Anthony P. Keinath

Genetic diversity among watermelon (Citrullus lanatus and Citrullus colocynthis) accessions

Genetic diversity was estimated among 42 U.S. PlantIntroduction (PI) accessions of the genusCitrullus (of these, 34 PIs are reported tohave disease resistance), and 5 watermelon cultivars, using 30RAPD primers. These primers produced 662 RAPD markers that could berated with high confidence. Based on these markers, geneticsimilarity coefficients were calculated and a dendrogram wasconstructed using the unweighted pair-group method witharithmetic average (UPGMA). The analysis delineated threemajor clusters. The first cluster consisted of a group of fivewatermelon cultivars, a group of C.lanatus var. lanatusaccessions, and a group of C.lanatus var. lanatusaccessions that contained some C.lanatus var. citroidesgenes. The second cluster consisted of the C.lanatus var. citroidesaccessions, while the third cluster consisted of theC. colocynthis accessions.The two C. lanatus clustersdifferentiated from each other and from the C.colocynthis cluster at the level of 58.8%and 38.9% genetic similarity, respectively. Assessment ofgenetic diversity among accessions that have been reported to havedisease resistance indicated that resistance to either anthracnose,downy mildew, powdery mildew, or watermelon mosaic virus is foundamong all major groups of Citrullus PIs.Additionally, resistance to gummy stem blight or Fusarium wilt mayexist among C. lanatus var.citroides PIs. This study demonstrates thatmolecular markers can be useful in assessing genetic diversity, andin sorting Citrullus PIs into phylogeneticgroups prior to their evaluation for disease or pestresistance.

Evaluation of fungicides for prevention and management of powdery mildew on watermelon

Powdery mildew of watermelon, caused by Podosphaera (sect. Sphaerotheca) xanthii, has been present throughout the southern United States since at least 1999. This study evaluated the efficacy of contact and systemic fungicides and a biofungicide (Bacillus subtilis) for preventing powdery mildew and curative applications of systemic fungicides for managing powdery mildew. The systemic fungicides azoxystrobin, pyraclostrobin, and myclobutanil were applied preventatively in alternation with mancozeb or B. subtilis during the first part of the season. In other treatments, mancozeb was applied preventively and the systemic fungicides myclobutanil, benomyl, and azoxystrobin were added curatively after powdery mildew was detected. Alternating preventative applications of mancozeb with azoxystrobin was one of the most effective fungicide combinations tested to prevent and manage this disease. Curative applications of systemic fungicides generally were less effective than preventative applications. Curative treatments that included tank mixes of fungicides required a greater amount of fungicide material than any preventative treatment. In both years, four and five applications were made with the Melcast melon disease forecaster and a weekly schedule, respectively, before powdery mildew was detected. Mean weight of individual fruit from nonsprayed plots was significantly lower than mean weight of fruit across all sprayed plots (P<0.003). Powdery mildew decreased yields significantly (R2=61.5, P=0.001 for the linear regression of area under the disease progress curve with weight of marketable fruit over both years), but no fungicide treatment consistently increased marketable yield compared with the nonsprayed control.

Simultaneous Detection of Acidovorax avenae subsp. citrulli and Didymella bryoniae in Cucurbit Seedlots Using Magnetic Capture Hybridization and Real-Time Polymerase Chain Reaction

To improve the simultaneous detection of two pathogens in cucurbit seed, a combination of magnetic capture hybridization (MCH) and multiplex real-time polymerase chain reaction (PCR) was developed. Single-stranded DNA hybridization capture probes targeting DNA of Acidovorax avenae subsp. citrulli, causal agent of bacterial fruit blotch, and Didymella bryoniae, causal agent of gummy stem blight, were covalently attached to magnetic particles and used to selectively concentrate template DNA from cucurbit seed samples. Sequestered template DNAs were subsequently amplified by multiplex real-time PCR using pathogen-specific TaqMan PCR assays. The MCH multiplex real-time PCR assay displayed a detection threshold of A. avenae subsp. citrulli at 10 CFU/ml and D. bryoniae at 10(5) conidia/ml in mixtures of pure cultures of the two pathogens, which was 10-fold more sensitive than the direct real-time PCR assays for the two pathogens separately. Although the direct real-time PCR assay displayed a detection threshold for A. avenae subsp. citrulli DNA of 100 fg/microl in 25% (1/4 samples) of the samples assayed, MCH real-time PCR demonstrated 100% detection frequency (4/4 samples) at the same DNA concentration. MCH did not improve detection sensitivity for D. bryoniae relative to direct real-time PCR using conidial suspensions or seed washes from D. bryoniae-infested cucurbit seed. However, MCH real-time PCR facilitated detection of both target pathogens in watermelon and melon seed samples (n = 5,000 seeds/sample) in which 0.02% of the seed were infested with A. avenae subsp. citrulli and 0.02% were infested with D. bryoniae.

Effect of Incorporation of Brassica spp. Residues on Population Densities of Soilborne Microorganisms and on Damping-off and Fusarium Wilt of Watermelon

Incorporating Brassica spp. residue to reduce populations of soilborne fungi and manage damping-off and Fusarium wilt of watermelon (Citrullus lanatus var. lanatus) was studied in two field experiments. Treatments included incorporating flowering Brassica napus cv. Dwarf Essex canola or B. juncea cv. Cutlass mustard and laying black polyethylene mulch at incorporation or 1 month after incorporation, methyl bromide, and a nontreated control. In both years, glucosinolates were identified and quantified in the shoots and roots of the flowering plants. In both years, the total concentration of glucosinolates incorporated per square meter was significantly higher for B. juncea than for B. napus. Isothiocyanates were inconsistently detected in the amended soils and none were detected more than 12 days postincorporation. After incorporation in 2004 and 2005, amended plots had higher populations of Fusarium oxysporum and Pythium spp. than the methyl bromide treatment, and in some treatments, populations were higher than in the control. Fluorescent Pseudomonas spp. were not suppressed in amended soils, and their populations were significantly higher in some amended treatments than those in methyl bromide-treated soils or nontreated control soils. Incidence of damping-off and severity of Fusarium wilt on seedless watermelon cv. Tri-X 313, which is susceptible to Fusarium wilt, were not consistently lower in brassica-amended soils or methyl bromide-treated plots than in nontreated control plots. Therefore, under spring conditions and methods used in this study, neither biofumigation nor methyl bromide fumigation in coastal South Carolina was an effective disease management tool for two soilborne pathogens of watermelon.

Sensitivity of Populations of Phytophthora capsici from South Carolina to Mefenoxam, Dimethomorph, Zoxamide, and Cymoxanil

In summer and fall 2003, Phytophthora blight and crown rot, caused by Phytophthora capsici, was found in three fields each of summer squash and pepper on three farms in two counties in South Carolina. Although this disease had been confirmed previously in the state, five of these outbreaks were in fields thought to be free of P. capsici. The objectives of this study were to determine whether isolates of P. capsici in South Carolina were sensitive to mefenoxam and to determine baseline sensitivities to dimethomorph, zoxamide, and cymoxanil, fungicides recently registered to control Phytophthora blight. Of 120 isolates tested for sensitivity to mefenoxam at 100 mg/liter, 8 isolates were resistant (relative colony diameter [RCD] > 90% of nonamended control), 60 isolates were sensitive (RCD < 30%), and 52 isolates were intermediately sensitive. Only sensitive isolates were found in two fields in which no mefenoxam-containing fungicides had ever been used. Intermediately sensitive or resistant isolates were found in the four fields in which mefenoxam had been applied previously. In all, 15 to 61 isolates were tested for sensitivity to dimethomorph, zoxamide, and cymoxanil. The concentrations at which RCD, percent cyst germination, and relative zoospore production were reduced to 50% (EC50 values) for mycelial growth were 0.19 ± 0.02 (± standard deviation) mg/liter for dimethomorph, 0.50 ± 0.50 mg/liter for zoxamide, and mostly >50 mg/liter for cymoxanil. EC50 values for zoospore cyst germination were 0.07 ± 0.02 mg/liter for dimethomorph and >50 mg/liter for cymoxanil. EC50 values for zoospore production were 0.63 ± 0.42 mg/liter for dimethomorph, 0.47 ± 0.51 mg/liter for zoxamide, and <50 mg/liter for cymoxanil. Sensitivity values obtained in this South Carolina study can be used as a comparative baseline to monitor shifts in sensitivity to the fungicides mefenoxam, dimethomorph, zoxamide, and cymoxanil in populations of P. capsici.

Soil amendment with cabbage residue and crop rotation to reduce gummy stem blight and increase growth and yield of watermelon

Three cropping sequences, watermelon-cabbage-soil solarization-watermelon, watermelon-wheat-soybean-watermelon, and 3 years watermelon, were evaluated for the effect on gummy stem blight and watermelon fruit yield. The 3-year experiment was conducted three times, first in the fall of 1991 through the summer of 1993, then twice in the fall of 1993 through the summer of 1995, with one of these being a second cycle in the same plots as the first test. Cabbage-solarization (P ≤ 0.07) and the wheat-soybean double crop (P 6.35 kg) and total healthy fruit compared with the nonsolarized treatments. Marketable yields of cv. Charleston Gray were 59.4, 35.4, and 39.4 kg of watermelon per 15 m of row in plots cropped to cabbage-solarization, watermelon, and wheat-soybean, respectively, the preceding year. Yield of watermelons weighing <6.35 kg was greater (P ≤ 0.04) after cabbage amendment and solarization than after the other two cropping sequences for both experiments in 1995. In 1994, thermotolerant fungi increased in solarized plots amended with cabbage residue and remained significantly (P < 0.01) higher in these plots than in nonsolarized plots the following year. Growth promotion and fruit yields in amended, solarized plots were not associated with changes in soil mineral nutrients, plant parasitic nematodes, or soil temperatures. Incorporating cabbage residue into mulched soil can increase growth and yield of watermelon.

Resistance to Benomyl and Thiophanate-methyl in Didymella bryoniae from South Carolina and New York

An initial collection of 7 isolates of Didymella bryoniae were grown on media amended with 0, 1, 3.2, 10, 31.2, or 100 mg benomyl per liter. Four isolates grew at all five concentrations of benomyl, but the other 3 isolates did not grow at concentrations > 1 mg/liter. Colony diameter of the four resistant isolates was reduced by 50% at 33.1 mg benomyl per liter, relative to growth on nonamended medium. Of 394 isolates tested, 182 isolates were resistant to benomyl; 178 of these resistant isolates were from South Carolina, 1 was from New York, and 3 were from Florida. Of 196 isolates grown on medium amended with 100 mg/liter thiophanate-methyl, 95 were sensitive and 101 were resistant. Essentially all isolates that were resistant to benomyl were resistant to thiophanate-methyl. In greenhouse tests, watermelon plants were sprayed with 0, 1.5, 15, 150, or 1,500 mg benomyl per liter and inoculated 1 day later with either a sensitive or a resistant isolate of D. bryoniae. Relative percent leaf area diseased was greater (P≤0.02) for the resistant isolate than for the sensitive isolate at ≥1.5 mg benomyl per liter. The occurrence of pathogenic, benzimidazole-resistant D. bryoniae in the eastern United States may reduce the effectiveness of benzimidazole fungicides for gummy stem blight management.

Internal Transcribed Spacer Regions 1 and 2 and Random Amplified Polymorphic DNA Analysis of Didymella bryoniae and Related Phoma Species Isolated from Cucurbits.

ABSTRACT Didymella bryoniae (anamorph Phoma cucurbitacearum) is the causal agent of gummy stem blight, although other Phoma species are often isolated from cucurbit plants exhibiting symptoms of the disease. The molecular and phylogenetic relationships between D. bryoniae and these Phoma species are unknown. Isolates of D. bryoniae and Phoma obtained from cucurbits grown at various geographical locations in the United States were subjected to random amplified polymorphic DNA (RAPD) analysis and internal transcribed spacer (ITS) sequence analysis (ITS-1 and ITS-2) to determine the molecular and phylogenetic relationships within and between these fungi. Using RAPD fingerprinting, 59 isolates were placed into four phylogenetic groups, designated RAPD group (RG) I, RG II, RG III, and RG IV. D. bryoniae isolates clustered in either RG I (33 isolates), RG II (12 isolates), or RG IV (one isolate), whereas all 13 Phoma isolates clustered to RG III. There was greater than 99% sequence identity in the ITS-1 and ITS-2 regions between isolates in RG I and RG II, whereas isolates in RG III, P. medicaginis ATCC 64481, and P. exigua ATCC 14728 clustered separately. On muskmelon seedlings, a subset of RG I isolates were highly virulent (mean disease severity was 71%), RG II and RG IV isolates were slightly virulent (mean disease severity was 4%), and RG III isolates were nonpathogenic (disease severity was 0% for all isolates). The ITS sequences indicate that RG I and RG II are both D. bryoniae, but RAPD fingerprints and pathogenicity indicate that they represent two different molecular and virulence subgroups.

Effect of Protectant Fungicide Application Schedules on Gummy Stem Blight Epidemics and Marketable Yield of Watermelon

The watermelon cultivar Royal Star was grown in fall 1996, spring and fall 1997, and spring 1998 and treated with 2.52 kg a.i./ha of the protectant fungicides mancozeb or chlorothalonil. Spray application schedules used in the experiments included two initiation times, sprays every 7, 10, or 14 days, and two termination times. Severity of gummy stem blight, caused by Didymella bryoniae, was high in fall 1996 and 1997, moderate in spring 1997, and low in spring 1998. In each experiment, fungicide applications reduced the area under the disease progress curve (AUDPC), percent leaf area diseased at the end of the season, time to reach 25% disease severity, or all three disease measurements relative to the nonsprayed control. In the 1996 and both 1997 experiments, 7-day spray intervals provided more effective disease control than 14-day intervals. In general, initiating sprays early reduced gummy stem blight compared with delayed sprays, but spray termination times did not affect AUDPC. In both fall experiments, fungicide applications increased yield of marketable fruit over the no-fungicide control. A 7-day spray interval increased marketable weight compared with a 14-day interval only in fall 1996. Weight of marketable fruit did not differ among treatments in either spring experiment. Differences in disease control among treatments often did not correspond to differences in marketable yields.

AFLP analysis of a worldwide collection of Didymella bryoniae

Didymella bryoniae (anamorph Phoma cucurbitacearum) is an ascomycete that causes gummy stem blight, a foliar disease that occurs on cucurbits in greenhouses and fields throughout the world. In a previous study using RAPD analysis, little genetic diversity was found among isolates of D. bryoniae from New York and South Carolina, USA. Here we report the use of amplified fragment length polymorphism (AFLP) analysis to assess the genetic variation within a worldwide collection of D. bryoniae, 102 field and greenhouse isolates from ten states in the USA (California, Delaware, Florida, Georgia, Indiana, Maryland, Michigan, Oklahoma, South Carolina, and Texas) and seven other countries (Australia, Canada, China, Greece, Israel, Sweden, and The Netherlands) were examined. Seven different AFLP primer-pair combinations generated 450 bands, of which 134 were polymorphic (30%). Using cluster analysis, two groups and a total of seven subgroups were delineated. Representative isolates varied in their virulence on muskmelon and watermelon seedlings, but the degree of virulence was not strongly associated with AFLP groupings. However, isolates from the northern USA grouped separately from isolates originating from the southern USA.