Genetic Diversity and Fungicide Sensitivity of Cytospora plurivora on Peach

Abstract

Cytospora plurivora D.P. Lawr., L.A. Holland & Trouillas has been associated with recent premature peach tree decline in South Carolina, but very little is known about the pathogen or chemical control options. Ninety-three C. plurivora isolates were collected in 2016 and 2017 from 1-year-old peach wood and symptomatic scaffold limbs, respectively, from orchards in six towns in South Carolina. Six unique genotypes were identified based on substantial ITS1-5.8S-ITS2 sequence variability and classified G1 to G6. Three of the genotypes (G2, G3, and G6) were isolated in high frequency in multiple locations of both years. In addition to the genotypic variation, multiple phenotypes were observed between and within genotype groups. Species identity was determined using additional gene loci: ACT, TUB, and EF, and isolates were found to belong to C. plurivora for all genotype groups. All tested genotypes were sensitive to thiophanate-methyl (FRAC 1) but exhibited slightly lower sensitivity to propiconazole and difenoconazole (both FRAC 3). Boscalid, fluopyram (both FRAC 7s), azoxystrobin, and pyraclostrobin (both FRAC 11s) were ineffective in vitro at inhibiting mycelial growth of C. plurivora genotypes. Field inoculation of peach and nectarine trees revealed that all genotypes developed twig cankers with differences in virulence. G1 was most virulent, and G6 was least virulent. This study provides a link between the C. plurivora genetic variability and virulence and provides fungicide sensitivity information that could be used to improve disease management practices.

Cytospora species can be important pathogens causing twig blight and cankers on peach trees in the United States (Adams et al. 2002; Hammar et al. 1989; Willison 1933). Commonly referred to as Cytospora canker, Leucostoma canker, or perennial canker, the disease can reduce the fruit-bearing wood available on peach trees which, ultimately, can reduce the yield produced from a tree or an orchard. Cytospora was first described by Ehrenberg (1818); currently there are more than 600 described epithets on a varied range of woody hosts, according to Index Fungorum. Although it is the asexual state of Leucostoma, Valsa, and several related genera, Cytospora became the preferred and protected name under the International Code of Nomenclature for algae, fungi, and plants due to Cytospora being the earliest recorded name (Rossman et al. 2015).

Cytospora species infect through openings in woody tissues such as a pruning wound, winter damage, or leaf scars around nodes (Luepschen et al. 1979; Tekauz and Patrick 1974). Initial symptoms of disease include brown and sunken canker formation around the infection site which then turns dark and dry with age (Willison 1933). Dieback occurs once the disease is advanced and causes the blockage and interruption of vascular tissues (Tekauz and Patrick 1974). The infected area later produces pycnidia that sporulate orange or yellow cirri, which provide the inoculum that may infect additional branches and neighboring trees (Hildebrand 1947). In western Colorado in the United States, it was reported that 100% of the peach-producing orchards have Cytospora canker, causing a total annual peach yield loss of 20% (Miller 2017). Currently, cultural management strategies aimed at reducing tree stress are suggested for management of Cytospora canker. Research of current chemical products as a control method is lacking.

Identification of pathogenic species is needed to optimize research on control strategies and provide insights into the underlying biology of the causal fungus. This is especially difficult for species that exhibit a cryptic nature in their physical traits, such as Cytospora species, making identification by morphology alone near impossible, especially in a field setting. Due to the overlap in morphological characteristics, past methods of Cytospora species identification have relied on host association as well (Adams et al. 2005). However, these identification methods have shown to be problematic because a single species of Cytospora can be found on different hosts while multiple Cytospora species may exist on a single host (Adams et al. 2005; Pan et al. 2020). One study described upwards of 28 different species found on Eucalyptus (Adams et al. 2005). Current methods rely on a combination of morphology and phylogeny, with multiple gene loci used to strengthen identification to the species level beyond that which standard fungal barcoding genes can provide (Raja et al. 2017; Tekpinar and Kelmer 2019). The barcoding region for fungi, internal transcribed spacer (ITS), is not ideal for some groups due to narrow or lack of barcoding gaps for this region (Raja et al. 2017). The use of secondary, protein-encoding loci greatly enhances species-level delineations, with one study promoting the use of beta-tubulin (TUB) and translation elongation factor 1-α (EF1α) for Cytospora species identification (Lawrence et al. 2017). Establishing new species-level descriptions is hindered due to the lack of data in sequence libraries for Cytospora species. However, there are continued efforts to isolate, identify, and, when necessary, describe new species within the Cytospora genus (Fan et al. 2015; Lawrence et al. 2017, 2018).

For years, Cytospora pathogens have been viewed as opportunistic (Hildebrand 1947; Willison 1936). Additionally, they have also been associated with bacterial canker, caused by Pseudomonas syringae, with pathologists suggesting that bacterial canker is the primary cause of dieback (Ritchie and Clayton 1981). The differences in symptoms between the two types of cankers occurs along the growing margins under bark tissue, with Cytospora canker exhibiting ring-like growth and bacterial canker exhibiting irregular marginal growth. Whether the pathogens are primary or opportunistic, they can rapidly spread through the wooden tissue of peach trees, which causes yield loss for growers. Previous studies have suggested that the disease is mainly an issue in the northern and western United States; however, research by Froelich and Schnabel (2018) showed widespread occurrence in South Carolina peach orchards. Cytospora species were isolated in the highest frequency of all twig blight pathogens recovered (Froelich and Schnabel 2018). This has raised concerns about the impact that a potential Cytospora outbreak may have on the state’s economy, as South Carolina is the second highest peach producer in the country, behind California, growing approximately 76,500 tons valued at around 101 million dollars in 2020 (USDA 2020). Previous research has also suggested the high genetic diversity of Cytospora, particularly within C. plurivora (Adams et al. 2002; Lawrence et al. 2018). This could have a significant impact on disease management because growers are combating a highly diverse pathogen. The objectives of this study were to evaluate the genetic diversity and virulence of C. plurivora isolates collected from South Carolina and to assess their sensitivities to fungicides used in peach production.

Materials and Methods

Sample collection and pathogen isolation

Cytospora plurivora isolates were collected in 2016 from orchards in six locations in South Carolina, including the towns Chesnee, Greer, McBee, Mountain Rest, Ridge Spring, and York (Fig. 1). Necrotic tissue from underneath the periderm of 1-year-old peach wood exhibiting pycnidia was excised and surface sterilized with 5% sodium hypochlorite for 1 min and washed in sterilized distilled water for 1 min. Wood discs were plated on potato dextrose agar (PDA), prepared by the manufacturer’s (Difco) instructions, and incubated at 22°C in the dark until mycelia developed. Isolates from single hyphae were cultured, purified, and stored on filter paper as described previously (Hu et al. 2011a). A separate collection of C. plurivora was acquired in South Carolina in 2017 from Chesnee, Greer, McBee, Ridge Spring, and York, but in contrast to using 1-year-old twigs as was done in 2016, isolates from scaffold limbs exhibiting severe dieback (greater than 80% but less than 100% of the limb being dead) were collected. Bark samples 10 × 10 cm in size were excised from the scaffold limbs to include the margin of necrotic and asymptomatic cambial tissue. The bark underwent similar processing as described above, and single-hyphal isolates were purified and stored on filter paper. Fig. 1 and Supplementary Table S1 contain the number of isolates by location. Isolates were selected for further analysis based on differences of culture morphology including radial growth, color, and margins of mycelium, as described in Baker (2022).

Fig. 1.

Map of South Carolina with sampling locations and genotype proportions.

DNA extraction, polymerase chain reaction (PCR) amplification, and sequencing

DNA was extracted according to a previous protocol (Chi et al. 2009), but mycelia were grown on PDA plates covered with ultra-clear cellophane (Research Products International) to make removal easier, and plastic pestles and microfuge tubes were used in place of an electric grinder. Initial PCR was conducted to amplify the ITS region 1, ribosomal 5.8S subunit, and ITS region 2 using primers ITS1-F and ITS4 (White et al. 1990). Subsequent PCR of selected Cytospora isolates was conducted to amplify additional loci of the actin (ACT) ACT-512F/ACT-783R (Carbone and Kohn 1999), beta-tubulin 2 (TUB2) Bt2a/Bt2b (Glass and Donaldson 1995), translation elongation factor 1-α (EF) EF1-688F/1251R (Alves et al. 2008), and calmodulin (CAL) primers CAL-228F/CAL-737R (Carbone and Kohn 1999) regions. PCR conditions followed manufacturer recommendations. PCR products underwent Sanger sequencing at the Arizona State University DNA Lab. These sequences were searched with the Basic Local Alignment Search Tool (nBLAST) from the National Center for Biotechnology Information to assign fungal species identities. Genious (version 11.0.4, Biomatters) was used for sequence assembly, alignment, and construction of phylogenetic trees for the ITS region. Molecular Evolutionary Genetics Analysis (version 10) was used for sequence assembly, alignment, and construction of phylogenetic trees for the ACT, TUB2, CAL, and EF gene regions. A partition homogeneity test using PAUP v 4.0 was conducted to see if the individual datasets could be combined. Accession numbers for all South Carolina isolates are listed in Supplementary Table S2. Reference isolates included in the phylogenetic trees were obtained from published research on C. plurivora and represent worldwide genetic variation of the species (Supplementary Table S3; Adams et al. 2002; Arhipova et al. 2011; Fotouhifar et al. 2010; Kepley and Jacobi 2000; Lawrence et al. 2018; Singh et al. 2007).

Fungicide sensitivity assay

The half maximal effective concentration (EC₅₀) was used to determine sensitivities of C. plurivora isolates to fungicides from four Fungicide Resistance Action Committee (FRAC) mode-of-action (MOA) codes commonly used by growers in South Carolina peach orchards. Thiophanate-methyl (methyl benzimidazole carbamates [MBC], FRAC 1, Topsin M70WDG); propiconazole and difenoconazole (demethylation inhibitors [DMI], FRAC 3, Tilt and Inspire, respectively); boscalid and fluopyram (succinate dehydrogenase inhibitors [SDHI], FRAC 7, Endura and Velum, respectively); and pyraclostrobin and azoxystrobin (quinone outside inhibitors [QoI], FRAC 11, Cabrio and Abound) were used in the assay. These fungicides were selected because they were registered for disease management of peach in the United States, with thiophanate-methyl being shown to be effective against Cytospora species (Miller et al. 2019). Malt extract agar (MEA; 10 g malt extract, 15 g agar, 1 liter H₂O) was amended with the fungicides from FRAC 1,3, and 11. For the FRAC 7 fungicides, minimal media (MM; 10 g glucose, 1.5 g K₂HPO₄, 2 g KH₂PO₄, 1 g [NH₄]₂SO₄, 0.5 g MgSO₄·7H₂O, 2 g yeast extract, 12.5 g agar, 1 liter H₂O) was used (Hu et al. 2011b). Salicylhydoxamic acid was also added to media for FRAC 11 to inhibit the alternative oxidase pathway. An initial study examined the sensitivity of five of the six C. plurivora genotypes from the 2016 collection to 0.1, 1, and 10 μg/ml of each active ingredient (ai). Three 4-mm mycelial plugs were extracted from the edges of 1-week-old cultures and placed on each fungicide-amended and nonamended control plate and incubated at 22°C in the dark until the three colonies were almost touching. At this point, the colonies were measured, and the average diameters were recorded. Subsequently, a final assay was conducted using refined concentrations based on the assessment of the preliminary measurements. The concentrations 0.1, 0.3, 1, 3, and 10 μg/ml ai were used for testing the thiophanate-methyl (FRAC 1); 0.1, 0.3, 1, 3, 10, and 30 μg/ml ai for difenoconazole and propiconazole (FRAC 3); 3, 10, 30, and 100 μg/ml ai for boscalid and fluopyram (FRAC 7); and 0.01, 0.03, 0.1, 0.3, 1, and 3, 10, 30, and 100 μg/ml ai for pyraclostrobin and azoxystrobin (FRAC 11). The assay was repeated.

In-field virulence

Virulence of C. plurivora genotypes was assessed on 7-year-old ‘Redgold’ nectarines and 6-year-old ‘Coronet’ peaches. Four trees of each cultivar were inoculated on November 16, 2018, each with six isolates representing genotypes G1 to G6. This experimental approach allowed for isolates to be compared in an environment of equal tree physiology and stress level. Two- or three-year-old wood was used for inoculation, and each isolate was inoculated on a single branch per individual tree with a total of 24 inoculations overall. Using an adapted protocol from Henriquez et al. (2006) and Barakat and Johnson (1997), 6-mm PDA plugs taken from 8-day-old cultures were placed into holes drilled into 2- or 3-year-old wood using a battery-powered hand drill with a 6-mm drill bit (Craftsman V20 20-Volt Max ½-in Cordless Drill). Limbs were surface sterilized before holes were drilled, and plugs with actively growing mycelium or without fungus (control) were inserted into holes using sterile tools. Inoculated branches were then wrapped with Parafilm. The Parafilm was removed after 48 h, and canker length was measured after 4 months. The bark was carefully removed from each inoculation site with a knife to expose the entire canker, and measurements of canker length were taken with a digital caliper.

Statistical analysis

JMP (SAS Institute, Inc., Cary, NC, U.S.A.) was used for all statistical modeling. A full-factorial analysis of variance was conducted to determine the significance of fungicide, genotype, isolate, and their interactions to EC₅₀. Fisher’s LSD test was used to compare EC₅₀ values between statistically significant (α = 0.05) factors.

Results

A total of 93 C. plurivora isolates were collected over two years from peach twigs and scaffold limbs of South Carolina orchards (Fig. 1; Supplementary Table S1). In 2016, 57 of the 93 C. plurivora isolates were collected from twigs in five of the six locations sampled (none were collected from the Ridge Spring location that year). In 2017, 36 of the 93 isolates were collected from scaffold limbs in four of the five locations sampled (i.e., Chesnee, Greer, McBee, and Ridge Spring).

DNA sequence analysis of a subset of isolates from all regions sampled revealed highest sequence identity (99.5%) to C. plurivora, a species described only recently (Lawrence et al. 2018). A phylogenetic tree of isolates selected from both years at random from each location showed high genetic diversity of C. plurivora isolates in the ITS1-5.8S-ITS2 region (Fig. 2) (deleted: representative isolates for the phylogenetic tree presented in Fig. 2 are a subset of isolates selected from another, larger phylogenetic tree that included all of the sequences from the 2016 and 2017 collections) as well as high diversity of culture morphology (Fig. 3). The tree presented in this paper was created to give an accurate portrayal of genetic variation, and isolates were chosen for Fig. 2 from both years to represent all locations. First, two large clusters were identified (Cluster 1 and Cluster 2). A total of six groups (three from each cluster) were identified based on a neighbor-joining phylogenetic tree and designated as genotypes 1 through 6 (G1 to G6; Fig. 2). Surprisingly, South Carolina isolates clustered together with C. plurivora isolates from other states and from other countries, indicating high genetic diversity. Culture morphology was variable, both within and among genotypes (Fig. 3).

Fig. 2.

Neighbor-joining phylogenetic tree based on the ITS1-5.8S-ITS2 region of Cytospora plurivora isolates collected from South Carolina in 2016 and 2017. Bolded isolates were obtained from the National Center for Biotechnology information Nucleotide BLAST (NCBI nBLAST) and were selected from published research to reflect the geographical locations of C. plurivora worldwide.

Fig. 3.

Culture morphologies of Cytospora plurivora genotypes (G1 to G6) on potato dextrose agar showing variability within and among genotypes.

The partition homogeneity test found that three of the four additional loci, ACT, TUB2, and EF, could be combined (P = 0.1170; α = 0.05). The maximum likelihood (ML) analysis used a best fit model of Kimura-2-parameter-model with gamma distribution, K2 + G, for both the combined and ACT trees. The combined alignment with three gene regions (Fig. 4) produced a tree of 28 nucleotide sequences that contained 1,908 sites, of which 630 (33%) were parsimony informative. The maximum parsimony (MP) analysis produced six parsimonious trees of 255 steps with a consistency index (CI) of 0.5231, retention index (RI) of 0.7163, and a rescaled consistency index (RC) of 0.4269. The ML analysis for the ACT alignment (Fig. 5) produced a tree of 38 nucleotide sequences that contained 342 sites, of which 98 (28%) were parsimony informative. The MP analysis produced four equally parsimonious trees of 210 steps with a CI of 0.5932, a RI of 0.7608, and a RC of 0.4513. The study isolates clustered together with C. plurivora, with supporting bootstrap values of 99% for ML and 99% MP for both the combined and the ACT trees. The CAL tree showed similar clustering but was not taxonomically informative due to the lack of reference sequences deposited in GenBank.

Fig. 4.

One out of six equally parsimonious trees generated by maximum parsimony analysis of the three-gene region (ACT, TUB, and EF-1) Cytospora combined data set. Numbers above and below the slash represent likelihood and parsimony bootstrap values from 1,000 replicates. Values with an asterisk were <70% for the bootstrap analysis. Ex-type isolates are indicated in bold. ACT = Actin, TUB = β-tubulin, EF1 = translation elongation factor 1-α, G = Greer, MB = McBee, MR = Mountain Rest, Y = York, CI = consistency index, RI = retention index, RC = rescaled consistency index, and KARE = Karney Agricultural and Research Extension.

Fig. 5.

One out of four equally parsimonious trees generated by maximum parsimony analysis of the ACT region. Numbers above and below the slash represent likelihood and parsimony bootstrap values from 1,000 replicates. Values with an asterisk were <70% for the bootstrap analysis. Ex-type isolates are indicated in bold. ACT = Actin, G = Greer, MB = McBee, MR = Mountain Rest, Y = York, KARE = Karney Agricultural and Research Extension, CI = consistency index, RI = retention index, and RC = rescaled consistency index.

The most frequently isolated genotype was G6, which was found in Chesnee, Greer, Ridge Spring, and Mountain Rest (Supplementary Table S1). The second most frequently isolated genotype was G2, which was found in each of the six locations. The two least frequently found genotypes were G5 (one isolate) and G4 (three isolates), which were found in McBee (G5), Greer (G4), and Ridge Spring (G4; Supplementary Table S1). Isolate colony morphology variation was observed within and between genotypes (Fig. 3).

Differences in sensitivity of C. plurivora isolates to various fungicides were observed (P = 0.002), but fungicide sensitivity was not genotype specific (Table 1). Across genotypes tested, the EC₅₀ values for thiophanate-methyl (FRAC 1) ranged from 0.021 to 0.513 μg/ml ai (Table 1). EC₅₀ values within genotypes for this fungicide were generally consistent with some variation occurring among isolates of G3. The EC₅₀ values for difenoconazole (FRAC 3) ranged from 0.120 to 2.912 μg/ml ai, and for propiconazole. it ranged from 0.394 to 2.879 μg/ml. EC₅₀ values for thiophanate-methyl were significantly lower compared with those for both DMI fungicides, but no significant difference was found between EC₅₀ values for difenoconazole and propiconazole. Boscalid, fluopyram, azoxystrobin, and pyraclostrobin inhibited mycelial growth at the lowest concentrations examined, but none of the higher doses arrested mycelial growth completely. Growth inhibition at 100 μg/ml boscalid or fluopyram ranged from 12 to 52%. Growth inhibition for azoxystrobin and pyraclostrobin at 0.01 and 3 μg/ml ranged from 55 to 70% and 70 to 90%, respectively. For that reason, the minimum inhibitory concentration (MIC) values rather than EC₅₀ values were calculated. No fungicide data was obtained from G4 genotype isolates.

Table 1.

Sensitivity of Cytospora plurivora genotypes to fungicides from four FRAC groups

Genotype group	Isolate	EC50 values (μg/ml)			MIC values (μg/ml)y
		Thiophanate-methyl (FRAC 1)	Difenoconazole (FRAC 3)	Propiconazole (FRAC 3)	Boscalid (FRAC 7)	Fluopyram (FRAC 7)	Pyraclostrobin (FRAC 11)	Azoxystrobin (FRAC 11)
G1	MB-3-16	0.352	0.853	0.598	>100	>100	>3	>3
	MB-15-16	0.516	0.596	0.817	>100	>100	>3	>3
	G-19-16	0.185	0.120	0.394	>100	>100	>3	>3
G2	C-26-16	0.416	1.122	1.716	>100	>100	>3	>3
G3	MB-2-16	0.096	1.779	2.116	>100	>100	>3	>3
	MB-8-16	0.021	1.666	2.967	>100	>100	>3	>3
	MR-13-16	0.467	1.311	1.123	>100	>100	>3	>3
	Y-17-16B	0.464	1.949	1.821	>100	>100	>3	>3
G5	MB-11-16	0.361	0.184	0.832	>100	>100	>3	>3
G6	C-27-16	0.317	0.752	1.551	>100	>100	>3	>3
	G-13-16	0.335	0.874	1.569	>100	>100	>3	>3
	MR-10-16	0.347	2.912	2.879	>100	>100	>3	>3
		bz	a	a

^yEC₅₀ values could not be calculated due to nonlinear dose responses at the concentrations tested; MIC = minimum inhibitory concentration.

^zLetters indicate significant differences of combined EC₅₀ values between fungicides that reached complete inhibition of mycelia growth (α = 0.05).

With regard to the field test examining canker formation of C. plurivora genotypes on nectarine and peach trees, the variances between treatments of the two independent field experiments were equal according to Bartlett and Levine tests (P ≥ 0.05), and thus the data of the two trials are presented as a merged dataset (Fig. 6). G1 was the most virulent genotype on peach and nectarine trees, followed by G3 (Fig. 6). The least virulent was G6, which was not found in all locations but was the most frequently isolated genotype in four of the six locations sampled (Figs. 1 and 6; Supplementary Table S1).

Fig. 6.

Virulence of Cytospora plurivora genotypes G1 to G6 on ‘Redgold’ nectarines and ‘Coronet’ peaches. Drilled holes with fungus-free plugs served as the control. Different letters indicate significance at P ≤ 0.05 according to Tukey’s HSD test.

Discussion

Our survey of South Carolina peach orchards in 2016 and 2017 revealed that the majority of the South Carolina isolates share a 99.5% similarity with that of C. plurivora, a species described from olive (Olea europaea) in California and given this epithet due to the plethora of hosts from which it was isolated (Lawrence et al. 2018). Identification was based on sequencing data of three protein encoding loci (ACT, TUB2, and EF), which have been shown to provide greater species determination beyond standard barcoding, especially for Cytospora species (Lawrence et al. 2017; O’Donnell et al. 2015). The fourth gene region, CAL, has been suggested for use in identifying species from the Diaporthales but was the least taxonomically informative due to the lack of ex-type species sequences deposited in GenBank specifically for Cytospora (Lawrence et al. 2017; Nouri et al. 2019; Yang et al. 2020). While not beneficial to this study, the CAL sequences still provide insight by showing a clustering pattern similar to that found in the other phylogenetic trees presented in this study, further strengthening the phylogenetic relationship of the South Carolina isolates (Baker 2022). A multigene analysis combined with morphological comparisons is preferred when examining Cytospora isolates of varying genetic background. While not present in this study, the morphology of these study isolates, including color, overall growth form, shape of growing region and margins, and conidia and conidiophore measurements, was examined and found to verify the species identity of C. plurivora (Baker 2022; Lawrence et al. 2018).

Genetic diversity among sampled isolates from South Carolina was initially based on ITS sequence variations. Intraspecific variation can vary with some species having little to no variation and others having extremely high variation (Nilsson et al. 2008). In the case of some Cytospora species, the understanding is that an inherent intraspecific variation exists (Adams et al. 2002), but to what extent is unknown. The variations within highly diverse genotypes can create a challenge when attempting to identify the pathogen to the species or genotype level. The ITS region has been considered the universal DNA barcode for the kingdom Fungi (Schochetal. 2012), although it is not ideal for Cytospora species.

While C. plurivora has already been reported from other parts of the United States, this study has expanded upon its geographical distribution to now include the state of South Carolina. This is not the first instance of Cytospora species being reported within the state. Previous accounts have reported Cytospora in South Carolina as early as the 1960s on Gordonia and Hydrangea (USDA 1960). The Cytospora isolates from peach were found as a result from an extensive survey into the causes of twig blight, indicating that C. plurivora has likely been present but has merely gone overlooked. Given the highly cultivated and managed nature of peach orchards, it is also likely that Cytospora cankers were removed before an outbreak could occur, eliminating the need for earlier studies into this pathogen. The findings from Froelich and Schnabel (2018) indicate the importance of conducting such a survey at regular intervals to remain vigilant to fungal organisms that pose a potential risk to vital crops systems.

The sequences of international isolates being intermixed within our local population is yet another reminder how genetically diverse a local population of C. plurivora can become. In addition to the genetic diversity, we also observed substantial morphological differences both between and within genotype groups. A study of more than 400 isolates from Michigan peach orchards revealed a relatively large number of cryptic species, maternal lines, and frequent incompatibility within maternal lines. The study indicated large genetic variation in L. persoonii and concluded that sexual recombination is common (Wang et al. 1998). The alternative explanation is that genotypes from other states such as Michigan or New York or from other countries such as Iran or South Africa (Fig. 2) were introduced. However, to the best of our knowledge, there is no exchange of tree material between United States nurseries and those in other countries or continents. Commercial peach trees generated at local nurseries with rootstocks and scions from United States plants (including trees) intended for propagation require a foreign phytosanitary certificate in advance according to the United States Customs and Border Protection (https://www.cbp.gov/).

Genotypes G2, G3, and G6 were isolated in high frequency in both years, indicating that these may be more virulent compared with other, less-frequently isolated genotypes. The virulence data we collected on peach and nectarine in the field confirmed G2 and G3 to be more virulent than G4, G5, and G6. However, G6 was the genotype with the least virulence. This genotype was only found in four of the six locations. It is possible that we collected in areas with strong presence of G6 and that, in general, it is less frequent compared with other genotypes. A larger sampling size of hundreds of isolates would shed light on this hypothesis. Whether the variation in phenotypic colony morphology among genotypes described in this study is associated with virulence was not investigated in this study.

We identified differences in fungicide efficacy between active ingredients but found no significant effect of the genotype. We attributed boscalid, fluopyram, azoxystrobin, and pyraclostrobin tolerance (as determined by MIC values) to an intrinsic trait of the species and not to a result of fungicide selection because of the relative insensitivity to these active ingredients across all genotypes and the lack of sensitive baseline isolates. Thiophanate-methyl was the most effective inhibitor of C. plurivora mycelial growth, followed by difenoconazole and propiconazole. Most previous studies have focused on the sensitivity of C. plurivora to multisite fungicides. One older study found ferbam (FRAC M03) and sulfur (FRAC M02) to be effective in inhibiting spore germination in vitro (Dhanvantari 1968). Another found that benomyl (FRAC 1) and captafol (FRAC M04) lessened Cytospora canker disease incidence when it was applied in both fall and spring (Northover 1992). However, the efficacy of captafol was not always confirmed (Grosclaude 1985). While captafol and benomyl can no longer be used on peach due to changes in regulations and many peach pathogens are resistant to benomyl (Bernstein et al. 1995; Penrose and Koffman 1977), active ingredients with identical MOAs to benomyl are still available. A more recent study confirmed high susceptibility of Cytospora leucostoma (syn. L. persoonii) to thiophanate-methyl and difenoconazole and also acknowledged incomplete inhibition of mycelium in laboratory assays by pyraclostrobin + boscalid (Miller et al. 2019).

In conclusion, C. plurivora is a widespread, genetically diverse pathogen of twigs and scaffold limbs in South Carolina peach orchards. Only FRAC 1 and 3 group fungicides had activity that may be sufficient for field control.