A PHENETIC ANALYSIS OF COLLETOTRICHUM GLOEOSPORIOIDES ISOLATES FROM SELECTED HOST PLANTS

Colletotrichum gloeosporioides is a ubiquitous fungus which infects a wide variety of plants in tropical, sub-tropical and temperate regions. This fungus accounts for substantial economic losses through out the world via both preharvest and postharvest diseases. A phenetic analysis of C. gloeosporioides isolated from Capsicum frutescens, Carica papaya, Mangifera indica, Persea americana, Ficus religiosa and Hevea brasiliensis was carried out to identify sub-specific populations. A total of 40 isolates from these six host species were used. The overall similarity among different isolates of C. gloeosporioides was determined using culture, conidial and appressorial characteristics. According to the resulting phenogram, fungal isolates had divided into two distinct groups at the initial stage separating C. papaya isolates from the rest of the isolates . The subsequent branching has lead to separation of C. gloeosporioides isolates of different hosts into distinct groups. A high degree of similarity was observed among the isolates obtained from C. frutescens , H. brasiliensis and F . religiosa. Similarly, isolates of P. americana and M . indica appear to be morphologically more similar to each other. Further, the study confirms the cross infection potential of some C. gloeosporioides isolates and the presence of host specific populations


INTRODUCTION
Colletotrichum gloeosporioides (Penz.) Penz. & Sacc. (teleomorph Glomerella cingulata (Stonem.) Spauld. & von Schrenk.) is a ubiquitous fungus which infects a wide variety of plants (Sutton, 1992) in tropical, sub-tropical and temperate regions. The pathogen can cause a number of diseases, anthracnose, leaf spot and seedling blight in cereals, legumes, ornamentals, vegetables and fruits. The most significant damage of this fungus occurs upon its attack on fruiting stage, leading to the incidence of anthracnose (Bailey et al., 1992).
The taxonomy of C. gloeosporioides has gained much attention due to the broad host range that this fungus exhibits. Since the plant pathogens are often named after the host plant, there are 17 acknowledged generic synonyms for Colletotrichum while there are about 600 synonyms for C. gloeosporioides. Von Arx in 1957 changed this system to a more systematic form via using morphological traits of different fungi. As a result Sutton in 1992, reduced the number of Colletotrichum species to nine; C. gloeosporioides, C. crassipes, C. lini, C. destructivum, C fuscum, C. fusarioides, C. phyllachoroides, C. paludosum, C. atramentarium, C. graminicola and C. dematium.
Morphological, growth, physiological and molecular differences have been used for taxonomic studies of Colletotrichum (Hong et al., 2008). The morphological characters which can be used for taxonomic purposes are limited to cultural, conidial and appressorial characters (Gunell and Gubler, 1992). Physiological characters that are commonly used for this pathogen are growth rate, virulence, germ-tube elongation and rate of appressorium formation.
Polymorphism in ribosomal RNA (rDNA) and mitochondrial DNA (mtDNA) has been used to assess the variability among populations of C. gloeosporioides that infect tropical fruits. These studies revealed that C. gloeosporioides isolates of Persea americana vary in rDNA and mtDNA banding patterns while isolates of Mangifera indica exhibit a similar rDNA and mtDNA banding patterns, independent from their __________________________________________ *Corresponding author's email: pubudug83@gmail.com geographic origin (Mills et al., 1992b). In addition, arbitrarily primed PCR (AP-PCR) technique can also be used for taxonomic studies of this fungus (Freeman et al., 1995).
Further, C. gloeosporioides isolates of different hosts consist of the cross inoculation capacity (Alahakoon et al., 1994). P. americana, M. indica and Nephelium lappaceum were most susceptible while Garcinia mangostana and Syzygium samarangense were least susceptible to C. gloeosporioides isolates of other crops. Although, variation on pathogenicity of C. gloeosporioides on different fruits has been observed through laboratory experiments (Freeman and Shabi, 1996) no field experiments on intact fruits have been carried out.
Therefore, a phenetic analysis of C. gloeosporioides isolates from different host plants was carried out with the objectives of identifying sub-specific levels or geographic populations whereas a cross inoculation study was done to distinguish cross infection potential of different isolates of C. gloeosporioides.

Isolation and growth of Colletotrichum:
Similar varieties of each of Capsicum frutescens, Carica papaya, M. indica and P. americana fruits exhibiting anthracnose symptoms were collected from a fruit stall. Diseased leaves of Ficus religiosa and Hevea brasiliensis with characteristic necrotic spots were field collected.
The fungus was isolated from symptomatic tissues of each host via cut separation of small square pieces of diseased tissues. These were surface sterilized in 0.5% chlorox (NaOCl) solution for 3-5 minutes followed by rinsing with sterile distilled water. The tissues were aseptically transferred onto PDA medium.
Pure cultures were maintained by subculturing at 14 day intervals. Conidia suspensions of each isolate were prepared by flooding 14 day old sporulating cultures with sterile distilled water and the mycelium was gently rubbed with a sterile glass spreader in order to dislodge conidia. The resulting suspension was filtered through glass wool. A loopful of conidia suspension was streaked across a thin plate of tap water agar and the plates were incubated at room temperature for about 18 hours. Monoconidial isolates were obtained from germinating spores on water agar and each isolate was given an acronym (C. frutescens -Cf; C. papaya -Cp; M. indica -Mi; P. americana -Pa; F. religiosa -Fr and H. brasiliensis -Hb).
Morphological studies: Seven day old starter cultures were examined under light and stereomicroscopes in order to identify suitable characters with variations in character status, for the phenetic analysis of C. gloeosporioides. This was repeated three times in order to overcome any deviations.
Colony characteristics: The increase in colony diameter was assessed by first demarcating the margins of the colonies each day and then measuring the diameter along two perpendicular axes. Colony colour was described using the degree of pigmentation of the colonies. In addition, nature of the colony margin, elevation of the colonies, presence /absence of sectoring and nature of the fungal mycelium were also recorded.
Conidial characters: Suspensions of conidia of each isolate were prepared and the concentration was adjusted to 1x10 7 conidia/ml using a haemocytometer. The conidia were examined under light microscope and the length and width of 100 conidia per isolate were measured using an eye piece graticule at x100 magnification. In addition, the shape of conidia and presence or absence of visible conidial masses were also recorded.
Characters of appressoria: Suspensions of conidia (1 x 10 5 conidia/ml) were prepared and 10 µl of the suspensions were placed on separate sterile glass slides and the slides were incubated in a moisture chamber for 10 -12 hours. Growth of appressoria was monitored and at the end of the incubation period a drop of cotton blue lacto phenol was added to arrest further development of the germinating conidia and stain the fungal structures. Finally the appressoria were examined and the length and width of 75 appressoria per slide were measured as described above and average lengths and widths were determined. In addition, shape and colour of appressoria were also recorded.
Cross inoculation studies: Diseased fruits of P. americana, M. indica and Colletotrichum infected leaves of H. brasiliensis of clone RRIC 100 were obtained. Pathogens were isolated on PDA as described previously. Conidia suspensions were prepared from 5 day old cultures. Drops (15 µl) of conidia suspensions from each of three isolates were applied on to ripe fruits of P. americana, M. indica, C. papaya and copper brown young H. brasiliensis leaves of clone RRIC 100 (Table 1). Controls were treated with 15µl drops of sterile distilled water and four replicates were used for each treatment. All the treated and control fruits were incubated at 100% RH and at room temperature (28 o C) until the characteristic anthracnose lesions developed. After six days of incubation, lesion diameter was measured.
The phenetic analysis: The characters with variation and non-overlapping character states were identified and coded into a data matrix. A cluster analysis and a Principal Component Analysis (PCA) was carried out using PC-ORD software.

RESULTS AND DISCUSSION
A total of 40 C. gloeosporioides isolates were studied for morphological and physiological data. The different characters and character states identified and coded for the phenetic analysis are given in Table 2.

Morphological studies:
Colony characteristics: The colony colour ranged from intense white to highly intense blackish ash. A distinct sequence of colour changes especially at the central region, was evident in C. gloeosporioides isolates obtained from C. papaya (Fig.1a) and F. religiosa (Fig. 1c) compared to the rest of the isolates. The lower surface of the cultures did not undergo a significant variation in colour with time.
Several different types of mycelia were found within the C. gloeosporioides isolates as dense aerial, even felted mat and compact aerial mycelium in tufts. Most of the colonies were grouped under dense aerial (Fig. 1c) and sparse aerial mycelium in tufts (Fig.1a) type while the even felted mat type was observed only in isolates obtained from C. frutescens (Fig.1b).
The elevation of the colonies also differed among the isolates. The highest elevation (>0.5 cm) was observed in the isolates obtained from C. papaya, M. indica and F. religiosa whereas the P. americana colonies showed the least elevation (<0.3 cm). The highest average growth rate ranged between 0.79 -1.1 cm/ day was shown by isolates obtained from C. papaya whereas the least value ranged between 0.56 -0.7 mm/day was shown by M. indica isolates.
Conidial morphology: The conidia shape was basically cylindrical with obtuse ends (Fig. 2a), except in isolates obtained from H. brasiliensis that consisted of few cylindrical conidia with obtuse ends and narrowed centre (Fig. 2b). The conidia lengths varied from 11.92 -15.67 µm among isolates whereas the conidia widths were more or less similar in most isolates. The highest mean conidia length (15. ± m) was in the isolates obtained from F. religiosa where G. cingulata, the teleomorphic state being the causal agent of the leaf spot disease. The least mean conidial length (11.92 ± 0.5 m) was recorded by isolates obtained from H. brasiliensis.
Characters of appressoria: During the present study all the isolates produced appressoria at the end of the germ-tube. The intensity of brown colour of the appressoria of some isolates increased drastically with time due to melonization. The colour of appressoria at the end of the incubation period was evaluated since under laboratory conditions C. gloeosporioides has proved to form three types of appressoria as hyaline, slightly pigmented and melanized (Estrada et al., 2000). In this study the most intense brown colour was shown by M. indica isolates whereas the faintest brown colour was shown by the P. americana isolate. However, non of the isolates produced hyaline appressoria.

Figure 2. Different conidia types observed among C. gloeosporioides isolates obtained from H. brasiliensis: (a) cylindrical conidia with obtuse ends and (b) cylindrical conidia narrowing at the centre (x 400).
Several differences were observed in isolates with respect to the morphology of appressoria. Most of the isolates consisted of lobed or round appressoria (Figs.3a and b). Ovate type of appressoria were produced only with C. frutescens isolate (Fig. 3c). The appressoria dimensions ranged from 5.0-6.8 m in width and 7.37 -11.05 m in lengths respectively, which is consistent with those of Du et al. (2005).

Cross inoculation studies: C. gloeosporioides
isolates when inoculated developed the largest lesions (8.5-7.4 mm diameter) on H. brasiliensis leaves, whereas the smallest lesions (4.1 mm) were observed in fruits of P. americana (Fig. 4). Isolates obtained from infected P. americana, M.
indica and H. brasiliensis produced lesions on all three hosts, C. papaya, M. indica and H. brasiliensis. In contrast, only the isolate obtained from P. americana could produce the characteristic anthracnose lesions on its original host, P. americana.
Phenetic analysis: Phenetic analysis includes numerical evaluation of the affinity/similarity between taxonomic units and ordering of these units into taxa on the basis of their affinities (Mayr, 1965).   . indica, (b) C. papaya, (c) H.  brasiliensis and (d) G. cingulata isolate from F. religiosa (10 x10 x 40). The 40 isolates separated into two distinct groups at the initial stage, separating C. papaya isolates (cluster A) from the rest of the isolates (cluster B). The formation of such a distinct groups may have resulted from the considerable morphological diversity that these isolates exhibited compared to other isolates. For example, wooly nature of the mycelium and intense blackish ash colour at the lower surface were observed only in these isolates. Further, conidia length of more than 13.99 m was recorded only from C. papaya isolates. A previous study conducted with molecular data has also obtained similar results supporting the separation of Colletotrichum isolates from different hosts (Mills et al., 1992b).
Considering the 25% similarity level, two major phenetic groups can be identified where cluster C includes isolates from C. frutescens, F. religiosa and H. brasiliensis while the other (cluster D) includes isolates from M. indica and P. americana. The unity of the isolates of C. frutescens, F. religiosa and H. brasiliensis indicates the presence of a high degree of similarity among them. These isolates share similar growth rates (0.7 -0.99 mm/day) and similar colony colour (off white) at the lower surface.
On the other hand, the isolates from M. indica and P. americana were grouped together to form cluster D, suggesting a close morphological similarity between these two sets of isolates. This may be due to the similarity among these isolates in colony colour at the upper surface which was of intense white colour, absence of sequence of colony colour change, conidia length (>13.99 m) and width (<3.5 m).
Cluster C has further separated into two groups at around 60% similarity level separating C. frutescens isolates (cluster E) from those of F. religiosa (cluster F) and H. brasiliensis. C. frutescens isolates consisted of white coloured, even felted mat of mycelium while the other two isolates consisted of moderate blackish ash and dense aerial mycelium. Further, in C. frutescens isolates the sequence of colour change at the upper surface was not observed unlike the other two sets of isolates. These features may have led to the separation of C. frutescens isolates, forming a distinct cluster.
Similarly, cluster D has separated into two groups at 40% similarity level, separating isolates of P. americana (cluster H) and M. indica (cluster G). P. americana isolates had moderate blackish ash colour lower colony surface whereas in M. indica the lower surface colony colour was off white. In addition, there were differences in the colonial elevation, colour of appressoria etc. between these two sets of isolates.
However, it is well established that the C. gloeosporioides isolates from M. indica show low levels of genetic variability in contrast to C. papaya and P. americana isolates. Both RFLP analysis of rDNA and Polymerase Chain Reaction (PCR) amplification of variable intergenic spacer region of the rDNA cluster have indicated that C. gloeosporioides isolates from M. indica do not exhibit geographically related groupings (Mills et al., 1992a(Mills et al., , 1992b. This may be either due to the introduction of the pathogen from a single source at wide scale or due to the selection pressure towards a single pathogen type. Cluster F has two groups around 75% similarity level separating isolates of F. religiosa (cluster I) and H. brasiliensis (cluster J) from each other. The elevation of F. religiosa colonies was more than 0.5 mm while the elevation of H. brasiliensis isolates was 0.31-0.5mm. Further, the appressoria colour of F. religiosa isolates was brown whereas in H. brasiliensis it was blackish ash. The origin of clusters I and J may have been due to these features.
Colletotrichum gloeosporioides isolates from different hosts have formed distinct groups within the phenogram. This may be due to the unique morphological and physiological character combinations which the different isolates possessed. This is in agreement with the findings of an earlier study of ribosomal DNA (rDNA) of C. gloeosporioides isolated from diverse hosts including P. americana, M. indica, C. papaya and Hevea brasiliensis covering different regions of the world, using RFLP (Mills et al., 1992b). The resulting rDNA repeat unit sizes were more or less similar for one particular host species. Such grouping is believed to be due to the separate evolution of C. gloeosporioides populations due to either ecological barriers or the rare sexual reproduction of this fungus. Both these situations are expected to limit the mixing of genes of distinct populations (Mills et al., 1992b). Identification of this type of grouping is important not only for the development of disease management practices but also for the investigation of resistant cultivars via plant breeding programmes (Manners et al., 1992). Further, the host specificity can be used for the benefit of field of agriculture. For example, C. gloeosporioides f.sp. malvae which is host specific has been used for biological control of Malva pussila, which is a common farmyard and green house weed (Mortensen, 1988). However, it has also been proved that the isolates obtained from the same host are not identical but there are variations among them with respect to biochemical and molecular characteristics (Maymon et al., 2006).
Despite forming distinct host specific groupings, C. gloeosporioides has the ability to cause cross infections (Freeman and Shabi, 1996). The cross infection potential of C. gloeosporioides has been studied under laboratory conditions (Alahakoon et al., 1994;Freeman and Shabi, 1996) where C. gloeosporioides isolates were found more virulent on the host which they were originally isolated from. The success of cross infection is proved to be dependant on the genetic adaptability of the fungus once in contact with a new host as well as the cultivar variety of the host specie. Crops such as avocado, mango, papaya etc. are often grown in mixed plantations. Therefore, it is important to have a sound knowledge of the host range and the cross inoculation capability of the fungus under natural conditions (Freeman et al., 1998).

Principal component analysis:
Principal component analysis was carried out using the data matrix resulting from the phenetic analysis. Different combinations of the first six dominant eigenvectors were examined. The first three axes together explain 77.74% of the total variance (Table 3). Upon consideration of the eigenvector values, the characters such as colony colour both at upper and lower surfaces, nature of the mycelium, colour of appressoria and average growth rate per day appear to contribute substantially towards the resolution among different isolates.
The best dispersion is given by the first two axes, which together explain 64.75% of the total variation (Fig. 6). The horizontal axis supports the separation of isolates obtained from C. papaya and M. indica from each other whereas, the vertical axis exhibits separation of isolates of H. brasiliensis from those of F. religiosa further strengthening the data obtained from the cluster analysis.