Morphological and Molecular Characterization of Pseudocercospora lythracearum on Lagerstroemia spp. based on Korea and Japan Isolates, Including the First Report on L. fauriei in Korea

Abstract

Pseudocercospora lythracearum is a widely distributed foliar pathogen of Lagerstroemia species, causing Cercospora leaf spot. It was first described from the United States. However, although the nucleotide sequences of the internal transcribed spacer (ITS) region, actin (actA) and translation-elongation factor (tef1) genes retrieved from two isolates of Ps. lythracearum were publicly available in GenBank, the sequence data from the original type specimen and other Pseudocercospora spp. found on Lagerstroemia spp. are lacking, making its molecular comparison difficult. Consequently, the taxonomic circumscription of the fungus has relied largely on morphology. In this study, we aimed to clarify molecular-phylogeny of the fungus within the genus Pseudocercospora, including samples obtained from L. indica in Korea and Japan, L. subcostata in Japan, and L. fauriei in Korea. A multilocus phylogenetic analysis was conducted using a combined dataset of ITS, actA, tef1, and the second-largest subunit of DNA-directed RNA polymerase II (rpb2) sequences. Comparative analyses with Pseudocercospora species associated with Lagerstroemia further supported the identification of our isolates as Ps. lythracearum. Multilocus analyses consistently placed all isolates in a well-supported clade together with previously identified Ps. lythracearum isolates from Korea (CBS 131925) and Japan (MUCC 890). This study represents the taxonomic clarification of Ps. lythracearum, and the first morphologically and molecularly confirmed occurrence of Ps. lythracearum on L. fauriei in Korea.

1. Introduction

Pseudocercospora lythracearum (Heald & F.A. Wolf) X.J. Liu & Y.L. Guo was originally described as Cercospora lythracearum Heald & F.A. Wolf based on specimens collected from Lagerstroemia indica L. in Texas, USA, in 1911 [1]. Subsequent taxonomic studies transferred this species to the genus Pseudocercospora, in accordance with the original morphological description, in which the conidiophores were characterized by rounded apices with minute, barely discernible conidiogenous loci, and no distinct scars or hila on the conidia. These descriptions are consistent with the diagnostic characters used to distinguish Pseudocercospora from Cercospora, in which conidiogenous scars and conidial hila are distinctly thickened, and have led to the reclassification of numerous Cercospora-like taxa into Pseudocercospora [2].

Ps. lythracearum has been reported on multiple Lagerstroemia species including L. indica in Brazil, Brunei Darussalam, China, Hong Kong, India, Japan, and Korea [3–9]; L. macrocarpa Wall. ex Kurz in Thailand [10]; L. parviflora Roxb. in India [7]; L. speciosa (L.) Pers. in India and Venezuela [7,11]; and L. subcostata Koehne in Japan [12]. Cercospora leaf spot caused by Ps. lythracearum is considered as the second most damaging fungal disease affecting Lagerstroemia species, after powdery mildew [13,14]. Several other Pseudocercospora species have been reported from Lagerstroemia hosts worldwide, including Ps. lagerstroemiae (R.C. Rajak, Soni & G.P. Pathak) Kamal, Ps. lagerstroemiae-parviflorae H.S.G. Rao, S. Narayan & Bhartiya, and Ps. paramignyae (Thirum. & Chupp) Y.L. Guo on L. parviflora, Ps. lagerstroemiae-lanceolatae U. Braun & Crous on L. lanceolata Wall. ex C.B. Clarke, all reported from India [7], Ps. lagerstroemiigena Goh & W.H. Hsieh on L. speciosa, and Ps. lagerstroemiae-subcostatae (Sawada) Goh & W.H. Hsieh on L. subcostata, both reported in China and Taiwan [15].

To date, Ps. lythracearum is the only Pseudocercospora species reported from Lagerstroemia spp. for which molecular data are currently available in public sequence repositories. Recent extensive taxonomic studies on the genus Pseudocercospora included two isolates of Ps. lythracearum from Korea (CBS 131925, on L. indica) and Japan (MUCC 890, on L. indica) [12,16]. Although these examined isolates of Ps. lythracearum from Korea and Japan were shown to be genetically similar, the morphological description of the species was not provided in those studies. Moreover, direct molecular comparison with the authentic materials is currently not possible due to the lack of ex-type materials, such as the sequence data from the epitype. Hence, a critical comparison with the original description and related taxa, based on an integrative approach that includes detailed morphological characterization, molecular phylogenetic relationships, and ecological niches (i.e., host plant), is necessary to clarify the taxonomic position of Ps. lythracearum. Therefore, the present study aims to reassess Ps. lythracearum based on collected specimens from Lagerstroemia spp. in Korea and Japan through morphological characterization and multilocus phylogenetic analysis.

2. Materials and methods

2.1. Specimen collection

Leaf spot symptoms resembling infections caused by cercosporoid fungi have been observed continuously on Lagerstroemia spp. in Korea and Japan. On subsequent surveys in Korea in 2025, symptomatic specimens of L. fauriei Koehne, infected by Pseudocercospora species, were first collected. Collected specimens from Korea were kept in the herbarium collection of Korea University (KUS, Seoul, South Korea) and Kookmin University (KMU, Seoul, South Korea). Specimens from Japan were maintained in the herbarium of the Graduate School of Bioresources, Mie University (TSU-MUMH, Tsu, Japan). Detailed information regarding sample sources, including specimen voucher numbers, collection dates and locations, strain numbers, and GenBank sequence accession IDs, is presented in Table 1.

Table 1.

Information on isolates of Pseudocercospora on Lagerstroemia spp. used in this study.

Voucher specimen number	Host plant	Collection date	Collection location	Isolate accession number	GenBank accession number
					ITS	actA	tef1	rpb2
KUS-F15114	Lagerstroemia indica	September 18 1998	Korea	–	–	–	–	–
KUS-F18001	L. indica	October 29 2000	Korea	–	–	–	–	–
KUS-F18012	L. indica	October 30 2000	Korea	–	–	–	–	–
KUS-F19899	L. indica	October 15 2003	Korea	CBS 131925	GU269740	GU320442	GU384452	–
KUS-F22427	L. indica	October 29 2006	Korea	KACC 42649	PX955663	PX971572	PX963120	PX963115
KUS-F23624	L. indica	August 29 2008	Korea	KACC 43950	PX955664	PX971573	PX963121	PX963116
KUS-F26908	L. indica	September 10 2012	Korea	–	–	–	–	–
KUS-F28104	L. indica	September 11 2014	Korea	–	–	–	–	–
KUS-F28219	L. indica	September 23 2014	Korea	–	–	–	–	–
KUS-F34841	L. indica	October 19 2025	Korea	KACC 411337	PX955660	PX971569	PX963117	PX963112
KMU-F0094 (=KUS-F34943)	L. fauriei	October 30 2025	Korea	KACC 411339	PX955661	PX971570	PX963118	PX963113
KMU-F0095 (=KUS-F34944)	L. fauriei	November 8 2025	Korea	KACC 411338	PX955662	PX971571	PX963119	PX963114
–	L. indica	September 10 1994	Japan	MUCC 890, MAFF237786	GU269741	GU320443	GU384453	–
–	L. subcostata	October 1 1993	Japan	MUCC 1202, MAFF236964	PZ240335	PZ244876	PZ263166	PZ263163
–	L. subcostata	October 1 1994	Japan	MUCC 1232, MAFF237185	PZ240336	PZ244877	PZ263167	PZ263164
–	L. subcostata	Unknown	Japan	MUCC 1401, MAFF410017	PZ240337	PZ244878	PZ263168	PZ263165

CBS: Centraalbureau voor Schimmelcultures, Westerdijk Fungal Biodiversity Institute; KACC: Korean Agricultural Culture Collection, Rural Development Administration, Wanju, South Korea; KMU: The mycological collection of Kookmin University; KUS: The herbarium of Korea University; MAFF: The Research Center of Genetic Resources, National Agriculture and Food Research Organization, Tsukuba, Japan; MUCC: Culture Collection, Laboratory of Plant Pathology, Mie University, Tsu, Mie Prefecture, Japan; ITS: internal transcribed spacer; actA: actin; tef1: translation elongation factor 1-alpha; rpb2: DNA-directed RNA polymerase II second-largest subunit.

Specimens included in the molecular analyses are shown in bold; the remaining specimens were examined morphologically only.

2.2. Morphological observations

Morphological characterization and measurements were performed on fresh specimens using an Olympus BX50 microscope (Olympus, Tokyo, Japan). To ensure data accuracy, a minimum of 30 measurements were taken for each major diagnostic structure. Microscopic photographs for identification were performed using a Zeiss AX10 microscope integrated with an AxioCam MRc5 camera (Carl Zeiss, Oberkochen, Germany).

2.3. Fungal isolation

Selected specimens were subsequently subjected to molecular analyses, and those included in the molecular dataset are indicated in bold in Table 1. Monoconidial isolates were obtained by suspending conidia from young lesions in sterile water, which were then streaked on 2% potato dextrose agar (PDA; Difco, Detroit, MI) supplemented with 100 mg/L streptomycin sulfate. The plates were incubated at 25 °C for three days until the conidia germinated. Subsequently, subculture was performed by transferring the germinated conidia to fresh PDA without antibiotics to obtain pure cultures. Colony color characteristics were identified using the Rayner color chart [17]. The isolates obtained from Korea were deposited in the Korean Agricultural Culture Collection (KACC, Wanju, South Korea), and those from Japan were preserved in the culture collection of the Laboratory of Phytopathology, Mie University (MUCC, Tsu, Japan) or the Genebank, National Institute of Agrobiological Sciences (MAFF), Tsukuba, Japan, with accession numbers assigned accordingly (Table 1). Specimens not included in the molecular analyses were retained as morphological reference specimens.

2.4. DNA sequencing

Genomic DNA was extracted from isolates obtained from two-week-old cultures using the MagListo™ 5M Kit (Bioneer, Daejeon, South Korea) according to the manufacturer’s instructions. A total of eight isolates were included in the molecular analyses, consisting of three Korean isolates from L. indica, three Japanese isolates from L. subcostata, and two Korean isolates from L. fauriei. The three Japanese isolates from L. subcostata had previously been examined morphologically by Crous et al. [12] but were not included in their molecular analysis. These isolates were therefore included in the present study for multilocus molecular analysis. The eight isolates were molecularly analyzed based on multilocus sequences including the internal transcribed spacer (ITS) region and partial sequences of the actin (actA), translation elongation factor 1-alpha (tef1), and RNA polymerase II second-largest subunit (rpb2) genes [12,18]. PCR products were purified and sequenced by Bionics (Seoul, South Korea) using the primer pairs V9G/ITS4, ACT-512F/ACT-783R, EF1-728F/EF2, and P_RPB2-F/P_RPB2-R or RPB2-5f2/fRPB2-7cR, respectively, with the following previously described protocols [19–23]. The sequencing results were assembled and aligned using the MUSCLE command in MEGA11 software [24]. All resulting sequences have been deposited into GenBank with the accession numbers listed in Table 1.

2.5. Molecular analyses

The data matrix was constructed by aligning the consensus sequences against sequences of closely related Pseudocercospora species from GenBank. The analysis was performed using a combined multilocus dataset (ITS, actA, tef1, rpb2) assembled using SequenceMatrix [25]. For comparison, Trochophora simplex (CBS 214744) served as an outgroup [26]. The phylogenetic trees were constructed based on three analysis methods. The maximum-likelihood (ML) analysis was performed using IQ-Tree version 3.1.1 [27]. The concatenated dataset was partitioned by locus into ITS (1–486), actA (487–736), tef1 (737–1187), rpb2 (1188–1876), and the best-fit substitution model for each partition was selected using ModelFinder based on the Bayesian information criterion (BIC). The selected models were HKY + I + R2 for ITS, TPM2u + I + G4 for actA, TN + F + G4 for tef1, and TPM3u + I + G4 for rpb2. Branch support was evaluated using 1000 ultrafast bootstrap (BS) replicates. Maximum parsimony (MP) analysis was performed on PAUP*4.0a with a heuristic search and a tree-bisection algorithm (100 random replicates) [28]. Branch robustness was tested through 1000 BS replications [29]. The best evolutionary model for Bayesian inference (BI) analysis was determined using MrModelTest v.2.2 [30], which determined the best model as SYM + I + G for ITS, HKY + I + G for actA and tef1, and GTR + I + G for rpb2 based on Akaike information criterion (AIC). Markov Chain Monte Carlo analysis was run on MrBayes v.3.2.7 [31] for 1,000,000 generations with a sampling frequency of every 1000 generations and a burn-in of 25%. The consensus tree was visualized using MEGA7 software [32]. Final results that are considered statistically significant are indicated on the relevant branches and nodes by BS ≥ 70% for MP and ML analyses, and a posterior probability (PP) of at least 0.90 for BI analysis. The resulting phylogenetic tree and corresponding alignment were deposited in TreeBASE.

3. Results

3.1. Morphological characteristics

The morphological characteristics of our specimens show relatively consistent characteristics. Leaf spots were amphigenous, 3–10 mm in diameter, circular to irregular, dark brown to light brown, vein-limited (Figures 1(A,B) and 2(A–C)). Caespituli amphigenous, grayish, solitary (Figure 2(D)). Mycelium internal, 1.25–2.5 μm in width, branched, hyaline, smooth to rough, thin-walled. Stromata well-developed, globular, 25–60 µm in diameter, pale brown to dark brown. Conidiophores aggregated in dense fascicles, arising from the upper cells of a brown stroma, pale brown to pale olivaceous, not branched, 7.5–46.4 × 2.0–5.0 µm, subcylindrical, indistinct 0–3-septate, occasionally slightly geniculate and swollen or constricted at the septa, smooth-walled; conidiogenous cells terminal, integrated, proliferating sympodially or percurrently, 2.0–3.9 × 1.3–2.6 µm, unbranched, rounded to truncated at the apex, with inconspicuous conidiogenous loci (Figures 1(C) and 2(E,F)). Conidia sporulating holoblastically, solitary, hyaline to pale olivaceous, (15.1–)21.4–65.5 × 2.3–3.75 µm, smooth, narrowly obclavate to cylindrical obclavate, straight or strongly curved, obtuse to rounded at the apex, short obconically truncated at the base, 2–8-septate, with unthickened and not darkened hila (Figures 1(D,E) and 2(G,H)). Two-week-old colonies grown on PDA at 25 °C were 18–21 mm in diameter, the obverse surface of colonies was pale gray to grayish, dense aerial mycelium, undulate to entire margin. The reverse surface was pale olivaceous-brown to grayish olivaceous, slightly darker toward the center (Figures 1(F) and 2(I)). The morphological characteristics of the examined specimens are similar to those of the original description of C. lythracearum (≡Ps. lythracearum) [1].

Figure 1.

Cercospora leaf spot disease on Lagerstroemia indica caused by Pseudocercospora lythracearum. (A, B) Leaf spot symptoms; (C) stromata and conidiophores; (D, E) conidia; (F) two-week-old colonies of Ps. lythracearum on PDA at 25 °C. Scale bar indicates 10 µm.

Figure 2.

Cercospora leaf spot disease on Lagerstroemia fauriei caused by Pseudocercospora lythracearum. (A–C) Leaf spot symptoms; (D) close-up view of grayish conidial masses; (E) stromata; (F) conidiophores; (G, H) conidia; (I) two-week-old colonies of Ps. lythracearum on PDA at 25 °C. Scale bar indicates 10 µm.

To further confirm the identity of our specimens as Ps. lythracearum, the morphological characteristics observed were compared with those of other Pseudocercospora species reported from Lagerstroemia spp. (Table 2). Our specimens are clearly distinguishable from Ps. lagerstroemiae-subcostatae by having well-developed stromata (25–60 µm diameter vs. absent or poorly developed), shorter conidiophores (7.5–46.4 µm long vs. 10.0–60.0 µm long), and generally shorter and narrower conidia ((15.1–)21.4–65.5 × 2.3–3.75 µm vs. 25.0–70.0 × 2.0–4.0 µm) [33]. Our specimens are also distinguishable from Ps. lagerstroemiae-parviflorae by the much shorter conidiophores (vs. 105–204 µm) and narrower conidia (vs. 3–5 µm) [34]. Furthermore, although Ps. lagerstroemiae-lanceolatae is morphologically close to Ps. lythracearum, it differs from our specimens in producing superficial secondary mycelium with solitary conidiophores, longer and more septation in conidiophores (10–100 µm long, up to 4–5-septate), and longer conidia (30–100 µm long) [35]. Our specimens also can be clearly distinguished from Ps. lagerstroemiigena by the absence of external secondary mycelium and secondary conidiophores [36]. In addition, due to the lack of accessible reference describing Ps. paramignyae on L. parviflora from India, our specimens were compared with the available description of Ps. paramignyae on Evodia sp. from China [37]. Our specimens are clearly distinguishable from Ps. paramignyae in having smaller stromata (vs. 30–80 µm), longer conidiophores (vs. 5.0–15.0 × 2.0–4.0 µm), and shorter conidia (vs. 25–100 µm). Conidia of our specimens are narrowly obclavate to cylindrical, whereas Ps. paramignyae are described as acicular to narrowly cylindrical and often strongly curved. However, molecular comparison was not possible due to the unavailability of the molecular data of those above-mentioned fungal species.

Table 2.

Morphological comparison of Pseudocercospora lythracearum specimens examined this study with other related species.

Fungus	Leaf spot	Mycelium	Stroma	Conidiophore	Conidium			Ref.
					Size (µm)	Septum	Shape
Ps. lythracearum	Amphigenous; caespituli amphigenous	Immersed	Well-developed; globular; 25–60 µm diameter	7.5–46.4 × 2.0–5 µm; indistinct 0–3-septate; occasionally slightly geniculate	(15.1–)21.4–65.5 × 2.3–3.75	2–8-septate	Narrowly obclavate to cylindrical obclavate; apex obtuse to slightly rounded; base short obconically truncate	This study
Ps. lagerstroemiae-subcostatae	Amphigenous; caespituli amphigenous	Immersed	None or up to 55 µm wide	10.0–60.0 × 3.0–5.0 µm; 1–3-septate; 0–2-geniculate	25.0–70.0 × 2.0–4.0	2–7-septate	Cylindrical to obclavate-cylindrical; apex obtuse; base subtruncate to short obconically truncate	[33]
Ps. lagerstroemiae-parviflorae	N/A	N/A	Developed	105–204 µm	3–5 wide	N/A	N/A	[34]
Ps. lagerstroemiae-lanceolatae	N/A	Secondary mycelium superficial	10–40 µm diameter	10.0–100.0 × 3.0–6.0 µm; (0–)1–4(5)-septate	30.0–100.0 × 3.0–4.0	3–8-septate	Obclavate-cylindrical; apex obtuse to subacute; base obconically truncate	[35]
Ps. lagerstroemiigena	Amphigenous; caespituli amphigenous	Secondary mycelium superficial	Immersed; subglobular; 15–40 µm diameter	10.0–30.0 × 2.0–3.0 µm; 0–1-septate; 0–2-geniculate; densely fasciculate or borne singly on the external mycelial hyphae	10.0–95.0 × 2.0–3.0	1–9-septate	Cylindrical to filiform or cylindric-obclavate; apex subacute to subobtuse; base obconic to obconically truncate	[36]
Ps. paramignyae	Amphigenous; caespituli hypogenous	Immersed	Subepidermal; globular; 30–80 µm diameter	5.0–15.0 × 2.0–4.0 µm; aseptate	25.0–100.0 × 2.0–3.0	3–10-septate	Acicular to narrowly cylindrical; apex subacute to obtuse; base obconically truncate to truncate	[37]
Ps. rhododendricola	Amphigenous; Caespituli epiphyllous	N/A	Immersed; 30–60 µm diameter	18.0–26.0 × 2.4–3.0 µm; generally continuous (rarely 1–2-septate); occasionally geniculate	54–96 × 2.0–2.5	3–6-septate	Filiform; apex sub-conical; base slightly tapered and more or less truncate	[38]

N/A: data not available.

3.2. Phylogenetic analysis

Among the four loci examined from eight isolates, ITS and rpb2 gene sequences showed the lowest levels of nucleotide variation. In the ITS region, six isolates (KACC 43950, KACC 411337, KACC 411338, MUCC 1202, MUCC 1232, and MUCC 1401) shared identical sequences and showed 100% identity to Pseudocercospora sp. on Macroptilium atropurpureum (OQ793671) in BLASTn search results. The remaining two isolates (KACC 42649 and KACC 411339) were also identical to each other, differing by only two nucleotide positions (guanine to adenine and a single nucleotide insertion absence to adenine) and showed 99.85% similarity to Pseudocercospora sp. on Piper sarmentosum (OQ793672). Similarly, the rpb2 sequences were highly conserved, with four isolates (KACC 43950, KACC 411337–411339) shared identical sequences and showed 100% identity with Ps. variicolor (LC744598) and Pseudocercospora sp. (PP404652, on Lagerstroemia sp.; MG930772, on Lonicera vidalii). The remaining isolates, differed from these isolates by a single nucleotide substitution (thymine to cytosine), resulting in 99.68% similarity to the same reference sequences.

Various mutations among the examined isolates were observed in both actA and tef1 loci. For actA gene, six isolates (KACC 43950, KACC 411337, KACC 411338, MUCC 1202, MUCC 1232, MUCC 1401) shared identical actA sequences (221 bp) and showed 100% identity with Pseudocercospora sp. on Populus spp. (KM982623, KM982605, KM982631), Ps. lythracearum (GU320442), and Ps. guianensis (GU320424). In contrast, KACC 411339 was identical to Ps. nandinae (GU320452, KX462568), while KACC 42649 was identical to Ps. glauca (GU320419), Ps. ceratoniae (LC599414), and Ps. cotini (LC599416). Sequence comparisons indicated that the six identical isolates differed from KACC 411339 with eight nucleotides and from KACC 42649 with 10 nucleotides, whereas KACC 42649 and KACC 411339 differed from each other with four nucleotides. The tef1 gene showed the highest level of sequence variations. Three isolates (KACC 43950, KACC 411338, KACC 411339) shared identical sequences and showed 99.03% similarity to Ps. pancratii (GU384470) and 98.71% to Ps. lythracearum (GU384452). Two isolates (KACC 411337, MUCC 1202) were 100% identical to Ps. rhapisicola (GU384482, LC599483). The remaining three isolates (KACC 42649, MUCC 1232, MUCC 1401) were identical to Ps. rhododendricola (ON997542, ON997541), Ps. lythracearum (GU384452), Pseudocercospora sp. on Pyracantha angustifolia (JQ325002), and Ps. pyracanthae-fortuneanae (KX853068). The KACC 411339 isolate group was differentiated from the KACC 411337 isolate group by 29 nucleotide positions and from the KACC 42649 isolate group by four positions, whereas KACC 411337 group and KACC 42649 group differed by 27 nucleotides. These results indicate that the individual loci showed varying levels of sequence genetic divergence, with ITS and rpb2 genes being relatively conserved, while actA and especially tef1 genes showed higher levels of sequence variability.

A phylogenetic tree based on ML analysis was constructed using a multilocus dataset composed of 49 taxa and 1876 characters (http://purl.org/phylo/treebase/phylows/study/TB2:S32587). The matrix consisted of 561 parsimony-informative sites, 1099 constant sites, and 777 variable sites. The resulting phylogenetic analyses based on ML, MP, and BI consistently supported the identification of isolates obtained in this study as Ps. lythracearum (Figure 3). In the multilocus phylogeny, the isolates obtained in this study formed a well-supported clade together with previously identified Ps. lythracearum isolates from Korea (CBS 131925) and Japan (MUCC 890), supported with high BS values from ML and PP from BI analysis (100/0.9). Although sequence comparisons of each locus revealed varying levels of genetic divergence, all isolates were consistently clustered within the Ps. lythracearum clade. The observed internal branching within this clade likely reflects locus-specific variability rather than interspecific divergence. Furthermore, our isolates also clustered together with a Pseudocercospora sp. isolate from Lagerstroemia sp. collected in the USA (CPC 14341). Although direct morphological comparison could not be done because of no detailed morphological information for this isolate in the published data [16], CPC 14341 provides an additional phylogenetic point of comparison for Pseudocercospora associated with Lagerstroemia. However, its taxonomic identity cannot be determined based on its geographic origin alone, and further morphological and molecular data are needed to clarify its relationship to Ps. lythracearum.

Figure 3.

A maximum-likelihood (ML) tree of Pseudocercospora spp. based on a concatenated ITS, actA, tef1, and rpb2 dataset (49 sequences; 1876 characters). Isolates obtained in the present study are highlighted in bold. Node support values include ML and maximum parsimony (MP) bootstrap values (>70%) and Bayesian posterior probabilities (≥0.90). Sequences derived from type or reference material are indicated by superscript. HT (from the holotype or ex-type), NT (from the neotype or ex-neotype), REF (from reference material), and ET (from the epitype or ex-epitype).

Pseudocercospora rhododendricola (RSP-2, RSP-3) was placed as a phylogenetically closely related taxon with Ps. lythracearum in the analysis. Morphological comparison with the original description of Ps. rhododendricola showed that our examined specimens of Ps. lythracearum are clearly distinguishable from the latter species by having amphigenous caespituli (vs. entirely epiphyllous), longer and wider conidiophores (vs. 18–26 × 2.4–3 µm), and shorter conidia (vs. 54–96 µm) (Table 2) [39]. On the other hand, similar to Ps. lythracearum, although the distribution of Ps. rhododendricola is known in Asian countries including China, Indonesia, Hong Kong, and Singapore [35,38,40], the limited number of samples examined prevents a definitive discussion of the species delimitation between Ps. rhododendricola and Ps. lythracearum.

4. Discussion

The plant genus Lagerstroemia L. (Lythraceae, Myrtales) is an ornamental woody plant that is phylogenetically positioned in the Rosids group of core Eudicots [41]. The genus consists of approximately 60 species and is widely distributed from tropical to northern temperate regions, with centers of diversity in southern and eastern Asia and northern Australia [13,42,43]. Despite this diversity, only a small fraction of the genus is used for ornamental purposes [44], with L. indica and L. fauriei being the most widely utilized species [45]. To date, more than 200 Lagerstroemia cultivars have been recognized, most of which are derived from the two parental species or their hybrids [45].

Among the diseases affecting Lagerstroemia spp. and their hybrids, powdery mildew caused by Erysiphe australiana (=E. lagerstroemiae) is regarded as the most widespread and economically important disease, followed by Cercospora leaf spot [13,44,46]. Although Cercospora leaf spot generally does not result in severe deterioration of plant health, it can reduce ornamental value by affecting foliage quality and autumn coloration [13].

The variation observed in both morphology and phylogeny of Ps. lythracearum isolates likely reflects genetic diversity within the species. This diversification may be related to the diversity and wide distribution of Lagerstroemia hosts, especially in East Asia. Thus, the results of this study clarify the taxonomic placement of the Korean and Japanese isolates associated with Cercospora leaf spot on Lagerstroemia spp. and support their identification as Ps. lythracearum.

Despite the horticultural significance of L. fauriei, knowledge of fungal pathogens associated with this species remains limited [47]. Cercospora leaf spot on L. fauriei has been reported previously from several countries [13,44,46]. However, these studies were mainly focused on the disease report, and did not provide detailed morphological descriptions or molecular data for the taxonomical study of the fungus.

In this study, Ps. lythracearum was identified as the species associated with Cercospora leaf spot on Lagerstroemia spp. in Korea and Japan, based on the integrated taxonomical concept, combined morphological characteristics on symptomatic specimens and multilocus phylogenetic analyses with multiple isolates. Moreover, the diversification of morphological characteristics and lineages of Pseudocercospora species having a limited host range in the original habitats of their host plants was revealed. It may suggest a potential scenario of host–pathogen co-speciation. This study also represents the first morphologically and molecularly confirmed association of Ps. lythracearum with L. fauriei in Korea, contributing to a clearer understanding of Pseudo­cercospora diversity on Lagerstroemia spp. and providing a foundation for future taxonomic and disease management studies.

Funding Statement

This research was supported by the Global-Learning & Academic Research Institution for Master’s·PhD Students, and Postdocs (LAMP) Program of the National Research Foundation of Korea (NRF) grant funded by the Ministry of Education (No. RS-2024-00443714). The laboratory grant, Institute for Fermentation, Osaka (IFO), and KAKENHI 26K08895 supported CN.

Disclosure statement

No potential conflict of interest was reported by the author(s).