Characterization of sooty blotch and flyspeck fungi on mango (Mangifera indica L.) in Peninsular Malaysia

Abstract

Sooty blotch and flyspeck (SBFS) disease complex, characterized by black smudges and clusters of spots on fruit surfaces, is commonly observed on mango. This study aimed to characterize the SBFS fungi associated with mango in selected orchards in Peninsular Malaysia using morphological, molecular, and pathogenicity approaches. A total of 87 symptomatic fruits were collected from eight orchards in Perak, Pahang, and Perlis during sampling conducted between October and November 2019. The number of colonies per fruit (CPM) and the percentage of fruit surface covered by SBFS were recorded. Thirty-one purified fungal cultures representing four genera (Zasmidium, Peltaster, Exophiala, and Scolecobasidium) were identified using morphological characters. Phylogenetic analysis of the ribosomal DNA sequences of the internal transcribed spacer (ITS) and large subunit (LSU) regions delineated three genera (Zasmidium, Exophiala, and Scolecobasidium) and a previously unknown Peltaster-like species. Twelve isolates were confirmed as SBFS pathogens using a modified version of Koch’s postulates on asymptomatic fruit with two inoculation methods (1) in situ in an orchard and (2) on excised fruit. Significant interactions were observed between states and infection levels (CPM and surface coverage) (P = 0.0001), as well as between states and fungal genera (P = 0.004). This study provides the first comprehensive characterization of the taxonomic diversity and pathogenicity of SBFS fungi on mango in selected regions of Malaysia, providing a foundation for future nationwide studies.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-38319-8.

Introduction

The sooty blotch and flyspeck (SBFS) disease complex causes dark blotches and clusters of black dots on the fruit surface, which reduces fresh-market value by degrading fruit appearance¹. SBFS colonies are commonly found on the leaves of trees and shrubs², as well as on fruits with waxy surfaces, such as apple^3–5, banana^6,7, mango^8,9, pawpaw¹⁰, and persimmon¹¹. Xu et al.¹² examined SBFS pathogen-host interactions using Peltaster fructicola E.M. Johnson, T.B. Sutton & Hodges on apple and coined the term ‘ectophytic’ to describe the parasitic lifestyle of SBFS fungi; SBFS fungal colonies degrade the epicuticular wax layer of apple fruit but do not penetrate the epidermal cell wall.

Of the >100 described and putative species of SBFS fungi, most belong to the order Capnodiales with relatively few species in Chaetothyriales¹. Most SBFS fungi belong to the phylum Ascomycota, primarily within the class Dothideomycetes (order Capnodiales), while a few species occur in Eurotiomycetes and Basidiomycota, such as Wallemia sebi⁵. Genera frequently associated with SBFS include Zasmidium, Peltaster, Microcyclospora, Microcyclosporella, Phaeothecoidiella, Houjia, Scolecobasidium, Devriesia, and Cyphellophora¹. Comparative studies across different countries, such as the United States, China, Germany, and Brazil indicate that Zasmidium and Peltaster species are among the most widespread and dominant members of the complex^1,3,4. Although regional variation exists in species assemblages, Zasmidium citri-griseum, Peltaster fructicola, and several Exophiala and Scolecobasidium species are frequently detected from fruit surfaces of temperate and tropical hosts, including apple, banana, and mango^4,7,9. These findings confirm that the SBFS complex represents one of the most diverse and globally distributed groups of ectophytic fruit-colonizing fungi¹. While certain SBFS species have a broad geographic distribution, others are restricted to specific regions, influenced by local climatic conditions and orchard management practices^4,13.

Mango production contributes substantially to both local and export markets in Malaysia, with production volumes of 1,751  metric tons (mt), 1,642 mt, and 2,151 mt in 2022 in the states of Perak, Pahang, and Perlis, respectively¹⁴. SBFS was recently reported on mango in Perak⁹. In the United States (U.S.), growers often divert SBFS-infected apples to processed products, resulting in revenue losses of up to 90% of their fresh-market value¹⁵. Mango growers in Malaysia encounter similar challenges. For example, a mango grower in Perak reported that the premium quality of Chok-Anan mango can be sold for the price of Malaysian Ringgit (MYR) 6/kg. However, due to SBFS infection degrading the fruit’s appearance, the selling price could be reduced to MYR 3/kg (personal communication).

The diversity of SBFS fungi has been documented in the U.S⁴., Turkey⁵, Norway¹³, Spain¹⁶, Germany¹⁷, and India¹⁸. However, information on the distribution and the SBFS species composition on mango in Malaysia is sparse. Knowledge of the species composition and etiology of SBFS disease on mango is an essential precursor for developing appropriate control strategies to mitigate SBFS infection. Therefore, the objective of this study was to characterize the SBFS fungal diversity of SBFS on mango in several Malaysian states: Perak, Pahang, and Perlis.

Results

Sooty blotch and flyspeck disease on mango

A total of 87 mango fruits showing symptoms of the SBFS disease complex were collected from eight orchards across three states in Malaysia: Perak, Pahang, and Perlis. The affected fruit exhibited black blotches and, in some cases, small black dots on the surface (Fig. 1).

Fig. 1.

Mango fruit infected with SBFS disease observed during orchard sampling in three states.

All sampling locations were dry during the sampling period. The mean temperature and relative humidity (RH) recorded across all sampling sites ranged from 27 to 28 °C and 90 to 93%, respectively (Table 1). Figure 2 presents representative views of the sampling locations from each state, highlighting observable differences among the sampling locations, such as mango orchards located near oil palm plantations in Perak, open-field orchards in Pahang, and bagged fruit management practices in Perlis.

Table 1.

Mean colonies per mango and mean percentage area of infection associated with different SBFS fungal genera causing SBFS disease on mango in Malaysia.

State	Location	Mean temperature (°C)a; Mean RH (%)b	No. of mango fruit collected	Variety	Mean CPMc	Mean percentage area of SBFS infection (%)	Fungal genera recovered
Perak	4°08’05.0"N, 100°45’35.4"E	27.9; 91.3	15	Chok Anan	420	49.2	Peltaster-like, Exophiala
	4°08’25.0"N, 100°45’47.8"E		8	Chok Anan	270	50.8	Peltaster-like, Zasmidium, Scolecobasidium
	4o08’18.3"N, 100o45’48.1"E		4	Chok Anan	125	32.1	Exophiala
Pahang	3o30’49.8"N, 102o34’01.2"E	26.8; 89.9	10	Chok Anan	285	29.2	Zasmidium
	3o33’10.6"N, 102o15’03.7"E		17	Chok Anan	195	29.1	Zasmidium
Perlis	6o28’35.7"N, 100o14’00.9"E	28.4; 92.5	19	Harumanis	129	7.8	Zasmidium
				Chok Anan	147		Zasmidium
	6o31’14.9"N, 100o10’14.3"E		6	Chok Anan	180	13.4	Zasmidium
	6o29’55.2"N, 100o11’15.6"E		8	Chok Anan	142	6.6	Zasmidium

^aMean temperature from different sampling locations within the same state. ^bMean RH from different sampling locations within the same state. ^cMean colonies per mango collected from three states in Malaysia.

Fig. 2.

Representative sampling locations of mango orchards in Malaysia. (A) Orchard in Perak bordering oil palm plantations (arrow). (B) open-field orchard in Pahang. (C) Mango fruit covered with fruit bags in Perlis (arrows).

Mycelial types on mango

Four mycelial types were observed under the microscope on all the 87 collected SBFS-infected mango fruits. Mycelial types were: (1) Ramose—sclerotium-like bodies of interconnecting cells that were larger, shinier, and more convex at the centre of the colony than at the margin, and dark thread-like mycelial mats; (2) Punctate—dark brown, multi-hyphal stranded sclerotium-like bodies uniformly distributed across the mycelial mats; (3) Fuliginous—homogenous mycelial mats with abrupt to feathery colony borders; and (4) Ridged honeycomb—clumps and ridges of mycelia occasionally appearing like honeycomb clusters, or unorganized sclerotium-like structures (Fig. 3).

Fig. 3.

Four different mycelial types observed on the SBFS-infected mango fruit collected in Perak, Pahang, and Perlis. (A) Ramose. (B) Punctate. (C) Fuliginous. (D) Ridged honeycomb.

The most predominant mycelial types observed on mango were ramose and punctate, followed by fuliginous and ridged honeycomb. The visual appearance of each of the mycelial types was consistent across observed mango fruits in all three states.

Fungal identification and phylogenetic analysis

Based on morphological evidence, the 31 total isolated SBFS fungi were attributed to four SBFS fungal genera: Zasmidium (15 isolates), Peltaster (12 isolates), Exophiala (3 isolates), and Scolecobasidium (1 isolate). Of the 2,106 mango peels, 31 pure cultures were obtained (1.47% success). These isolates were obtained from punctate and ramose mycelial types, whereas isolations from fuliginous and ridged honeycomb did not yield pure cultures. Twelve isolates, representing each of the four SBFS fungal genera were selected based on morphology on media: Zasmidium (6 isolates); Peltaster (3 isolates); Exophiala (2 isolates); and Scolecobasidium (1 isolate) for subsequent study.

Morphological characteristics and phylogenetic analyses of rDNA sequences supported the delineation of the 12 representative isolates into distinct SBFS fungal genera. Each genus is described and illustrated in the subsequent subheadings.

Taxonomic placement based on LSU rDNA sequences

The maximum likelihood (ML) and Bayesian inference (BI) phylogenetic trees of the LSU rDNA placed the SBFS isolates in the present study with previously identified SBFS species and other non-SBFS species. The LSU rDNA gene tree grouped all six SBFS fungal isolates together with the previously identified Z. citrigriseum U. Braun & Crous from Thailand and the U.S., with high bootstrap values (maximum likelihood bootstrap (MLBS) = 100%; Bayesian posterior probabilities (BPP) = 0.99) (Fig. 4).

Fig. 4.

One of 2 most parsimonious trees obtained from partial sequences of the LSU sequences of SBFS species and other ascomycetes. Numbers shown at a node represents bootstrap values (only bootstrap values of >50% are shown). The first number was obtained from ML, and the second number was from BI methods. Tree length = 913 steps; consistency index = 0.638554; retention index = 0.888739; and the composite index = 0.567508. Taxa denoted by GenBank accession, followed by species name or isolate number, and country of origin (CN-China, CO-Colombia, DE-Germany, IN-India, IT-Italy, MY-Malaysia, NL-Netherlands, SI-Slovenia, TH-Thailand, TR-Turkey, US-USA, UY-Uruguay, ZA-South Africa). Isolates from the present study are shown in bold. Previously identified members of the SBFS complex are denoted with *. The tree is rooted to Peziza vesiculosa.

However, the three representative isolates previously identified morphologically as Peltaster did not group with any previously described Peltaster species: P. fructicola, P. gemmifer, and P. cerophilus Jana Frank & Schroers (Fig. 4). Two isolates from the present study, namely isolates L24A and L27A, were grouped together with the undescribed genera from China ‘Pseudopeltaster’ (MLBS = 99%; BPP = 0.96). The other isolate, L111C was uniquely grouped together but distinctly from both the ‘Pseudopeltaster’ clade and several other previously described Peltaster species (Fig. 4).

Exophiala species that have been previously described were aligned with the present two isolates. Isolate L22F was grouped with E. spinifera McGinnis from China, isolated from humans with high bootstrap values (MLBS = 90%; BPP = 0.56). Isolate L13A was grouped with E. bergeri Haase & De Hoog from the U.S., with high bootstrap values (MLBS = 90%; BPP = 1.0) (Fig. 4). Moreover, the phylogenetic analyses using rDNA sequences, grouped the present isolate, L26A with O. cordanae Samerpitak, Crous & De Hoog, isolated from dead leaves in Colombia, India, and South Africa, with high bootstrap values: LSU (MLBS = 100%; BPP = 0.98) (Fig. 4).

Taxonomic placement based on ITS rDNA sequences

Zasmidium spp. were most frequently occurring SBFS species and were recovered from mango. A total of six Zasmidium isolates were clustered with Z. citrigriseum (MLBS = 74%; BPP = 0.89) from Thailand and the U.S. (Fig. 5). The ML ITS trees from 19 taxa were 181 steps (Consistency index (CI) = 0.718519, Retention index (RI) = 0.872054, Composite index = 0.688971) (Fig. 5).

Fig. 5.

One of 2 most parsimonious trees obtained from partial sequences of the ITS sequences of Zasmidium species. Numbers shown at a node represents bootstrap values (only bootstrap values of >50% are shown). The first number was obtained from ML, and the second number was from BI methods. Tree length = 181 steps; consistency index = 0.718519; retention index= 0.872054; and the composite index = 0.688971. Taxa denoted by GenBank accession, followed by species name or isolate number, and country of origin (AU-Australia, CN-China, CO-Colombia, ID-Indonesia, IN-India, KR-South Korea, MY-Malaysia, TH-Thailand, US-USA, ZA-South Africa). Isolates from the present study are shown in bold. Previously identified members of the SBFS complex are denoted with *. The tree is rooted to Teratosphaeria cryptica.

The ITS rDNA strongly distinguished all three isolates of the undescribed Peltaster-like species and were 390 steps (CI = 0.682320, RI = 0.928394, Composite index = 0.654636) (Fig. 6). The present Peltaster-like species did not group with other previously described Peltaster species. Two isolates were grouped with undescribed ‘Pseudopeltaster’ clade (MLBS = 99%; BPP = 1.0), and one isolate was sister to the ‘Pseudopeltaster’ clade (MLBS = 100%; BPP = 1.00) (Fig. 6).

Fig. 6.

One of 3 most parsimonious trees obtained from partial sequences of the ITS sequences of Peltaster-like species and ‘Pseudopeltaster’ species. Numbers shown at a node represents bootstrap values (only bootstrap values of >50% are shown). The first number was obtained from ML, and the second number was from BI methods. Tree length = 390 steps; consistency index = 0.682320; retention index = 0.928394; and the composite index = 0.654636. Taxa denoted by GenBank accession, followed by species name or isolate number, and country of origin (CN-China, DE-Germany, MY-Malaysia, SI-Slovenia, TR-Turkey, US-USA). Isolates from the present study are shown in bold. Previously identified members of the SBFS complex are denoted with *. The tree is rooted to Capnodium coffeae.

Similarly, the ITS rDNA analyses grouped two isolates in the present study with either E. spinifera from Egypt and Germany (MLBS = 96%; BPP = 1.0) or E. bergeri (MLBS = 70%; BPP = 0.78) from Canada and Brazil (Fig. 7). The present isolates were also grouped with Exophiala species causing SBFS disease on apples in Germany (accession number: KP400577, KP400576, KP400575). The 6 ML ITS rDNA trees from 18 taxa had 269 steps (CI = 0.655172; RI = 0.829684; and composite index = 0.613781).

Fig. 7.

One of 6 most parsimonious trees obtained from partial sequences of the ITS sequences of Exophialaspecies Numbers shown at a node represents bootstrap values (only bootstrap values of >50% are shown). The first number was obtained from ML, and the second number was from BI methods. Tree length = 269 steps; consistency index = 0.655172; retention index = 0.829684; and the composite index = 0.613781. Taxa denoted by GenBank accession, followed by species name or isolate number, and country of origin (BR- Brazil, CA-Canada, DE-Germany, EG-Egypt, JP-Japan, MY-Malaysia, US-USA, UY- Uruguay). Isolates from the present study are shown in bold. Previously identified members of the SBFS complex are denoted with *. The tree is rooted to Cladophialophora carrionii.

The phylogenetic analyses of ITS rDNA sequences grouped the present isolate, L26A with O. cordanae Samerpitak, Crous & De Hoog, isolated from dead leaves in Colombia and India with high bootstrap values ITS (MLBS = 99%; BPP = 0.52) (Fig. 8). The ITS rDNA trees delineated one putative species and were 330 steps (CI = 0.84158, RI = 0.886525, Composite index = 0.800559) (Fig. 8).

Fig. 8.

One of 4 most parsimonious trees obtained from partial sequences of the ITS sequences of Ochroconis species Numbers shown at a node represents bootstrap values (only bootstrap values of >50% are shown). The first number was obtained from ML, and the second number was from BI methods. Tree length = 330 steps; consistency index = 0.84158; retention index = 0.886525; and the composite index = 0.800559. Taxa denoted by GenBank accession, followed by species name or isolate number, and country of origin (CN-China, CO-Columia, DE-Germany, IN-India, MY-Malaysia, NL-Netherlands, TH-Thailand). Isolates from the present study are shown in bold. Previously identified members of the SBFS complex are denoted with *. The tree is rooted to Fusicladium ramoconidii.

Class: Dothideomycetes; order: Capnodiales; genus: Zasmidium

Six isolates of Zasmidium species showed felty, olivaceous-grey colonies with smooth margins and homogeneously dark green reverse on potato dextrose agar (PDA) (see Supplementary Fig. S1 online). Growth diameter on malt extract agar (MEA) at 4 weeks was 3.7 ± 0.13 cm. Conidiophores were solitary, mononematous, erect to flexuous, sometimes curved, unbranched, thick-walled, 1–3-septate, cylindrical, geniculate, and brown. Conidia (5.93‒)13.57‒14.71(‒26.77) × 1.62–1.67 μm were solitary in simple, short to long cylindrical, with some ends swollen straight to somewhat curved, aseptate, light brown, thin-walled, and verruculose (see Supplementary Fig. S1 online). On the surface of the mango fruit, all isolates exhibited ramose mycelial type.

Class: Dothideomycetes; order: Capnodiales; genus: Peltaster

Three previously undescribed Peltaster-like species were isolated from colonies showing a punctate mycelial type. On PDA, colonies of these species initially exhibited yeast-like growth during the early stage, which later transitioned into filamentous growth. The colonies were felty, olivaceous to black, with uneven and undulated margins, a wrinkled surface, and a raised to mounded appearance. The reverse side of the colonies was uniformly black (see Supplementary Fig. S2 online). Microscopic observation grouped the three isolates into two morphotypes. Morphotype I (isolates L24A and L27A) showed a colony diameter of 0.84 ± 0.08 cm after one month of incubation on MEA at room temperature. Conidia measured (1.49‒)2.31‒2.36(‒3.25) × 1.24–1.26 μm, produced laterally on hyphal cells from intercalary conidiogenous loci that developed as short, undifferentiated hyphal stubs. The conidia were unicellular, cylindrical, hyaline, and commonly found clustered along the hyphae near conidiogenous openings. These openings were phialide-like, with short necks and slightly thickened walls (see Supplementary Fig. S2 online). No long conidiophores were observed.

In contrast, morphotype II (isolate L111C) of the undescribed Peltaster-like species produced conidia from a single fertile locus along the hyphal cells, a characteristic similar to that of previously described Peltaster species. Long conidiophores were present, bearing conidia clustered at their tips. The conidia measured (2.92‒)4.07‒4.11(‒5.68) × 2.26–2.32 μm, were hyaline, and cylindrical in shape (see Supplementary Fig. S3 online).

Class: ChaetothyriomyExophialacetes; order: Chaetothyriales; genus: Exophiala

The species isolated from mango in the present study showed a punctate mycelial type. Based on colony and microscopic characteristics, the two isolates were grouped into two morphotypes. Morphotype I (isolate L22F) showed colonies that initially appeared slimy or mucoid at the centre on PDA, later developing aerial mycelia and becoming slightly raised with age. The colonies were olivaceous grey to brownish black, with an olivaceous black reverse. After one month of incubation on MEA, the average colony diameter was 3.30 ± 0.24 cm. The conidiophores were micronematous (similar in structure to vegetative hyphae), simple or branched, erect, long, spine-like, brownish, and septate. Conidia measured (2.57‒)3.49‒3.49(‒5.14) × 1.75–1.77 μm, were obovoidal and produced both at the tips of conidiogenous cells and intercalary positions along the hyphae (see Supplementary Fig. S4 online).

Morphotype II (isolate L13A) formed colonies with an average diameter of 2.68 ± 0.13 cm after one month on MEA. The conidiophores were long and septate. Young cultures produced numerous yeast-like budding cells, along with larger, inflated, subglobose to broadly ellipsoidal cells that developed short torulose (swollen) hyphae. Chlamydospores were present, subspherical in shape. Conidia measured (6.00‒)9.93‒10.34(‒13.95) × 5.86–5.92 μm, were hyaline, smooth, thin-walled, and broadly ellipsoidal (see Supplementary Fig. S5 online).

Class: Dothideomycetes; order: Venturiales; genus: Scolecobasidium

One isolate exhibited ramose mycelial type on the mango fruit surface and showed colony appearance on PDA was dark olivaceous-brown, and the reverse was brownish black. The growth diameter after one month on MEA was 3.50 ± 0.14 cm. Hyphal coils with rope-like hyphal strands were present in this fungal species. Conidiophores were mostly cylindrical to acicular, arising from hyphae, unbranched, with septa (see Supplementary Fig. S6 online). Conidia (1.39‒)2.05‒2.06(‒2.72) × 1.61–1.62 μm, were solitary, sub-hyaline to hazel brown, verruculose, obovoid to extensively fusiform.

Pathogenicity test

Koch’s postulates were fulfilled and confirmed the pathogenicity of all 12 representative isolates, causing typical SBFS disease symptoms on inoculated mango fruit in both in situ and excised fruit inoculation methods (see Supplementary Fig. S1‒S6 online). Each representative SBFS fungal species consistently showed SBFS disease symptoms associated with either punctate or ramose mycelial types in both inoculation methods. The fruit treated as control treatments were asymptomatic. Re-isolated fungal isolates from both inoculation methods were confirmed to be identical to the original isolates in DNA sequence analysis.

Statistical analysis

Three Perak orchards showed the most severe SBFS infection, with the highest mean percentage area of SBFS infection at 47% (27 collected fruits), followed by Pahang at 29% (27 collected fruits), and Perlis at 9% (33 collected fruits) (Fig. 9B). The mean number of the total CPM at each of the eight sampling locations in Perak, Pahang, and Perlis ranged from 125 to 420 (Table 1).

Fig. 9.

The CPM and percentage area of SBFS infection on mango fruit, reflecting the SBFS disease severity observed in different states. (A) CPM against state. (B) Percentage area (%) of SBFS infection against state. Means within a bar graph followed by the same letter are not significantly different (P < 0.05).

Significant differences were observed in both CPM (P = 0.0001) and percentage area of SBFS infection (P = 0.0001) among the different states (Perak, Pahang, and Perlis). Comparing the three states, Perak showed the highest SBFS diversity with the recovery of three genera (Zasmidium, Exophiala, and Scolecobasidium) and a previously unknown Peltaster-like species. In contrast, Pahang and Perlis states recorded only one SBFS fungal genus (Table 1). Perak had the highest mean CPM (332.1) and percentage area of SBFS infection on mango (47%) compared to Pahang (228.4 and 29%, respectively), and Perlis (143.2 and 9%, respectively) (Fig. 9).

Significant interactions were observed for both ‘States × SBFS genera’ (P = 0.004), suggesting that multiple factors may have influenced the composition of SBFS assemblages in the present findings. Perak state recorded the highest number of recovered SBFS isolates from infected mango fruits and had the most diverse identified SBFS taxa (Table 1).

Discussion

This study delineated three genera (Zasmidium, Exophiala, and Scolecobasidium) and a previously unknown Peltaster-like species. Based on morphological characteristics, this Peltaster-like species showed both shared and distinct features that differentiate it from the previously described P. fructicola (Johnson et al., 1996), P. cerophilus¹⁹, and P. gemmifer²⁰. Similar findings were reported by Tham et al.⁹ on Peltaster-like species causing SBFS disease on mango in Perak. Peltaster species have been widely reported in the eastern and midwestern United States^3,4,17. Previously, ‘Pseudopeltaster’ was reported in China in 2015 based on SBFS symptoms on fruits and plants (Chen, Sun, and Gleason, unpublished data). As ‘Pseudopeltaster’ remains undescribed and lacks morphological description, it should be considered a provisional or informal taxon until further studies are conducted. The unique rDNA sequences of our isolates suggest that more detailed analysis is needed to properly delineate these putative Peltaster-like species. For example, Rosli et al.²⁰ described P. gemmifer based on morphology supported by phylogenetic analysis using three loci (ITS, LSU, and mrSSU). Therefore, side-by-side morphological comparisons with other Peltaster species and additional molecular analyses using genes such as elongation factor 1-alpha, actin, and beta-tubulin are required to fully describe and classify the Peltaster-like species found in this study^9,20,21.

In addition, this is the second report of a member of the SBFS complex in the Venturiales, Ochroconis cordanae. Hao et al.⁷ reported a closely related SBFS on banana, Scolecobasidium musae G.Y. Sun & Lu Hao in China. Ochroconis is synonymous to Scolecobasidium and are cosmopolitan, inhabiting soil, water, air, plant, and vertebrates²².

Zasmidium citrigriseum was the most common SBFS assemblage in the present study, associated with the ramose mycelial type. This species is previously reported to cause leaf spots and lesions on leaves but has not been reported as a SBFS component²³. Zasmidium angulare Batzer & Crous and Z. litsease, were reported as causal agents of SBFS in China, on apples and the petiole of soft bollygum, Litsea glutinosa^24,25.

The ramose mycelial type on mango was also found associated with two Exophiala species. The Exophiala genus has been reported to cause several SBFS mycelial types, such as ridged honeycomb and fuliginous^4,5,17. Despite intra-genus differences in mycelial types, the consistency of mycelial type within species plays a crucial role in the initial field-based diagnosis of SBFS and the preliminary attribution of SBFS putative species³. While E. bergeri is not commonly reported as a phytopathogenic fungus, E. spinifera has been associated with human cutaneous illnesses^26,27. These two species, previously not recognized as members of SBFS species complex, were found on mango fruit. In the Lower Elbe region of Germany, several other Exophiala species made up a very small proportion of the SBFS complex and were uniquely found in limited locations on few colonies per apple without morphological descriptions on media¹⁷.

Differences in the distribution of the identified SBFS fungal genera on mango were observed among the three surveyed states (Perak, Pahang, and Perlis). Johnson et al.²¹ reported that P. fructicola was widely distributed across the Southeast, Midwest, and Northeast regions of the U.S., whereas other SBFS fungi, such as Stomiopeltis species, P. sessilis, and G. polystigmatis, were confined to specific regions⁴. However, the biogeographic patterns of SBFS species distribution on mango in Malaysia and other parts of Southeast Asia remain largely unexplored. The composition of SBFS fungi is known to differ among geographic regions and according to orchard management practices¹⁶. In larger countries such as the U.S., a five-year surveillance study in Iowa apple orchards revealed that Peltaster spp. was a minor component of the SBFS complex, except when the total duration of wetness from petal fall to harvest exceeded 700 h. Some studies also showed that a few species, such as Peltaster and Dissoconium, can grow at 35 °C, although most SBFS fungi exhibit optimal mycelial growth between 20 °C and 25 °C³. Therefore, further analysis assessing the correlation between environmental factors in the Malaysian climate and the composition of SBFS on mango is needed to support the differences observed in the present study.

Management of SBFS generally combines both cultural and chemical control measures. Cultural practices such as pruning aim to reduce favorable conditions for infection, while the removal of fruit mummies (overwintering fruits that harbor SBFS inoculum) can lower disease risk. The use of fruit bags to physically protect fruits from airborne pathogens is common in China and Japan and was also observed in Perlis during this study. This practice has been reported to reduce SBFS incidence, as shown in Kentucky orchards¹. Batzer et al.¹⁶ reported that abandoned (unsprayed) orchards harbor higher species diversity compared to orchards managed with intensive fungicide programs and cultural practices. Similarly, a four-year survey of apple orchards in northern and southern Germany recorded 19 named and putative SBFS species, with species diversity being higher in abandoned orchards than in organic ones, while no SBFS colonies were found in conventionally sprayed orchards¹⁷. A genus-level survey of 35 apple orchards in Norway also found seven SBFS genera, reporting lower SBFS diversity in organic orchards compared to abandoned ones¹³.

Although some differences were observed among the locations in the present study, future work should focus on analyzing correlations between SBFS diversity, orchard management practices, and environmental conditions. For example, Gleason et al.¹. documented that SBFS fungi can grow on leaves and twigs of many plant species, acting as reservoir hosts for the inoculum. Typically, SBFS inoculum travels only a few hundred meters from these reservoir hosts at orchard borders to initiate infections. Therefore, further studies, such as spore trap monitoring in mango orchards adjacent to other crops like oil palm (as observed in Perak), could help verify the potential movement of SBFS inoculum between host plants. The proximity of an oil palm plantation next to the mango orchard in Perak may have contributed to the diverse SBFS assemblage found there. Although this study did not directly confirm the dissemination of SBFS spores from bordering oil palm trees into mango orchards, Gleason et al.¹ and William & Sutton¹⁵ suggested that various plant species bordering orchards could act as reservoir hosts, spreading SBFS inoculum across hundreds of meters. Additionally, differences in fungicide sensitivity among SBFS species may influence their prevalence in managed orchards²⁸. Species that are more sensitive to commonly used fungicides may be rare or absent in regularly sprayed orchards¹³. Subsequent experiment, such as screening the isolates collected in this study for fungicide sensitivity could therefore help identify effective fungicides for managing SBFS on mango.

On mango, SBFS was first reported by Martinez et al.²⁹; the study elaborated on the association of several SBFS species on mango without confirming pathogenicity. Later, in Japan, Ajitomi et al.⁸ identified Stomiopeltis species and confirmed it as the SBFS fungus causing the disease on mango in Japan. Tham et al.⁹, reported the first record of SBFS occurrence on mango (variety: Chok Anan) in Malaysia, caused by species associated with SBFS fungal genus Peltaster. More recently, Ganesan et al.¹⁸. recorded several SBFS-associated fungal species on mango in India, including Pseudocercospora sp., Pallidocercospora sp., Zasmidium sp., and Passalora sp. In China, a wider diversity of SBFS fungi has been reported across several genera and host plants. These include Dissoconium, Uwebraunia, and Ramichloridium²⁴; Scolecobasidium⁷; Scleroramularia¹⁸; Cyphellophora¹⁸; and Dissoconium, Mycosphaerella, and Peltaster³⁰.

Mango fruit with a high percent coverage area of SBFS infection (>50%) were frequently observed to produce overlapping colonies, comprising several mycelial types, which made isolating pure SBFS fungi challenging³. From the total of 2,106 mango peels exhibiting punctate, ramose, fuliginous, and honeycomb mycelial types, only the punctate and ramose mycelial types were successfully cultured, resulting in a total of 31 SBFS fungal isolates. Isolating SBFS fungi from the surface of the hosts is often unsuccessful due to the ectophytic nature of SBFS fungi¹³. Moreover, compared to other common non-ectophytic fungi, the standard surface sterilization process for fungal isolation typically compromises the viability of SBFS fungi on culture media¹. Additionally, the majority of SBFS species do not readily sporulate on fruit surfaces. Even if the SBFS species successfully grow on growth media, their slow growth rate in vitro causes other fast-growing fungi to outgrow them on the culture plate¹. Batzer et al.³ reported that SBFS fungi typically require at least one month to develop visible colonies in vitro, during which their morphological structures become clearer for identification and suitable for DNA extraction. Given the relatively low isolation success rate (1.47%), future work should consider modifying the surface sterilisation protocol to improve fungal recovery. Adjustments such as using lower disinfectant concentrations, shorter exposure times, or brief rinsing in sterile water before plating may help preserve the viability of surface-residing SBFS fungi. Previous studies on endophytic and surface-dwelling fungi have shown that optimizing sterilisation steps, particularly by reducing sodium hypochlorite concentration and exposure duration can significantly enhance isolation success³¹. However, this has not yet been proven for SBFS fungi and therefore requires further assessment under laboratory conditions.

Although this study was limited to selected orchards in several states of Peninsular Malaysia and the molecular characterization was restricted to the ITS and LSU gene regions, the findings provide important baseline data on SBFS fungi associated with mango in Malaysia. To further elucidate potential regional trends in SBFS biodiversity, expanded studies involving a greater number of sprayed and unsprayed orchards, as well as the inclusion of additional states in Peninsular Malaysia with diverse orchard management practices, would be highly beneficial. Subsequent research should also focus on improved delineation of SBFS fungal species on mango in Malaysia by incorporating additional gene loci to strengthen molecular resolution and species delimitation. In addition to ITS and LSU regions, sequencing of other protein-coding genes, such as translation elongation factor 1-alpha (TEF), RNA polymerase II subunits (RPB1 and RPB2), actin (ACT), beta-tubulin (TUB2), and calmodulin (CAL) have been shown to be useful for resolving phylogenetic relationships among SBFS fungi^6,20. Integrating these loci in future studies will provide a more robust framework for accurate species identification and a clearer understanding of SBFS diversity on mango in Malaysia.

Materials and methods

Sample collection, fungal isolation, morphological observations

A total of 87 mango fruits exhibiting SBFS symptoms were collected from eight orchards located in three states of Malaysia (Perak, Pahang, and Perlis) during sampling conducted from October to November 2019 (Table 1). Permission to collect mango samples was granted by the relevant authorities, including the Jabatan Pertanian Negeri Perak, Pahang, and Perlis. During sample collection, the temperature and RH at each location were measured using a UT333 Uni-T^® mini temperature and RH meter (UNI-TREND Technology China Co., Ltd, China). The mean values of temperature and RH for each location were then recorded. Each fruit was placed in an individual bag and stored at room temperature until further observation in the laboratory. For each fruit, the number of colonies per mango (CPM) was recorded. The percentage of fruit surface area covered by SBFS colonies was estimated using standard area diagrams developed by Spolti et al.³² Mycelial types were assigned following descriptions by Batzer et al.³ with modifications. Each SBFS colony and its subtending portion of the peel were excised from fruit and photographed under dissecting microscope (Olympus BX41, Japan) equipped with image processing software (Olympus Software Imaging Solutions, Germany).

Isolation of SBFS fungi followed protocols of Batzer et al.³ and Mayfield et al.⁵. After fruit were washed under flowing tap water for 30 min and allowed to dry, peels containing the SBFS colonies were excised and pressed between paper towels until dry before being transferred to acidified water agar (AWA; 20 g agar, 1.25 mL lactic acid, 1 L distilled water) to inhibit bacterial growth¹⁶. Plates were incubated for 3 to 6 weeks at room temperature (23 to 25 °C) under intermittent light until mycelial growth in the media was observed. New growth was then transferred onto PDA (39 g HIMEDIA^® PDA, 1 L with distilled water). Morphological observations, including colony growth diameter and colony appearance, were made on MEA (15 g malt extract, 20 g Bacto™ agar, 1 L distilled water) which were incubated in the dark at room temperature for 1 month¹⁷. Conidial shapes and sizes (length and width) were recorded for 30 spores from each culture on synthetic nutrient-poor agar (SNA; 1 g KH2PO4, 1 g KNO3, 0.5 g MgSO4.7H2O, 0.2 g dextrose, 0.2 g sucrose, 20 g Bacto™ agar, 1 L distilled water)^3,20,33.

PCR amplification, sequencing, and phylogenetic analysis

A total of 12 representative isolates of putative SBFS fungi, exhibiting distinct morphological characteristics were subjected to molecular analysis. Fungal DNA was extracted from 4- to 6-week-old PDA cultures grown at room temperature (23 to 25 °C) under intermittent ambient light Batzer et al.³. A small amount of mycelium was sampled using a pipette tip and transferred to a 1.5-mL microcentrifuge tube containing 100 µl PrepMan^® Ultra Sample Preparation Reagent (Thermo Fisher Scientific, USA), following the manufacturer’s instructions. Two regions of the nuclear ribosomal DNA were amplified using the following primer pairs, respectively: internal transcribed spacer (ITS), ITS1 (5-TCCGTAGGTGAACCTGCGG-3)/ITS4 (5-TCCTCCGCTTATTGATATGC-3)³⁴ and large subunit (LSU), LROR (5-GTACCCGCTGAACTTAAGC-3)/LR5 (5-TCCTGAGGGAAACTTCG-3)³⁵. The total volume of the 25-µl PCR reaction mixture contained 12.5 µl of EconoTaq PLUS GREEN 2X Master Mix (Lucigen, Madison, WI, USA), 0.5 µM of each forward and reverse primer, 1.0 µl of DNA template, and 11.0 µl sterile distilled water. Both loci were amplified using the following PCR thermal cycling (Peltier thermal cycler PTC-100, MJ Research, USA) conditions: initial denaturation at 94 ℃ for 95 s followed by 35 cycles of denaturation at 94 °C for 35 s, annealing at 49 ℃ for LSU; and at 52 ℃ for ITS for 60 s, and final extension at 72 ℃ for 2 min.

Amplification of PCR products was verified by running 4 µl of the PCR product in 1% agarose gel using gel electrophoresis, operated at 80 V, 400 mA for 90 min before the gel was visualised under an ultraviolet transilluminator using the Molecular Image Gel Documentation System (Gel Dox XR, Bio Rad, USA). All PCR products were sent to MyTACG Bioscience Enterprise (Selangor, Malaysia) for purification and DNA sequencing. Molecular Evolution and Genetic Analysis (MEGA 7.0) software was used to align the forward and reverse sequences for each locus to acquire the consensus sequence and manually aligned using BioEdit³⁶. Consensus sequences were inspected and manually corrected where necessary. Sequence alignments were performed using the ClustalW algorithm implemented in MEGA 7.0, and minor manual adjustments were made in BioEdit to improve alignment accuracy³⁶. Ambiguously aligned regions and terminal gaps were trimmed manually. The finalized consensus sequences were subjected to nBLAST searches in the GenBank database (National Center for Biotechnical Information, Bethesda, MD, USA). The aligned ITS sequences of the mango isolates and selected sequences from previously identified fungi from GenBank were subjected to ML and BI methods.

The ML analysis was performed in MEGA 7.0³⁶. Gaps were treated as ‘missing data’, and characters were unordered and treated with equal weight. Bootstrapping was performed using 1,000 replicates. The best-fit nucleotide substitution model was determined automatically by MEGA for each dataset. The ITS sequences of Exophiala species were analyzed under the Kimura 2-parameter model with a Gamma distribution (K2 + G), while the ITS sequences of Peltaster species were analyzed using the Tamura 3-parameter model with a Gamma distribution (T92 + G). The LSU sequences of SBFS species and other reference strains of ascomycetes were analyzed using the Tamura–Nei model with a Gamma distribution and proportion of invariant sites (TN93 + G + I). For the ITS sequences of Scolecobasidium species, the Tamura 3-parameter model (T92) was applied, whereas the ITS sequences of Zasmidium species were analyzed using the Tamura–Nei model with a Gamma distribution (TN93 + G).

The BI analysis was employed to infer the phylogenetic relationships among the fungal isolates using MrBayes v3.2.7³⁷. The optimal nucleotide substitution model for each dataset was selected based on the Akaike Information Criterion (AIC), as determined in MEGA³⁶. Distinct substitution models were applied to each dataset according to the gene region and taxon analysed. For the LSU dataset (all representative species), the Tamura–Nei model with invariant sites and gamma-distributed rate variation was selected. For the ITS region, specific models were used for: Zasmidium species (TN93 + G), Peltaster species (T92 + G), Exophiala species (K2 + G), and Scolecobasidium species (T92). Two independent runs, each comprising four Markov chains, were executed for 2,000,000 generations, with trees sampled every 1,000 generations. Convergence was confirmed by monitoring the average standard deviation of split frequencies, which was accepted below 0.01. The initial 25% of sampled trees were discarded as burn-in. The initial 25% of sampled trees were discarded as burn-in, and a 50% majority-rule consensus tree was generated from the remaining trees to estimate posterior probabilities (PP) of each clade. The resulting phylogenetic tree was visualized and edited using FigTree v1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/). Outgroup taxa for analyses of the ITS and LSU were Teratospaeria cryptica, Capnodium coffeae, Cladophialophora carrionii, Trimmatostroma abietis, Fusicladium ramoconidii, and Peziza vesiculosa. The list of representative isolates and sequences of SBFS fungi used in this study is provided in Supplementary Table S1 online. The alignments and gene trees were deposited in TreeBASE (submission ID: S32381; www.treebase.org).

Pathogenicity test

All 12 representative isolates were grown on PDA for one month at 23 to 25 °C under intermittent ambient light to prepare for conidial suspensions. Suspensions were made by flooding the PDA plates with sterile distilled water, and a metal spatula was used to scrape the surface of the colonies. The spore suspension was adjusted to 5 × 10⁵ conidia mL^{− 1} using a hemocytometer^38,39. All spore suspensions were prepared 12 h prior to application and stored in the refrigerator at 5 °C until inoculation⁴⁰.

Each representative isolate was subjected to pathogenicity tests on mature, asymptomatic mango fruit (cvs. Chok Anan and Harumanis) using two methods - excised and in situ - following protocols of Gao et al.⁶ and Rosli et al.⁴⁰ with modifications. A mango (cv. Chok Anan) orchard in Perak (4°08′28.1” N, 100°45′39.6” E) was used to perform the in situ inoculation (Fig. 10). Approximately 14 days after fruit set and prior to inoculation, mango fruit were bagged (https://shopee.com.my/Pembalut-Pelam-Pembalut-buah-Putih-berlilin-50-pcs-Sarung-buah-Fruit-Cover-Vegetable-Cover-i.91615702.4515719957) for one month to prevent colonization by ambient SBFS inoculum⁴⁰. No fungicide sprays were applied at any time.

Fig. 10.

Mature, healthy, and disease-free mango fruits used for pathogenicity test. (A) Mature mango fruits in the orchard. (B) In situ inoculation. (C) Excised inoculation.

A total of 31 bagged fruit were selected on two trees designated for inoculation. These mango fruits were arbitrarily assigned to each of the 12 representative SBFS fungal isolates using coloured strings tied to the closest branch (Fig. 10B). Before inoculation, all fruit bags were removed, and the fruit were surface sterilised with 70% ethanol and allowed to dry for approximately 2 min. The inoculum (approximately 0.5 mL) was sprayed once onto each of the two hemispheres of the mango fruit using 50 mL-spray bottles, according to the assigned coloured strings. Each SBFS isolate was inoculated onto two fruit, one per tree. Each fruit was gently rotated during spraying to ensure even coverage of the inoculum. A tray was used to collect inoculum runoff during the inoculation process before the inoculated fruits were re-covered with new fruit bags. Sterile distilled water was sprayed as a control treatment, and five non-inoculated mango fruit remained bagged until harvest as an additional control treatment. Inoculated fruit were left overnight before removing the fruit bags the next day (except for the five fruit that were bagged until harvest) and left on the tree for approximately four weeks until harvest. Positive infection was confirmed by visible mycelial growth on the inoculated fruit surface. A total of 31 mango fruit were used. The in situ inoculation in Perak was conducted twice and the final repetition was performed on cv. Harumanis mango variety in Perlis (6°23’40.0"N, 100°08’40.4"E).

Excised fruit inoculation included 26 mature (approximately 1 month until harvest), asymptomatic fruit (cv. Chok Anan), harvested with 15 to 30 cm of attached stem to reduce postharvest fungal infection by Botryosphaeriaceae species⁴¹. In the laboratory, fruit were rinsed with tap water for about 2 min, surface sterilised with 70% ethanol using a paper towel, and allowed to dry for approximately 2 min. Each spore suspension was swabbed once onto both hemispheres of each of two fruit using sterile cotton balls, and the fruit were rotated to ensure uniform inoculation. Sterile distilled water was swabbed as a control. All inoculated fruit were kept in sealed trays with a water-filled 9-cm-diameter Petri plate and under intermittent light at 23 to 25 °C for 3 to 4 weeks (Fig. 10C). Positive infection was determined based on the presence of visible mycelial growth on the fruit surface. The inoculation trial on cv. Chok Anan was conducted twice, and the third repeat of the trial was conducted on cv. Harumanis. At the end of each trial, all inoculated mango fruit were observed for SBFS colonies; colonies were re-isolated, and their identities were confirmed based on morphology characters and comparing the ITS sequences with the original inoculated isolates.

Statistical analysis

The percentage area of SBFS infection, CPM, and number of putative SBFS species on the collected mango fruit from the orchards in Perak, Pahang, and Perlis were subjected to statistical analysis using Statistical Package for the Social Sciences (SPSS) software version 26 (IBM Corp., Armonk, NY, USA).

The normality of the data was assessed using the Kolmogorov-Smirnov test. If the significance level was greater than 0.05, normality was assumed. Depending on this assessment of normality, either parametric or non-parametric tests were used⁴². One-way analysis of variance (ANOVA) was used, followed by Tukey’s Honestly Significant Difference (HSD) post hoc test for single step multiple comparison. It was used to determine the similarities and differences of total CPM among the different states. In addition, Kruskal-Wallis test was used to analyse the interaction between the percentage area of SBFS colonies infection severity with the different states and between SBFS genera between states. All P-values < 0.05 were considered statistically significant for all analyses⁴².

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

K.X.T. performed carried out data collection, analysis, and drafted the manuscript. K.S.G. supervised MEGA software analyses and prepared data visualizations. M.F.M. assisted with research design and statistical analysis. J.C.B. supervised phylogenetic analyses and validated results. M.L.G. contributed to writing and critical review. H.R. managed project administration and secured funding. All authors reviewed and approved the manuscript.

Funding

This research was supported by the Fundamental Research Grant Scheme (FRGS/1/2024/WAS04/USM/02/5) from the Ministry of Higher Education, Malaysia, and the Bridging GRA Research Grant (304/PBIOLOGI/6316604) from Universiti Sains Malaysia (USM). The first author received Graduate Student Financial Assistance (Gra-Assist) from the Institute of Postgraduate Studies, USM, for her Master’s program.

Data availability

The sequence data of this study have been deposited in the GenBank and the accession numbers are presented in **Supplementary Table S1 online**.

Declarations

Competing interests

The authors declare no competing interests.