From July to October 2010, rubber leaves exhibiting typical symptoms of blight were collected in 3 regions - Selangor (Sungai Buloh), Perak (Sungai Biong), and Johor (Segamat, Kota Tinggi) - of Peninsular Malaysia. The climatic conditions in these areas are typically tropical, with a mean annual temperature of 26°C, average humidity of 80%, annual average rainfall between 2000-2500 mm (higher for East Malaysia than Peninsular Malaysia) with the northeast monsoon from October to March (wet period), and a southwest monsoon from May to September (dry period).

Samples were collected from rubber trees representing 7 clones in all 3 locations. Ten leaves were cut per tree, and all the samples were placed in paper bags and transported to the laboratory where they were processed after 1 day. Thin sections (5 mm in diameter) of infected tissues were cut with a sterile scalpel, and their surfaces were disinfected via immersion in a 70% ethanol solution for 1 min, rinsing with sterile distilled water, and drying the tissues on sterile blotting papers (Chmlab group, Spain). The disinfected pieces of leaves were placed in petri dishes containing potato dextrose agar (PDA) supplemented with 1 mg·ml⁻¹ streptomycin to minimize bacterial growth. The plates were incubated at 25°C under continuous white fluorescent light for 3 days. Once the fungal colonies were large enough to be examined, the hyphal tips of each colony were transferred to a new petri dish and were incubated as described above to obtain pure cultures. All the isolates obtained were maintained in the culture collection of the Department of Plant Protection, Faculty of Agriculture, Universiti Putra Malaysia (UPM), Malaysia.

Among the 33 isolates obtained, two strains originating from each geographical region and that have been proved to be virulent genotypes were used in this study. On the eighth day of growth, colour (using a colour scale) (Rayner, 1970), growth, and the presence or absence of conidiomata were recorded for colonies incubated at 26°C on PDA. To study the formation of pycnidia, the same isolates were cultured on sterile pine. The needles were autoclaved twice at 120 °C for 20 min and then placed on petri dishes containing 5 ml of molten water agar, which served as an adhesive. Petri dishes were incubated as described above for 12-15 days. Three petri dishes were used per isolate and medium, and the experiment was performed twice. The length and width of 50 conidia per isolate were measured at 400× magnification under a light microscope (Nikon Model Eclipse E200). The conidia extracted from pycnidia in water agar and pine needles were used for conidial measurement studies, which provided minimum and maximum sizes. Conidial colours, shapes, and septation were also recorded.

For each isolate, 6 mycelial plugs of a 10-d-old culture grown at 26°C were transferred into Erlenmeyer flasks containing 50 ml Difco potato dextrose broth (24 g/l). The flasks were shaken at 80 rpm at room temperature (28 ± 2°C) for 5 days by using an orbital shaker. The mycelia were filtered through a layer of cheesecloth and washed twice with distilled water. Dried mycelium (250 mg) was mechanically ground to a fine powder in liquid nitrogen by using a pre-chilled mortar and pestle. DNA extraction was performed using the method described by Moller et al. (1992). The concentration of the resulting DNA was determined using an ND-1000 spectrophotometer (NanoDrop 2000; Thermo Scientific, USA). The original DNA was then diluted in sterile distilled water to a concentration of 1 μg/μl and used in further reactions.

In order to completely characterize the isolates, the oligonucleotide primers ITS1 (5-TCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) (White et al., 1990) were used to amplify and sequence the internal transcribed spacer regions (ITS1 and ITS2) as well as the complete 5.8S rRNA gene. To obtain additional sequence information, another locus was also sequenced. The unknown locus (BotF15) containing microsatellite repeats was amplified with the primers BOT15 (5′-CTGACTTG TGACGCCGGCTC-3′) and BOT16 (5′-CAACCTGCTCA GCAAG CGAC-3′) (Slippers et al., 2004). To amplify each gene region, polymerase chain reaction (PCR) mixtures consisting of the following components were used: 1 μl of each primer (10 mM); 0.5 μl dNTPs (10 mM); 2 μl of MgCl₂ (25 mM); 2.5 μl of 10 mM reaction buffer containing MgCl₂ (25 mM), (Fermentas, USA); 0.5 U of Taq polymerase (Vivantis, Shah Alam, Malaysia); 5 μl of DNA; and 12.5 μl of sterile distilled water. The following amplification conditions were used: initial denaturation at 94°C for 3 min; 35 cycles of denaturation at 94°C for 1 min, annealing at 60°C for 1 min, and extension at 72°C for 2 min; and a final extension at 72°C for 10 min. PCR products (5 μl) were run on 2% agarose gels in a 1× TAE buffer solution (40 mM Tris, 20 mM acetic acid, and 1 mM EDTA) at 80 V for 45 min at room temperature. The gel was stained with ethidium bromide, and the bands were visualized under UV light and photographed using a gel documentation system (GeneSnap, ver 6.03, Syngene Laboratories). The size of the amplified DNA fragment was determined using molecular weight markers (GeneRuler 100 bp DNA Ladder and 1 kb DNA ladder marker [Fermentas, USA]).

The PCR products were purified using a GeneJET PCR Purification Kit (iNTRON biotechnology, Korea). The resulting amplicons were sequenced in both directions by NKH Biosciences Solutions, Kuala Lumpur, Malaysia, by using an automated sequencer. The nucleotide sequences were read and edited with MEGA version 4.0 (Tamura et al., 2007). The sequences of the strains isolated in this study were compared with those from other species in the Botryosphaeriaceae family that are available in GenBank (National Center for Biotechnology Information database). The sequences were aligned using the computer program MEGA version 4.0. Phylogenetic Analysis Using Parsimony (PAUP) version 4.0b10 (Swofford, 2003) was used to perform phylogenetic analyses for the dataset comprising sequences obtained from this study and with those of closely related species in the N. ribis/parvum species complex available from GenBank in order to determine phylogenetic relationships within individual and combined data sets. Gaps were treated as a fifth character, and all the characters were unordered and of equal weight.

To determine the relative pathogenicity of the fungal isolates, a single rubber leaf was set apart in a 15-cm-diameter petri dish and arranged by using a completely randomized design. One isolate was chosen per location (SB30, SK20, and SK10), and 2 rubber clones (RRIM 2002 and PB350) were used in the trial. For each isolate, a conidial suspension was prepared from 12-d-old cultures grown on PDA plates, and the concentration was adjusted to 3 × 10⁶ spores/ml using a haemocytometer. The leaves were inoculated with micropipettors by placing 10 droplets of the spore suspension onto the abaxial side. The leaves were then placed under continuous fluorescent light (20°C) under humid conditions, with 70-80% relative humidity. Four replicates of leaves were sprayed with a spore suspension of each isolate per rubber clone, and control leaves were inoculated with distilled water. The blighted leaves were assessed 5, 10, and 15 days after inoculation by using a scale (0-4), which represented the following: 0, no visible infection; 1, light lesion around inoculation point lacking spores; 2, large dark lesion without sporulation; 3, large lesion and weak sporulation; and 4, very strong sporulation on the upper and lower surfaces of the leaf. Leaves representing inoculated isolates or uninoculated controls were selected for fungal re-isolations from the edges of the lesion. This experiment was repeated twice for each isolate, and the mean number of lesions per leaflets in each experimental unit (4 leaflets) was computed and expressed as a disease severity index (DSI). The disease index (DI) for each leaf was calculated using the following formula (Ma and Michialides, 2002):

where i is severity (0-4) and Ni is the number of leaves with the severity of i.

The DI for each leaf in each isolate was calculated, and DI < 1 and 1 ≤ DI < 2 were used for classification as resistant (R) and susceptible (S), respectively. Data from all the experiments were analysed using the SPSS statistical package version 17.

Fungal isolation from the margins of infected leaves (Fig. 1A) resulted in the recovery of 68 isolates, including 33 resembling members of the Botryosphaeriaceae family, which were fast growers with a white mycelium that turned either dark greenish-grey or greyish within a few days. Overall, 1 type of fungal species belonging to Botryosphaeriaceae was identified by its culture and conidial morphology. These fungi were recovered from all clones commonly grown in Malaysia, including RRIM2024, RRIM901, RRIM2023, RRIM2025, PB350, PB260, and KT39/35. The highest incidence of these isolates was in Johor (16), followed by Perak (10) and Selangor (7). Multiple isolates from Botryosphaeriaceae were obtained from most clones in all of the rubber-growing regions studied.

The characteristics that led to the identification of this isolate were in agreement with previously reported descriptions (Nyaka et al., 2012) shown in Fig. 1B. Colonies were characterized by quick colony growth (16.5-17.0 mm/d), which covered the petri plate after 5 days. In addition, the white fluffy mycelium became slightly yellow or brownish on the undersides after 4 days. By day 5, the centres became light brown and appressed; the white fluffy mycelium was present only at the margins. Furthermore, pycnidia appeared within 7-12 days on PDA (Fig. 1C). Conidia produced on pine needles were initially hyaline and aseptate and became light brown and 1- or 2-septate with age. In addition, the conidia were ellipsoidal with round apices and truncate bases of 16.5-24.0 × 3.5-6.5 μm and a mean L/W ratio of 3.9 μm. No teleomorph structures were formed. The colony and conidial characteristics were typical of Neofusicoccum sp., as described on the Botryosphaeria website (http://www.crem.fct.unl.pt/botryosphaeria_site/).

All 33 isolates belonging to the Botryosphaeriaceae family were selected for ITS sequence comparisons to obtain a broad indication of their identities and select isolates for data sets used in the final analyses. We further refined the relationships and identifications of our isolates by performing phylogenetic analyses based on BotF15 sequences. DNA extraction and PCR were successfully performed for both gene regions selected. PCR fragments for the ITS region were determined to be approximately 550 bp in size, whereas those for BOTF15 were approximately 350 bp. BLAST searches against the NCBI (www.blast.ncbi.nlm.nih.gov) database confirmed that the isolates collected represented species in the Botryosphaeraiceae family and indicated that isolates from rubber trees were most closely (99%) related to N. parvum.

Data sets containing ITS and BotF15 sequences were compiled using sequences obtained from isolates of the Botryosphaeraiceae family in Malaysia and those from the N. ribis/parvum species complex. The ITS data set comprised a total of 18 isolates that is, sequences from 6 isolates (SB30, SB32 SK10, SK20, SB41, SB42) obtained from rubber trees, and 12 sequences from the N. ribis/parvum species complex obtained from the GenBank. The BotF15 data set consisted of 18 isolates, 6 of which were from rubber trees and 12 from the N. ribis/parvum species complex, representing the most complete database of BotF15 sequences in this family (Begoude et al., 2010; Pavlic et al., 2009).

Of the 568 characters in the ITS data set, 475 were constant and 9 were parsimony-informative. Maximum parsimony (MP) analyses generated 100 identical trees (TL = 98, CI = 0.969, RI = 0.786, and RC = 0.762). Isolates from the rubber tree were grouped in a single well-supported clade with N. ribis. During the analyses of the BotF15 data set, the sequences consisted of 364 characters with 356 constants and 7 parsimony-informative characters. MP analyses yielded one of the most parsimonious trees (TL = 8, CI = 1, RI = 1, and RC = 1). The tree generated from using both ITS and BotF15 sequences (data not shown) also clustered the isolates from the rubber tree within the N. ribis clade with strong statistical support.

Concordance among the ITS and BotF15 datasets was confirmed by the calculated incongruence length difference (I = 0), which suggested a lack of conflict between these gene genealogies. The datasets were combined, and a total of 867 bases was generated for the combined ITS and BotF15 datasets. Of these, only 16 characters were parsimony-informative. After heuristic searches, one of the most parsimonious trees of 19 steps (CI = 1, RI = 1, and RC = 1) was obtained (Fig. 2). The consensus tree obtained from the combined analysis of ITS and BotF15 sequences showed that the isolates collected from diseased rubber leaves formed a well-supported clade with N. ribis.

All the isolates were capable of infecting leaves, resulting in variable degrees of lesions. The isolates and clones exhibited highly significant differences (p < 0.001) (Table 1). Isolate SK20 produced the largest lesions, followed by isolates SK10 and SB30. Quantitative analysis of these lesions (disease index) indicated a considerable variation in susceptibility to N. ribis. The pathogenicity trials showed the high susceptibility of clone PB350 and the low susceptibility of RRIM2002 (Fig. 3). The leaves showed small light spots 5 day after inoculation. After 10 days, they became slightly brown. By 15 days, undulate margins appeared with pycnidia on the leaf surface and the leaves turned completely brown. Re-isolations from lesions on the inoculated leaves resulted in the recovery of the inoculated fungi.