ResearchFree Access icon

Optimizing a Seedling Screening Method for Anthracnose Resistance in Watermelon

    Authors and Affiliations
    • Edgar Correa1 2
    • Kevin Crosby1
    • Subas Malla1 2
    1. 1Department of Horticultural Sciences, Texas A&M University, College Station, TX 77840
    2. 2Texas A&M AgriLife Research and Extension Center, Uvalde, TX 78801

    Published Online:


    Among three races of Colletotrichum orbiculare, which cause anthracnose of cucurbits, screening for race 2 resistance was studied under greenhouse conditions at various inoculum concentrations, and plants were rated on different days postinoculation (DPI). The objectives of this study were optimizing inoculum concentration and phenotyping DPI for seedling resistance. Five inoculum concentrations were compared (2.5 × 104, 5 × 104, 1 × 105, 2.5 × 105, and 5 × 105 conidial spores/ml). Four watermelon genotypes (‘Black Diamond’, ‘Charleston Gray 133’, PI 543210, and PI 189225) and two cucumber genotypes (‘Marketer’ and ‘H19 Little Leaf’) were evaluated. Disease was recorded on the percentage of cotyledon area lesion, severity of hypocotyl lesion (SHL), severity of petiole lesion (SPL), and percentage of leaf area lesion (PLL), as well as a disease index (INDX) from 5 to 14 DPI. There was a significant difference among genotypes and inoculum concentrations. The resistant PI 189225 was significantly different (P < 0.05) from the highly susceptible PI 543210. Inoculum 1 × 105 spores/ml was at par with 5 × 105 and 2.5 × 105 but significantly different from 5 × 104 and 2.5 × 104 for AUDPS PLL, AUDPS INDX, AUDPS SPL, and AUDPS SHL. Inoculum at 1 × 105 spores/ml was optimal to differentiate germplasm. A genotype plus genotype-by-environment biplot showed that PLL was a representative rating. A single PLL rating on 9 DPI would optimize resources for screening a large set of germplasm for anthracnose resistance in a watermelon breeding program.

    Cucurbit anthracnose caused by Colletotrichum orbiculare (Berk. & Mont) Arx syn. C. lagenari has been a major fungal pathogen on watermelon (Citrullus lanatus [Thunb.] Matsum. & Nakai) since the early 20th century (Gardner 1918; Meier 1920; Orton 1917). The disease occurs throughout the eastern and southern United States (Florida, Georgia, Alabama, Texas, South Carolina, and North Carolina), where frequent warm temperatures (20 to 32°C), rainy conditions, and high humidity favor the spread of spores and spore germination (Monroe et al. 1997; Norton et al. 1995). The fungus is hemibiotrophic and can infect the whole plant (leaves, vines, and fruits) of cucurbits. Symptoms associated with anthracnose are black to brown irregular spots found on the leaves, oval tan lesions on the stem, and brown sunken lesions on fruits. A resistant host usually shows chlorotic hypersensitivity reaction on the leaf. Fruit lesions are viewed as blemishes that downgrade watermelon fruits to unmarketable quality. Infested fields have lower marketable yields, resulting in a monetary loss to growers. Keinath (2018) showed how detrimental unchecked anthracnose can be. A 10 t/ha decrease of yield was observed between a water control and several fungicide treatments.

    As a hemibiotrophic fungus, C. orbiculare has two nutritional stages. The biotrophic stage produces specialized intracellular bulbous-shaped hyphae to obtain nutrients, secretes effector proteins to suppress the plant immune system, and produces sexual acervuli (De Silva et al. 2017; Kubo et al. 2016). During the necrotrophic stage, hyphae are narrow; they grow rapidly and break down the host cell wall to consume nutrients (De Silva et al. 2017). Currently, three races of the anthracnose fungus are of concern to watermelon production. Initial studies compared different isolates on ‘Congo,’ ‘Charleston Gray,’ ‘Fairfax,’ ‘Garrison,’ and ‘New Hampshire Midget’ (Goode and Winstead 1957). All genotypes were susceptible to one isolate, which was established as race 2.

    Winstead et al. (1959) evaluated two- to three-true-leaf-stage watermelon seedlings at 2.5 × 103 to 1 × 104 conidial spores/ml to test anthracnose resistance. In 1980, two publications reported several plant introductions (PIs) with race 2 anthracnose resistance: PI 270550, PI 326515, PI 271775, PI 271779, PI 203551, PI 299379, and PI 189225 (Sowell et al. 1980; Suvanprakorn and Norton 1980). Seedlings were inoculated with 2 × 104 conidial spores/ml and rated at 7 days postinoculation (DPI). A later study evaluated several C. orbiculare isolates from across the United States and determined three major vegetative compatibility groups (Wasilwa et al. 1993). The study inoculated seedlings at cotyledon stage 4 days after emergence with 8 × 104 spores/ml and rated them from 3 to 9 DPI. Boyhan et al. (1994) verified the resistance of PIs and identified another race 2 resistant accession (PI 512385), using 5 × 104 spores/ml and evaluation at 14 DPI. Recently, Keinath (2015) identified the race of new isolates with 3-week-old seedlings inoculated with 5 × 105 spores/ml and rated them at 14 DPI. Evaluation of 1,408 USDA-ARS Citrullus PIs for race 2 anthracnose resistance was conducted with 1 × 105 spores/ml on 3-week-old seedlings and rated at 7 DPI (Patel 2019).

    Inconsistencies in inoculum concentration, seedling age, and incubation time were found in previous studies. Additionally, responses of PIs may not be reliable if the inoculum concentration or time postinoculation are not sufficient. This study was conducted to optimize a protocol for seedling screening for anthracnose resistance in watermelon. The objectives of the study were to optimize inoculum rate, disease rating, and incubation time for disease rating.

    Material and Methods

    Plant material.

    Seedlings to be inoculated with C. orbiculare race 2 and a water control were grown for 3 weeks in 24-cell trays (Growers Solutions, Cookeville, TN) during the spring and summer of 2019 at Uvalde, TX (29.2097°N, 99.7862°W). The seeds were sown on April 22 and 24, 2019, into Proline C/B growing mix (Jolly Gardener, Poland Spring, ME) and grown in the greenhouse at 25 to 30°C. They were inoculated on May 13 and 15, 2019.

    The genotypes evaluated included an anthracnose disease differential set (Wasilwa et al. 1993), two cucumber varieties (‘Marketer’ and ‘H19 Little Leaf’), and two watermelon genotypes (‘Black Diamond’ and ‘Charleston Gray 133’). In addition, two watermelon PIs, the susceptible PI 543210 and the resistant PI 189225, were included in our tests. Seeds of Marketer, H19 Little Leaf, and Black Diamond were obtained from Eden Brothers (Arden, NC), and Charleston Gray 133 was obtained from Willhite Seed (Poolville, TX). Seeds of PI 543210 and PI 189225 were obtained from the USDA-GRIN database and increased locally by manual selfing in the greenhouse.

    Culture and inoculation.

    The isolate WmColl4 of C. orbiculare race 2 was originally recovered from foliage of the commercial hybrid watermelon ‘Mardi Gras’ in South Carolina (Keinath 2015). Inoculum was grown on half-strength potato dextrose agar (BD Biosciences, Franklin Lakes, NJ) for 7 to 14 days. Spores were harvested via scraping with distilled water and poured through cheesecloth into a beaker. A hemocytometer was used to determine the initial concentration of the inoculum, which was diluted with distilled water containing Tween 80 (10 µl in 100 ml) to the final inoculum concentrations of 2.5 × 104, 5 × 104, 1 × 105, 2.5 × 105, and 5 × 105 spores/ml. Two- to four-true-leaf seedlings were inoculated in the greenhouse with a CO2 sprayer at 30 PSI (discharge rate of 500 ml/min). Inoculum was sprayed to cover the leaves for 5 s duration across the seedling tray without runoff. Inoculated seedlings were immediately moved into a humidity chamber and kept for 48 h at approximately 100% relative humidity at 22 to 24°C without light. Trays were moved to the greenhouse under natural light, with a day temperature of 30°C and night temperature of 25°C, until disease rating.

    Disease rating.

    Seedlings were rated daily from 5 to 14 DPI. Each seedling was given an individual rating based on different parameters: percentage of leaf area lesion (PLL) and percentage of cotyledon area lesion (PCL), with 0% indicating no lesion and 100% indicating death of the true leaf or cotyledon. The severity (0 to 6) of hypocotyl lesion (SHL) was rated with 0 indicating no lesion and 6 indicating the hypocotyl was completely dried, and severity (0 to 5) of petiole lesion (SPL) was rated with 0 indicating no lesion on the petiole and 5 indicating the petiole was completely dry and dead (Figs. 1 and 2). There was also a whole plant rating that was given to the seedlings based on a disease index (INDX) developed by Patel (2019). The INDX was a cumulative rating given on incidence and severity at different plant parts based on their importance. Parameters measured as a percentage were rated in increments of 5%. The number of true leaves was accounted for as well. Individual true leaves from each seedling were rated for PLL, and the average of all the true leaves on the plant was used as the PLL for that plant. A genotype’s rating was based on the mean of four plants (one plot). The area under the disease progress stair (AUDPS; Simko and Piepho 2012) was calculated on R software (R Core Team 2014; version 3.6.2) using the ‘agricolae’ (de Mendiburu 2014) package function ‘audps’ for PLL, PCL, SHL, SPL, and INDX.

    FIGURE 1

    FIGURE 1 Colletotrichum orbiculare on watermelon hypocotyl rating scale for severity: 0 = no lesion observed; 1 = water-soaked lesion appears faintly; 2 = water-soaked lesion is more apparent, a bit larger; 3 = lesion is orangish and concaves inward; 4 = lesion is highly concaved inward and starts to spread around stem; 5 = the whole stem is infected and the plant starts to bend over; and 6 = the vascular system of the plant is exposed and the stem begins to dry.

    Download as PowerPoint
    FIGURE 2

    FIGURE 2 Colletotrichum orbiculare on watermelon petiole/stem rating scale for severity: 0 = no lesion observed; 1 = water-soaked lesion; 2 = small necrotic spot; 3 = elongated longer brownish black lesion; 4 = lesion is highly concaved inward and starts to spread around the stem; and 5 = the whole stem is infected and the plant starts to bend over.

    Download as PowerPoint

    Experimental design.

    Treatments were arranged in a factorial arrangement of a randomized complete block design with two replications. The first factor, inoculum, had six levels (water control, 2.5 × 104, 5 × 104, 1 × 105, 2.5 × 105, and 5 × 105 spores/ml). The second factor, genotype, had six levels (Marketer, H19 Little Leaf, Charleston Gray 133, Black Diamond, PI 189225, and PI 543210). The third factor, seeding date, had two levels (April 22 and 24, 2019). Disease ratings for seedlings were the dependent variables. There were four plants per replication, with a total of 16 plants evaluated. A single plant was considered an experimental unit, and disease rating was an average over four plants, which was considered as the plot.

    Data analysis.

    AUDPS was calculated for the various traits. Because the water control level for inoculum did not show disease symptoms, data were not used for analysis. Due to immune and highly resistant responses from cucumber lines, H19 Little Leaf and Marketer, and nonparametric disease rating for SHL and SPL traits, there was a bimodal distribution for AUDPS SHL and AUDPS SPL. After removing cucumber lines from the dataset, Kolmogorov–Smirnov and/or Shapiro–Wilk tests failed to reject the null hypothesis for the normality test. Hence, parametric statistical analyses were conducted only on watermelon genotypes. A mixed-model analysis of variance (ANOVA) for the data used JMP (Pro 14.0.0) with seeding date and replication nested within seeding date as random factors and genotypes and inoculum concentrations as fixed factors. The least significant difference was used to compare factor means. Correlation coefficients between traits were calculated using JMP (Pro 14.0.0). Genotype plus genotype-by-environment (GGE) biplot analysis (Yan and Tinker 2006) used the ‘GGEBiplotGUI’ package (Frutos et al. 2014) with options scaling = standard, centered by = tester-centered G+GE, and SVP = symmetrical (Yan and Kang 2003) on R software (R Core Team 2014; version 3.6.2).


    Disease symptoms.

    On the most susceptible line (PI 543210), seedlings from inoculated treatments showed disease symptoms starting from 3 DPI and the complete death on 10 DPI, whereas seedlings from the water control treatment did not show disease symptoms during the trial period. Inoculation of seedlings caused water-soaked lesions that turned brown to black on the hypocotyl, cotyledons, leaves, and stems. Faint water-soaked lesions were visible on the true leaves of susceptible PI 543210 as early as 3 DPI at the highest inoculum concentrations, 5 × 105 spores/ml, but not resistant PI 189225 (Fig. 3). Disease parameters were recorded at 5 DPI when susceptible Black Diamond and PI 189225 had initial lesions at the lower inoculum concentrations. In cucumber cultivars, H19 Little Leaf showed no anthracnose lesions early on, and Marketer showed local hypersensitive reaction, became chlorotic, and started to dry out, but it had no necrotic and systemic lesions as on the watermelon cultivars. At 1 × 105 spores/ml, a general upward trend for SHL and PCL was observed (Fig. 4).

    FIGURE 3

    FIGURE 3 A, PI 543210 (left) and PI 189225 (right) seedlings at 3 days postinoculation (DPI) with Colletotrichum orbiculare race 2. B, Overhead view of seedlings at 8 DPI with 1 × 105 spores/ml, from left to right: ‘Black Diamond’, ‘Marketer’, ‘H19 Little Leaf’, PI 543210, PI 189225, and ‘Charleston Gray 133’.

    Download as PowerPoint
    FIGURE 4

    FIGURE 4 Progression with time across days postinoculation (DPI) of watermelon (‘Black Diamond’, PI 543210, PI 189225, ‘Charleston Gray 133’) and cucumber genotypes (‘Marketer’, ‘H19 Little Leaf’) for A, percentage of leaf area lesion (PLL); B, disease index (INDX); C, severity of petiole lesion (SPL); D, severity of hypocotyl lesion (SHL); and E, percentage of cotyledon area lesion (PCL) evaluated at 1 × 105 spores/ml.

    Download as PowerPoint

    ANOVA for AUDPS and disease progress.

    ANOVA showed a highly significant (P < 0.01) difference for genotype and inoculum concentration but not for genotype × inoculum interaction for all traits (Table 1). PI 543210 had the highest disease ratings for all traits, indicating the most susceptible genotype, and was significantly (P < 0.05) different from other genotypes for all traits except AUDPS SHL, where it was at par with Black Diamond and Charleston Gray 133 (Table 2). In contrast, PI 189225 was the most resistant genotype, as exhibited by the lowest disease severity for all traits except AUDPS SHL, showing the highest disease severity. The resistant genotype was significantly different (P < 0.05) from the remaining genotypes for AUDPS PLL, AUDPS INDX, AUDPS SPL, and AUDPS SHL. On inoculum factor, the level 1 × 105 had disease severity at par with 5 × 105 and 2.5 × 105 for all traits, but it was significantly (P < 0.05) different from levels 5 × 104 and 2.5 × 104 for AUDPS PLL, AUDPS INDX, AUDPS SPL, and AUDPS SHL (Table 2). The lowest inoculum level, 2.5 × 104, produced the least disease severity for all traits, and the disease ratings were significantly (P < 0.05) different from other levels for AUDPS PLL, AUDPS INDX, and AUDPS SPL.

    TABLE 1 Analysis of variance of interaction between watermelon genotypes and inoculum concentration of Colletotrichum orbiculare race 2z

    TABLE 2 Effect of Colletotrichum orbiculare race 2 on genotype and spore concentration on progress of anthracnose in watermelonx

    The trend for cucumber genotypes showed a different progression than that of the watermelon genotypes. Extremely low to no disease severity was observed for H19 Little Leaf, based on AUDPS PLL, AUDPS INDX, AUDPS SPL, and AUDPS SHL at 1 × 105 spores/ml (Fig. 4). AUDPS INDX, SPL, and SHL for Marketer were near zero at all concentrations. Marketer had a lower disease severity for AUDPS PLL, compared with all watermelon genotypes, but it did increase across concentrations similar to the resistant watermelon genotype PI 189225 (Figs. 4 and 5).

    FIGURE 5

    FIGURE 5 Progression of watermelon and cucumber genotypes’ percentage of leaf area lesion (PLL) from anthracnose at A, 5.0 × 105; B, 2.5 × 105; C, 1.0 × 105; D, 5.0 × 104; and E, 2.5 × 104 spores/ml.

    Download as PowerPoint

    GGE biplot analysis and correlation.

    GGE biplot analysis uses principal component analysis and displays results on graphs. The length of the vector indicates discriminatory ability of the traits. The longer the vector length is for the trait, the more discriminatory the trait is. The closer the trait vector is to the abscissa, the more representative the trait will be. Discriminativeness versus representativeness GGE biplot graphs for AUDPS traits are shown (Fig. 6). Genotypes were entries and AUDPS traits were testers for GGE biplot analysis. The longest vector length, AUDPS PLL, was the most discriminating trait. AUDPS SPL and AUDPS SHL had the lowest vector length, indicating that they were the least discriminating traits. AUDPS INDX had the lowest cosine angle with the abscissa, indicating that the trait was the most representative among the five traits. The angle between two trait vectors represents a correlation between traits. If the angle is small (acute), the traits will have a high positive correlation. The vector angle was lower between AUDPS PLL and AUDPS INDX compared with AUDPS PLL and AUDPS PCL, indicating a high correlation between the former two traits. A significant and high correlation (r = 1.0, P < 0.01) was observed between AUDPS PLL and AUDPS INDX (Table 3). A positive correlation was observed among AUDPS PLL, AUDPS INDX, AUDPS PCL, and AUDPS SPL. In contrast, AUDPS SHL was negatively correlated with all traits.

    FIGURE 6

    FIGURE 6 Genotype plus genotype-by-environment biplot showing discrimitiveness versus representativeness view of four genotypes for area under the disease progress stairs (AUDPS) of the following traits: percentage of leaf area lesion (PLL), disease index (INDX), severity of petiole lesion (SPL), severity of hypocotyl lesion (SHL), and percentage of cotyledon area lesion (PCL).

    Download as PowerPoint

    TABLE 3 Pearson correlation among AUDPS traits (N = 4) averaged across all factors for Colletotrichum orbiculare race 2 in watermelonz

    The disease rating on each DPI from the most discriminative trait, AUDPS PLL, was used for GGE biplot analysis. A discriminativeness versus representativeness GGE biplot for PLL on different DPIs was estimated in Figure 7. PLL on DPI 8 and DPI 9 had the longest vector length, indicating that they were the most discriminating. The vector PLL DPI 9 had the smallest vector angle with the abscissa and was the most representative among different PLL DPIs. The correlation between AUDPS PLL and different PLL DPIs was significant (P < 0.001) and high (Table 4). The highest correlation (r = 1.0) was observed between AUDPS PLL and PLL DPI 9, followed by AUDPS PLL and PLL DPI 10 (r = 0.99).

    FIGURE 7

    FIGURE 7 Genotype plus genotype-by-environment biplot showing discrimitiveness versus representativeness view of four genotypes for percentage of leaf area lesion on different days postinoculation (DPI).

    Download as PowerPoint

    TABLE 4 Correlation between AUDPS PLL and PLL averaged across all factors at different days postinoculation (DPI) (from 5 to 14 DPI) (N = 4)z


    Based on our preliminary test, we independently evaluated seedling damage from anthracnose on cotyledons, hypocotyls, petioles, and true leaves. True leaf petiole infection would expand and cut off water and nutrients to the true leaves, leading to their death. It is important to avoid inoculum runoff during inoculation to minimize the confounding effect of severe infection on hypocotyl and petiole. The traits evaluated did not show an equal level of importance in differentiating resistant versus susceptible genotypes, or seedlings may be exhibiting differential responses on different parts of the seedling.

    The GGE biplot displayed that AUDPS PLL was the most discriminating trait. ANOVA results also showed that the trait was able to differentiate genotypes into four groups. The resistant genotype PI 189225 had the lowest disease severity, whereas the susceptible genotypes Charleston Gray 133 and Black Diamond differed from each other as well as from the most susceptible PI 543210. Martyn and McLaughlin (1983) working with Fusarium wilt (Fusarium oxysporum f. sp. niveum) previously classified watermelon into disease resistance categories (resistant, moderately resistant, moderately susceptible, and highly susceptible). Our results indicate that the AUDPS PLL could be similarly used when evaluating seedlings for anthracnose resistance.

    The AUDPS PCL had a higher vector length on the GGE biplot; however, the trait could not significantly distinguish between resistant PI 189225 and susceptible line Charleston Gray 133. Those two genotypes may have cotyledon leaf resistance. During this study, we did not evaluate watermelon seedlings at the cotyledon stage (Wasilwa et al. 1993) or cotyledon and true leaf stage, similar to the one mentioned for downy mildew in broccoli (Farnham et al. 2001). Host plants may respond with different resistant genes after infection on different plant parts. In Fusarium head blight disease in wheat, there were five types of resistance: type I (incidence), type II (severity), type III (deoxynivalenol toxin), type IV (Fusarium-damaged kernel), and type V (yield) (Mesterházy 1995; Miller et al. 1985; Schroeder and Christensen 1963). Farnham et al. (2001) reported different downy mildew resistance for cotyledon and true leaf in broccoli. Future study on evaluating disease resistance at cotyledon and seedling stages separately or in combination might elucidate if the resistance expressed on cotyledons and true leaves differs.

    AUDPS SHL showed a low discriminating ability on the GGE biplot. The trait divided genotypes into two groups, unlike other traits for which groupings were up to four. Interestingly, resistant genotype PI 189225 had the highest AUDPS SHL and was significantly different from the remaining genotypes. Seedling hypocotyl infection leads to death. It is important to have resistant genotypes exhibiting not only resistance for PLL, SPL, and PCL but also for SHL. The results indicate that the resistant seedling might die during the seedling and/or adult stages before maturity due to infection in the hypocotyl region. Our results agreed with Winstead et al. (1959), who reported the death of resistant seedlings before maturity due to infection on the stem at the soil line. They rescued resistant seedlings by grafting a resistant scion onto a rootstock. For disease resistant plant selection, it is important to select not only for true leaf resistance but also for hypocotyl resistance. Otherwise, the leaf resistant plant might require grafting. The results demonstrated that SHL is as important as PLL.

    Our results showed that disease severity varied among different inoculum rates. The inoculum rate 2.5 × 104 showed the least disease, whereas 5 × 105 had the highest disease. The inoculum rate 1 × 105 was significantly different from 5 × 104 and 2.5 × 104 but not from 2.5 × 105 and 5 × 105 for all traits except AUDPS PCL. When applying 1 × 105 spores/ml to seedlings, watermelon genotypes can be categorized into highly susceptible, moderately susceptible, and resistant (Fig. 5). The difference between highly resistant and moderately susceptible could be due to minor modifier genes involved in anthracnose susceptibility previously theorized (Love and Rhodes 1988). AUDPS PLL shows that disease progression for PI 543210 was faster than that of Charleston Gray 133 and Black Diamond, indicating infection and spread can occur at a quicker rate. Hence, 1 × 105 spores/ml seemed an optimum concentration to evaluate 3-week-old seedlings for resistance to anthracnose in a greenhouse. However, we observed reduced disease severity in the winter season compared with the summer season. Wasilwa et al. (1993) also reported the same trend. A preliminary test is recommended to optimize parameters for disease evaluation based on location and time of year.

    In a breeding program where many lines need to be screened, could the number of disease ratings be reduced? GGE biplot analysis showed that PLL rating on 9 DPI was the most representative. Correlation analysis showed that there was a high association between AUDPS PLL and PLL, for all DPI. Any single rating of PLL at 5 DPI to 14 DPI could replace AUDPS PLL. The correlation between AUDPS PLL and PLL on 9 DPI was the highest (r = 1.0). It can be inferred from our results that a single disease rating of PLL on 9 DPI could replace AUDPS PLL, similar to Bai and Shaner (1996), in which they also mentioned that a single rating could replace AUDPC rating for Fusarium head blight in wheat. A single rating would optimize resources when screening seedlings for anthracnose resistance.


    Inoculum concentration of 1 × 105 spores/ml was identified to be sufficient to differentiate seedlings for C. orbiculare race 2 resistance. Disease rating of PLL is a good measure for anthracnose severity rating when evaluating watermelon seedlings. To optimize resources in a breeding program, multiple disease severity ratings could be replaced with a single disease rating of PLL at 9 DPI. Different seedling true leaf stages were not evaluated but could be important in obtaining consistent results. The study identified a difference within susceptible genotypes, a higher level of severity and a moderate level. Cotyledon leaf necrosis might be different from resistance to stem, petiole, and true leaf necrosis. A future study may be needed to understand the relationship between anthracnose cotyledon leaf necrosis and true leaf necrosis in watermelon. A further look at the different sources of resistance for adult plant resistance and watermelon yield output would show the potential of resistance integration.


    We thank Filomena Hernandez at the Uvalde AgriLife Research and Extension Center for assistance with tray preparation, seeding, and maintenance of seedlings. We also thank Dr. Anthony Keinath (Clemson University, Clemson, SC) for providing Colletotrichum orbiculare race 2 isolate.

    The author(s) declare no conflict of interest.

    Literature Cited

    Current address for E. Correa: Hazera Seed USA Inc., 6601 Lyons Rd, Coconut Creek, FL 33073.

    Funding: Research was supported by USDA Hatch Project TEX09665.

    The author(s) declare no conflict of interest.