Break Thru^® MSO, Evonik^®, Essen, Germany), placed in a rotary incubator shaker with diameter orbit (or throw) of 28 mm (SL-223–5, Solab^®, Piracicaba, SP, Brazil) and agitation speed of 248 rpm, with constant temperature at 28 ± 1°C for 30 min. The flasks were subsequently taken to ultrasound water bath for 5 min to promote the detachment of conidia from the substrate (see the protocol details in Supplementary Figure S1). After these steps, a serial dilution was performed and the conidia were counted in a Neubauer (hemocytometer) chamber under a light microscope (Leica^® MD250, Germany) with 400 × magnification. The results were expressed in conidia/g-dry-matter (gDM) of rice previously considering the initial rice moisture (average of 39% w/w). The experiment was carried out in a completely randomized design (CRD) involving two treatments with six replicates each, obtaining n = 12 replicates for each treatment, since the entire experiment was repeated twice on different occasions using a new fungal inoculum.

In this study, the protein sources studied in the nutritional composition of the culture medium were as follows: (1) Yeast extract (10.4% nitrogen and 40% carbon; Acumedia/NEOGEN^®, Lansing, MI, United States), (2)- cottonseed flour (9.4% nitrogen and 40% carbon; Pharmamedia^®, ADM Co., Decatur, IL, United States), (3) corn bran (9.3% nitrogen and 40% carbon; Protenose^®, Ingredion, Mogi Guaçu, SP, Brazil), and (4) soybean meal (8.5% nitrogen and 40% carbon; Baker’s soyflour^®, ADM Co., Decatur, IL, United States). All sources initially contained 36-g carbon/L, had initial pH value of 6.0 and a C:N ratio of 10:1 for the MS formation and a preferred C:N ratio of 50:1 for the conidia formation. The 14-day-old sporulated PDA cultures of C. rosea were used to prepare liquid pre-cultures as described in the previous section. A sterile stock solution of 25% (w/v) dextrose monohydrate (35% carbon; Cerelose^®, Ingredion^®, Mogi Guaçu, SP, Brazil) was used in all fermentation studies as the main carbon source (Table 1). The experiment was carried out in a CRD including eight treatments with three replicate flasks each, and repeated three times on different dates.

All liquid media (100 ml per shake flask) tested in this fermentation study were inoculated with a standard spore suspension (10% inoculum volume) obtained from the liquid pre-culture. Briefly, the inoculation procedure applied to all liquid media was performed by adding 10 ml of the suspension (pre-culture) that contained 1 × 10⁷ conidia/ml into baffled Erlenmeyer flasks (250 mL) with stainless steel caps and filled with 90 ml of medium. These liquid cultures were incubated in an orbital rotary shaker (248 rpm) at 28 ± 1°C and 12:12 h photoperiod for 7 days. The counting of spores and MS were carried out from the second to the seventh day of the fermentation. To determine the concentration of the submerged conidia, a 1-ml aliquot of liquid culture was removed from the flasks and successively diluted for the counting in a Neubauer chamber. The quantification of MS was performed on a glass slide covered with a cover slip (24 × 50 mm). The fermented broth underwent a 10^–1 dilution and 100 μL was added directly to the slide. Only MS with at least a diameter greater than 50 μm were computed, while the submerged conidia ranged in size from 3.86 to 8.21 μm (mean ± SE: 5.93 ± 0.22 μm) in length and from 2.89 to 4.71 μm (mean ± SE: 3.76 ± 0.14 μm) in width. The results were expressed in submerged conidia/ml and MS/ml.

The determination of the viable propagules was performed based on the counting of the colony forming units (CFU) in Petri dishes (90 × 15 mm) that contained 20 ml of PDA amended with 0.01% (v/v) of Triton X-100^® (Synth^®) and 0.001% (w/v) of chloramphenicol (Sigma^®). After the serial dilution of the samples obtained from the liquid cultures, an aliquot of 50 μL of the suspension of C. rosea from the 10^–7 dilution was transferred to the medium. The plates were incubated in a growth chamber at 25 ± 2°C with a photoperiod of 12:12 h, and the number of CFU was computed after 4 days. To carry out the drying of the biomass produced, on the seventh day of the fermentation, the entire volume of the liquid culture was mixed with 5% of diatomite (Diatom^® M45, Brazil) and filtered (slow filtration with filter paper; pore size <12 μm, 80 g/m²; Whatman^®) through a Büchner funnel coupled to a Kitasato to dewater the biomass-diatomite mixture for the next step involving air drying. The drying was carried out in a horizontal tray chamber with controlled airflow (relative humidity of the purged air: 18–50%) for 15 h at 22 ± 2°C until reaching the final moisture content less than 5% (w/w) (Kobori et al., 2015). After drying, the materials were ground to fine granules (<2 mm) as also referred here as air-dried microgranules (see details in Supplementary Figure S2).

The study of C. rosea parasitism using different propagules on S. sclerotiorum sclerotia was carried out with 24 Petri dishes (90 × 15 mm) that contained 30 g of autoclaved dry soil in each plate, with moisture adjusted to 100% of field capacity, using sterile distilled water. The dark mature sclerotia of Sclerotinia were collected from 25-day-old PDA cultures and used in bioassays. After the distribution of 12 sclerotia per plate, 10.5 ml of C. rosea suspension that contained 16 μl of surfactant solution (0.02% of Break-thru MSO, Evonik^®, Essen, Germany) were evenly applied to the soil in each plate. For control, 10.5 ml of distilled water was applied in the same way as mentioned above. The plates were incubated for 14 days in a growth chamber at 25 ± 1°C with 12:12 h photoperiod. After that period, all sclerotia were removed from the soil and then superficially disinfested with 70% ethanol (1 min), 2% sodium hypochlorite (NaOCl) (1 min), and rinsed three times in sterile distilled water prior to plating them on a selective medium. Twelve surface-sterilized sclerotia were transferred to Neon-S medium [1 L of water; 40 g of PDA (Difco^®), 50 mg of bromophenol blue, 50 mg of chloramphenicol, and 50 mg of free acid 2,4-D)] per plate (20 ml medium, 90 × 15 mm), and then incubated for 7 days to evaluate viability (Ferraz et al., 2011). The viability was carried out by observing the biochemically induced color change in the Neon-S medium due to the pathogen germination, i.e., viable or germinated sclerotia change the color of the medium from purple to yellowish due to the production of oxalic acid (Napoleão et al., 2006; Ferraz et al., 2011). In addition, we also determined the C. rosea presence by assessing its outgrowth on sclerotia, as a surrogate to confirm the ability of this bioagent to penetrate, parasite, and kill the target pathogen structure (Rodríguez et al., 2011). In this bioassay, we tested the following four treatments: Control with distilled water, suspension of C. rosea conidia from solid fermentation on rice, suspension of C. rosea conidia from liquid fermentation (medium M3 in Table 1), and suspension of C. rosea MS from liquid fermentation (medium M7 in Table 1). The inoculum density was standardized to 1 × 10⁶ CFU/g of sterile dry soil, which was obtained with air-dried granules that contained 6 × 10⁸ CFU/g, except for the treatment with MS, which delivered 2.5 × 10⁵ MS/g of dry soil. The assays were performed twice on different dates, with six replicates for each treatment. An illustrative flowchart of this experimental procedure is described in Supplementary Figure S3.

The performance of submerged liquid fermentation process of C. rosea for mass production of submerged conidia was assessed and validated using a 3-L laboratory benchtop bioreactor (New Brunswick™ BioFlo/CelliGen^® 115, Eppendorf^®, New Brunswick, NY, United States) with the medium coded as M3 (Table 1): 50:1 C:N ratio, 36 g carbon/L, pH 6.0, initial inoculum density of 5 × 10⁶ submerged conidia/ml, 9.1% (w/v) dextrose monohydrate and 0.84% (w/v) soybean meal (Bakers souflour^®). The liquid pre-culture was the same as previously described (Table 1). The medium was inoculated with a 4-day-old liquid pre-culture to deliver a final concentration of 5 × 10⁶ conidia/ml using 180 ml (10% v/v) of a suspension prepared with sterile distilled water and adjusted to 5 × 10⁷ submerged conidia/ml, considering a total working volume of 1.8 L. The initial pH of the medium was adjusted to 6.0 without any control during the cultivation process. The initial parameters after inoculation were: agitation speed of 400 rpm, gas-flow of 1.0 L/min (0.56 vvm), and temperature of 28 ± 0.5°C. The mechanical agitation was propelled with two Rushton (flat blade) impellers. Antifoam solution at 0.05% (v/v) (Break-thru^® AF 9903, Evonik Operations GmbH, Essen, Germany) was pumped into the culture broth whenever needed to prevent foaming. Temperature, dissolved oxygen level (DO), gas-flow and agitation speed were monitored across time, and culture samples were taken every 24 h until the fourth day to determine the submerged spore concentration, the CFU and the pH. Then, both the agitation speed and the gas-flow (purged with filtered atmospheric air inlet) were altered during the fermentation process as a means to avoid DO dropping below 15%. This fermentation process was carried out in batch cultures and independently repeated on six different dates.

The 4-day-old culture broth that contained submerged conidia and mycelium (without any microsclerotium biomass) was harvested, centrifuged (10,000 rpm for 20 min at 10°C), and the resulting fresh biomass (20.0 g) was evenly mixed with 37.0 g (61.6% w/w) of diatomaceous earth (DE) [Diatom^® M45; bulk density, 280–360 g/L; fine powder < 0.12 mm (or < 150 mesh); Diatom Mineração Ltda., Mogi das Cruzes, SP, Brazil] and 3.0 g (5% w/w) of organosilicon-based dispersant (Break-thru SD260, Evonik Operations GmbH, Essen, Germany) to a final weight of 60-g formulated fungal biomass. The drying process followed the procedure described in Kobori et al. (2015) and is the same in the section “Submerged liquid fermentation to selection of nitrogen sources.” After 14–16 h of air-drying, the air-dried microgranular formulation held a final moisture content in the range of 2.6 – 3.6% (w/w) from different bioreactor fermentation batches (n = 6). After that, this fungal preparation was vacuum packed with a 5-layer barrier bag made of nylon-poly plastic (15 × 22 cm, Equapack Embalagens, São Paulo, SP, Brazil) and then cold stored at 4–6°C until use in bioassays to further evaluate the efficacy against sclerotia of S. sclerotiorum and whitefly nymphs. Before each bioassay, the air-dried submerged conidia had their viability checked on PDA medium, and it was expressed in CFU/g. For the assessment of myceliogenic germination, a sample of 0.03 g of each formulation batch was sprinkled over the surface of 2% (w/v) agar-water medium poured in Petri plate (90 × 15 mm) and incubated for 24 h at 25 ± 1°C and 12:12 h photoperiod. The percent viability was determined by randomly counting 100 fungal microgranules under a stereomicroscope at 40× magnification and deemed germinated in the presence of hyphal outgrowth. After an additional 6 days of incubation, the sporulated microgranules exhibiting profuse fungal outgrowth and conidiation were harvested using sterile 0.04% Tween 80^® solution and then serially diluted to proceed with counts on a Neubauer chamber in order to compute the spore production per gram of this air-dried microgranular formulation (Supplementary Figure S2).

The air-dried microgranular formulation that contained mainly submerged conidia of C. rosea retained viability of 6 × 10⁸ CFU/g. The inoculum loads tested to deliver active viable propagules to the soil were 1 × 10⁴, 1 × 10⁵, and 1 × 10⁶ CFU/g soil. The whole bioassay was repeated twice on different occasions with a total of 12 replicates per treatment. The methodology was the same as described before in the section “Bioassay of C. rosea vs. S. sclerotiorum.”

The whitefly colony of B. tabaci biotype B or Middle East–Asia Minor I (MEAM1) was reared on cabbage (Brassica oleracea L., cv. Manteiga, TopSeed^®, SP, Brazil) and on Jack bean (Canavalia ensiformis L., Piraí Sementes Ltda., Piracicaba, SP, Brazil) in a screenhouse under natural environmental conditions. The bean plants cv. Pérola were grown in the potted-soil inside a screenhouse and infested with the whitefly adults for 24 h to allow oviposition underside the leaves. After laying their eggs, the whitefly adults were removed from the bean plants by blowing them away (Silva et al., 2019). After approximately 12–14 days, the whitefly nymphs had reached the first to the second instars. The air-dried submerged conidia produced by the liquid culture in the bioreactor were tested against the first to the second instar nymphs to assess their pathogenicity and virulence using a concentration-mortality bioassay. The four concentrations of air-dried microgranular formulation of C. rosea (2.2 × 10⁹ CFU/g) prepared with 0.04% Tween 80^® were tested at: 5 × 10⁶, 1 × 10⁷, 5 × 10⁷, and 1 × 10⁸ CFU/ml. The sterile aqueous 0.04% Tween 80^® solution was sprayed as control. The potted bean plants bearing leaves infested with the whitefly nymphs were individually sprayed with a hand-held dual-gravity airbrush assembled to a benchtop spray-tower loaded with 300 μL of each spore concentration (Silva et al., 2019). After spraying, these potted bean plants were incubated in a growth chamber set to 26 ± 2°C and 12:12 h photoperiod with a data logger (HOBO^® U12-012, Sigma Sensors, São José dos Campos, SP, Brazil) to track-record the hourly temperature and the relative humidity inside this chamber throughout the experimental course. According to the data logger recordings during the 7 days of incubation, the mean temperature and relative humidity were 26.8°C (25.7–27.8°C) and 84.2% (30.7–96.4%), respectively. The number of nymphs per leaf varied greatly and ranged from 4 to 34, with the majority above 10 nymphs per leaf. The bioassay was repeated twice on different dates and each treatment had a total of seven replicates. The dead and live nymphs were recorded after 5 and 7 days of incubation. Dead nymphs, exhibiting typical symptoms of mycosis by C. rosea, appeared shriveled, dehydrated, and generally with orange-like or pinkish color.

The production data of aerial conidia, submerged conidia, and MS were fitted to generalized linear models with negative binomial distribution with or without fixed effects interaction in the linear predictor. When a significant effect was detected, the treatment means were compared by Tukey HSD at p < 0.05.

The whitefly mortality proportion, obtained across increased concentrations of C. rosea submerged conidia, was fitted to a two-parameter log–logistic model with binomial distribution using the “drc package” in R (Ritz et al., 2015). In addition, the median lethal concentrations for the mortality rates recorded on days 5 and 7 post-spraying were estimated and then compared by the z-score test and slopes by the t-Student test. Similarly, the data generated on proportion of non-viable or parasitized sclerotia of S. sclerotiorum after exposure to different concentrations of spore-treatments with C. rosea were fitted to the same model as described for the whitefly mortality, including the concentration of C. rosea inoculum as the fixed effect in the linear predictor. An estimated LC₅₀ of C. rosea submerged conidia was also computed based on the antagonism exerted over sclerotia.

All statistical analyses were performed in the R statistical software environment (R Core Team, 2015¹).

The highest C. rosea conidial concentration was observed after 7 days of cultivation in Erlenmeyer flasks filled with autoclaved wet rice grains and inoculated with submerged conidia produced by liquid culture. Under these conditions, C. rosea produced 1.1 × 10⁹ conidia/gDM. The inoculum of the suspension made with aerial conidia grown on PDA medium gave an average of 8.1 × 10⁸ conidia/gDM. Nonetheless, there was no significant difference in conidial yield on rice grains due to the inoculum origin (χ² = 1.26, p = 0.26), which means that solid-state fermentation of C. rosea on rice grains can be performed with any inoculum type of C. rosea either grown on PDA or in liquid medium (Figure 1).

The submerged liquid fermentation carried out in shake flasks of C. rosea allowed the production of both submerged conidia and MS under appropriate C:N ratio, nitrogen source, and cultivation time. Notably, the interactions nitrogen source vs. C:N ratio and nitrogen source vs. cultivation time significantly influenced the production of submerged conidia, indicating that soybean meal as the nitrogen source with 50:1 C:N ratio on the seventh day of cultivation reached the maximum spore yield by liquid culture (p < 0.001, Supplementary Table S1). Clearly, the 50:1 C:N ratio had a remarkable impact on the increased yields of submerged conidia than 10:1 C:N ratio, in which the majority of liquid cultures attained spore production peak by day 7 of cultivation, although concentrations more than 1 × 10⁹ submerged conidia/ml were only achieved by day 7 of cultivation (Figure 2A). The nitrogen source also played a crucial role in the spore production, in which soybean meal followed by cottonseed flour and yeast extract stood out as the best, whilst corn bran exhibited the lowest performance. The highest submerged conidial production was achieved after 7 days of fermentation using soybean meal with 50:1 C:N ratio cultures that resulted in 1.41 × 10⁹/ml, followed by cottonseed flour corresponding to 1.03 × 10⁹/ml.

It is noteworthy that MS production was significantly influenced by three factors (p = 0.0067, Supplementary Table S1), the combination of soybean meal as the nitrogen source, 10:1 C:N ratio and 6 days of cultivation, resulting in a yield of 1.35 × 10⁴/ml by submerged cultures of C. rosea. However, from day 4 onward, the amount of MS did not significantly increase with longer cultivation times. Interestingly, yeast extract was the second-best nitrogen source that yielded 1.34 × 10⁴ MS/ml after 7 days of fermentation. With 50:1 C:N ratio, liquid cultures attained yields less than 3 × 10³/ml, whilst the opposite was found with higher nitrogen content imposed by lower C:N ratio that boosted MS yields over 1 × 10⁴/ml, regardless of nitrogen source and cultivation time (Figure 2B). Among cultures grown under 10:1 C:N ratio, the production peak of MS during fungal growth was reached between days 3 and 4 of cultivation, regardless of nitrogen source. Notably, the best nitrogen sources at 10:1 C:N ratio to produce MS followed the order listed as follows: Soybean meal, cottonseed flour, yeast extract, and the minor impact was attributed to the corn bran.

Overall, at a fixed amount of high carbon content (36 g/L), the C:N ratio clearly showed to dramatically affect propagule formation and production by submerged cultures of C. rosea, as higher C:N ratio favored the production of submerged conidia, whereas lower C:N ratio was preferred for MS development. This indicates that higher nitrogen content was crucial for fungal vegetative growth that leads to optimal MS production (Figure 2A).

All submerged cultures of C. rosea grown with 10:1 C:N ratio had a thicker aspect due to the higher nitrogen content promoting more vegetative growth and increased MS production (Figure 2C). The microsclerotia (MS) were typically a thread of mycelial aggregates usually larger than 50 μm and were dark pigmented (Figure 2D). In contrast, liquid cultures with 50:1 C:N ratio portrayed more liquid aspect or less viscosity due to less mycelial biomass and showed an increased production of submerged conidia of variable sizes (range of 3.86 to 8.21 μm in length and 2.89 to 4.71 μm in width), which were formed by conidiogenous cells known as phialides (Figure 2D).

To expand the production scale of C. rosea, the fermentation was carried out in an automated benchtop bioreactor, using the medium that contained soybean meal 0.84% (w/v), C:N ratio of 50:1, and initial pH of 6.0 (medium M3, Table 1), as it provided the greatest yield of submerged conidia under liquid culture in shake flasks.

Among the six fermentation batches performed, the pH initially set to 6.0 slightly dropped to 5.4–5.8 during the entire fermentation course (Table 2). To maintain the dissolved oxygen level above 15% throughout the fermentation course, the agitation speed increased from 400 to 700 rpm at the last day (day 4), in conjunction with the increase of sparged air (gas-flow) that varied from 1.0 to 2.0 L/min. Liquid cultures of C. rosea conducted in bioreactors depicted similar trends of DO level, although some variation was noted between these DO profiles across production batches with two typical patterns described in Figure 3B. As for the fermentation parameters, there was a rapid decrease in oxygen availability especially on the first day of cultivation across four fermentation batches, with oxygen consumption being more pronounced within 24 h post-inoculation. After 48 h of cultivation, it was noticed the dissolved oxygen level went back up to 50%, which also reflected a decrease in the specific growth rate from day 3 to day 4 (Table 2). Nevertheless, the concentration of the submerged conidia increased over time with a peak at day 4 corresponding to 1.1 × 10⁹/ml (Table 2 and Figure 3A). In turn, the number of total viable propagules (CFU) slightly increased from day 2 to day 4, resulting in a maximum yield of 4.35 × 10⁸ CFU/ml (Table 2 and Figure 3A). The variation in production of submerged conidia and total viable propagules from batch to batch also occurred and is described in Table 2. The microscopic observations also confirmed the progress of submerged conidial production during the fermentation course, as notably by day 4 depicting the highest conidial density.

The number of viable propagules (mostly submerged conidia) in the air-dried microgranular formulation of C. rosea was estimated to contain 2.0 × 10⁹ CFU/g. Since we used the M3 medium in the bioreactor studies, this growth environment stalled the formation of MS. At last, the air-dried microgranular formulation upon 24 h rehydration rendered 100% myceliogenic germination, followed by a conidial production of 3.4 × 10¹⁰ conidia/g microgranules at day 7 post-incubation, indicating a large number of spores released by this fungal formulation.

All C. rosea propagules tested (i.e., aerial conidia, submerged conidia, and MS) against S. sclerotiorum parasitized 100% of sclerotia, even after surface sterilization of such structures, as noted by unchanged color of the Neon-S medium; whereas the untreated control exhibited all sclerotia viable as noted by the presence of profuse myceliogenic germination and color change from purple to yellowish of the Neon-S medium, which is indicative of oxalic acid production by the pathogen (Supplementary Figure S4). Furthermore, all non-viable sclerotia exhibited a profuse outgrowth of C. rosea, which strongly suggests that this bioagent internally colonized and parasitized sclerotia, regardless of its propagule type tested as inoculum (Supplementary Figure S4). These results indicate that C. rosea propagules (aerial conidia, submerged conidia, and MS), derived from either liquid or solid fermentation, were highly effective in constraining the myceliogenic germination of S. sclerotiorum sclerotia.

In the follow-up experiment, there was a significant effect of C. rosea inoculum rate on sclerotial parasitism of S. sclerotiorum (χ² = 33.98, p < 0.0001), in which all fungal concentrations differed from the untreated control group and induced sclerotial parasitism from 44.4 to 80.7% in a proportional manner (Supplementary Figure S5). It is noteworthy that the highest inoculum rate of C. rosea (1 × 10⁶ CFU/g of soil) caused the strongest reduction in sclerotial viability. The relation of sclerotial parasitism rate with the inoculum load of C. rosea applied to the soil arena was properly described with a two-parameter binomial log-logistic model (Figure 4A), in which the concentration required to suppress or parasitize 50% of the sclerotia was estimated to be 3.23 × 10⁴ CFU/g of soil (Table 3).

The healthy and viable Sclerotinia sclerotia from untreated control group were confirmed by the yellowish color produced by the Neon medium due to a reaction with oxalic acid secreted by the fungus (Figure 4B), whereas the submerged conidia of C. rosea were able to inhibit the myceliogenic germination of S. sclerotiorum sclerotia by impeding the color change in the Neon medium that remained purple (Figure 4C).

The insect bioassay revealed that C. rosea air-dried submerged conidia were pathogenic to early whitefly nymphs. The whitefly mortality significantly increased with incubation time and fungal concentration tested (interaction time × concentration: χ² = 10.14, p = 0.0015), resulting in a range of 31–67.5% and 33–76.2% dead nymphs after 5 and 7 days of exposure, respectively. The two concentration-dependent mortality curves were different from each other (χ² = 6.0, p = 0.0497), indicating that the whitefly mortality was relatively higher by day 7 than day 5 of evaluation, particularly when nymphs were exposed to higher fungal inoculum rates (Figure 5).

With respect to the median lethal concentrations and the mortality curve slopes, the whitefly mortality had similar slopes but differed in terms of LC₅₀ values attained by different evaluation days, in which day 7 resulted in lower LC₅₀ than day 5 (Table 4). In this line, the susceptibility of the whitefly nymphs to the C. rosea submerged conidia increased with time exposure, which resulted in a reduction of LC₅₀ by 42.3% from day 5 to day 7 post-spraying (p = 0.048, Table 4). The dead nymphs due to infection with C. rosea usually exhibited symptoms of shriveled and orange-like appearance and signs of mycelial mats growing out of insect cadavers (Supplementary Figure S6).