Onion leaves showing SLB symptoms were collected from the field of the Division of Vegetable Science, ICAR-Indian Agricultural Research Institute, New Delhi. The infected leaves were washed properly under running tap water to remove dirt and soil. Surface sterilization of the diseased leaves was performed by immersing them in sodium hypochlorite (NaOCl) solution at 2% (v/v) solution for 1–2 min and then washed twice with sterile distilled water. After drying on sterilized tissue paper, leaves were fragmented into small pieces of approx. 5–6 mm². Fragments of leaves were cultured on sterile potato dextrose agar media (PDA, HiMedia) in plastic petri plates (90 mm dia × 15 mm depth) and incubated at 25 ± 2°C for 24–48 h. Emerging fungal colonies were closely monitored for their growth pattern, and with the help of a sterile needle, the hyphal tip of the fungal colony was subcultured on the fresh potato dextrose agar media. These subcultured Petri plates were kept at 25 ± 2°C temperature for 7–10 days under a 14-h light and 10-h dark period. The fungus was identified based on morphological characteristics (conidia and conidiophore shape and size; Ellis, 1971) and the internal transcribed spacer (ITS) sequence. DNA was extracted from the fungal mycelial mat using the CTAB method (Cullings, 1992), and the ITS region was amplified using primers ITS1 (5′-TCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) (White et al., 1990). The amplified product was sequenced directly by outsourcing (Barcode Biosciences, New Delhi) and identified using NCBI BLAST. The sequence was submitted to NCBI, and the GenBank number was obtained. The pure culture of S. vesicarium was utilized for the preparation of 1 × 10⁵ ml⁻¹ spore suspensions for inoculation.

The field screening experiment through artificial inoculation was conducted by spraying the spore suspension in the experimental field of the Division of Vegetable Science, ICAR-IARI, New Delhi (77°09′27″E, 28°38′23″N, and altitude 228.61 m above the mean sea level). Seeds of 157 onion accessions, consisting of breeding lines, landraces, open-pollinated varieties (OPVs), and hybrids ( Supplementary Table 1 ), were sown in the nursery during the second week of October 2020. The transplanting of plants to the experimental plot was done on the third week of December 2020 in a randomized complete block design with three replications. Sixty plants per accession were transplanted with a row spacing of 15 cm and a plant spacing of 10 cm. Recommended cultural practices were followed, and a nitrogen: phosphorus: potassium (NPK) fertilizer was applied at 110:40:60 kg/ha. Plants of highly susceptible accession number AKON068 were planted in every third row and around the border of the experiment plot for an additional source of inoculum for infection. For ensuring uniform soil moisture throughout the experimental plot, flood irrigation was applied at 7–10-day intervals throughout the experimentation period. Manual removal of weeds was carried out at 15-day intervals until the neck-fall stage. The experimental plot was surrounded by a 2.50-m-high translucent plastic sheet, and the application of fungicides was restricted in and around the experimentation site.

Morphological traits such as plant height (PH, cm), leaf length (LL, cm), leaf width (LW, cm), number of leaves per pseudostem (NOL), pseudostem length (PsL, cm), pseudostem width (PsW, mm), polar diameter (P, mm), equatorial diameter (E, mm), neck thickness (N, mm), bolting percentage, days to maturity (DTM), average bulb weight (ABW, g), gross yield (GW, t/ha), and marketable yield (MY, t/ha) were recorded. Gross yield was calculated according to the following formula: Yield (tons/ha) = (Yield in kg per plot × 666.67)/Number of plants planted. Marketable yield was calculated using the same formula, and bulbs having an equatorial diameter greater than 3.50 cm were counted as marketable. The dry matter (DM) of the bulbs was assessed according to Nieuwhof et al. (1973) and expressed in percentage (%). Total soluble solids (T.S.S) were determined by using a hand refractometer (model, PAL-3) and expressed in °B.

Genotypes showing a moderately resistant and susceptible reaction were identified under field conditions ( Supplementary Table 2 ). Bulbs of the same genotypes were screened for Stemphylium blight under controlled conditions since SLB is also a devastating disease for seed crop too. The experiment was conducted inside the automated greenhouse of the Division of Vegetable Science, ICAR-Indian Agricultural Research Institute, New Delhi. This experiment comprised 20 previously identified genotypes (15 moderately resistant, four susceptible, one highly susceptible) during the field screening. The experiment was laid out in a completely randomized design with two replications. Five bulbs of each accession were planted in each pot filled with sterilized cocopeat, perlite, and vermiculite in the ratio of 2:1:1, respectively. The onion plants were kept at 18°C–22°C temperature under ambient light conditions and watered every 3–5 days. Plants were fertilized after 15 days of planting. Thereafter, every week a water-soluble fertilizer (NPK 19:19:19) at the rate of 5 g/l was applied.

For field screening, onion plants were inoculated 60 days after transplanting. A conidial suspension of 1 × 10⁵ ml⁻¹ was prepared by serial dilution technique and sprayed using a knapsack sprayer in the evening followed by flood irrigation to maintain the optimum relative humidity (RH). Onion plant leaves were sprayed with fine mist of water twice after 8–10 h of inoculum spraying to maintain leaf wetness for facilitating better conidium germination. The field was irrigated repeatedly at 2-day intervals until the disease appeared. The fungus was re-isolated from the infected leaves, and its identity was confirmed through morphological observations. In greenhouse screening, a similar concentration of conidial suspension was used as mentioned. For maintenance, a relative humidity of<90% and leaf wetness for 8–10 h using automated misters were applied.

Scoring of the disease incidence was done by using a 0–5-point scale ( Table 1 ) (Sharma, 1986). Disease assessment was done by taking 10 plants per accession and per replicate into consideration at an interval of 10 days after the inoculation and repeated thrice. The infected areas of each leaf were observed visually. Percent disease incidence (PDI) was calculated (Wheeler, 1969) using the following formula:

where N₁ to N₅ represent the total number of leaves falling under the 1–5 scales, respectively. Based on their PDI values, each onion accession was classified into six disease reaction classes: immune (I<5), resistant (5.1 ≤R≥ 10.0), moderately resistant (10.1 ≤MR≥ 20.0), moderately susceptible (20.1 ≤MS≥ 40.0), susceptible (40.1 ≤S≥ 60.0), and highly susceptible (HS >60.0) (Pathak et al., 1986). The area under the disease progress curve (AUDPC) was calculated using the trapezoidal method according to a formula given by Madden et al. (2007).

where yi is an assessment of a disease (percentage) at the ith observation, ti is the time (in days) at the ith observation, and n is the total number of observations.

For biochemical analyses, leaves were collected from onion bulbs grown in pots under greenhouse conditions. To perform biochemical analyses in leaf tissues, leaf samples were frozen in liquid nitrogen. Leaf samples were collected before inoculation and later on 10-day intervals at 10, 20, and 30 days after inoculation with the S. vesicarium pathogen.

The total phenolic content of onion leaf was estimated spectrophotometrically using the Folin–Ciocalteu (FC) reagent method developed by Singleton and Rossi (1965). Absorbance was taken using a UV-1900i UV-Vis double-beam spectrophotometer (SHIMADZU, Japan) at 750 nm. Total phenol content was expressed in ‘mg gallic acid equivalent (GAE)’ per g of fresh weight.

Pyruvic acid was estimated spectrophotometrically through the method of Anthon and Barrett (2003). Standards were prepared by using sodium pyruvate solution. The reading was carried out in a spectrophotometer at 515 nm using the UV-1900i UV-Vis double-beam spectrophotometer (SHIMADZU, Japan). Pyruvic acid was expressed as µmole g⁻¹ sample.

Catalase activity estimation was done according to Aebi (1984). The H₂O₂ oxidation was assayed spectrophotometrically at 240 nm every 30 s for 2 min, and activity was computed by calculating the amount of H₂O₂ decomposed. A standard curve drawn with the known concentration of H₂O₂ was used to determine the initial and final contents of hydrogen peroxide. Catalase activity was measured in μmol H₂O₂/min/mg protein.

The peroxidase activity was assayed by using guaiacol as the substrate by following the method described by Chance and Maehly (1955). The guaiacol peroxidase activity was determined spectrophotometrically by measuring the increase in absorbance at 470 nm by the conversion of guaiacol to tetraguaiacol due to its oxidation. The molar extinction coeffcient of tetraguaiacol was taken as 26.6 mM⁻¹ cm⁻¹. One unit of enzyme activity is defined as the formation of 1 µmol product of tetraguaiacol by the enzyme-catalyzing action per minute at 30°C.

Total soluble protein estimation was done by the Bradford reagent method (Bradford, 1976). The absorbance was recorded using a spectrophotometer at 595 nm (Bradford Protein assay, USA) and expressed in mg per gm of fresh weight. For the preparation of protein standard, 1 mg/ml bovine serum albumin was dissolved in distilled water and utilized as the stock solution.

The analysis of variance and comparison of means were analyzed using HAU-OPSTAT software (http://hau.ernet.in/OPSTAT). Genetic parameters including the genotypic and phenotypic variance, genotypic and phenotypic coefficients of variance, heritability (broad sense), and expected genetic advance (GA) were performed using package variability 0.1.0 (Popat et al., 2020) through R software (ver. 4.2.1). The mean data for the recorded traits were used for calculating Pearson’s correlation coefcient by using ggplot 2 (R 2016) in R Studio (2021).

The infected onion leaves started to exhibit initial symptoms of SLB after 7–10 days of inoculation under field conditions. Initial symptoms included small tan to brown water-soaked, oval lesions on the tips and center of outer leaves. Necrosis started from the top and advanced downward on both sides of the lesions. Sometimes only tan necrosis occurred on the tip of leaves, which is not associated with any lesion formation and progressed toward the bottom of the leaves. In later stages, these symptoms advanced to dark-brown to black sporulation and were more evident on older plants. On outer leaves, black and purple target spot lesions with black sporulation were also evident occasionally ( Figure 1 ). The isolate was identified as Stemphylium vesicarium on the basis of cultural characteristics and morphology of conidia and conidiophores. Whitish with light gray center, filiform, whitish margin colonies of Stemphylium vesicarium were observed on PDA medium ( Figure 2 ). The observed conidia were light brown to brown in color, ovoid to oblong with zero to five transverse septa and one to five longitudinal septa, ranging 15–23 × 40–49 μm ( Figure 3 ). Conidiophores were straight or occasionally one-branched with a swollen apex. The obtained ITS sequence showed 100% sequence homology with NCBI accessions of S. vesicarium (MZ452063). The generated sequence was submitted to the NCBI database (OP473965).

Descriptive statistics of 16 morphological and biochemical traits indicated the presence of significant variation among onion genotypes under disease pressure ( Supplementary Table 2 ). Approximately, a normal distribution was observed for most of the traits apart from bolting percentage and days to maturity ( Figures 4A, B ). Among various morphological traits, the plant height varied from 44.16 to 86.70 cm with a mean value of 70.31 cm ( Supplementary Table 1 ). The plant height was recorded maximum in genotype SBO023 (86.70 cm) whereas the minimum was observed in SBO157 (44.16 cm). Similarly for leaf length, SBO023 (75.79 cm) recorded the highest value and minimum leaf length was recorded in SBO157 (35.43 cm). The leaf width was maximum in genotype SBO115 (2.07 cm) and minimum in SBO046 (0.53 cm) followed by SBO057 (0.54 cm). The number of leaves was maximum in SBO078 (11.11) and minimum in SBO083 (5.51). The genotype SBO091 (21.82 mm) followed by SBO109 (19.16 mm) exhibited the highest value for pseudostem length, whereas SBO06 (8.84 mm) had minimum pseudostem length. For trait pseudostem diameter, the maximum value was recorded for SBO038 (17.98 mm), whereas the minimum was observed in SBO132 (4.75 mm). Minimum neck thickness was recorded in SBO056 (0.38 cm), whereas the maximum was found in SBO080 (1.27 cm). The equatorial diameter of bulb was maximum in SBO148 (60.41 mm) and minimum in SBO091 (39.05 mm) with an average value of 51.20 mm. The genotype SBO148 (58.04 mm) was found have the highest value for polar diameter, whereas the minimum value was recorded for SBO131 (37.72 mm). A wide variability for average bulb weight was observed. The average bulb weight varied from 39.20 to 99.80 g with an average of 69.71 g. The highest average bulb weight was recorded in SBO148 (99.80 gm), whereas the lowest was observed in SBO038 (39.20 gm). The gross yield was recorded maximum in SBO020 (41.61 t/ha), whereas the lowest was observed in SBO076 (10.11 t/ha) followed by SBO116 (11.44 t/ha) and SBO138 (13.44 t/ha). The average marketable yield varied from 6.00 to 38.00 t/ha with an average of 22.22 (t/ha). The highest marketable yield was recorded in SBO071 (38.00 t/ha) followed by SBO120 (37.28 t/ha), and the least was recorded in SBO119 (6.00 t/ha). The minimum days to maturity were recorded for two genotypes SBO044 (97.00 days) and SBO073 (97.00 days), whereas the maximum days to maturity were observed for genotype SBO124 (131.00 days) followed by SBO116 (130.50). TSS ranged from 6.82 to 15.88 (°B) and showed a significant variation among genotypes. Maximum TSS was recorded in genotype SBO069 (15.88°B), whereas the minimum was found in genotype SBO145 (6.82°B). The average values of dry matter content was 10.76%. Maximum dry matter content was found in genotype SBO025 (15.26%) followed by SBO138 (14.96%), whereas the minimum dry matter content was recorded in genotype SBO090 (5.75%). Nearly 80 genotypes were present in a set of 157 genotypes which did not bolt at all, whereas the maximum bolting percentage was found in three genotypes SBO093, SBO118, and SBO119 (17.50%).

The analysis of variance was estimated for all the 17 traits which showed highly significant mean squares at P< 0.01 and 0.001, indicating variation among the 157 onion genotypes ( Table 2 ). A wide range of values among the genotypes for all traits under consideration also indicates the presence of variation. Moreover, mean, standard errors, and coefficients of variation were also estimated which again confirmed considerable genetic variability among the genotypes. The phenotypic variance (σ²P) for all the traits was higher than the genotypic variance (σ²G). The phenotypic variance ranged from 0.03 (NT) to 128.94 (ABW), whereas the genotypic variance ranged from 0.03 (NT) to 126.93 (ABW). PCV and GCV values more than 20% are regarded as high, values ranging between 10% and 20% as medium, and values less than 10% as low (Deshmukh et al., 1986). GCV values ranged from 5.35% (Polar diameter) to 142.33% for the trait percent bolting. Similarly, the PCV values ranged from 8.28% for polar diameter to 150.03% for percent bolting. GCV values were lower than those of PCV values for all the traits. Traits such as leaf width, pseudostem diameter, neck thickness, percent bolting, gross yield marketable yield, and percent disease incidence were recorded with high GCV and PCV values. The highest GCV was observed for trait percent bolting (142.33%) followed by leaf width (29.57%) and PDI (28.70%). The medium GCV was recorded for traits such as leaf length, number of leaves, pseudostem length, total soluble solids, dry matter content, and average bulb yield. The plant height, equatorial diameter, polar diameter, and days to maturity were grouped under the traits having low GCV. The lowest GCV was recorded for polar diameter (5.35%) followed by equatorial diameter (7.11%) and days to maturity (8.52%). Similarly for traits such as percent bolting (150.03%), leaf width (29.68%), marketable yield (33.32%), gross yield (29.95%), PDI (29.14%), neck thickness (27.31%), and pseudostem diameter (24.38%), a high PCV was observed, whereas four traits, viz., plant height (9.63%), equatorial diameter (8.57%), polar diameter (8.28%), and days to maturity (8.57%), were recorded with the lowest PCV values. Medium PCV was recorded in five traits, namely, leaf length, number of leaves, pseudostem length, total soluble solids, dry matter content, and average bulb weight. The traits including polar diameter, percent bolting, gross yield, and marketable yield recorded a high difference among GCV and PCV values as compared with the rest of the traits which have a very small (<1) difference in their GCV and PCV values. Heritability values higher than 80% were considered as very high, values between 60% and 79% were moderately high, 40%–59% were medium, and values less than 40% were low (Singh, 2001). Heritability in a broad sense was also estimated for all traits, which ranged from 41.77% (polar diameter) to 99.51% (pseudostem width). Traits such as plant height, leaf length, leaf diameter, number of leaves, pseudostem length, pseudostem width, neck thickness, total soluble solids, dry matter content, bolting percentage, average bulb weight, days to maturity, and PDI were observed to have the highest heritability. Only one trait, i.e., equatorial diameter, was recorded with moderately high heritability. The remaining traits such as polar diameter, gross yield, and marketable yield were found to have medium heritability. Estimation of genetic advance for various traits ranged from 0.35% to 23.39% and was found highest for average bulb weight (23.39%). Moderate genetic advance was observed for traits such as days to maturity, PDI, plant height, and leaf length, whereas all the remaining traits were found to have low genetic advance.

Field evaluation at 10, 20, and 30 DAI showed a wide phenotypic response. The mean values for the percent disease incidence of SLB among the 157 genotypes at 30 DAI ranged from 13.20% to 68.00% ( Supplementary Table 3 ). The highest SLB incidence was observed in SBO131 (68.00%), and the lowest was recorded in SBO113 (13.20%). Only four categories of disease reaction existed among the 157 onion genotypes ( Figure 5A ). Out of 157 genotypes, 16 genotypes (10.20%) exhibited moderately resistant reactions whereas 107 genotypes (68.10%) were found to be moderately susceptible. Thirty-three genotypes (21.00%) were categorized as susceptible, and one (SBO131) genotype (0.70%) was found to be highly susceptible at 30 DAI ( Figure 5B ). None of the genotypes were found to be immune or resistant against SLB in this study. Out of 16 moderately resistant genotypes, three were open pollinating varieties (OPV), five were inbred lines, seven were breeding lines, and one was hybrid. The maximum numbers of moderately resistant (7), moderately susceptible (36), and susceptible (13) genotypes were found among breeding lines. Among all genotypes belonging to OPVs, breeding lines, inbred lines, and hybrids, only one genotype belonging to the OPV group exhibited a highly susceptible reaction ( Table 3 ). The area under the disease progress curve ranged from 230.00 to 1012.00 and was recorded minimum in SBO113 (230.00) and maximum in SBO131 (1012.00) ( Figures 6A, B ). The results for the lowest and highest mean values of PDI and AUDPC were similar, but they were not identical in ranking of the genotypes, e.g., ‘SBO076’ had a PDI of 14.80 and was ranked in second place on PDI basis but was ranked fourth based on the AUDPC scale ( Supplementary Table 3 ).

Concerning the phenotypic correlation coefficient, the results revealed that there was a significant positive correlation present between PDI and AUDPC (p< 0.0001). On the other hand, PDI was found to be significantly negatively correlated with leaf length (r = -0.20*), number of leaves (r = -0.17*), pseudostem width (r = -0.31***), neck thickness (r = -0.21**), and dry matter (r = -0.29***) ( Table 4 ). The correlation between plant height and leaf length was highly significant and positive (r = 0.94***). Traits like leaf width, number of leaves, neck thickness, pseudostem length, pseudostem width, equatorial diameter, polar diameter, gross yield, and marketable yield were also found to be significantly positively correlated with plant height. A significant but very low positive correlation was observed among leaf length and traits such as leaf width, number of leaves, pseudostem width, neck thickness, polar diameter, average bulb weight, marketable yield, and gross yield. Leaf width was found to be significantly positively correlated with number of leaves, pseudostem length, pseudostem width, neck thickness, average bulb weight, and days to maturity. A significant positive correlation was recorded among number of leaves and traits like pseudostem width, neck thickness, days to maturity, and gross yield. Pseudostem length was found to be significantly negatively correlated with equatorial diameter (r = -0.17*), whereas pseudostem width was observed to have a significant positive correlation with neck thickness and days to maturity. Neck thickness and days to maturity were recorded to have a significant positive correlation among them. Equatorial diameter and polar diameter of bulb were significantly positively correlated with average bulb weight, gross yield, and marketable yield. A negative correlation was also found among polar diameter and TSS (r = -0.23**). The average bulb weight showed a significant positive correlation with traits like marketable yield (r = 0.57**) and gross yield (r = 0.56**), whereas it was found to be negatively correlated with TSS (r = -0.21**). A significant positive correlation was observed among TSS and dry matter content (r = 0.35**), whereas TSS was found to be negatively correlated with marketable yield (r = -0.23**) and gross yield (r = -0.30**). The bolting percentage was also recorded as significantly negatively correlated with marketable yield (r = -0.27**) and gross yield (r = -0.28**).

For screening and biochemical characterization at different intervals under controlled conditions, 20 genotypes were selected. Genotypes were selected based on their phenotypic response to SLB under artificial inoculated field conditions. The results pertaining to the screening and biochemical characterization at different day intervals under controlled conditions are presented in Table 5 . The leaves showed initial symptoms of SLB after 7 days, and the first observation was recorded at 10th DAI. The lowest PDI was recorded in genotype SBO113 (4.73%), whereas the highest was observed in SBO131 (11.56%) at 10 DAI. Similarly at 20 DAI, genotype SBO113 (10.66%) followed by SBO097 (11.15%) and SBO116 (13.28%) showed minimum values for PDI, whereas the maximum PDI was recorded in SBO131 (30.51%) followed by SBO100 (25.33%). At 30 DAI, genotypes SBO113 (22.86%) followed by SBO097 (25.12%) exhibited the lowest PDI values, whereas SBO131 (62.22%) followed by SBO132 (59.54%) was recorded with the highest PDI. Biochemical compounds were also estimated before inoculation and after inoculation at the 10-day interval up to 30 days ( Table 5 ). The accumulation of different biochemical compounds showed an increasing trend and was found highest at 20 days except protein content ( Figure 7 ). Protein content was highest at the initial stage, but it kept on decreasing after inoculation. Before inoculation, the total foliar phenol ranged from 0.96 to 1.45 mg GAE/gm fw. The maximum accumulation of total foliar phenol content was recorded in genotype SBO023 (1.45), and the minimum was observed in SBO112 (0.96). Pyruvic acid in leaves was estimated highest in SBO116 (2.68 μmol/gm fw), whereas the lowest was recorded in SBO112 (1.05 μmol/gm fw). The catalase activity was found to be maximum in SBO113 (2.18 μmol H₂O₂/min/mg Prot.), whereas the minimum was observed in SBO134 (0.61 μmol H₂O₂/min/mg Prot.). Peroxidase activity was recorded maximum in genotype SBO097 (7.76 μmol/min/mg Prot.), and minimum activity was observed in SBO077 (3.21 μmol/min/mg Prot.). The maximum total protein level in leaves was recorded in genotype SBO041 (5.60 mg/gm fw). After 10 days of inoculation, the maximum total foliar phenol was observed in SBO041, and the minimum was recorded in SBO100. The genotype SBO097 showed the highest accumulation of pyruvic acid (4.45) and peroxidase (9.58). Accumulation of pyruvic acid and peroxidase activity was found to be minimum in SBO132. Catalase activity and total protein were recorded maximum in SBO113, whereas the lowest values for catalase activity were recorded in SBO131, and for total protein, it was observed in SBO134. The genotype SBO113 showed the highest accumulation for total foliar phenol, catalase, and protein, whereas genotype SBO131 recorded the lowest accumulation of the same at 20 days. Also, concentrations of pyruvic acid and peroxidase were found maximum in genotype SBO097, whereas the minimum was observed in SBO132. Similarly at 30 DAI, genotype SBO113 was found to have the highest concentration of total foliar phenol, catalase, and total protein. The pyruvic acid and peroxidase accumulation was also recorded maximum in SBO097. The genotype SBO131 recorded the lowest concentration of total foliar phenol, catalase, and peroxidase, whereas genotypes SBO132 and SBO134 were observed with the lowest accumulation of pyruvic acid and total protein, respectively.

The Pearson correlation coefficient revealed a significant negative correlation between PDI and biochemical parameters during stage II ( Figure 8 ). A significant negative correlation was found between PDI and TFPC (r = -0.64**). On the other hand, a significant positive correlation was found among the biochemical parameters. TFPC exhibited a highly significant correlation with protein content (r = 0.82***) and catalase (r = 0.71***) in the positive direction. Pyruvic acid content was found to have a highly significant positive correlation with guaiacol peroxidase (r = 0.80**) and catalase (r = 0.75***). Catalase was found to be significantly positively correlated with guaiacol peroxidase (r = 0.58**) and protein content (r = 0.71**). The correlation between guaiacol peroxidase and protein content was found to be non-significant in this study.

During the seed-to-bulb stage (stage I), under artificially inoculated field conditions, 15 genotypes were selected as moderately resistant and the PDI ranged from 13.20 to 19.20 whereas five genotypes with a PDI range of 49.20–68.00 were selected. These genotypes, in bulb form, were again screened under controlled conditions to assess their resistance ability under the bulb-to-seed stage. It was observed that all the genotypes categorized as moderately resistant (MR) in the first stage were categorized as susceptible in the bulb-to-seed stage. PDI ranged from 22.86 to 35.27 in the genotypes categorized as MR earlier. In the susceptible category, the PDI range was 49.20–68.00 in the seed stage, and in the bulb stage, the PDI ranged from 51.20 to 62.22 ( Figure 9 ).