﻿Exploring ascomycete diversity in Yunnan II: Introducing three novel species in the suborder Massarineae (Dothideomycetes, Pleosporales) from fern and grasses

﻿Abstract This article presents the results of an ongoing inventory of Ascomycota in Yunnan, China, carried out as part of the research project series “Exploring ascomycete diversity in Yunnan”. From over 100 samples collected from diverse host substrates, microfungi have been isolated, identified and are currently being documented. The primary objective of this research is to promote the discovery of novel taxa and explore the ascomycete diversity in the region, utilising a morphology-phylogeny approach. This article represents the second series of species descriptions for the project and introduces three undocumented species found in the families Bambusicolaceae, Dictyosporiaceae and Periconiaceae, belonging to the suborder Massarineae (Pleosporales, Dothideomycetes). These novel taxa exhibit typical morphological characteristics of Bambusicola, Periconia and Trichobotrys, leading to their designation as Bambusicolahongheensis, Periconiakunmingensis and Trichobotryssinensis. Comprehensive multigene phylogenetic analyses were conducted to validate the novelty of these species. The results revealed well-defined clades that are clearly distinct from other related species, providing robust support for their placement within their respective families. Notably, this study unveils the phylogenetic affinity of Trichobotrys within Dictyosporiaceae for the first time. Additionally, the synanamorphism for the genus Trichobotrys is also reported for the first time. Detailed descriptions, illustrations and updated phylogenies of the novel species are provided, and thus presenting a valuable resource for researchers and mycologists interested in the diversity of ascomycetes in Yunnan. By enhancing our understanding of the Ascomycota diversity in this region, this research contributes to the broader field of fungal taxonomy and their phylogenetic understanding.

A comprehensive study of the genera in Pleosporales was carried out by Zhang et al. (2012), based on morphological studies of the type specimens coupled with phylogenetic analyses.Consequently, the taxonomic treatment of numerous Pleosporales was updated by various authors, based on polyphasic taxonomic approaches, mainly using morphology-phylogeny-based taxonomy (Ariyawansa et al. 2014(Ariyawansa et al. , 2015a, b;, b;Phookamsak et al. 2014Phookamsak et al. , 2015;;Tanaka et al. 2015;Thambugala et al. 2015;Boonmee et al. 2016;Jaklitsch and Voglmayr 2016;Jaklitsch et al. 2016aJaklitsch et al. , b, 2018;;Su et al. 2016;Chen et al. 2017;Hashimoto et al. 2017;Wanasinghe et al. 2017a, b).Even though novel taxa of Pleosporales have been dramatically increasing over the last ten years after the taxonomic circumscription provided by Zhang et al. (2012) and Hyde et al. (2013), there is still over a quarter of the total known species lacking molecular data and/or reliable phylogenetic markers for clarifying the placements in Pleosporales.
Yunnan is known as one part of the 36 global biodiversity hotspots where over 17,000 species of vascular plants are known, including highly endemic species (Feng and Yang 2018;Cai et al. 2019).Highly diverse environments and geographical distribution, as well as flourishing vegetation, have shown the Province to be one of the richest sources of fungi, covering over 40% of the known species in China (Feng and Yang 2018;Liu et al. 2018).Feng and Yang (2018) estimated a species number of fungi existing in Yunnan Province, based on the ratio of local vascular plants and fungi (1:6) following the suggestion of Hawksworth (2001).With this estimation, Yunnan may harbour over 104,000 fungal species; of which only 6000 described species have been reported from the Province, including approximately 3000 species of Ascomycota and Basidiomycota (Feng and Yang 2018).
The present study aims to introduce three novel pleosporalean species from Yunnan, based on morphological characteristics and phylogenetic evidence coupled with the differences in nucleotide pairwise comparison amongst closely-related species.

Sample collection, isolation, morphological examination and preservation
Samples were collected from Yunnan Province, China during 2016-2021 at three different collecting sites: Honghe (rice terraces), Kunming (botanical garden) and Xishuangbanna (secondary forest).Specimens were collected during the rainy (September) and dry seasons (January and April) and brought to the laboratory in sealed plastic Ziploc bags for further observation and examination.The samples were observed and axenic cultures, via single spore isolation, were obtained within 1-2 weeks after collection.Single spore isolation was performed using the spore suspension technique (Senanayake et al. 2020).Two sets (five spores per set) of the germinated spores were placed separately on to freshly sterilised potato dextrose agar (PDA) medium and incubated under normal day/night light conditions at room temperature (15-25 °C depending on the rainy and dry seasons).Culture characteristics, growth and sporulation in vitro were observed and recorded after one and four-week intervals.
Macro-morphological features, such as ascomata and fungal colonies visualised on host substrates, were observed using an Olympus SZ61 series stereomicroscope and photo-captured by a digital camera.Micro-morphological features were examined by differential interference contrast (DIC) microscopy using a Nikon ECLIPSE Ni-U compound microscope and images captured with a Nikon DS-Ri2 camera.The mucilaginous sheath that covered the ascospores was checked by staining with India Ink and the fungal centrum was stained using Congo red for checking the clearity of conidiophores and conidiogenous cells.Lactoglycerol was added to preserve important morphological features on permanent slides.All morphological features were measured using Tarosoft (R) Image FrameWork version 0.9.7.and photographic plates were edited and combined using Adobe Photoshop CS6 software (Adobe Systems Inc., San Jose, CA, USA).
Axenic living cultures were preserved in PDA and sterilised double-distilled water (ddH 2 O) at 4 °C for short-term storage and long-term glycerol storage at -20 °C and -80 °C, respectively.Ex-type living cultures were deposited at the collection of Rungtiwa Phookamsak housed at Honghe Center for Mountain Futures (RPC) and duplicated in the Culture Collection of the Herbarium of Cryptogams Kunming Institute of Botany, Academia Sinica (KUNCC), Kunming, China and Mae Fah Luang University Culture Collection (MFLUCC), Chiang Rai, Thailand.The type specimens were preserved with silica gel and deposited in the Herbarium of Cryptogams Kunming Institute of Botany Academia Sinica (KUN-HKAS), China.Index Fungorum numbers (http://www.indexfungorum.org; accessed on 25 May 2023) were obtained for the newly-described taxa.

DNA extraction, PCR amplification and sequencing
Fungal genomic DNA was extracted from fresh mycelia using the Biospin Fungus Genomic DNA Extraction Kit (BioFlux, Hangzhou, China) following the procedure from the manufacturer.The genomic DNA was also extracted from ascomata using a Forensic DNA Kit (Omega, Norcross, GA, USA) in case the fungus could not be obtained from the pure culture.Amplicons were generated by polymerase chain reaction (PCR) using five phylogenetic markers, including the internal transcribed spacers region of ribosomal DNA (ITS; ITS1-5.8S-ITS2), the partial 28S large subunit nuclear ribosomal DNA (LSU), the partial 18S small subunit rDNA (SSU), the partial RNA polymerase II second largest subunit (rpb2) and the partial translation elongation factor 1-alpha (tef1-α).The ITS region was amplified with the primer pair ITS4 and ITS5 (White et al. 1990), the LSU region with LR0R and LR5 (Vilgalys and Hester 1990), the SSU region with NS1 and NS4 (White et al. 1990), the rpb2 region with fRPB2-5F and fRPB2-7cR (Liu et al. 1999) and the tef1-α region with EF1-983F and EF1-2218R (Rehner and Buckley 2005).The component of PCR reaction was performed in a total volume of 25 μl, containing 2 μl DNA template (30-50 ng/μl), 1 μl of each forward and reverse primer (10 μM), 12.5 μl Master Mix (mixture of EasyTaqTM DNA Polymerase, dNTPs and optimised buffer; Beijing TransGen Biotech Co., Ltd., Chaoyang District, Beijing, China) and 8.5 µl of double-distilled water (ddH 2 O).The thermal cycle of PCR amplification for ITS, LSU, SSU, rpb2 and tef1-α was set up following Phookamsak et al. (2014Phookamsak et al. ( , 2023)).PCR products were purified and sequenced by using PCR primers at TsingKe Biological Technology (Kunming City, Yunnan Province, China).The quality of raw sequence data was checked and trimmed of low-quality segments with BioEdit 7.1.3.0 (Hall 1999).The consensus sequences of the newly-generated strains were assembled using SeqMan Pro version 11.1.0(DNASTAR, Inc.Madison, WI, USA) and submitted to the GenBank database to further encourage accession within the scientific community.

Sequence alignments and phylogenetic analyses
The newly-generated sequences were subjected to the nucleotide BLAST search tool on the NCBI website for checking the correctness of species identification and searching for closely-related taxa that were further included in the sequence alignment dataset.Reference sequences from relevant publications and BLAST results of the closely-related species were downloaded from Gen-Bank to supplement the datasets (Tables 1-3).Three datasets were prepared to construct the phylogenetic trees for clarifying phylogenetic relationships of the novel taxa in Bambusicolaceae (Table 1), Dictyosporiaceae (Table 2) and Periconiaceae (Table 3).The individual gene dataset was aligned using MAFFT v.7 (Katoh et al. 2019) and improved manually where necessary in Bioedit 7.1.3.0 (Hall 1999).The alignments of individual gene datasets were prior analysed by Maximum Likelihood (ML) for checking the congruence of tree topologies and further combined into a multigene dataset.Phylogenetic analyses were performed, based on ML and Bayesian Inference (BI) analyses.
Maximum Likelihood (ML) implemented by the Randomised Axelerated Maximum Likelihood (RAxML), was performed in RAxML-HPC v.8 on the XSEDE (8.2.12) tool via the online web portal CIPRES Science Gateway v. 3.3 (Miller et al. 2010) using default settings, but adjusted with 1000 bootstrap replicates and a gamma-distributed rate variation of a general time reversible model (GTR) was applied.The BI analyses were conducted by MrBayes on XSEDE v. 3.2.7avia the same web portal as in ML, with two parallel runs.The best-fit model of nucleotide substitution was determined by MrModelTest v. 2.3 (Nylander et al. 2008).Six simultaneous Markov chains were run for 1-5 million generations, but stopped automatically when the critical value for the topological convergence diagnostic reached 0.01.Trees were sampled every 100 th generation.The initial 10% of sample trees were treated as burn-in (estimated by Tracer v. 1.7; Rambaut et al. (2018)) and discarded.The remaining trees were used to calculate the posterior probabilities in the majority rule consensus tree.The phylograms were visualised using Figtree v. 1.4.0 (Rambaut and Drummond 2012) and backbone trees were laid out and edited in Adobe Illustrator version 20.0.0.software (Adobe Systems Inc., San Jose, CA, USA).

Phylogenetic analyses
In this study, three phylogenetic analyses were conducted to clarify the phylogenetic placements of our new taxa within the Bambusicolaceae (Analysis 1), Dictyosporiaceae (Analysis 2) and Periconiaceae (Analysis 3), as follows:

Analysis 1
The Bambusicola species tree was constructed using a sequence dataset of the concatenated ITS, LSU, rpb2, SSU and tef1-α of all Bambusicola species, as well as representatives of other related genera.A total of 37 strains were included, with two strains of Pseudotetraploa bambusicola (CGMCC 3.20939) and P. curviappendiculata (JCM 12852) as the outgroup.Primarily, phylogenetic analysis of the concatenated LSU, SSU and ITS sequence dataset was conducted, based on ML and compared with the multigene phylogenetic analysis (the concatenated ITS, LSU, rpb2, SSU and tef1-α sequence dataset).Phylogenetic analysis, based on the concatenated LSU, SSU and ITS gene regions, showed a similar topology with the concatenated ITS, LSU, rpb2, SSU and tef1-α gene regions and were not significantly different (data not shown).Hence, multigene phylogenetic analysis of the concatenated ITS, LSU, rpb2, SSU and tef1-α gene regions was selected to represent the phylogenetic relationships of the new species with other closely-related species in Bambusicolaceae.The aligned dataset contained 4929 characters, including gaps.Phylogenetic relationships were inferred by conducting analyses using both ML and BI methods.The best-scoring RAxML tree was selected to represent the relationships amongst taxa, with a final likelihood value of -29592.797597(Fig. 1).The matrix contained 1905 distinct alignment patterns, with a 22.83% proportion of gaps and completely undetermined characters.The estimated base frequencies of A = 0.243583, C = 0.258293, G = 0.271748, T = 0.226375; substitution rates AC = 1.393909,AG = 2.806593, AT = 1.064133,CG = 1.193703,CT = 6.412290,GT = 1.000000; gamma distribution shape parameter α = 0.589535; Tree-Length = 1.823129.For BI analysis, GTR + I + G was selected as the best-fit model by AIC in MrModelTest for each gene (ITS, LSU, rpb2, SSU and tef1-α).Six simultaneous Markov chains were set to run for 1,000,000 generations, but stopped at 25,000 generations because the convergence diagnostic hit the stop value, resulting in 251 total trees.The first 10% of trees were discarded as the burn-in phase of the analyses and the remaining trees were used for calculating posterior probabilities in the majority rule consensus tree, of which the final average standard deviation of split frequencies at the end of total MCMC generations was 0.005298.Multigene phylogenetic analyses demonstrated that all genera of Bambusicolaceae formed well-resolved clades (up to 98% ML, 1.00 PP; Fig. 1) in the pres-

Analysis 2
The Trichobotrys tree was constructed using sequence data from ITS, LSU, SSU and tef1-α.A total of 61 strains of taxa in Dictyosporiaceae and closely-related families (viz.Didymosphaeriaceae, Lentitheciaceae, Morosphaeriaceae, Sulcatisporaceae and Trematosphaeriaceae) were included, with Phaeosphaeria oryzae (CBS 110110) and Phaeosphaeriopsis glaucopunctata (MFLUCC 13-0265) (Phaeosphaeriaceae) as the outgroup.Primarily, phylogenetic analysis of the concatenated LSU, SSU and ITS sequence dataset was conducted, based on ML and compared with phylogenetic analysis of the concatenated ITS, LSU, SSU and tef1-α sequence dataset.Phylogenetic analysis, based on the concatenated LSU, SSU and ITS sequence dataset, showed a similar topology with the concatenated ITS, LSU, SSU and tef1-α sequence dataset and were not significantly different (data not shown).Hence, multigene phylogenetic analysis of the concatenated ITS, LSU, SSU and tef1-α gene regions was selected to represent the phylogenetic relationships of Trichobotrys sinensis sp.nov.with other closely-related species in Dictyosporiaceae.The aligned dataset contained 3729 characters, including gaps.Phylogenetic relationships were inferred by conducting analyses using both ML and BI methods.The best-scoring RAxML tree was selected to represent the relationships amongst taxa, with a final likelihood value of -28366.415110(Fig. 2).The matrix contained 1566 distinct alignment patterns, with a 39.19% proportion of gaps and completely undetermined characters.The estimated base frequencies of A = 0.239629, C = 0.244575, G = 0.269426, T = 0.246371; substitution rates AC = 1.123110,AG = 2.634717, AT = 1.787337,CG = 0.836519, CT = 6.160493,GT = 1.000000; gamma distribution shape parameter α = 0.461486; Tree-Length = 3.107341.For BI analysis, GTR + I + G was selected as the bestfit model by AIC in MrModelTest for each gene (ITS, LSU, SSU and tef1-α).Six simultaneous Markov chains were run for 4,085,000 generations, resulting in 40,851 total trees.The first 10% of trees were discarded as the burn-in phase of the analyses and the remaining trees were used for calculating posterior probabilities in the majority rule consensus tree, of which the final average standard deviation of split frequencies at the end of total MCMC generations was 0.009998.
Multigene phylogenetic analyses of the concatenated ITS, LSU, SSU and tef1-α demonstrated that all representative families formed well-resolved clades in the present study.Our new isolate grouped with two unnamed Gregarithecium sp.(strains GMB1217 and MFLUCC 13-0853), with high support in ML and BI analyses (99% ML, 100 PP; Fig. 2) and clustered with Trichobotrys effusus (strains 1179, HNNUZCJ-94, FS524, SYSU-MS4729 and DFFSCS021) with high support (100% ML, 1.00 PP; Fig. 2) in Dictyosporiaceae.Gregarithecium sp.(strains GMB1217 and MFLUCC 13-0853) is unpublished and showed to be conspecific with our new isolate.Therefore, our new isolate is introduced as Trichobotrys sinensis, based on phylogenetic evidence coupled with morphological characteristics.Trichobotrys formed a highly-supported subclade with Gregarithecium (99% ML, 1.00 PP; Fig. 2) in the present study.However, these two genera are represented by different morphs.Therefore, the congeneric status of these two genera is doubtful in the study pending future study.

Analysis 3
The Periconia species tree was constructed using sequence data from ITS, LSU, SSU and tef1-α of all taxa in Periconiaceae and other related families (viz.Lentitheciaceae, and Massarinaceae).A total of 71 strains were included, with Morosphaeria ramunculicola (KH 220) and M. velatispora (KH 221) as the outgroup.The aligned dataset contained 3646 characters, including gaps.The best-scoring RAxML tree was selected to represent the relationships amongst taxa, with a final likelihood value of -19141.848334(Fig. 3).The matrix contained 1265 distinct alignment patterns, with a 32.87% proportion of gaps and completely undetermined characters.The estimated base frequencies of A = 0.239678, C = 0.253426, G = 0.268914, T = 0.237981; substitution rates AC = 1.751555,AG = 3.051838, AT = 1.900841,CG = 1.359429,CT = 9.411951, GT = 1.000000; gamma distribution shape parameter α = 0.505775; Tree-Length = 1.483987.For BI analysis, GTR + I + G was selected as the best-fit model by AIC in MrModelTest for each gene (ITS, LSU, SSU and tef1-α).Six simultaneous Markov chains were run for 555,000 generations, resulting in 5551 total trees.The first 10% of trees were discarded as the burn-in phase of the analyses and the remaining trees were used for calculating posterior probabilities in the majority rule consensus tree, of which the final average standard deviation of split frequencies at the end of total MCMC generations was 0.009941.
Specimen examined.China.Yunnan Province: Honghe Hani and Yi Autonomous Prefecture, Honghe County, rice terraces, on dead culm of bamboo, 26 Jan 2021, R. Phookamsak BN06 (KUN-HKAS 129042, holotype).Notes: As the axenic culture is not active, the sequences of SSU and rpb2 were obtained from genomic DNA extracted from ascomata and dried culture.
Culture characteristics.Colonies on PDA reaching 23-25 mm diam.after two weeks at room temperature (20-30 °C).Colony dense, circular, flattened, slightly raised, surface smooth, edge fimbriate, velvety, with fairly fluffy at the margin; colony from above, white to white-grey, separated from the centre by greenish-grey radiating near the margin; colony from below, pale yellowish to cream at the margin, deep green near the margin, with dark green concentric ring, separating the margin from greenish-grey to dark green centre; slightly produced light yellowish pigment tinted agar.
Distinguishing Periconia kunmingensis from other Periconia species, based on morphological features alone, presents challenges.However, differentiation can be achieved by considering variations in the sizes of conidiophores, conidiogenous cells and conidia, as well as the number of conidiophores originating from the stromatic, swollen part of the conidiophores, septation characteristics and the occurrence and origin of the host.A comprehensive morphological comparison is provided in Table 4.

Discussion
This paper, in the series "Exploring ascomycete diversity in Yunnan", presents three novel taxa in the suborder Massarineae (Pleosporales), viz.Bambusicola hongheensis (Bambusicolaceae), Periconia kunmingensis (Periconiaceae) and Trichobotrys sinensis (Dictyosporiaceae).The novelties of these taxa were well-justified, based on morphological characteristics and phylogenetic evidence, as well as the differences in nucleotide pairwise comparison of reliable genes amongst closely-related taxa.This provides a better fundamental knowledge of the taxonomic framework of ascomycetes in this region.Bambusicola hongheensis is justified, based on multigene phylogeny and the differences in nucleotide pairwise comparison of the ITS region with closely-related species.Monkai et al. ( 2021) mentioned that many Bambusicola species have similar morphology, but these species can be distinguished, based on multigene phylogeny and they also recommended the use of the rpb2 gene for delineating species level of Bambusicola.Unfortunately, the rpb2 sequence did not distinguish B. hongheensis from B. loculata in the present study; however, the ITS region of B. hongheensis showed > 1.5% nucleotide differences amongst the closely-related species viz.B. loculata, B. massarinia and B. triseptatispora.This provides adequate justification for the species' novelty following the recommendation of Jeewon and Hyde (2016).
Although many Bambusicola species are morphologically somewhat similar, it is notable that they can also be distinguished by their represented asexual morphs that are easily sporulated in vitro as well as on natural substrates.For instance, coelomycetous asexual morphs of B. massarinia and B. triseptatispora sporulated in vitro; of which B. massarinia can be distinguished from B. triseptatispora in having pale brown, 1-septate, cylindrical conidia (Dai et al. 2012).Whereas conidia of B. triseptatispora are light brown, 3-septate, cylindrical to cylindrical-clavate (Dai et al. 2017).Unfortunately, the asexual morphs of B. hongheensis and B. loculata have not yet been determined.Hence, further studies on their asexual morphs sporulated in vitro should be carried out for a better understanding through their sexual-asexual reproduction, as well as gaining criteria of species delineation.
Trichobotrys sinensis is morphologically typical of Trichobotrys.Trichobotrys was previously classified into Ascomycota genus incertae sedis (Wijayawardene et al. 2022b).Although the sequence data of the type species of Trichobotrys is currently unavailable, the inclusion of available sequence data along with our new species that morphologically align well with Trichobotrys in the phylogenetic analyses, provides compelling evidence supporting the placement of Trichobotrys within the Dictyosporiaceae.This information contributes to our understanding of taxonomic relationships and highlights the need for further studies to explore the molecular characteristics and genetic diversity of Trichobotrys species within the Dictyosporiaceae.
Synanamorph is the term of use for fungal taxa producing two or more different asexual morphs which were often linked by the sporulation in culture (Wijayawardene et al. 2021a(Wijayawardene et al. , 2022c)).Many fungal taxa have been reported for their synanamorphism, such as Botryosphaeria with dichomera-like in vitro and Neofusicoccum (as Fusicoccum) (Barber et al. 2005), Barbatosphaeria fagi (≡ Calosphaeria fagi) with ramichloridium-like and sporothrix-like asexual morphs (Réblová et al. 2015) and Synnemasporella with sporodochial and pycnidial asexual morphs on natural hosts (Fan et al. 2018).The formation of two or more different morphs in a single species has led to misidentification and the distinct morphs have been somehow counted as different species (Wijayawardene et al. 2021a(Wijayawardene et al. , 2022c)).It has further caused problems in the dual nomenclature of pleomorphic fungi that proposed one name for one fungus (McNeill and Turland 2012;Rossman et al. 2015).Interestingly, Trichobotrys sinensis formed two different asexual morphs, one in nature (as Trichobotrys) and another in vitro (pycnidial coelomycetous asexual morph) which is the first report of the synanamorphism for the genus Trichobotrys.This new finding provides insight into pleomorphism which is essential in further revision of taxonomic boundaries and easing of existing complications.It is noteworthy that Trichobotrys formed a well-resolved clade with Gregarithecium in the present phylogenetic analyses.Unfortunately, the sexual morph of Trichobotrys has not yet been determined.Similarly, the asexual morph of Gregarithecium has also not yet been reported.Hence, the sexual-asexual connection between Gregarithecium and Trichobotrys is doubtful pending future study.
Periconia kunmingensis is introduced in this paper, based on its morphology and phylogeny.Morphologically, P. kunmingensis fits well with the generic concept of Periconia and its phylogenetic affinity is also well-clarified within Periconiaceae.It is noteworthy that the ITS region could not be used to separate P. kunmingensis from other closely-related species, including P. cookei and P. verrucosa, based on the nucleotide pairwise comparison.Whereas, the ITS sequences of P. delonicis, P. elaeidis and P. palmicola are unavailable.The interspecific variation amongst these species may be questionable.However, the rpb2, and tef1-α gene regions which have sufficient genetic variation can be used to distinguish these species.Nevertheless, the rpb2 gene of most Periconia species is unavailable.Therefore, the sequences of protein-coding genes (e.g.rpb2 and tef1-α) are acquired to offer reliable phylogenetic markers for species delineation.
Over the past five years, the number of newly-described fungal species has been rapidly increasing in Yunnan.Several novel and interesting ascomycetes were described and illustrated from various host plants and on different substrates and habitats.Many studies of ascomycetous taxonomy on specific host substrates have become essential and challenging for mycologists across the region.For instance, D.N. Wanasinghe and his colleagues (2018-2022) carried out research studies on fungal biogeography and published over 40 novel taxa of wood-inhabiting fungi, as well as other substrates in this region (Bao et al. 2019;Wanasinghe et al. 2020Wanasinghe et al. , 2021;;Yasanthika et al. 2020;Mortimer et al. 2021;Ren et al. 2021;Wijayawardene et al. 2022a;Maharachchikumbura et al. 2022).Simultaneously, S. Tibpromma andher colleagues (2018-2022) have also carried out research studies of fungal taxonomy and diversity on various host plants, such as agarwood, coffee, Pandanus, para rubber and tea plants.They introduced 20 novel taxa from Pandanus (Tibpromma et al. 2018), while taxonomic studies on the other plants (approximately 45 novel species on agarwood, coffee and para rubber) are pending (S.Tibpromma, personal data information).A comprehensive study of freshwater Sordariomycetes in Yunnan has been carried out by Luo et al. (2018aLuo et al. ( , b, 2019) ) who introduced more than 50 novel taxa and reported more than 75 Sordariomycetes species in Yunnan.Even though these studies unravelled a substantial number of ascomycetes in Yunnan, there is still a huge gap of knowledge in hitherto undescribed novel taxa in this region.If considering only the plant and fungal ratio, many of the so far fungal taxonomic studies on land plants have underestimated these in Yunnan, especially on those economic and horticulture plants.Hence, the inventory of ascomycetes on these land plants will be interesting in further research studies.

Conclusion
In conclusion, this study introduces three novel species in the suborder Massarineae (Pleosporales): Bambusicola hongheensis, Periconia kunmingensis and Trichobotrys sinensis.These species were found as saprobes in different habitats, with B. hongheensis and P. kunmingensis occurring in terrestrial environments, while T. sinensis was discovered in a freshwater stream.Notably, the presence of Trichobotrys in a freshwater habitat is a significant finding, as it aligns with other aquatic lignicolous species within the family Dictyosporiaceae.The novelty of B. hongheensis is supported by multigene phylogeny and nucleotide pairwise comparison, although further genetic analysis is recommended.Differentiation between Bambusicola species can also be achieved through the examination of their asexual morphs.Trichobotrys sinensis, morphologically typical of Trichobotrys, is phylogenetically placed within Dictyosporiaceae and highlights the need for additional studies on molecular characteristics and genetic diversity within the genus.The observation of synanamorphism in T. sinensis adds complexity to its morphological identification and taxonomic boundaries.The introduction of Periconia kunmingensis is supported by its morphology and phylogenetic affinity within the family Periconiaceae, although the use of protein-coding genes is recommended for reliable species delineation.This study contributes to our understanding of ascomycete diversity in Yunnan and emphasises the importance of such investigations to enhance our knowledge of newly-discovered taxa.

Figure 1 .
Figure 1.Phylogram of the best-scoring ML consensus tree of taxa in Bambusicolaceae and Occultibambusaceae.The new isolate is indicated in blue.Isolates from type materials are in bold.The ML ultrafast bootstrap and Bayesian PP values greater than 60% and 0.90 are shown at the nodes.

Figure 2 .
Figure 2. Phylogram of the best-scoring ML consensus tree of Trichobotrys species in Dictyosporiaceae and closely-related families viz.Didymosphaeriaceae, Lentitheciaceae, Morosphaeriaceae, Sulcatisporaceae and Trematosphaeriaceae.The new isolate is indicated in blue.Isolates from type materials are in bold.The ML ultrafast bootstrap and Bayesian PP values greater than 70% and 0.95 are shown at the nodes.
Bambusicolaceae was first introduced by Hyde et al. (2013) to accommodate Bambusicola with B. massarinia being the type species.Subsequently, another three genera were accommodated in this family viz.Corylicola (Wijesinghe et al. 2020), Leucaenicola (Jayasiri et al. 2019) and Palmiascoma

Figure 3 .
Figure 3. Phylogram of the best-scoring ML consensus tree of taxa in Periconiaceae and the closely-related families Lentitheciaceae and Massarinaceae.The new isolate is indicated in blue.Isolates from type materials are in bold.The ML ultrafast bootstrap and Bayesian PP values greater than 50% and 0.95 are shown at the nodes.

Table 1 .
Species details and GenBank accession numbers used in phylogenetic analysis of Bambusicola species (Bambusicolaceae, Pleosporales).The new sequences are indicated in bold and the ex-type strains are indicated by superscript "T".Missing sequences are indicated by "-".

Table 2 .
Species details and GenBank accession numbers used in phylogenetic analysis of taxa in Dictyosporiaceae (Pleosporales).The new sequences are indicated in bold and the ex-type strains are indicated by superscript "T".Missing sequences are indicated by "-".

Table 3 .
Species details and GenBank accession numbers used in phylogenetic analysis of Periconia species (Periconiaceae, Pleosporales).The new sequences are indicated in bold and the ex-type strains are indicated by superscript "T".Missing sequences are indicated by "-".

Table 4 .
Morphological comparison of Periconia kunmingensis with other related species.A novel species is indicated by black bold.