Research results and discussion:

1.      Determine the sources and major infection periods of Colletotrichum (bitter rot fungi) in PA apple orchards.

Determine if bitter rot fungi overwinter on outer bud scales

Apple twigs with approximately 75 buds were collected from ‘Honeycrisp’  block on March 19^th 2019, and incubated on a wire rack in a plastic tub with wet paper towel in the bottom. The twigs were individually visually checked for orange C. fioriniae indicative of C. fioriniae on May 9^th. There was abundant black rot on most of the twigs, along with occasional Phomopsis and unidentified mycelia. The only spore mass that even resembled C. fioriniae turned out to be Fusarium when observed under the microscope. No Colletotrichum spores were observed. The tubs had been placed in front of the ventilation unit and needed to be frequently re-moistened. Whether this intermittent drying had any impact on Colletotrichum sporulation is unknown.

In 2018 we had found Colletotrichum infections on a few buds that comprised less than 1% of the total. Since we only tested about 75 buds in 2019, we might have missed that low incidence of infection. It is also entirely possible that other saprophytic fungi overwhelmed the Colletotrichum infections. In personal communication with other scientists, we learned that plating of outer bud scales on ½ strength PDA agar might be a better detection method. In 2017 we had vigorously surface disinfested buds and then plated the bud scales and found no Colletotrichum. It is possible that Colletotrichum only survives on outer and not inner bud scales, and our vigorous surface disinfestation killed any Colletotrichum that was present.

Determine if spore sources are mostly in the tree canopy, splashed up from the orchard floor, or from nearby forested areas, and the timing and quantity of spore dispersal from those sources

Both C. fioriniae specific primer sets targeting the GAPDH and β-tubulin genes produced good amplification of both the standard curve samples and environmental samples (Figure 1a and 1c). In environmental samples with high spore concentrations the melt curve looked clean (Figure 1b). However, at lower concentrations, most environmental samples had secondary peaks that can indicate non-specific binding (Figure 1d). While not all secondary peaks are actual indicators of non-specific binding, samples with secondary peaks should be run on an agarose gel to confirm a single product (https://www.idtdna.com/pages/education/decoded/article/interpreting-melt-curves-an-indicator-not-a-diagnosis). Since secondary peaks in a Sybr-Green based assay would raise red flags for most peer reviewers, and since running every sample on agarose gel to confirm a single product is time-consuming, it was decided to go with the more specific (and also more expensive) TaqMan probe based assay.

After considering the time and money it would take to design and test a TaqMan assay that is specific to C. fioriniae, it was decided to go with the C. acutatum species-complex specific assay of Debode et al (2009). While this assay could hypothetically detect any or all of the 20-odd species within the C. acutatum species complex, it would most likely detect only or mostly C. fioriniae, since that is the only species within the C. acutatum species complex that we have found in our orchards at Penn State FREC. Spore trap samples collected in 2018 before the start of this project funding were also analyzed with the assay from Debode et al (2009). Spores were collected 16 times from May through October in 2018, 19 times from April through November in 2019, and 19 times from February through November in 2020. Two out of the 16 collections in 2018 were removed from analysis due to contamination with copper sulfate that had been added as an experimental biocide, and 2 of the 2018 and 1 of the 2019 collections were removed from analysis because of contamination in the untreated control DNA extraction sample. This left 12 collections in 2018, 18 collections in 2019, and 19 collections in 2020, all at 23 locations (5 forest, 18 orchard).

Some samples were missing from these collection timepoints due to spills, broken bottles, etc. Samples that were above the range of the standard curve were counted as the maximum standard curve value of 10⁷ conidia equivalent per sample. The 1060 samples showed spores are dispersed throughout the growing season (Figure 2). The start of high numbers of spores in the spring coincides with bud break and leaf formation, while the decline of spore dispersal in the fall coincides with leaf drop (Figure 2).  The forest and untreated inoculated apple tree spore traps had the highest spore counts, while the fungicide treated apple spore traps had the lowest counts. These results showing spore dispersion throughout the season is somewhat bad news for management of bitter rot, as it indicates that fungicide applications will also have to be season long.

Comparison of the conidia equivalents across 2018, 2019, and 2020 showed that within the orchard, fungicide treated trees had lower numbers that untreated trees, and the numbers in the forest were higher than the orchard (Figure 3). An ANOVA followed by a Tukey-Kramer test showed the forest samples had more spores than the orchard samples, and the fungicide treated orchard samples had less than the untreated samples at α = 0.01 (Figure 3).

The finding that more spores are being dispersed in the forest than in the orchard was unexpected. This led to the question of where in the forest the spores were coming from. Based on earlier findings  that C. fioriniae can be an endophyte in various broadleaf forest plants (Marcelino et al. 2009), the leaves of several forest understory plants were collected, and a freezing method based on Børve & Stensvand (2017) and Mertely & Legard (2004) was used to detect endophytic Colletotrichum species infections in leaves. Leaves were surface disinfested by submersion in 70% ethanol to break the water surface tension, rinsed in deionized water, submersed in 10% (0.5% NaClO) bleach for 40s, rinsed in deionized water, submersed in 70% ethanol for 20s, and rinsed again in deionized water. They were then placed on a rack in a plastic tub (Figure 4, top) with wet paper towel to maintain high humidity, frozen solid in a -20 or -80 freezer to kill the leaves, and incubated at room temperature (21-23C) for two weeks to allow endophytic fungi to sporulate. Deionized water was sprayed on the leaves as needed to maintain moist conditions. The results were abundant orange Colletotrichum conidia masses on many of the leaves (Figure 4, bottom). A few pure fungal cultures were obtained from these spore masses, and sequencing of intron 1 of the GAPDH gene (using primers GDF1 5’-GCCGTCAACGACCCCTTCATTGA-’3 and GDR1 5’-GGGTGGAGTCGTACTTGAGCATGT-’3 (Templeton et al. 1992)) showed that the sequences were identical to those from C. fioriniae from apple. Apple leaves were then also tested, and a number of which also produced abundant orange Colletotrichum conidia masses (Figure 5).

To make sure this wasn’t just a rare, localized occurrence, in 2019 over 1,138 leaves of 24 forest plant species were collected during the months of June to September from the orchard and forested area surrounding the spore traps at FREC, and from the nearby Michaux State Forest (MSF), mostly within a kilometer of N 40.034086, W -77.342036. Permit number SFRA-1920 was obtained from the Bureau of Forestry of the Pennsylvania Department of Natural Resources for the collection of leaves from MSF. This area of MSF was heavily logged in the 1800s and saw commercial activity up to the 1970s, but has been forested since then, and is at least 4 straight-line kilometers from the nearest agricultural fields. Plant species were identified based on morphology as per Rhoads & Block (2007).

The leaves were tested via the freezing method described above. After incubation, photographs were taken of the leaves for calculation of % leaf surface area covered by orange spore masses.  Using ImageJ software (Schneider et al. 2012), the outline of the leaf was manually traced and the pixels of leaf area obtained. The area containing orange spore masses was then traced to obtain the pixels of leaf area with orange spores and divided by the total leaf area to obtain a percentage of leaf area with orange spores (Figure 6). The overall results showed orange conidial masses on 43% of leaves, and quantification of spore mass area showed 30% of leaves had more than 1% leaf area with orange spores, with high variability among plant species (Figure 7). Levene’s test of variances of leaf area with orange spores by plant species showed they were not equal (p < 0.0001) and the general linear model in SAS showed that location, plant species, and the interaction of location and plant species were all highly significant predictors of percent leaf area with orange spores, with p-values of < 0.001. Given the in-equality of variances and the difficulties of separating the effects of location and location-plant species interactions from the effect of plant species alone, no means separation test based on plant species was performed. The results are instead shown as a dot-plot of percent leaf area with orange spores arranged by plant species, where every leaf with more than 1% leaf area with orange spores is shown as an individual dot (Figure 7).

This confirmed that Colletotrichum species are abundant as leaf endophytes both in farm woodlots and large forests, and raised the question of which Colletotrichum species these are. To answer that, 97 orange conidial masses were randomly selected from leaves of 20 plant species and the conidial masses streaked onto a Petri dish of ½ strength PDA (Potato Dextrose Agar, Difco, Franklin Lakes NJ). The resulting fungal growths were single-spore or hyphal-tip isolated to form genetically uniform fungal cultures. DNA was extracted from cultures with the NucleoSpin Microbial kit (Macherey Nagel, Bethlehem, PA) and intron 1 of the GAPDH gene was sequenced. After it was observed that many leaves also had areas of black spores, such as those in the Aspen photo in Figure 6, four  black spored cultures were obtained for identification and comparison. The non-Colletotrichum fungal cultures were identified by sequencing the ITS gene using primers ITSF_KY02 5’-CTTGGTCATTTAGAGGAAGTAA-’3 (Toju et al. 2012) and ITS4 5’-TCCTCCGCTTATTGATATGCCAG-’3 (White et al. 1990).

Amplification, sequencing, and phylogenetic analysis of the GAPDH gene revealed that 87 isolates clustered with C. fioriniae, 2 with C. nymphaeae, 2 by themselves but closest to C. salix, and 3 by themselves but closest to C. aenigma.  Three orange spored isolates that did not amplify with GAPDH were identified with ITS as Gnomoniopsis paraclavulata. Of the 4 black spored isolates, one amplified with GAPDH and matched C. fioriniae, while ITS identified the others as Diaporthe (Phomopsis) spp. and Pestalotiopsis spp. (Table 1). The most likely explanation for the single black-spored isolate that was identified as C. fioriniae (FREC121, Table 1) is that since it was from a leaf that had both orange and black conidia, and few C. fioriniae conidia were inadvertently plated out with the black conidia for single-spore isolation, and one of them was randomly or subconsciously selected.  The 3 non-Colletotrichum orange-spored isolates were identified with ITS as Gnomoniopsis paraclavulata. These 3 isolates came from MSF from the closely related white and chestnut oaks. G. paraclavulata is described as living on white oak leaves, with orange conidia masses and conidia shapes nearly identical to those of Colletotrichum, although the conidia are distinctly smaller (Sogonov et al. 2008). Since C. fioriniae was also isolated from leaves of white and chestnut oaks, it should be assumed that a certain percentage of the leaf area with orange spores shown in Figure 7 of these 2 plant species is from G. paraclavulata. The identity of the black-spored isolates as Diaporthe (Phomopsis) spp. and Pestalotiopsis spp. makes sense, as these are widely distributed and well known leaf endophytes. The main result though, was that most of the isolates were C. fioriniae (figure 8).

To confirm identity to species, a subset of 13 C. fioriniae isolates, both C. nymphaea isolates, both C. salix isolates and 2 isolates that resemble C. aenigma, along with isolates from fruit and reference isolates from other studies were chosen for multilocus sequence analysis using additional 4 to 7 genes. For isolates in the C. acutatum species complex the genes were Tub2, calmodulin, and GS using primers T1 5’AACATGCGTGAGATTGTAAGT’3 (O’Donnell & Cigelnik 1997) and TubR1 5’-TTCTGGACGTTGCGCATCTG-’3 (Hu et al. 2015), CL1C 5’-GAATTCAAGGAGGCCTTCTC-’3 and CL2C 5’-CTTCTGCATCATGAGCTGGAC-’3 (Weir et al. 2012), and GSF1 5’-ATGGCCGAGTACATCTGG-’3 and GSR1 5’-GAACCGTCGAAGTTCCAGGAA-’3 (Stephenson et al. 1997) respectively. For isolates in the C. gloeosporioides species complex the additional genes ApMat and APN2 were used with primers CgDL_F6 5’-AGTGGAGGTGCGGGACGTT-’3 and CgMAT1_F2  5’-TGATGTATCCCGACTACCG-’3, and ColDL-F3 5’-GGGAGAAGCGAACATACCA-’3 and CgDL-R1 5’-GCCCGACGAGCAGAGGACGTAGTC-’3, respectively (Rojas et al. 2010). For a few unresolved isolates the actin gene was also sequenced using primers Act512F 5’-ATGTGCAAGGCCGGTTTCGC-’3 and Act783R 5’-TACGAGTCCTTCTGGCCCAT-’3 (Carbone & Kohn 1999).

PCR conditions were denaturation at 94C for 5 min followed by 35 cycles of denaturation at 94C for 30s, annealing at 55C for all genes except APN2 and Actin (58C) for 30s, elongation at 72C for 30s for GAPDH, 45s for Actin, 60s for calmodulin, GS, and ITS, 80s for ApMat, and 120s for Tub2, with a final elongation of 72C for 7 min. The PCR product was purified and sequenced by Eurofins Genomics (Louisville, KY). DNA sequences were concatenated with CodonCode Aligner 7.0 (Centerville MA), aligned with MEGA (Kumar et al. 2018), manually checked for accuracy,  and trimmed to uniform lengths.  Individual genes were BLAST searched on GenBank to match with previously described species, sequences of the type strains were added to the DNA alignments. A maximum parsimony phylogeny of concatenated genes was created with PAUP version 4 (Swofford 2003) using heuristic search, keeping optimal trees only, starting trees for branch swapping by stepwise addition and swapping on best only, using random stepwise addition with 10 reps, 0 seed, and holding 1 tree at each step, and branch swapping with the TBR algorithm using a reconnection limit of 8 and swapping on best trees only, followed by a bootstrap analysis to 1,000 replicates.

The phylogeny of the isolates in the C. acutatum species complex showed that the C. fioriniae and C. nymphaeae isolates from plant leaves are indistinguishable from those isolates from fruits, in some cases having 100% sequence identity across the 4 genes (figure 9). Isolates FREC145 and FREC146, while genetically distinct, were closest to C. salix. Isolates FREC139 and FREC142 grouped with and yet remained distinct from a clade that contained C. aenigma, C. hebeiense, and C. conoides (data not shown). BLAST search of each gene found 100% identity at the GAPDH loci to unidentified isolate CT03 from Peony leaves (Garfinkel & Chastagner 2018) and at the ApMat and APN2 loci to unidentified isolate Coll887 from diseased cranberry fruit (Doyle et al. 2013). Doyle et al. 2013 designated Coll887 as C. sp.indet.C., so we designated our matching species likewise.

To evaluate the ability of these isolates to cause bitter rot in apple, Koch’s postulates were performed on 7 C. fioriniae isolates that had previously been isolated from the fruits of grape, peach, and pear, and from the leaves of Amur honeysuckle, poison ivy (from both FREC and MSF) and wild grape.  ‘Golden Delicious’ apples were surface disinfested by spraying with 70% ethanol and air drying, a ~5x5 mm wound was made in the surface of the apple at its equator, mycelial plugs of the isolates were placed in the wound, and incubated for 2 weeks at 25°C and 24 hour photoperiod. Lesions were measured and photographed, and fungal colonies were isolated from the outer edge of the lesion. The fungal colonies were identified to species by sequencing of the intron in the GAPDH gene as described above.  

After wound inoculation the 7 C. fioriniae isolates all created typical bitter rot symptoms in the ‘Golden Delicious’ apples. The fungi isolated from the edge of the necrotic lesion resembled C. fioriniae in culture, and DNA extraction and amplification of the GAPDH gene showed 100% sequence identity to the original isolates.

To check for the abundance of endophytic infections in apple leaves, leaves were collected from 2 commercially fungicide treated trees and 1 untreated tree in 2018, and 3 commercially fungicide treated trees and 5 untreated trees in 2019. A few leaves of trees with fungicide treatments had endophytic infections, while endophytic infections were abundant in leaves of untreated trees (Figure 10). For 4 of the untreated trees in 2019, bitter rot incidence was obtained from 25 randomly selected apples at harvest and one month post-harvest and recorded as cumulative incidence. When compared with Student’s t-test, fungicide vs. non-fungicide treated leaves were different at p = 0.01, and a paired t-test of the 4 trees with both leaf and fruit data were different at p = 0.07 (figure 10).

These results indicate that the C. fioriniae strains that are common as leaf endophytes are identical to the C. fioriniae strains that are causing bitter rot of apple. This raises the question of whether forests and/or fence rows adjacent to orchards are sources of inoculum in orchards. If this were the case, we hypothesized that bitter rot incidence would be higher in the area of the orchard closest to the forest, with decreasing bitter rot incidence farther away. While we did not test this directly on the research farm, we surveyed apple growers, asking them about the patterns of bitter they see in their orchards. The grower survey was distributed in 2019 to apple growers through email, and in paper form at Penn State winter tree fruit meetings. Participants were asked to check a box if any forest or fence-row tree were growing within 1,000 feet of their apple orchard and if the pattern of bitter rot in orchards was uniformly distributed, a gradient, or in random “hot spots”.  

Of the 27 apple growers that filled out the survey on bitter rot distribution, 16 reported having forest and/or fence row trees within 1,000 feet of their orchard. A clear majority report bitter rot was either uniformly distributed or in “hot spots” with only one reporting a gradient across the orchard (Figure 11). The grower that reported a gradient also checked the boxes for uniformly distributed and “hot spots” and did not report forest and/or fence row trees within 1,000 feet of the orchard.

Given the evidence discussed above, we propose an infection cycle for C. fioriniae in the forest (Figure 12). It assumes that reproduction is dominated by asexually produced, rains-plashed conidia. Endophytic infections in leaves are hypothesized to be the main site for infections throughout the growing season. In fruit bearing plants, maturing infected fruits are also a factor. Buds, fruit mummies, fruit scars, and cankers serving as overwintering sites. This infection cycle is similar to previously published infection cycles (Everett et al. 2018) but with a greater emphasis on endophytic infections in leaves. It is also an infection cycle and not necessarily a disease cycle, only becoming a disease cycle on plants that are stressed or have susceptible fruits.

This hypothesized infection cycle has several implications for C. fioriniae as a plant pathogen. Since it is endemic in the forest canopy, eradication from nearby areas would be impractical.  However, the spatial limitations of rain-splash dispersal mean that forests are not regular sources of inoculum and likely only serve as initial inoculum sources during extreme rain and wind events, after which C. fioriniae becomes established in the agricultural area. This also lessens the risk of spreading fungicide resistance, since the source of inoculum is from non-fungicide treated areas. The abundance of endophytic leaf infections on agricultural plants means that eliminating weeds as potential inoculum sources may have minimal effect, since most of the inoculum is coming from endophytic infection in the host plant itself. There could be a potential for early season fungicide applications to reduce the amount of endophytic leaf infections as a way to reduce inoculum throughout the season. Finally, given that C. fioriniae is an endophyte on many plants for which is has not been identified as a pathogen, it appears that genetic resistance to necrotrophic infections is widespread, and that breeding for genetic resistance in agricultural plants may be the best long-term management strategy.

It should be reiterated that the proposed infection cycle is a hypothesis, to be tested and adjusted as more research results become available. The results presented in this research inform our management of C. fioriniae as a plant pathogen, and our understanding of its ecological role in forests. It adds to the growing body of evidence that the main ecological role of C. fioriniae is as a leaf endophyte of numerous broadleaf plants.

Detect the timing and incidence of dormant infections in apples using the freezing method.

Incidence of infection and bitter rot in 2019 was low overall, with no clearly defined period during which apples were resistant to infection (Figure 13). In 2020 only Captan was applied as a protectant fungicide instead of both Captan and Mancozeb as in 2019, and bitter rot incidence was higher overall, although final bitter rot incidence was still below 20% in all treatments (Figure 13). The results from this two-year study suggested there might be slight differences in apple susceptibility to bitter rot throughout the season, but that there is no period during which apples are resistant. Consequently, growers have to remain vigilant and apply fungicides to highly susceptible cultivars throughout the growing season, especially during warm wet weather.

A trend observed across both years was that infections were higher after the infection period, which was to be expected, and that bitter rot at and after harvest was lower than the incidence of infections after the infection period, which indicated that not all infections during the summer persist until harvest and cause bitter rot. A statistical comparison of combined 2019 and 2020 data was performed with ANOVA and a Tukey HSD test, not including the first two treatments of the season as those did not have pre-inoculation data or the negative controls which did not have an infection period. The results showed an increase of infection after the infection period (p < 0.0001), and a subsequent decrease from immediately after the infection period to final bitter rot incidence (Figure 14). The importance of this is not immediately obvious, but could be used to guide future research into practices that decrease the number of infections that persist to harvest. This decrease is unlikely to have come from post-infection period applications of fungicides, as the fungicides Mancozeb and Captan are strictly surface contact fungicides with no systemic activity.

2.      Track persistence of Colletotrichum (bitter rot fungi) and microbial succession in infected apples on the orchard floor.

Apples with bitter rot were placed out in the orchard in the fall of 2018 and 3 sample collections were taken of the rotting apples between October and March. The apples were almost completely degraded by May and no collection was taken that month. Apples with bitter rot were again placed out in the orchard in the fall of 2019, and 5 sample collections have been taken of the rotting apples between October and May. Percent dry matter of the samples across both years showed an initial rapid decrease in the fall after apples were placed on the ground, with slower decreases over the colder winter months (Figure 15).

Q-PCR tests of DNA extracted from the 2018/19 samples showed little decrease in total C. fioriniae DNA in decaying apples from October to March (Figure 16). The 2019/2020 samples included samples from apples with bitter rot in the tree canopy, which showed increases in C. fioriniae DNA as the rot spread through the apple with subsequent decreases from February to May (Figure 16). The other samples showed decreases in total C. fioriniae DNA in the period from February to May (Figure 16). These results demonstrated a decrease in C. fioriniae DNA as the apples decayed, as was expected, but also showed that apples on the ground the following spring still contained C. fioriniae DNA, albeit at much lower concentrations than in the fall. Whether this C. fioriniae DNA came from C. fioriniae that was still alive, or whether successional microbes had outcompeted and killed C. fioriniae and the DNA was simply residual DNA that had not yet degraded is not known. Given the abundance of C. fioriniae detected in spore traps in the spring, as shown in the previous section, it is possible that some of the C. fioriniae DNA detected in the decayed apples collected in the spring was not from the apple itself, but splashed onto the apple surface from above. Visual observations of the decayed apples revealed no orange spore masses indicative of C. fioriniae.

Metagenomic amplicon DNA sequencing of the fungal ITS1 barcoding region showed relatively high levels of Colletotrichum DNA in the initial samples, as was expected (Figure 17). The remaining fungi in the initial samples would have been mostly on the apple surface, as those apples were only starting to decay at that point. Subsequent samples show a rapid drop in the relative amounts of Colletotrichum DNA and rapid increases in other fungi, with spring collected samples from apples that were on the ground showing no Colletotrichum, while Colletotrichum persisted into the spring in samples from the tree canopy (Figure 17).

Classification of the fungal genera into filamentous or yeast categories showed that the vast majority of succession fungi that moved in after Colletotrichum were yeasts (Figure 18). This makes sense considering that yeasts grow well in conditions with lots of soluble sugars, which are abundant in mature apples. The biggest differences were between the apples on the ground and those in the tree canopy, with those in the tree canopy having relatively more filamentous fungi and a different population of yeasts (Figures 17 and 18).

There are several conclusions of this persistence and microbial succession of Colletotrichum in apples with bitter rot study. The successional fungi that move in after bitter rot starts the decay process are mostly yeasts. Yeasts rarely cause plant diseases and are generally considered beneficial. Some C. fioriniae DNA overwinters in apples with bitter rot, but for apples that are on the ground, the absolute and relative amounts of C. fioriniae DNA are low. Apples with bitter rot that remain in the tree canopy retain higher levels of C. fioriniae DNA and are likely sources for spores the following season. There were little differences between apples with bitter rot on the ground in grass-free herbicide strips, in grassed row middles, or in a nearby diverse plant mix. The key take-home message for apple growers is that apples with bitter rot should be removed from the tree canopy and tossed on the ground to decay. Exactly where on the ground the apples are tossed is unlikely to make a large difference.

References in this section: 

Børve, J., and Stensvand, A. 2017. Colletotrichum acutatum occurs asymptomatically on apple leaves. Eur. J. Plant Pathol. 147:943–948.

Carbone, I., and Kohn, L. M. 1999. A Method for Designing Primer Sets for Speciation Studies in Filamentous Ascomycetes. Mycologia. 91:553–556.

Debode, J., Van Hemelrijck, W., Baeyen, S., Creemers, P., Heungens, K., and Maes, M. 2009. Quantitative detection and monitoring of Colletotrichum acutatum in strawberry leaves using real-time PCR. Plant Pathol. 58:504–514.

Doyle, V. P., Oudemans, P. V., Rehner, S. A., and Litt, A. 2013. Habitat and Host Indicate Lineage Identity in Colletotrichum gloeosporioides s.l. from Wild and Agricultural Landscapes in North America. PLoS One. 8.

Garfinkel, A., and Chastagner, G. 2018. Identification of peony diseases in the Pacific Northwest and Alaska.

Hu, M.-J., Grabke, A., and Schnabel, G. 2015. Investigation of the Colletotrichum gloeosporioides species complex causing peach anthracnose in South Carolina. Plant Dis. 99:797–805.

Kumar, S., Stecher, G., Li, M., Knyaz, C., and Tamura, K. 2018. MEGA X: Molecular evolutionary genetics analysis across computing platforms. Mol. Biol. Evol. 35:1547–1549.

Marcelino, J. a P., Gouli, S., Parker, B. L., Skinner, M., Schwarzberg, L., and Giordano, R. 2009. Host plant associations of an entomopathogenic variety of the fungus, Colletotrichum acutatum, recovered from the elongate hemlock scale, Fiorinia externa. J. Insect Sci. 9:25.

Mertely, J. C., and Legard, D. E. 2004. Detection, Isolation, and Pathogenicity of Colletotrichum spp. from Strawberry Petioles. Plant Dis. 88:407–412.

O’Donnell, K., and Cigelnik, E. 1997. Two Divergent Intragenomic rDNA ITS2 Types within a Monophyletic Lineage of the Fungus Fusarium Are Nonorthologous. Mol. Phylogenet. Evol. 7:103–116.

Rhoads, A. F., and Block, T. A. 2007. The Plants of Pennsylvania: an Illustrated Manual. University of Pennsylvania Press.

Rojas, E. I., Rehner, S. A., Samuels, G. J., Van Bael, S. A., Herre, E. A., Cannon, P., et al. 2010. Colletotrichum gloeosporioides s.l. associated with Theobroma cacao and other plants in Panamá: multilocus phylogenies distinguish host-associated pathogens from asymptomatic endophytes. Mycologia.

Schneider, C. A., Rasband, W. S., and Eliceiri, K. W. 2012. NIH Image to ImageJ : 25 years of image analysis. Nat. Methods. 9:671–675.

Sogonov, M. V, Castlebury, L. A., Rossman, A. Y., Mejía, L. C., and White, J. F. 2008. Leaf-inhabiting genera of the Gnomoniaceae, Diaporthales. Stud. Mycol. 62:1–79.

Stephenson, S. A., Green, J. R., Manners, J. M., and Maclean, D. J. 1997. Cloning and characterisation of glutamine synthetase from Colletotrichum gloeosporioides and demonstration of elevated expression during pathogenesis on Stylosanthes guianensis. Curr. Genet. 31:447–454.

Swofford, D. L. 2003. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods).

Templeton, M. D., Rikkerink, E. H. A., Solon, S. L., and Crowhurst, R. N. 1992. Cloning and molecular characterization of the glyceraldehyde-3-phosphate dehydrogenase-encoding gene and cDNA from the plant pathogenic fungus Glomerella cingulata. Gene. 122:225–230.

Toju, H., Tanabe, A. S., Yamamoto, S., and Sato, H. 2012. High-coverage ITS primers for the DNA-based identification of ascomycetes and basidiomycetes in environmental samples. PLoS One. 7.

Weir, B. S., Johnston, P. R., and Damm, U. 2012. The Colletotrichum gloeosporioides species complex. Stud. Mycol. 73:115–180.

White, T. J., Bruns, S., Lee, S., and Taylor, J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In PCR Protocols: A Guide to Methods and Applications, New York: Academic Press, p. 315–322.