Since the 1980s a new disease has been affecting Australian lychee. Pepper spot appears as small, black superficial lesions on fruit, leaves, petioles and pedicels and is caused by Colletotrichum gloeosporioides, the same fungus that causes postharvest anthracnose of lychee fruit. The aim of this study was to determine if a new genotype of C. gloeosporioides is responsible for the pepper spot symptom. Morphological assessments, arbitrarily-primed PCR (ap-PCR) and DNA sequencing studies did not differentiate isolates of C. gloeosporioides from anthracnose and pepper spot lesions. The ap-PCR identified 21 different genotypes of C. gloeosporioides, three of which were predominant. A specific genotype identified using ap-PCR was associated with the production of the teleomorph in culture. Analysis of sequence data of ITS and β-tubulin regions of representative isolates did not group the lychee isolates into a monophyletic clade; however, given the majority of the isolates were from one of three genotypes found using ap-PCR, the possibility of a lychee specific group of C. gloeosporioides is discussed.
Lychee or litchi (Litchi chinensis ssp. chinensis) is an evergreen subtropical fruit tree native to southern China that is now grown in a number of subtropical and tropical countries for its desirable high value fruit (Nakasone & Paull, 1998). After harvest, lychee fruit are very susceptible to a range of pathogens. Worldwide, one of the most common postharvest diseases of lychee is anthracnose, caused by the fungal pathogen Colletotrichum gloeosporioides and occasionally C. acutatum (Coates et al., 2005). Anthracnose of lychee manifests as a browning of the pericarp, where the infection is generally limited and the aril (the fleshy part of the lychee) is usually not affected and only occasionally collapses (Fitzell & Coates, 1995). Infection takes place in the field but generally remains quiescent until after harvest (McMillan, 1994).
Colletotrichum gloeosporioides is a ‘group species’, meaning that elements within the group cannot be separated satisfactorily on the basis of classical criteria such as conidial shape and size (Sutton, 1980), but recent advances have been made with the use of molecular characterization and pathogenicity testing (Cai et al., 2009). Similarly, there have been difficulties in identifying subdivisions within C. acutatum; however, using primarily ITS and β-tubulin sequencing, Shivas & Tan (2009) divided C. acutatum into C. acutatum, C. simmondsii and C. fioriniae.
In 1982 a disease caused by C. gloeosporioides affecting lychee fruit that were still hanging on trees was first recorded and by 1989 this disease, known as pepper spot, had been found in all the major growing areas in Australia (Drew & Drew, 2001). This disease is so called because symptoms appear as small (<1 mm diameter) raised black lesions on the surface of fruit, leaves and petioles (Cooke & Coates, 2002). The pepper spot symptoms tend to be more severe on the lower rather than upper branches of affected trees (Drew & Drew, 2001). Although pepper spot does not affect the flesh, it still leads to downgrading of the fruit quality (Cooke & Coates, 2002). Production costs are also increased because of the expense related to sorting out the second grade fruit. Pepper spot affects a number of varieties including Kwai May Pink, Bengal, Salathiel, Wai Chee and Tai So (Drew & Drew, 2001). Unlike anthracnose, there does not appear to be a quiescent period during the C. gloeosporioides infection process because symptoms appear during fruit development in the orchard (Cooke & Coates, 2002).
There are indications that pepper spot may be present in other countries in addition to Australia. McMillan (1994) described anthracnose of lychees in Florida as developing from a quiescent infection that occurs when the fruit are small. He made note of ‘incipient infections, which may be pinpoint in size on growing fruit’. He then described these infections as progressing to coalesce into brown spots as the fruit soften. Symptoms matching the description of pepper spot are also present on fruit in Taiwan and China (Ann et al., 2004; Y. Diczbalis, Department of Employment, Economic Development and Innovation (DEEDI), personal communication).
Symptoms similar to pepper spot (i.e. preharvest, limited lesion type) caused by Colletotrichum spp. have also been recorded on other subtropical and tropical fruit crops. In Australia, avocado pepper spot (also caused by C. gloeosporioides) has been recorded on the avocado cultivar Hass (Willingham et al., 2000) and in South Africa on Hass and Pinkerton cultivars (Schoeman & Manicom, 2000). The symptoms on avocado are similar to those seen on lychee, where the lesions appear as small (0·1–0·5 mm diameter) black raised lesions on the surface of fruit, twigs and pedicels (Willingham et al., 2000).
In avocado, pepper spot symptoms are often associated with fruit sunburn (Schoeman & Manicom, 2000), tend to be on the northern, more sun-exposed side of trees (in the southern hemisphere; Willingham et al., 2000) and are associated with conditions of high rainfall, high minimum temperatures and high minimum humidity which increase the period of time the fruit is wet after rain (Schoeman & Manicom, 2000). There are also some indications that host factors affect pepper spot development of lychee. There is some evidence of an effect of tree nutrition where trees with pepper spot tend to have lower levels of calcium in the leaves than trees without pepper spot and the symptoms are generally more severe on the lower rather than upper branches of the affected tree (Drew & Drew, 2001).
Despite the evidence that host and environment play a role in the development of pepper spot, they may not be the actual cause of the disease. Instead, favourable host and environment combinations may be likely to increase expression of the disease. The fact that the disease has only been reported in Australia since the 1980s suggests that a new genotype of C. gloeosporioides may be responsible for the development of pepper spot. The aim of the current research was to determine if there is a subpopulation of C. gloeosporioides causing pepper spot distinct from those isolates which cause anthracnose.
Materials and methods
Three commercial orchards growing Kwai May Pink lychee were selected for sampling: one in northern New South Wales (Brooklet 28°43′S 153°31′E), during February 2003 and two in north Queensland (Mena Creek 17°40′S 145°55′E and Toobanna 18°42′S 146°08′E), both during December 2003. At each orchard five trees were arbitrarily sampled. Five mature fruit with visible pepper spot symptoms from each tree were collected and at least 20 other fruit without visible pepper spot were collected from the same tree. The additional fruit were incubated to encourage development of anthracnose.
Fungal isolates from pepper spot symptoms were obtained by surface sterilizing fruit with 70% v/v ethanol, air drying and then excising pepper spot lesions with a scalpel and placing onto streptomycin amended half strength potato dextrose agar (SPDA); four isolations were performed per fruit. To obtain isolates from tissues with anthracnose symptoms, intact fruit were incubated in a moist chamber (25°C, 80% relative humidity) until anthracnose symptoms appeared on the fruit (generally after 3–7 days of incubation). Fruit peel was surface sterilized by washing in 2% w/v sodium hypochlorite for 2 min followed by repeated rinsing in sterile distilled water (SDW). Isolations were carried out by taking tissue (c. 3 mm2) from the internal side of the fruit skin at the edge of the lesion and plating onto SPDA; four isolations were performed per fruit.
For each putative Colletotrichum isolate, monoconidial cultures were generated from subcultures by making a conidial suspension in SDW of which 250 μL was 16-streaked onto water agar (WA). WA plates were incubated in the dark at 25°C for 12 h and then examined using dissecting (Wild M8) and compound (Olympus BH-2) microscopes. Single germinated conidia were excised from the WA using a sterile scalpel and transferred onto half strength potato dextrose agar (PDA). Once the isolate was established on the PDA, cubes of agar were stored under SDW in 5 mL screw-cap glass vials. Following identification all isolates were submitted to the Agri-Science Queensland (BRIP) culture collection.
Conidial measurements and determination of teleomorph production in culture
Fresh cultures were initiated from water-stored isolates by plating one piece of agar onto PDA or SPDA, placed in the dark for 2 days and then transferred to a cabinet at 12 h near-UV light (wavelength 300–380 nm)/12 h dark at approximately 24°C. After 3–4 days, cultures that did not start to produce conidial masses were scored with a sterile scalpel to encourage sporulation. When sufficient orange/pink conidial masses had been produced the conidia were measured by mounting on a glass slide in lacto-fuchsin (0·1% w/v acid fuchsin in lactic acid) and examined at ×1000 magnification using a light microscope of which the measurement scale in the eye piece had been calibrated using a stage micrometer. Only conidia from conidiomata were examined, as conidia produced directly from mycelium are not considered representative of the conidia for the culture (Cox & Irwin, 1988). The length and width of 20 conidia per isolate were measured and the mean value was taken for each isolate. The shape of the apices (both obtuse, both pointed or one obtuse and one pointed) was also recorded for each measured conidium. Conidial measurements were compared with published descriptions (Mordue, 1971; Dyko & Mordue, 1979). The undersides of culture plates were also examined to determine whether any isolates caused a change in the pigmentation of the media on which they were growing.
In order to examine the ability to produce the teleomorph in culture, cultures sporulating on PDA were streaked onto fresh oatmeal agar (OMA) plates. Plates were kept in the dark for 2 days before being transferred to near-UV light (12 h light/12 h dark). After at least a month the PDA and OMA plates were inspected using ×10 magnification under a dissecting microscope to determine whether perithecia (olive/brown spherical fruiting bodies) had been produced. Where olive/brown coloured perithecia were apparently found they were picked off using a needle or scalpel blade, mounted on a glass slide in a drop of lacto-fuchsin and gently squashed using a cover slip to confirm presence of ascocarps containing ascospores.
Data of conidial measurements were subjected to analyses of variance (anova, GenStat 9th edn) to determine whether there were morphological differences among isolates with respect to initial symptom type and geographical location.
Mycelia of the 150 monoconidial isolates of Colletotrichum were obtained for DNA extraction by inoculating potato dextrose broth (PDB) and maintaining cultures in the dark for 3–5 days at 24°C with occasional shaking. Mycelia were harvested by filtering through Miracloth (Calbiochem) under vacuum and then frozen at −70°C until DNA was extracted. Each isolate was grown in broth and extracted on two separate occasions so that there were duplicates of each isolate.
DNA extraction was based on the methods of Bentley et al. (1994). Mycelium was ground to a fine powder in liquid nitrogen and incubated in 1 mL extraction buffer (2% (w/v) sodium dodecyl sulphate (SDS), 40 mm ethylenediaminetetraacetic acid (EDTA), 40 mm NaCl, 100 mm Tris, 25 mm diethyldithiocarbamic acid) for 2 h at 37°C. The suspension was extracted twice with an equal amount of phenol, then extracted with 1 mL chloroform. DNA was precipitated by pipetting supernatant into 1·2 mL ice-cold 100% ethanol and 60 μL 3 m sodium acetate (pH 5·2) and storing at −20°C overnight. A pellet of DNA was formed by centrifuging at 10 000 g for 30 min.
The absorbance at 260 nm and the ratio of absorbance at 260:280 nm (Smart Spec Plus Spectrophotometer; Bio-Rad) were determined in order to quantify the DNA extracted and ensure template was free of contaminating proteins.
Arbitrarily-primed PCR (ap-PCR) was undertaken for each isolate using the methods and primers described by Freeman et al. (1996). The primers used were 5′-GACACGACACGACAC-3′ (abbreviated to (GACAC)3) and 5′-CAGCAGCAGCAGCAG-3′ (abbreviated to (CAG)5). Reactions were performed in a 20 μL solution containing 50 ng isolate DNA, 5× Green GoTaq PCR Buffer (Promega), 0·2 mm (each) dATP, dCTP, dGTP, and dTTP, 1 μm primer (GACAC)3 or (CAG)5 and 1 U of Taq DNA polymerase (Promega).
Reactions were carried out in a PTC-100 Programmable Thermal Control thermal cycler (MJ Research). The conditions were 95°C for 2 min, 30 cycles of 95°C for 30 s, annealing at 48°C ((GACAC)3) or 60°C ((CAG)5) for 30 s and 72°C for 1·5 min. Reactions were then cooled to 4°C for 10 min and held at 16°C.
PCR products were visualized by gel electrophoresis in 15 × 15 cm 1·5% w/v agarose gels running with 0·5× TBE (100 V for 120 min). Gels were post-stained with ethidium bromide and visualized under UV light (254 nm) using a Gel Doc XR (Bio-Rad) which was also used to capture an electronic image for data scoring.
Gels were scored for the presence and absence of bands and a data matrix was created using the program quantity one (Bio-Rad). The data matrices for each primer were combined and NTSYSpc v. 2.1 (Rohlf, 2000) was used to create a similarity matrix using the Jaccard coefficient (number of common bands/total number of bands in the two isolates being compared; Jaccard, 1901). The data were clustered using the unweighted pair group methods, arithmetic average (UPGMA; Michener & Sokal, 1957). A dendrogram was created from the cluster using NTSYSpc v. 2.1.
DNA sequence analysis
In order to examine the relationship of the lychee isolates to Colletotrichum from other hosts and to complement ap-PCR, DNA sequence analysis was undertaken for representative isolates (Table 1) from the ap-PCR study. The internal transcribed spacer regions 1 and 2 (ITS1 and ITS2) and 5·8S region of the ribosomal DNA were amplified using the ITS5 (5′-GGAAGTAAAAGTCGTAACAAGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) primers (White et al., 1990). A portion of the β-tubulin gene was amplified using Bt2a (5′-GGTAACCAAATCGGTGCTGCTTTC-3′) and Bt2b (5′-ACCCTCAGTGTAGTGACCCTTGGC-3′) primers (Glass & Donaldson, 1995). Reactions were performed in a total volume of 50 μL and contained 100 ng DNA, 5×Colourless GoTaq PCR Buffer (Promega), 0·2 mm (each) dATP, dCTP, dGTP and dTTP, 1 μm each of primers (ITS5 and ITS4, or Bt2a and Bt2b) and 1 U of Taq DNA polymerase (Promega). Reactions were carried out in a C1000 thermal cycler (Bio-Rad) with the following conditions: 95°C for 5 min, 35 cycles of 95°C for 30 s, annealing of 58°C (ITS) or 60°C (β-tubulin) for 30 s, and 72°C for 1 min, with a final extension step of 72°C for 10 min.
Table 1. Isolate code, species, host origin, country of origin and GenBank accession number for Colletotrichum spp. isolates used in maximum likelihood and Bayesian analysis in this study. Selection of isolates sequenced in this study was made on the basis of ap-PCR genotype (Fig. 2)
aFor the isolates collected as part of this study the isolate code describes where isolate is from: AL – Brooklet (NSW), RL – Toobanna (NQ), GL – Mena Creek (NQ). PS indicates isolate is from a pepper spot lesion, AN from anthracnose. The last two digits are tree and fruit number. For example, ALPS11 is the isolate from Brooklet from a pepper spot lesion on tree 1, fruit 1.
bType specimen for C. gloeosporioides.
cBotryosphaeria dothidea included as an out-group.
PCR products were purified using High Pure PCR Product Purification Kit (Roche) following manufacturer’s instructions and then 10 μL of purified PCR product was run out on a 1·0% w/v agarose gel (0·5×TBE run at 100 V for 45 min, post-stained with ethidium bromide) against a molecular ladder standard (1 kb GeneRuler, Fermentas) to estimate product concentration. Samples were labelled and sequenced at the Australian Genome Research Facility (AGRF), Brisbane using an ABI 3730xl 96-capillary automated DNA sequencer (Applied Biosystems). Sequencing was done in both forward and reverse directions and resulting data were assembled into single contigs using vector NTI (Invitrogen).
A search was made of the GenBank database (http://www.ncbi.nlm.nih.gov/nucleotide/) for ITS and β-tubulin sequence data of isolates of C. gloeosporioides, C. acutatum and C. fragariae from lychee, citrus, mango, avocado, olive, capsicum and strawberry to include in the analysis with isolates sequenced in this study. Preference was given for isolates with both ITS and β-tubulin sequence data from published papers (Table 1). The sequence data from an isolate of Botryosphaeria dothidea from European ash was included as an out-group (Table 1).
Sequences were aligned using multiple sequence comparison by log-expectation (muscle v. 3.7; Edgar, 2004) using online software available from Methods and Algorithms in Bioinformatics (MAB) at The Montpellier Laboratory of Informatics, Robotics, and Microelectronics (LIRMM) (http://www.phylogeny.fr/version2_cgi/programs.cgi). Alignments were checked manually before curation with Gblocks v. 0.91b (Castresana, 2000), also available online from LIRMM. Sequence data for ITS and β-tubulin were combined and a maximum likelihood (ML) tree was constructed using the computer program RAxML v. 7.2.6 (Stamatakis, 2006) using the general time reversible (GTR) model of evolution (GTRMIX setting) and 1000 bootstrap replications; the resulting tree with bootstrap support values was viewed using FigTree v. 1.3.1. RAxML allowed the partitioning of data for each gene sequenced and also enabled inclusion in the analysis of accessions which were missing β-tubulin data.
Bayesian Markov chain Monte Carlo (MCMC) analysis (MrBayes v. 3.1, Ronquist & Huelsenbeck, 2003) was performed with the ITS and β-tubulin data with and without the addition of ap-PCR data. An advantage of Bayesian MCMC analysis is that data for each gene can be partitioned to allow for a different model of evolution for each gene. It also enables the inclusion of standard data, such as the presence and absence data of ap-PCR or morphology. Like the ML analysis of RAxML, Bayesian MCMC analysis is able to compensate for missing data, such as that of the missing β-tubulin data. For each analysis four chains were run: one cold and three heated with the temperature parameter of the analysis set at 0·25. Data were partitioned into ITS and β-tubulin with GTR model of evolution assumed, ap-PCR data were analysed under a standard discrete model, and all parameters estimated separately for each partition. The ML tree constructed using RAxML was used as a starting tree for each analysis. The analysis was run for 1 × 108 generations with tree sampling every 100 generations. The analysis was stopped when the standard deviation of split frequencies dropped below 0·03. Convergence was checked by examining the overlay plot for all four runs and ensuring the potential scale reduction factor was approaching one. The resulting consensus tree displaying posterior clade probabilities was viewed using FigTree v. 1.3.1.
Isolates were given a code which describes where each isolate is from: AL describes isolates from Brooklet (New South Wales, NSW), RL–Toobanna (Northern Queensland, NQ), GL–Mena Creek (NQ). PS indicates isolate is from a pepper spot lesion, AN from anthracnose. The last two digits are tree and fruit number. For example, ALPS11 is the isolate from Brooklet from a pepper spot lesion on tree 1, fruit 1.
Isolates were classified morphologically either as C. gloeosporioides based on the description by Mordue (1971) or as C. acutatum as per the description of Dyko & Mordue (1979). Isolates were considered to be C. gloeosporioides when most conidia had obtuse apices (Fig. 1a) and C. acutatum when the majority of the conidial apices were pointed (Fig. 1b). Further differentiation was made on the basis of C. gloeosporioides having conidia with an average width greater than 3·70 μm (Fig. 1a). Of the total 150 isolates collected, 141 were classified as C. gloeosporioides and nine as C. acutatum. Some of those designated as C. acutatum exhibited conidia that were also medianly constricted (Fig. 1b). All of the C. acutatum isolates (GLAN11, GLAN13, GLAN23, GLAN24, GLAN25, GLAN34, GLAN41, GLAN42 and GLAN51) were from anthracnose lesions on fruit collected from Mena Creek. None of the C. acutatum isolates caused a change in the pigmentation of the media on which they were growing.
Analysis of variance of the data from each of the C. gloeosporioides isolates showed no significant differences in length and length:width ratio but there were significant differences in conidium width. Conidia of isolates from pepper spot and anthracnose symptoms on fruit from Mena Creek were significantly wider than pepper spot isolates on fruit from Toobanna (Table 2). Conidia from Toobanna anthracnose isolates were significantly wider than those from Toobanna pepper spot isolates. Conidia from Mena Creek pepper spot isolates were significantly wider than conidia from anthracnose and pepper spot isolates from Brooklet (Table 2). When analysis was performed ignoring location and simply comparing pepper spot isolates to anthracnose isolates, there were no significant differences between the two groups based on conidial morphology (data not shown).
Table 2. Conidium measurements of Colletotrichum gloeosporioides isolates from lychee grouped according to symptom type and geographic location for isolates collected in this study
Average conidium length (μm)
Average conidium width (μm)
Length to width ratio
aLetters following means indicate significance at P =0·05. Due to the unequal number of replicates after the removal of the C. acutatum isolates data, the anova was performed using each location × symptom as a single treatment; n =16 for Mena Creek–anthracnose, n =25 for all other treatments. The average LSD is quoted in the table for comparisons of the average conidium width.
Brooklet – anthracnose
Brooklet – pepper spot
Mena Creek – anthracnose
Mena Creek – pepper spot
Toobanna – anthracnose
Toobanna – pepper spot
The production of the teleomorphic stage Glomerella cingulata was not limited to isolates from one symptom type or location. Eighteen of the 150 isolates produced asci and ascospores (Fig. 1c), seven of which were from pepper spot lesions and 11 from anthracnose lesions. The majority of the isolates that produced the teleomorph were from fruit collected at Mena Creek (14 isolates: GLPS12, GLPS13, GLPS22, GLPS24, GLPS44, GLPS54, GLPS55, GLAN14, GLAN21, GLAN22, GLAN31, GLAN43, GLAN45 and GLAN53). There were considerably fewer isolates producing the teleomorph from Brooklet (four isolates: ALAN33, ALAN35, ALAN41 and ALAN45) and none from Toobanna. None of the C. acutatum isolates produced the teleomorphic stage G. acutata under the conditions used in this study.
Clear, reproducible banding patterns were produced with both ap-PCR primers used. Amplification with (CAG)5 produced 41 polymorphisms (Fig. S1) whilst amplification with (GACAC)3 produced 44 polymorphisms. Similar dendrograms were obtained for each primer individually and thus combined data is presented. The combined data resulted in 85 polymorphisms used to identify 25 genotypes (Fig. 2). Some genotypes, ALAN12 and ALAN24 designated as Genotypes 3 and 4, were distinguished on the basis of one polymorphism.
The C. acutatum isolates grouped separately to the C. gloeosporioides isolates and contained four genotypes (Fig. 2). There was only 23% similarity between the C. acutatum isolates and the C. gloeosporioides isolates.
The ap-PCR analysis did not differentiate the pepper spot isolates from the anthracnose isolates. The majority of the isolates (73%) grouped together with 75% similarity to each other (Fig. 2, branch indicated by asterisk). Within this group were isolates from both pepper spot and anthracnose as well as isolates from all three locations and a total of nine genotypes (Genotypes 1–9), none of which had isolates that produced the teleomorph in culture. However, for all of the isolates there was grouping according to location. Every genotype except for Genotype 6 contained isolates from only one location. Genotype 6, which had by far the greatest number of isolates (60), still only contained isolates from north Queensland (Mena Creek and Toobanna), with no isolates from Brooklet.
Most other genotypes identified had very few isolates, most comprising only one isolate, except for Genotype 14 which contained 15 isolates all from Mena Creek. Of these, 14 isolates produced the teleomorph under laboratory conditions. Both pepper spot and anthracnose isolates were found in this group.
Single lychee trees often contained two and sometimes three genotypes including both C. acutatum and C. gloeosporioides.
DNA sequence analysis
The isolates selected for DNA sequence analysis were representative of the different clades of lychee isolates found using ap-PCR (Table 1; Fig. 2). Also included was a mango isolate of C. gloeosporioides (BRIP 28734). All isolates selected amplified with both primer sets (Bt2a and Bt2b; ITS4 and ITS5). The size of the product amplified with ITS5 and ITS4 was approximately 600 bp for the C. gloeosporioides isolates, with the C. acutatum isolate GLAN11 giving a slightly larger product (results not shown). Amplification with Bt2a and Bt2b resulted in a product of approximately 500 bp with again a slightly larger product amplified from GLAN11, the C. acutatum isolate (results not shown).
For the C. gloeosporioides isolates there were 15 polymorphic nucleotide sites (of 472 sites analysed) in the ITS region and 16 polymorphic nucleotide sites (of the 363 sites analysed) for the portion of the β-tubulin gene sequenced. Some pepper spot and anthracnose isolates had identical sequence data. β-tubulin sequence data were identical for ALPS11 (Genotype 1), ALPS15 (Genotype 1), ALAN11 (Genotype 1), ALAN12 (Genotype 3), RLAN11 (Genotype 6), RLPS24 (Genotype 6) and RLPS33 (Genotype 9; data not shown). ALPS11, ALPS15, ALAN11 and ALAN12 had identical ITS sequence data which differed from the ITS sequence data of RLAN11, RLPS24 and RLPS33 by only two nucleotides. The ITS sequence of the C. gloeosporioides isolate from lychee in Japan differed from the Australian lychee isolates ALPS11, ALPS15, ALAN11 and ALAN12 by three nucleotides.
blast searches for the ITS and β-tubulin data for GLAN11 (ITS and β-tubulin) and Cg2 from mango from Taiwan (ITS only) showed 99% similarity to the C. simmondsii type specimen as described by Shivas & Tan (2009). The blast searches for the C. gloeosporioides isolates sequenced in this study showed a range of similarity of 98–99% to the C. gloeosporioides type specimen for ITS and 97–98% for β-tubulin data.
When sequence data from each of the regions were analysed separately there was low statistical support for the clades, but maximum likelihood tree topology was similar for ITS and β-tubulin data (data not shown). When posterior probabilities were calculated using MCMC analysis, there was stronger statistical support for the tree generated with inclusion of ap-PCR data, hence this tree is presented (Fig. 3). The C. acutatum isolates formed a distinct clade with high posterior probability of 1, and the C. fragariae isolate also separated out with a high posterior probability of 1 (Fig. 3). There were two clear monophyletic clades formed from the rest of the C. gloeosporioides isolates; a group containing the mango isolates (posterior probability of 0·7) and a group containing the olive isolates (PT111, VM206), the citrus isolate (CBS953 97) from the Mediterranean and two lychee pepper spot isolates from Australia (RLPS45 and GLPS12) with a posterior probability of 0·9. The lychee C. gloeosporioides isolates did not form a monophyletic clade but the isolates from genotypes 1–9 found using ap-PCR (ALPS11, ALPS15, ALAN11, ALAN12, RLPS24, RLAN11 and RLPS33) formed a distinct clade with a posterior probability of 0·54. However, there was still insufficient resolution for the lychee isolates GLPS23, ALAN33, GLAN15, 239930 and the avocado isolate CG_388.
This study has confirmed that the majority of isolates collected from the three lychee orchards for both anthracnose and pepper spot symptoms were C. gloeosporioides. Analysis of conidial morphology and incidence of teleomorphic stage showed no distinctions between isolates from pepper spot with those from anthracnose. There were significant differences for conidium width between pepper spot and anthracnose isolates from different locations; however, these were not consistent across all locations. This is different to anthracnose of the tropical pasture legume Stylosanthes where isolates causing two different diseases can be distinguished on the basis of morphology (Irwin & Cameron, 1978). Type A anthracnose of Stylosanthes appears as lesions on stems, leaflets and flowers affecting a range of cultivars and conidia are uniform in size with obtuse ends. In contrast, Type B anthracnose only affects S. guianensis and manifests as a general necrosis of the terminal shoots and conidia are of variable size and tapered at one end (Irwin & Cameron, 1978).
Using ap-PCR, it was found that isolates of C. gloeosporioides causing pepper spot and anthracnose on lychee fruit are not genetically distinct from one another, and in some cases have identical ap-PCR genotypes. This is similar to the findings of Giblin (2006) where isolates of C. gloeosporioides from avocado pepper spot and anthracnose had identical DNA fingerprint patterns. It is likely that the appearance of pepper spot is due to the lychee plant responding differently to infection by the same pathogen, depending on host factors such as tree health and nutrition or environmental factors.
The ap-PCR identified numerous genotypes within the C. gloeosporioides isolates but only three genotypes were predominant. Of the 150 isolates in this study, 73% of the isolates grouped together into a large clade using ap-PCR, showing a similarity of about 75%. When representative isolates from this group were sequenced, the β-tubulin gene sequences were identical and there were only two single nucleotide differences for the ITS region. This kind of limited diversity has been observed in populations of C. gloeosporioides from mango, where studies have suggested that the population of mango C. gloeosporioides worldwide shows host specialization (Alahakoon et al., 1994). It is thought that this mango population has been carried endophytically in mango seed around the world (Hayden et al., 1994). The reason why the majority of the lychee C. gloeosporioides isolates show such limited diversity could be for a similar reason. Lychee plants are propagated clonally and it is possible that specific genotypes of C. gloeosporioides have been carried from farm to farm, on clonally propagated planting material. Pathogenicity testing and further sequencing needs to be done to resolve the place of the lychee C. gloeosporioides isolates in the C. gloeosporioides group and determine if there is a subpopulation of C. gloeosporioides associated with lychee.
Ap-PCR genotype appears to be related to the ability to produce the teleomorphic stage in culture. None of the closely related isolates from Genotypes 1 to 9 produced the teleomorph in culture, whilst 14 of the 15 isolates of Genotype 14 produced the teleomorph. The link between genotype and teleomorph production in culture has been found a number of times for C. gloeosporioides. This link was so strong for apple-infecting isolates of C. gloeosporioides that the authors speculated that the G. cingulata and C. gloeosporioides isolates represented two genetically isolated taxa (Gonzalez et al., 2006). In a study on C. gloeosporioides from subtropical fruit, Hayden et al. (1994) found isolates that produced the teleomorph in culture grouped together in RAPD analysis regardless of the host from which the isolate was collected. However, in a large study using DNA amplification fingerprinting of avocado-infecting C. gloeosporioides isolates, Giblin (2006) found no link between clade or genotype and teleomorph production.
Nine isolates of C. acutatum were identified on the basis of morphology and supported by the molecular characterization. These isolates were all from anthracnose lesions and were recovered from one farm out of the three assessed. The low incidence of collection of C. acutatum in this study is consistent with the observations of Johnson as reported by Coates et al. (2005) that C. acutatum plays a relatively minor role in lychee anthracnose and the main causal pathogen is C. gloeosporioides.Shivas & Tan (2009) recently subdivided C. acutatum into C. acutatum, C. fioriniae and C. simmondsii. A single isolate of C. acutatum was sequenced in the present study. When the ITS region of this isolate was blast searched it showed 99% similarity to C. simmondsii. This is unsurprising as GLAN11 (BRIP48724) collected as part of this study was included in the study of Shivas & Tan (2009).
The results of this paper demonstrate that the isolates of pepper spot from lychee cannot be distinguished from the isolates of anthracnose on the basis of conidial morphology, ap-PCR patterns nor on ITS and β-tubulin gene sequence data. Thus it appears that the development of pepper spot in lychee may be under the control of factors other than pathogen genotype, for example host and/or environmental factors.
The authors are grateful to Australian Fruit Producers (Brooklet), the Gatteras (Mena Creek) and Rowells (Toobanna) for access to their orchards and fruit for isolate collection. Alistair McTagget is thanked for his patience in teaching the use (and joys of) of RAxML and MrBayes. JA was a recipient of a Department of Agriculture, Fisheries and Forestry Science and Innovation Award; study leave was provided by the Department of Employment, Economic Development and Innovation. Two anonymous reviewers are acknowledged for their suggestions for improvements of the manuscript.