During the cultivation season between 2012 and 2013, five sweet persimmon orchards in the Gyeongnam province were visited. Three orchards were maintained with practical management methods (Daegok, Munsan, Yearee) and two orchards (Masan, Masan-2) were maintained with organic farming methods. Typical anthracnose disease symptoms were observed on twigs and fruit, and infected samples were collected and transported to the lab within three hours of collection. Disease lesions were cut from the surface of the fruit, with a depth less than 1 cm. The surface of each lesion was sterilized with 1% sodium hypochlorite followed by 70% ethanol and then the specimens were washed in sterilized distilled water three times. The specimens were dried for 15 min under a fume hood and were placed in water agar media (WA) at 28°C until spores developed. Conidia were observed using a microscope (Olympus BX15). From the five orchards, a total of 485 pathogens were isolated (Table 1).

Genomic DNA was extracted from fresh mycelium by the CTAB method (Ford et al., 2000). DNA from each isolate was amplified using the following species-specific primer pairs: Colg1 (5′-AACCCTTTGTGAACATACC-3′) and Colg2 (5′-CCCTCCGGATCCCAG-3′) (443-bp) for C. gloeosporioides; and Colg1 and CT 2 (5′-CTT TAA GGG CCT ACG TCA A-3′) (375-bp) for C. truncatum (Chen et al., 2006). The PCR conditions included an initial denaturation step at 94°C for 5 min, followed by 34 cycles of 94°C for 30 sec, 63°C for 30 s, and at 72°C for 1 min. A final extension step heated samples to 72°C for 6 min. The primer set, col2 (d5′-TTA CTA CGC AAA GGA GGC T-3′) and acut1 (d5′-CCG GAG GAA ACC AAA CTC TAT TTAC-3′) (462-bp) were used to detect C. acutatum, while the primer set, col1 (d5′-AAC CCT TTG TGA ACR TAC CTA-3′) and col2 (318-bp), were employed for all other Colletotrichum spp. (Martinez et al., 2003). The same PCR conditions were applied for the identification of C. acutatum and Colletotrichum spp. samples, except that the annealing temperature was 56°C for the col2 and col2 primers. The PCR mixtures for analysis included 10 μg of template DNA, 1U Taq DNA polymerase, 0.1 mM dNTPs, 2×PCR reaction buffer, and 10 pmol of each primer. The complete ITS region of the ribosomal DNA were amplified with the primers, ITS1 (5′-GCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3) (Kumar et al., 2005). Sequencing was subsequently performed by Solgent Co. (Daejon, Korea). Sequences were compared to previously deposited Colletotrichum spp. sequences in GenBank (www.blast.ncbi.nlm.nih.gov/).

Five fungicides were selected based on their mode-of-action and standard operating procedure (SOP) (Table 2). Thiophanate-methyl (Carbamate), dithianon (Quinone), tebuconazole (Triazole), fluazinam (Aniline), and picoxystrobin (Strobilurin) were evaluated for effective concentration 50 (EC₅₀) using a standard pathogen strain (C. gloeosporioides JH-1). Briefly, six-well plates were filled with PDA media containing amended with half-log scale concentrations of each fungicide (0 μg/ml, 0.1 μg/ml, 0.213 μg/ml, 0.457 μg/ml, 1 μg/ml, 2.13 μg/ml for fluazinam and thiophanate-methyl; 0 μg/ml, 0.213 μg/ml, 0.457 μg/ml, 1 μg/ml, 2.13 μg/ml, and 4.57 μg/ml for picoxystrobin, dithianon, and tebuconzaole) and a 4-mm plug of actively growing mycelium was placed at the center of each agar plate. Mycelium growth was measured every 24 h up to 5 days. EC₅₀ values were calculated using a Probit model with curve slope value.

The sensitivity of each isolate (n=150, with 30 isolates from each orchard) to the fungicides examined was tested by determining the EC₅₀ concentrations for each fungicide. Mycelia growth was measured 3, 5, and 7 days after the inoculation of fungi on the fungicide amended with 1/5×PDA at 28°C. Sensitivity or tolerance of all of the tested isolates was evaluated based on the mean growth value±standard deviation (SD).

The ITS region β-tubulin, and GS (intron 2) genes were amplified for PCR-RFLP analysis. The following primer sets were used: ITS1 (5′-GCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) (Martin et al., 2004) for ITS, TubG1 (5′-TCTCGATGTTATCCGCCG-3′) and TubGR (5′-TGAGCTCAGGAATGACG-3′) (MacKenzie et al., 2009) for β-tubulin, and GSF1 (5′-ATGGCCGAGTACATCTGG-3′) and GSR1 (5′-AACCGT CGAAGTTCCAC-3′) (Liu et al., 2007) for GS. PCR assays were performed using PCR PreMix reagent (Bioneer, Korea) with 10 pmol of each primer and 80 ng of fungal genomic DNA. The PCR conditions included an initial denaturation step at 94°C for 7 min, followed by 30 cycles of 94°C for 1 min, 45°C for 2 min, and 72°C for 3 min. A final extension step at 72°C for 7 min was also performed. The resulting PCR amplicons were subjected to enzymatic digestion with the restriction enzymes, AluI, HeaIII, TeqI, and MspI (NEB, UK) in a 20 μl total reaction volume according to the manufacturer’s protocol. After 2 h at 37°C, the enzymes digests were inactivated at 70°C for 15 min. The fragmented DNAs were separated on 3% agarose gels at 90 V or 200 mA for 40 min in 1× TAE buffer. The gels were stained with ethidium bromide to visualize the bands.

The fragmented products were scored according to the presence (score of 1) and absence (score of 0) of particular bands. These distances were used to cluster the isolates according to the unweighted pair group method with averages (UPGMA) to determine genetic distances among the isolates (Nei et al., 1973). A phylogenetic tree was generated using the XLSTAT program (www.xlstat.com).

From the Daegok, Yearee, Masan, Masan-2, and Munsan orchards, a total of 122, 49, 92, 82, and 140 anthracnose pathogens were isolated, respectively (Table 1). Four sets of species-specific primers were employed to identify the following pathogens: Colg1 and Colg2 for C. gloeosporioides, col1 and acut1 for C. acutatum, Colg1 and CT2 for C. tuncatum, and col1 and col2 for Colletotrichum spp. (Supplementary Fig. 1). Based on the PCR results obtained for the anthracnose pathogens isolated from each orchard, 156 were identified as C. gloeosporioides, 40 were identified as C. tuncatum, 15 were identified as C. acutatum, and 1 was identified as Colletotrichum spp. (Table 3). The conidia morphology for all of the isolates were typical of C. gloeosporioides, and included a cylindrical shape with a rounded end. The average spore diameter was 4.5 μm (Supplementary Fig. 2).

Effective concentration 50 (EC₅₀) values were calculated for five fungicides which are commonly used to control anthracnose disease in sweet persimmon orchards. As a standard, the strain, C. gloeosporioides JH-1, was used. The EC₅₀ values were 31.5 μg/ml for dithianon, 0.9 μg/ml for picoxystrobin, 3.2 μg/ml for thiophanate-methyl, 1.8 μg/ml for fluazinam, and 2.2 μg/ml for tebuconazole in a Probit model of mycelium growth (Table 3). The five tested fungicides also exhibited a reliable inhibition of mycelium growth against the standard strain (C. gloeosporioides JH-1) (Fig. 1).

Thirty isolates from each orchard (a total of 150 isolates) were randomly selected to be tested in the fungicide sensitivity assay. All tested concentrations of the fungicides were based on the EC₅₀ values determined for each fungicide. The mean mycelium growth diameters were 10.752 mm for picoxystrobin, 10.265 mm for dithianon, 2.929 mm for thiophanate-methyl, 1.269 mm for fluazinam, and 4.821 mm for tebuconazole. Fluazinam provided the best inhibition of pathogen growth. However, the sensitivities of the other fungicides were found to vary significantly against the pathogen isolates. Based on the EC₅₀ concentrations that were determined, the isolates showed tolerance against picoxystrobin and dithianon, yet were sensitive to thiophanate-methyl, fluazinam, and tebuconazole, regardless of their orchard of origin. It is of note that the Musan orchard has been severely damaged by anthracnose disease for the past couple years (data not shown). No resistance isolate against fluazinam was detected. It was also observed that the fungicide sensitivities of the organic orchard isolates differed from those of the practical orchard isolates (Fig. 2). Most of the Colletotrichum isolates exhibited minimal fungicide resistance, while thiophanate-methyl and fluazinam effectively inhibited pathogen growth in vitro regardless of each pathogen origin.

A total of 127 isolates were analyzed by PCR-RFLP to analyze genetic variations. Digests of the resulting ITS gene products with AluI, HeaIII, and MspI produced multiple bands in the isolates obtained from the practical and organic orchards (Supplementary Fig. 3). The AluI-ITS and MspI-ITS digests produced two bands (400 bp, 200 bp) in most of the samples analyzed. In contrast, the HeaIII-ITS digests generated three bands (300 bp, 200 bp, 100 bp), while digests of the practical orchard isolates with TaqI did not produce any fragmentation of the original product band (400 bp). For the organic orchard isolates, two band fragments (400 bp, 200 bp) were generated for the ITS regions digested with TaqI.

The β-tubulin PCR-RFLP patterns were similar to those of the ITS regions. The anthracnose isolates obtained from the practical orchards (Daegok, Munsan, Yearee) showed a simple pattern, while the isolates from the organic orchards (Masan and Masan-2) showed a diverse pattern (Supplementary Fig. 4). Greater diversity was observed in the analysis of the GS gene in the organic orchard and practical orchard isolates (Supplementary Fig. 5). For example, when the GS gene in the practical orchard isolates was digested with AluI, two bands (200 bp, 190 bp) were produced. When the GS gene product was digested with MspI, five bands (300 bp, 210 bp, 150 bp, 120 bp, 100 bp) were generated. The digests with HeaIII also generated five bands (300 bp, 210 bp, 150 bp, 130 bp, 100 bp). The digests containing TaqI only showed the original PCR product (1.5 kb) was present. Most of the GS regions that were amplified from the organic orchard isolates produced four bands following a digest reaction with TaqI (500 bp, 400 bp, 300 bp, 200 bp). Among the four restriction enzyme assays, MspI and HaeIII provided clearer resolution patterns than those observed following the digests with AluI and TaqI, regardless of the genes involved. The PCR-RFLP results showed that the isolates from the organic orchards exhibit more diverse patterns at the genetic level than the isolates from the practical orchards.

A phylogenetic analysis of the ITS-RFLP (Fig. 3a) and β-tubulin-RFLP (Fig. 3b) results showed that they are not distinguishable from the origin of the pathogen. In contrast, the phylogenetic tree that was generated from the GS-RFLP data shows reliable grouping between the pathogen isolates from the practical orchards and the organic orchards (Fig. 3c). However, the PCR-RFLP data did not distinguish the fungicide sensitive versus fungicide-resistant isolates (data not shown).