Novel saprobic Hermatomyces species (Hermatomycetaceae, Pleosporales) from China (Yunnan Province) and Thailand

Abstract During our survey of the diversity of woody litter fungi in China and Thailand, three Hermatomyces species were collected from dead woody twigs of Dipterocarpus sp. (Dipterocarpaceae) and Ehretiaacuminata (Boraginaceae). Both morphology and multigene analyses revealed two taxa as new species (Hermatomycesturbinatus and H.jinghaensis) and the remaining collections as new records of H.sphaericus. Hermatomycesturbinatus is characterized by 1) dimorphic conidia, having circular to oval lenticular conidia and 2) turbinate conidia consisting of two columns with two septa composed of 2–3 cells in each column. Hermatomycesjinghaensis is characterized by dimorphic conidia, having circular to oval lenticular conidia and clavate or subcylindrical to cylindrical conidia and consisting of one or two columns with 6–8 cells in each column. Phylogenetic analyses of combined LSU, ITS, tub2, tef1-α and rpb2 sequence data supports the placement of these new taxa within Hermatomycetaceae with high statistical support.


Introduction
Over the past few decades, the number of studies using a molecular-based approach to study microfungal diversity in the greater Mekong subregion (GMS) has increased rapidly, especially on freshwater and woody litter fungi from China (Yunnan Province) and Thailand (Hapuarachchi et al. 2019;Dong et al. 2020;Li et al. 2020;Monkai et al. 2020;Wanasinghe et al. 2020Wanasinghe et al. , 2021Mortimer et al. 2021). Hyde et al. (2018) reported that about 96% of fungi from Thailand are new to science. Feng and Yang (2018) estimated 104,000 fungal species currently exist in Yunnan Province, China; however, only about 6,000 are extant. Therefore, further studies need to be conducted to fill gaps in knowledge regarding the diversity, taxonomy and phylogeny of microfungi in the GMS. Supporting this obligation, we have begun to study plant-based ascomycetes in GMS. The current study accounts for hermatomyces-like ascomycetes recovered from the woody litter in China (Yunnan Province) and Thailand.
Hermatomyces was introduced by Spegazzini (1911) with H. tucumanensis as the type species. Doilom et al. (2017) accommodated Hermatomyces in Lophiotremataceae based on combined LSU, SSU, tef1-α and rpb2 sequence data. Later, Hashimoto et al. (2017) validated Hermatomycetaceae (Hermatomycetaceae Locq. 1984 was not validly published, Art. 39.1) to accommodate the genus Hermatomyces. This genus is known only by its asexual morph that is characterized by sporodochial conidiomata and dimorphic (lenticular or cylindrical) conidia of one or two types. The lenticular conidia are globose to subglobose, hyaline to pale brown peripheral cells with dark brown central cells, and the cylindrical conidia is hyaline, cylindrical to subcylindrical or turbinate and consisting of 1-4 columns of 2-12 cells (Spegazzini 1911;Tibpromma et al. 2016;Hashimoto et al. 2017;Hyde et al. 2019;Pem et al. 2019;Phukhamsakda et al. 2020).
Our investigation led to the discovery of three Hermatomyces species, including two novel species, on dead woody-based substrates. Morphological illustrations and multi-gene phylogenetic analyses with ML, MP and BI of combined LSU, ITS, tub2, tef1-α and rpb2 sequence data are used to confirm the phylogenetic placement of the novel species within Hermatomyces.

Sample collection, examination and isolation
Woody litter samples were collected from China (Yunnan Province) during the dry season (December 2019) and Thailand (Tak Province) during the wet season (August 2019). Samples were brought to the laboratory using plastic Ziploc bags. Fungal specimens were then examined using a stereomicroscope (Olympus SZ61, China). Pure cultures were obtained via single spore isolation on potato dextrose agar (PDA) following the methods described in Senanayake et al. (2020). Cultures were incubated at 25 °C for three weeks. Micro-morphological structures were photographed using a Nikon compound microscope (Nikon ECLIPSE Ni) fitted with a Canon (EOS 600D) digital camera. Measurements were taken using the Tarosoft

DNA extraction, amplification and sequencing
DNA extraction, amplification, sequencing, sequence alignment and phylogenetic analyses follow the methods of Dissanayake et al. (2020) with the following details. Two partial rDNA genes and three protein coding genes were used in our study, including internal transcribed spacer region (ITS) using primer pair ITS5/ITS4 (White et al. 1990), 28S large subunit nuclear ribosomal (LSU) using primer pair LR0R/ LR5 (Vilgalys and Hester 1990), translation elongation factor 1-alpha gene (tef1-α) using primer pair EF1-983F/EF1-2218R (Rehner and Buckley 2005), RNA polymerase II second largest subunit (rpb2) using primer pair fRPB2-5F/fRPB2-7cR (Liu et al. 1999) and β-tubulin (tub2) using primer pair T1/T22 (O'Donnell and Cigelnik 1997). Amplification reactions were performed in a total volume of 25 μL of PCR mixtures containing 8.5 μL ddH 2 O, 12.5 μL 2× PCR MasterMix (TIANGEN Co., China), 2 μL DNA template and 1 μL of each primer. The PCR thermal cycle program for LSU, ITS, tef1-α and rpb2 were set as described in Tibpromma et al. (2018). The PCR amplification condition of tub2 was set as denaturation at 94 °C for 3 minutes, followed by 35 cycles of denaturation at 94 °C for 45 seconds, annealing at 56 °C for 50 seconds and extension at 72 °C for 1 minute, with a final extension step at 72 °C for 10 minutes. PCR products were sent to the Qingke Company, Kunming City, Yunnan Province, China, for sequencing. Sequences were deposited in GenBank (Table 1).

Sequence alignment and phylogenetic analyses
Representative species used in the phylogenetic analyses were selected based on previous publications (Koukol et Phukhamsakda et al. 2020). Sequences were downloaded from GenBank (http://www. ncbi.nlm.nih.gov/), and their accession numbers are listed in Table 1. The newly generated sequences in this study were assembled by BioEdit 7.0.9.0 (Hall 1999). Individual gene regions were separately aligned in MAFFT v.7 web server (http://mafft.cbrc.jp/ alignment/server/) (Katoh et al. 2019). The alignments of each gene were improved by manually deleting the ambiguous regions plus gaps and combined using BioEdit 7.2.3. Final alignments containing LSU, ITS, tub2, tef1-α and rpb2 were converted to NEXUS format (.nxs) using CLUSTAL X (2.0) (Thompson et al. 1997) and processed for Bayesian and maximum parsimony analysis. The FASTA format was changed into PHYLIP format via the Alignment Transformation Environment (ALTER) online program (http://www.sing-group.org/ALTER/) and used for maximum likelihood analysis (ML). ML was carried out in CIPRES Science Gateway v.3.3 (http://www.phylo.org/ portal2/; Miller et al. 2010) using RAxML-HPC2 on XSEDE (8.2.12) (Stamatakis 2014) with the GTRGAMMA substitution model and 1,000 bootstrap iterations. Maximum parsimony analysis (MP) was performed in PAUP v. 4.0b10 (Swofford 2002) with the heuristic search option and Tree-Bisection-Reconnection (TBR) of branch-swapping algorithm for 1,000 random replicates. Branches with a minimum branch length of zero were collapsed, and gaps were treated as missing data (Hillis and Bull 1993).
Bayesian analysis was executed in MrBayes v.3.2.2 (Ronquist et al. 2012). The model of evolution was estimated using MrModeltest v. 2.3 (Nylander et al. 2008) via PAUP v. 4.0b10 (Ronquist and Huelsenbeck 2003). The SYM+I+G for LSU and ITS; HKY+I for tub2; GTR+I+G for tef1-α and rpb2 were used in the final command. Markov chain Monte Carlo sampling (MCMC) in MrBayes v.3.2.2 (Ronquist et al. 2012) was used to determine posterior probabilities (PP) (Rannala and Yang 1996;Zhaxybayeva and Gogarten 2002). Bayesian analyses of six simultaneous Markov chains were run for 2,000,000 generations and trees were sampled and printed to output at every 200 generations (resulting in 10,001 total trees). The first 25% of sampled trees were discarded as part of the burn-in procedure, the remaining 7,501 trees were used to create the consensus tree and the average standard deviation of split frequencies was set as 0.01. Phylogenetic

Hermatomyces iriomotensis KH 361
Hermatomyces megasporus CCF 5897  Figure 1. Phylogenetic RAxML tree based on analysis of a combined LSU, ITS, tub2, tef1-α and rpb2 and dataset. Bootstrap support values for ML and MP equal to or higher than 75% and Bayesian PP equal to or greater than 0.95 are shown at nodes. Hyphens (--) represent support values less than 75% / 0.95 BYPP. The ex-type strains are in bold and the new isolate in this study is in blue bold. The tree is rooted with Anteaglonium globosum (ANM 925.2) and A. parvulum (MFLUCC 14-0821). The scale bar represents the expected number of nucleotide substitutions per site. constant, 174 variable characters were parsimony-uninformative and 622 characters were parsimony-informative. The Kishino-Hasegawa test shows length = 1,388 steps with CI = 0.671, RI = 0.884, RC = 0.593 and HI = 0.329. The RAxML analysis of the combined dataset yielded a best scoring tree with a final ML optimization likelihood value of -13406.555506. Estimated base frequencies were as follows: A = 0.241874, C = 0.266701, G = 0.257552, T = 0.233873; substitution rates AC = 1.188604, AG = 4.826453, AT = 1.273226, CG = 0.855218, CT = 11.409386, GT = 1.00; gamma distribution shape parameter α = 0.16102.
In the phylogenetic tree obtained from ML, MP and BI analysis (Fig. 1) the maximum likelihood analysis resulted in trees largely with similar topology and clades as in the maximum parsimony and Bayesian analyses. The new species, Hermatomyces turbinatus, is sister to H. nabanheensis (KUMCC 16-0149) with high support (94% ML, 91% MP and 1.00 BYPP, Fig. 1). Hermatomyces jinghaensis is nested between H. trangensis and H. clematidis with a strongly supported monophyletic group (98% ML, 92% MP, 1.00 PP; Fig. 1). New isolates of H. sphaericus (KUMCC 20-0231; MFLUCC 21-0036) clustered with remaining H. sphaericus strains as a monophyletic group (Fig. 1). The topology of the phylogenetic tree is in accordance with recent phylogenetic studies discussing species in Hermatomycetaceae (Nuankaew et al. 2019;Phukhamsakda et al. 2020). Etymology. Referring to the turbinate shape of the conidia.
Known host and distribution. Dipterocarpus sp. (Thailand). Culture characteristics. Colonies on PDA, reaching 30-40 mm diam., after 3 weeks at 25-30 °C, circular, convex with papillate and radially furrowed at the center, rough, labate, crenate edge, fluffy, dense, gray black, in reverse darkens at the center, pale brown to gray at edge. Notes. Hermatomyces turbinatus is introduced as a new species based on its distinct morphology, which is supported by phylogenetic analyses. In the phylogenetic analyses, H. turbinatus is distinct from extant species in this genus and formed a sister clade to H. nabanheensis with strong support (94% ML, 91% MP, 1.00 PP; Fig. 1). Hermatomyces turbinatus differs from H. nabanheensis in having turbinate conidia with two columns, while H. nabanheensis has cylindrical conidia with one or two columns. Hermatomyces turbinatus has two conidial types, and its lenticular conidia are similar to H. tectonae in shape and size. However, the turbinate conidia of H. turbinatus have 2 columns of 2-3 cells in each column, while the turbinate conidia of H. tectonae have 2 columns of 3 cells in each column. We also compared the morphological characters of H. turbinatus to other species of Hermatomyces (Table 2). Despite no molecular data being available for the three species viz. H. dimorphus, H. uniseriatus and H. truncates, H. turbinatus nonetheless differs from these species in conidial characteristics (Table 2). Etymology. The species epithet "jinghaensis" refers to the location where the species was collected.
Known host and distribution. Tropical and subtropical regions of Central and South America, Africa, Asia, Oceania and North America. The species were found as saprobes on Acanthaceae, Apocynaceae, Arecaceae, Asteraceae, Dipterocarpaceae, Euphorbiaceae, Fabaceae, Lamiaceae, Leguminosae, Mimosaceae, Nyctaginaceae, Oxalidaceae, Pandanaceae, Pinaceae, Rhamnaceae, and Sterculiaceae (Zhang et al 2009;Koukol et al. 2018Koukol et al. , 2019. Culture characteristics. Colonies on PDA, reaching 35-40 mm diam., after 3 weeks at 25-30 °C, with circular, umbonate, fluffy, velvety, entire edge, a circular raised band, gray white, in reverse dark gray, black toward the center. Notes. The characters of our new strain of Hermatomyces sphaericus (KUMCC 20-0231, MFLUCC 21-0036) are similar to the type collection (K(M)-IMI 37763) in having gray black to black sporodochia, mononematous, pale brown, smooth, monoblastic, integrated, terminal, cylindrical, hyaline to pale brown conidiogenous cells and globose to subglobose conidia (Hughes 1953). A multigene phylogeny indicates that novel strains clustered within the H. sphaericus clade (Fig. 1). We name our strain (KUMCC 20-0231, MFLUCC 21-0036) as H. sphaericus, which has been reported from different plant families and genera (Koukol et al. 2018). However, we consider this might be a species complex that need further detailed studies. Our study provides the new host records of H. sphaericus on Dipterocarpus sp. (Dipterocarpaceae) and Ehretia acuminata (Boraginaceae), and updates sequence data for the new collections of H. sphaericus.

Discussion
This study introduces two new species of woody-based litter fungi; Hermatomyces jinghaensis from Yunnan, China and Hermatomyces turbinatus on Dipterocarpus sp. from Thailand. We also report for the first time two new records of H. sphaericus on Dipterocarpus sp. and Ehretia acuminata in China and Thailand.
Hermatomyces (Hermatomycetaceae) is different from other similar genera in its sporodochial conidiomata and in having one to two (lenticular and cylindrical conidia) unusual conidial types (Spegazzini 1911). All species of Hermatomyces have lenticular conidia with similar characteristics, whereas some species have cylindrical and turbinate conidia, which have greater variance in shape, size, number of columns and cells. Koukol et al. (2018Koukol et al. ( , 2019 have reported that multiple species may occur together on a single sample, a phenomenon we observed, which may complicate morphological identification and separation for culturing. Therefore, molecular sequence data are more reliable for the identification of Hermatomyces species (Tibpromma et al. 2016Nuankaew et al. 2019;Phukhamsakda et al. 2020). (1953), which may be the most widespread of species in Hermatomyces distributed across many subtropical and tropical regions worldwide (Wijayawardene et al. 2014;Doilom et al. 2017;Koukol et al. 2018Koukol et al. , 2019Hyde et al. 2019;Jayasiri et al. 2019;Nuankaew et al. 2019;Phukhamsakda et al. 2020). This species has been reported as saprobic on dead plant tissues of several host families (Tibpromma et al. 2016Doilom et al. 2017;Jayasiri et al. 2019). In addition, Koukol et al. (2018) reported that H. sphaericus (ARIZ: PS0053) was isolated from seeds of Apeiba membranacea (Malvaceae), suggesting this species could be an endophyte. Previous studies have indicated that H. sphaericus is not restricted to any single host (Koukol et al. 2018(Koukol et al. , 2019Jayasiri et al. 2019), whereas other species of Hermatomyces are saprobic on a limited number of hosts and are limited to specific regions (Rao and de Hoog 1986;Leão-Ferreira et al. 2013;Prasher and Prasher 2014;Hyde et al. 2016Hyde et al. , 2019Tibpromma et al. 2016Tibpromma et al. , 2017Tibpromma et al. , 2018Doilom et al. 2017;Hashimoto et al. 2017;Koukol et al. 2018Koukol et al. , 2019Nuankaew et al. 2019;Delgado et al. 2020;Phukhamsakda et al. 2020; Table 2). In this study, our new strains of H. sphaericus had slight morphological differences in lenticular conidia size compared to the type strains and other strains of H. sphaericus (Hughes 1953, Table 2). As reported by Koukol et al. (2018), H. sphaericus is a plurivorous species, and accordingly the phenotypic variation among strains could be influenced by environmental factors and culture conditions or it could have speciated in isolated polulations .

Hermatomyces sphaericus was introduced by Hughes
Species delineation in Hermatomyces, especially in the H. sphaericus clade, is subject to much controversy due to species inconsistency in morphological and phylogenetic status. Koukol et al. (2018) synonymized H. chromolaenae, H. saikhuensis and H. tectonae under H. sphaericus based on morphological and molecular comparisons and suspected that H. pandanicola could either be a hybrid species or incorrect sequences were used in the analysis. Koukol et al. (2019) considered that during isolation of H. biconisporus, a conidium of H. sphaericus might have been taken instead, leading to contamination when extracting DNA and the misinterpretation of its taxonomic placement. Phukhamsakda et al. (2020) further confirmed that H. biconisporus, H. pandanicola and H. sphaericus should be treated as the same species based on Genealogical Concordance Phylogenetic Species Recognition (GCPSR) analysis.
Hermatomyces had long been treated as "incertae sedis" within Ascomycota (Wijayawardene et al. 2012). Doilom et al. (2017) placed Hermatomyces in Lophiotremataceae baed on phylogenetic analyses, and consequently, Hashimoto et al. (2017) revised the family Lophiotremataceae based on morphological observations and phylogenetic analyses, and Hermatomyces was accepted in the family Hermatomycetaceae, as monophyletic. Recent studies and our study indicate Hermatomyces to be highly polyphyletic, and Hermatomyces morphology has evolved, which is mainly characterized by lenticular and cylindrical conidia ( Fig. 1; Koukol et al. 2018Koukol et al. , 2019Hyde et al. 2019;Phukhamsakda et al. 2020). Support for a single H. sphaericus species (Fig. 1) lacks internal statistical support and includes H. biconisporus, H. chromolaenae, H. pandanicola, H. saikhuensis and H. tectonae and we suspect that this is a species complex. Tibpromma et al. (2018) also noted that H. sphaericus could be a species complex including several species and did not accept the synonymy of H. saikhuensis and H. tectonae in H. sphaericus owing to their significant base-pair differences.
In this study, we combined two non-translated loci (LSU, ITS) and three protein-coding regions (tub2, tef1-α and rpb2) to carry out phylogenetic analysis for Hermatomyces species in order to validate phylogenetic placement of the taxa within Hermatomyces. In our phylogenetic analyses, H. tectonae, H. chromolaenae, H. biconisporus, H. pandanicola and H. saikhuensis grouped together with strains of H. sphaericus (PRC 4100, PRC 4104, PMA 116081). Hermatomyces saikhuensis and H. chromolaenae are characterized by one conidium type (lenticular) similar to H. sphaericus, however, they differ in the shape, color and size of conidia (Tibpromma et al. 2016 Table 2). Hermatomyces tectonae, H. biconisporus and H. pandanicola are characterized by dimorphic conidia which differ from H. sphaericus (Tibpromma et al. 2016Doilom et al. 2017;Koukol et al. 2018; Table 2). Hermatomyces sphaericus (PRC 4100, PRC 4104, PMA 116081) did not have a morphological description for inter-species comparison (Koukol et al. 2018). Further taxon sampling and more sequence data are needed to elucidate this clade.