Phylogenetic assessment and taxonomic revision of Halobyssothecium and Lentithecium (Lentitheciaceae, Pleosporales)

Our studies on lignicolous aquatic fungi in Thailand, Sweden, and the UK resulted in the collection of three new Halobyssothecium species (H. bambusicola, H. phragmitis, H. versicolor) assigned to Lentitheciaceae (Pleosporales, Dothideomycetes). Multi-loci phylogenetic analyses of the combined large subunit, small subunit, internal transcribed spacers of ribosomal DNA, and the translation elongation factor 1-alpha sequence data enabled a revision of the taxa assigned to Lentithecium and the transfer of L. cangshanense, L. carbonneanum, L. kunmingense, L. unicellulare, and L. voraginesporum to Halobyssothecium. Collection of an asexual morph of L. lineare and phylogenetic analysis confirmed its taxonomic placement in Keissleriella. Detailed descriptions and illustrations of H. bambusicola, H. phragmitis, and H. versicolor are provided.

Various studies have shown that Pleospora obiones/ Leptosphaeria discors are synonyms, but clearly do not belong in any of these genera (Khashnobish and Shearer 1996). Jones (1962), Cavaliere (1968), and Webber (1970) reported Leptosphaeria discors collections with larger ascospores than those by Crouan and Crouan (1867) indicating that there might be a second morphologically similar species.
In the present study, a phylogenetic tree of taxa in Lentitheciaceae was constructed based on sequence data of four loci (LSU, SSU, ITS, TEF1-α) to reevaluate the taxonomic status of Halobyssothecium and Lentithecium. The latest treatments and updated accounts of Lentitheciaceae in Dayarathne et al. (2018), Hongsanan et al. (2020), and Wijayawardene et al. (2020)

Materials and methods
Sample collection, morphological observation, and fungal isolation Samples of submerged decayed wood were collected from a freshwater stream in Chiang Mai, Thailand. Dead and decaying Halimione portulacoides was collected from Hayling Island bridge, Hampshire, UK. Drift culms and stems of Phragmites sp. were obtained from Sudersand and Kappelshamnsviken in Gotland, Sweden. The samples were observed using a stereomicroscope for the presence of fruiting bodies. Micromorphological features were photographed using a Motic SMZ 168 Series dissection microscope for fungal structures on the woody substrate while microscopic characters were documented using a Nikon Eclipse 80i microscope. Single spore isolation was used to obtain pure cultures and colonial characteristics described. Herbariumtype specimens are deposited in Mae Fah Luang University (MFLU). Ex-type and ex-paratype living cultures are deposited at Mae Fah Luang University Culture Collection (MFLUCC). The new species and combinations were registered in Faces of Fungi (http://www.facesoffungi.org/; Jayasiri et al. 2015) and Index Fungorum database (http://www.indexfungorum.org/ names/IndexFungorumRegisterName.asp).

DNA extraction, PCR amplification, and sequencing
Fungal mycelia from pure cultures grown in malt extract agar (MEA) for 30 days were scraped using a sterilized scalpel and kept in a sterilized 1.5 mL microcentrifuge tube. Genomic DNA was extracted using the Biospin Fungus Genomic DNA Extraction Kit (BioFlux®, China) following the manufacturer's protocol. Polymerase chain reaction (PCR) was used to amplify four markers: the large subunit (LSU), small subunit (SSU), internal transcribed spacers (ITS) of rDNA, and the translation elongation factor 1-alpha gene (TEF1-α). The LSU was amplified using the primers LR0R and LR5 (Vilgalys and Hester 1990). The SSU was amplified using the primers NS1 and NS4 (White et al. 1990). For ITS, primers ITS5 and ITS4 were used (White et al., 1990). TEF1-α was amplified using primers EF1-983F and EF1-2218R (Rehner 2001). Polymerase chain reaction was performed in a volume of 25 μl, which contained 12.5 μl of 2× Power Taq PCR Master Mix (Bioteke Co., China), 1 μl of each primer (10 pM), 1 μl genomic DNA, and 9.5 μl doubledistilled water (ddH 2 O). The PCR thermal cycle program for LSU, SSU, ITS, and TEF1-α amplification were as follows: initial denaturing step of 94°C for 3 min, followed by 40 cycles of denaturation at 94°C for 45 seconds, annealing at 56°C for 50 seconds, elongation at 72°C for 1 min, and final extension at 72°C for 10 min. Agarose gel electrophoresis was done to confirm the presence of amplicons at the expected molecular weight. PCR products were purified and sequenced with the primers mentioned above at a commercial sequencing provider (Beijing Qingke Biotechnology Co., Ltd). A BLASTn search of the newly generated sequences was carried out to exclude contamination and to search for related taxa in GenBank database (www.ncbi.nlm.nih.gov/blast/).

Phylogenetic analyses
The taxa table was assembled based on the closest matches from the BLASTn search results and from recently published data in Dayarathne et al. (2018) and Devadatha et al. (2020). Sequences generated from the four markers were analyzed along with other sequences retrieved from GenBank (Table 1). Four datasets, one for each marker, were aligned with MAFFT v. 7 using the web server (http://mafft.cbrc.jp/alignment/server; Katoh et al. 2019) with the following settings: L-INS-i tree-based iterative refinement methods, 20PAM/k = 2 scoring matrix for nucleotide sequences and 1.53 gap opening penalty. Alignment was further refined manually, where necessary, using BioEdit v.7.0.9.0 (Hall 1999). Aligned sequences were automatically trimmed using TrimAl v. 1.3 on the web server (http://phylemon.bioinfo.cipf. es/utilities.html). The online tool "ALTER" (Glez-Peña et al. 2010) was used to convert the alignment file to phylip and nexus formats. Phylogenetic analyses of both individual and combined gene data were performed using maximum likelihood (ML), maximum parsimony (MP), and Bayesian inference (BI).
Maximum parsimony (MP) analysis was performed using the heuristic search option with 1000 random taxa addition and tree bisection and reconnection (TBR) as the branchswapping algorithm in PAUP* 4.0b4 (Swofford 2002). All characters were unordered and of equal weight and gaps were treated as missing data. Maxtrees were unlimited, branches of zero length were collapsed and all multiple, equally parsimonious trees were saved. Clade stability was assessed using a bootstrap (BS) analysis with 1000 replicates, each with ten replicates of random stepwise addition of taxa (Hillis and Bull 1993     Maximum likelihood analysis was performed using RAxML-HPC2 on XSEDE on the CIPRES web portal (Stamatakis 2006(Stamatakis , 2014Stamatakis et al. 2008) (http:// www.phylo.org/portal2/; Miller et al. 2010). The GTR+ GAMMA model of nucleotide evolution was used. RAxML rapid bootstrapping of 1,000 replicates was performed. The best-fit evolutionary models for individual and combined datasets were estimated under the Akaike Information Criterion (AIC) using jModeltest 2.1.10 on the CIPRES web portal and each resulted to the GTR+I+G model (Nylander 2004;Darriba et al. 2012). Bayesian inference analyses were performed using MrBayes v. 3.2.6 on XSEDE at the CIPRES webportal (Ronquist and Huelsenbeck 2003), using the parameter setting of two parallel runs, four chains, the run for 4,000,000 generations at which point the standard deviation of split frequencies was below 0.01. Trees were sampled every 1,000 generations and all other parameters were left as default. Bayesian analysis resulted in 4,000 trees after the run wherein the first 1,000 trees, 25% of the total, were in the burn-in phase and were discarded. The remaining 3,000 trees were used to calculate the posterior probability (PP). Newly generated sequences have been deposited in GenBank (Table 1).

Genealogical concordance phylogenetic species recognition analysis
New species and their most closely related species were analyzed using the Genealogical concordance phylogenetic species recognition (GCPSR) model. A pairwise homoplasy index (PHI) (Bruen et al. 2006) test was performed in SplitsTree4 (Huson 1998;Huson and Bryant 2006) as described by Quaedvlieg et al. (2014). This was done to determine the recombination level within phylogenetically closely related species using a four-locus concatenated dataset for new species of Halobyssothecium. The test detects incompatibility between pairs of sites regarding whether there is genealogical history that can be inferred parsimoniously that does not involve any recurrent or convergent mutations. Pairwise homoplasy index below a 0.05 threshold (Фw < 0.05) indicates that there is significant recombination present in the dataset. The relationships between closely related species were visualized by constructing a split graph, using both the LogDet transformation and splits decomposition options.

Taxonomy
Culture characteristics: On MEA, colony circular with filamentous margin, reaching 25-30 mm diam. in 25 days at 25°C , brown to grayish brown from above, yellowish brown to dark brown from below, surface rough, dry, raised, with dense mycelia, edge filiform. Etymology: In reference to the host genus Phragmites, from which the species was isolated.
Notes: HKAS 102150. LSU, SSU, ITS, TEF1-α sequence data are available. Sexual morph: Undetermined. Asexual morph: Descriptions and illustrations refer to Hyde et al. (2016) Distribution: EGYPT, Sohag City, on decayed wood submerged in the River Nile   (2020) In the phylogenetic analysis (Fig. 1), Psedomurilentithecium camporesii does not cluster within Lentitheciaceae but forms a w e a k l y s u p p o r t e d c l a d e b a s a l t o L a t o r u a c e a e , Longipedicellataceae, and Trematosphaeriaceae. Broader taxon sampling, including other families in Pleosporales, is necessary to confirm its placement.
Keissleriella caudata (E. Müll.) Corbaz, Phytopathologische Zeitschrift 28 (4): 411 (1957) Preliminary phylogenetic analysis shows that Keissleriella caudata does not group with other Keissleriella species, but clusters instead with Corynespora species. Only ITS sequence data of K. caudata is available in GenBank with an accession number MH857034. BLAST analysis did not show any Keissleriella species in the first 100 closely related sequence data. A fresh collection of specimens and additional DNA sequence data are required to confirm its placement within Pleosporales.
Keissleriella rara was transferred to Lentithecium by Suetrong et al. (2009) together with K. cladophila and Massarina phragmiticola. The present phylogenetic analysis shows that Lentithecium rarum clustered in Keissleriella as sister taxon to K. trichophoricola Crous & Quaedvl. (Fig. 1). The same placement was observed also by Singtripop et al. (2015). Keissleriella linearis was transferred by Zhang et al. (2009b) to Lentithecium based on LSU and SSU sequence data. Keissleriella linearis, in common with other Keissleriella species, has short brown setae around the apex of the ascomatal ostiole, but Zhang et al. (2009b) opined that the presence of setae has little phylogenetic significance. In their phylogenetic analysis, other species and strains of Keissleriella were not included. Singtripop et al. (2015) reexamined the type specimen of L. lineare and transferred it to Keissleriella based on morphology and LSU sequence data, and this is in agreement with recent studies by Tanaka et al. (2015), Hyde et al. (2016) and the present study. However, Dayarathne et al. (2018) and Devadatha et al. (2020) placed L. lineare in the Lentithecium clade. The recent discovery of the asexual morph of L. lineare by Tibell et al. (2020) and the phylogenetic analysis based on the four-locus sequence dataset in the present study supports its taxonomic placement in Keissleriella.
The continuous discovery of novel fungal species has significantly contributed to the revision of fungal taxa (Arzanlou et al. 2007;Boonmee et al. 2011;Tanaka et al. 2015;Hashimoto et al. 2017;Hyde et al. 2018Hyde et al. , 2020a. Phylo ge ne tic a nal ysis of t he n ewly dis cov ere d Halobyssothecium species, including all the members of Lentitheciaceae, with molecular data supports the transfer of Lentithecium cangshanense, L. carbonneanum, L. kunmingense, L. unicellulare, and L. voraginesporum to Halobyssothecium. In the present placement, members of Halobyssothecium have brown and versicolored ascospores without sheath and hyaline conidia, while Lentithecium species possess hyaline ascospores with mucilaginous sheaths.