Morpho-Phylo Taxonomy of Novel Dothideomycetous Fungi Associated With Dead Woody Twigs in Yunnan Province, China

Within the field of mycology, macrofungi have been relatively well-studied when compared to microfungi. However, the diversity and distribution of microfungi inhabiting woody material have not received the same degree of research attention, especially in relatively unexplored regions, such as Yunnan Province, China. To help address this knowledge gap, we collected and examined fungal specimens from different plants at various locations across Yunnan Province. Our investigation led to the discovery of four species that are clearly distinct from extant ones. These taxonomic novelties were recognized based on morphological comparisons coupled with phylogenetic analyses of multiple gene sequences (non-translated loci and protein-coding regions). The monotypic genus Neoheleiosa gen. nov. (type: N. lincangensis) is introduced in Monoblastiaceae (Monoblastiales) for a woody-based saprobic ascomycete that possesses globose to subglobose or obpyriform ascomata with centric or eccentric, papillate ostioles, an ascomatal wall with thin-walled cells of textura globulosa, cylindric, pedicellate asci with an ocular chamber, and 1-septate, brown, guttulate, longitudinally striated, bicellular ascospores. Neoheleiosa has a close phylogenetic affinity to Heleiosa, nevertheless, it is morphologically dissimilar by its peridium cells and ornamented ascospores. Acrocalymma hongheense and A. yuxiense are described and illustrated as new species in Acrocalymmaceae. Acrocalymma hongheense is introduced with sexual and asexual (coelomycetous) features. The sexual morph is characterized by globose to subglobose, ostiolate ascomata, a peridium with textura angularis cells, cylindric-clavate asci with a furcate to truncate pedicel and an ocular chamber, hyaline, fusiform, 1-septate ascospores which are surrounded by a thick, distinct sheath, and the asexual morph is featured by pycnidial conidiomata, subcylindrical, hyaline, smooth, annelledic, conidiogenous cells, hyaline, guttulate, subcylindrical, aseptate conidia with mucoid ooze at the apex and with a rounded hilum at the base. Acrocalymma yuxiense is phylogenetically distinct from other extant species of Acrocalymma and differs from other taxa in Acrocalymma in having conidia with three vertical eusepta. Magnibotryascoma kunmingense sp. nov. is accommodated in Teichosporaceae based on its coelomycetous asexual morph which is characterized by pycnidial, globose to subglobose, papillate conidiomata, enteroblastic, annelledic, discrete, cylindrical to oblong, hyaline conidiogenous cells arising from the inner layer of pycnidium wall, subglobose, oval, guttulate, pale brown and unicelled conidia.


INTRODUCTION
Dothideomycetes is the largest and most ecologically diverse class of Ascomycota (Hongsanan et al., 2020a), consisting of 28,729 species (Kirk, 2019). This class comprises saprobes, human and plant pathogens, endophytes, epiphytes, lichens, lichenicolous, nematode-trapping and rock-inhabiting members Zhang J.F. et al., 2019;Hongsanan et al., 2020a). Hyde et al. (2013) provided the first comprehensive monograph of the families in Dothideomycetes. Since then, the taxonomies of Dothideomycetes have been updated with new taxa in several journal series, e.g., Fungal Diversity notes, Fungal planet description sheets, Mycosphere notes, Fungal Biodiversity Profiles, Fungal Systematics and Evolution, New and Interesting Fungi. Wijayawardene et al. (2014) provided an outline for the proposals of protection or suppression of generic names of Dothideomycetes. Consistent with the "one fungus-one name" concept, Rossman et al. (2015) provided recommendations for the nomenclature of pleomorphic genera in the class. Attributable to the continual changes of the taxa in this class, the taxonomy of Dothideomycetes is in a perpetual transitional state (Pem et al., 2019a), and as such, the outline of this class has been frequently revised (Wijayawardene et al., 2017. Recent publications by Hongsanan et al. (2020a,b) expanded the taxonomic concepts of families in the Dothideomycetidae, Pleosporomycetidae, and orders and families incertae sedis in Dothideomycetes. Hongsanan et al. (2020a,b) have accepted 38 orders and 210 families in Dothideomycetes. In order to provide a suitable platform to bring these data together, the website https://www.dothideomycetes.org was established by Pem et al. (2019a). Liu et al. (2017) proposed divergence time estimates as additional evidence for rearranging the internal classification of this class and this has been helpful to establish new families and species (Zhang S.N. et al., 2019, Bhunjun et al., 2021. The most recent order-level multi-gene phylogeny for Dothideomycetes is provided by Maharachchikumbura et al. (2021), which also introduced another two orders viz. Homortomycetales and Holmiellales, bringing the total number of orders to 40 in the class.
Monoblastiaceae is the only family in Monoblastiales comprising both lichenized and non-lichenized ascomycetes. Wijayawardene et al. (2020) accepted six genera in this family. Initiated by Hyde et al. (2020); Hongsanan et al. (2020b) synonymized Eriomycetaceae under Monoblastiaceae. Accordingly, this family currently includes 11 genera (Hongsanan et al., 2020b). The majority of these fungi grow on bark in tropical forests, but the family is also commonly found in leaf-inhabiting lichen communities (Aptroot and Sipman, 1993;Lücking, 2008). These foliicolous lichens can be useful in monitoring the environmental health of tropical forest ecosystems (Hongsanan et al., 2020b). Alcorn and Irwin (1987) introduced Acrocalymma to accommodate A. medicaginis, which was previously identified as Stagonospora meliloti, known for causing root and foliar rot of Medicago sativa in Australia. In the phylogenetic analyses of Trakunyingcharoen et al. (2014), Acrocalymma species (A. aquatica, A. cycadis, A. fici, A. medicaginis, and A. vagum) represented an undefined lineage in Dothideomycetes for which the family name Acrocalymmaceae was introduced. They also showed that Massarina walkeri and A. medicaginis are congeneric and thus introduced a new combination, A. walkeri. Recently, Jayasiri et al. (2019) introduced another species, Acrocalymma pterocarpi, to this family. Dong et al. (2020) introduced the most recent species, Acrocalymma bipolare, a freshwater species recovered from submerged wood in the Nile River, Egypt.
Teichosporaceae was established by Barr (2002) based on morphological similarities of Bertiella, Byssothecium, Chaetomastia, Immotthia, Loculohypoxylon, Moristroma, Sinodidymella, and the type genus Teichospora. However, Moristroma, Byssothecium and Bertiella were excluded from the family by Lumbsch and Huhndorf (2010). Jaklitsch et al. (2016) revised Teichosporaceae and accepted only Teichospora in the family. The family Floricolaceae was also synonymized under Teichosporaceae by Jaklitsch et al. (2016), and every genus within this family became a synonym of Teichospora. however this current taxonomic rearrangement needs to be verified with wider sampling. In addition subsequent outlines did not follow the monotypic nature of Teichosporaceae (Wijayawardene et al., 2018). Currently, Teichosporaceae contains thirteen accepted genera (Hongsanan et al., 2020a).
The present research paper introduces two new species in Acrocalymmaceae, one new genus and a new species in Monoblastiaceae and one new species in Teichosporaceae from fifteen specimens collected from Honghe, Kunming, Lincang, Qujing, and Yuxi in Yunnan Province, China.

Isolates and Specimens
Fungal materials were collected from various deciduous trees and dried woody litter in Yunnan Province, China during the dry season. Collected samples were brought to the laboratory in Zip lock plastic bags. Samples were examined with an Olympus SZ61 Series microscope. Single ascospore isolation was carried out following the method described in Senanayake et al. (2020). Germinated spores were individually transferred to potato dextrose agar (PDA) plates and grown at 20 • C in daylight. The living cultures were deposited at the Kunming Institute of Botany Culture Collection (KUMCC), Kunming, China, and duplicated at the China General Microbiological Culture Collection Center (CGMCC). Dry herbarium materials have been stored in the herbarium of Cryptogams Kunming Institute of Botany, Academia Sinica (KUN-HKAS). MycoBank numbers have been registered as outlined in MycoBank 1 .

Morphological Observations
In hand sections of the ascomata/conidiomata, which were mounted in distilled water, the following characteristics were evaluated: ascomata/conidiomata diameter, height, color, and shape; width of peridium; and height and diameter of ostioles. Length and width (at the widest point) of asci, ascospores, conidiophores and conidia were also measured. Images were captured with a Canon EOS 600D digital camera fitted to a Nikon ECLIPSE Ni compound microscope. Measurements were made with the Tarosoft (R) Image Frame Work program, and images used for figures were processed with Adobe Photoshop CS5 Extended version 10.0 software (Adobe Systems, United States).

DNA Extraction, PCR Amplifications, and Sequencing
Genomic DNA was extracted from the axenic mycelium as described by Phookamsak et al. (2017). When the spores failed to germinate in culture, DNA was extracted directly from the fruiting bodies of the fungus as outlined by Wanasinghe et al. (2018b). DNA to be used as templates for Polymerase chain reaction (PCR) were stored at 4 • C for use in regular work and duplicated at -20 • C for long-term storage.
The primers and PCR protocols for each gene were conducted by following Thiyagaraja et al. (2020a) and Wanasinghe et al. (2020). PCR was carried out at a volume of 25 µl, which contained 12.5 µl of 2× Power Taq PCR MasterMix (Bioteke Co., China), 1 µl of each primer (10 µM), 1 µl genomic DNA and 9.5 µl deionized water. The amplified PCR fragments were sent to a commercial sequencing provider (BGI, Ltd., Shenzhen, China). The nucleotide sequence data acquired were deposited in GenBank (Table 1).

Sequencing and Sequence Alignment
Sequences generated from different primers were analyzed along with sequences retrieved from GenBank (Tables 1-3). Sequences with high similarity indices were determined from a BLAST search to find the closest matches with taxa in Pleosporales, and from recently published data (Thambugala et al., 2015;Jaklitsch et al., 2016;Jayasiri et al., 2019;Hongsanan et al., 2020b). The multiple alignments of all consensus sequences, as well as the reference sequences were automatically generated with MAFFT v. 7 (Kuraku et al., 2013;Katoh et al., 2019) 2 , and improved manually when necessary using BioEdit v. 7.0.9.0 (Hall, 1999).

Phylogenetic Analyses
The single-locus datasets were examined for topological incongruences among loci for members of the analyses. The alignments were concatenated into a multi-locus alignment that was subjected to maximum-likelihood (ML) and Bayesian (BI) phylogenetic analyses.
The CIPRES Science Gateway platform (Miller et al., 2010) was used to perform RAxML and Bayesian analyses. ML analyses were made with RAxML-HPC2 on XSEDE v. 8.2.10 (Stamatakis, 2014) using GTR + GAMMA swap model with 1,000 bootstrap repetitions.
Evolutionary models for Bayesian analysis were selected independently for each locus using MrModeltest v. 2.3 (Nylander et al., 2008) under the Akaike Information Criterion (AIC) implemented in both PAUP v. 4.0b10 and GTR + I + G was selected as the best fit model for all three analyses. MrBayes analyses were performed setting GTR + I + G, 5 M generations, sampling every 1,000 generations, ending the run automatically when standard deviation of split frequencies dropped below 0.01 with a burn-in fraction of 0.25. ML bootstrap values equal or greater than 60% and BYPP greater than 0.95 are given above each node of every trees.

Phylogenetic Analyses
Three analyses were performed in this study; the first is an updated phylogeny of the genera in Monoblastiaceae (Figure 1), whereas the remaining two datasets represent taxa in Acrocalymmaceae (Figure 2) and genera treated in Teichosporaceae (Figure 3), respectively.
Monoblastiaceae SSU, LSU, ITS, tef 1, and mtSSU phylogeny (Figure 1): The alignment contained 25 isolates and the tree was rooted to Elsinoe centrolobii (CBS 222.50) and E. phaseoli (CBS 165.31). The final alignment contained a total of 4,062 characters used for the phylogenetic analyses, including alignment gaps. The RAxML analysis of the combined dataset yielded a best scoring tree with a final ML optimization likelihood value of −14276.596302. The matrix had 1,076 distinct alignment patterns, with 49.73% of undetermined characters or gaps. Parameters for the GTR + I + G model of the combined amplicons were as follows: Estimated base frequencies; A = 0.249541, C = 0.238483, G = 0.280978, The newly generated sequences are indicated in bold. NA: Sequence data not available in GenBank. The newly generated sequences are indicated in bold. NA: Sequence data not available in GenBank.
Frontiers in Microbiology | www.frontiersin.org  The newly generated sequences are indicated in bold. NA: Sequence data not available in GenBank.
T = 0.230998; substitution rates AC = 0.975103, AG = 2.399794, AT = 1.850660, CG = 1.423793, CT = 6.985604, GT = 1.000; proportion of invariable sites I = 0.411774; gamma distribution shape parameter α = 0.563137. The Bayesian analysis ran 125,000 generations before the average standard deviation for split frequencies reached below 0.01 (0.008726). The analysis generated 1,251 trees (saved every 100 generations) from which 939 were sampled after 25% of the trees were discarded as burnin. The alignment contained a total of 1,077 unique site patterns. Acrocalymmaceae SSU, LSU, and ITS phylogeny (Figure 2): The alignment contained 22 isolates and the tree was rooted to Boeremia exigua (CBS 431.74). The final alignment contained a total of 2,317 characters used for the phylogenetic analyses, including alignment gaps. The RAxML analysis of the combined dataset yielded a best scoring tree with a final ML optimization likelihood value of −5083.5038. The matrix had 265 distinct alignment patterns, with 31.12% of undetermined characters or gaps. Parameters for the GTR + I + G model of the combined amplicons were as follows: Estimated base frequencies; A = 0.250179, C = 0.212903, G = 0.27372, T = 0.263198; substitution rates AC = 2.467522, AG = 3.254238, AT = 3.774672, CG = 0.816185, CT = 11.170637, GT = 1.000; proportion of invariable sites I = 0.737361; gamma distribution shape parameter α = 0.95506. The Bayesian analysis ran 420,000 generations before the average standard deviation for split frequencies reached below 0.01 (0.009821). The analysis generated 4,201 trees (saved every 100 generations) from which 3,151 were sampled after 25% of the trees were discarded as burnin. The alignment contained a total of 266 unique site patterns.
Teichosporaceae SSU, LSU, ITS, tef 1, and rpb2 phylogeny (Figure 3) proportion of invariable sites I = 0.495477; gamma distribution shape parameter α = 0.4815. The Bayesian analysis ran 390,000 generations before the average standard deviation for split frequencies reached below 0.01 (0.009894). The analysis generated 3,901 trees (saved every 100 generations) from which 2,926 were sampled after 25% of the trees were discarded as burnin. The alignment contained a total of 1,162 unique site patterns.
The phylogenetic results obtained for each dataset are discussed where applicable in the descriptive notes below.

Notes
During our investigation on the diversity of woody-based Dothideomycetes in Yunnan Province, three isolates (HKAS 111907, HKAS 111908, HKAS 111909) were recovered from decaying woody litter in Honghe and Yuxi counties. Two of them had asexual morphs while the third specimen had a sexual morph. Conidiomata, conidiophores and conidia of these asexual fungi morphologically resemble the remaining taxa in Acrocalymma (Alcorn and Irwin, 1987;Zhang et al., 2012;Crous et al., 2014;  Trakunyingcharoen et al., 2014). The sexual morph was similar to Acrocalymma pterocarpi, which was the only known sexual morph in the genus, by its ascomata and asci features (Jayasiri et al., 2019). Phylogenetically, HKAS 111907, HKAS 111908, and HKAS 111909 are monophyletic with strong bootstrap supports (100% ML/0.99 BYPP, Figure 2). This clade has a sister relationship to Acrocalymma cycadis (CBS 137972 and Ct-LP55), but was not statistically supported. Within our new isolates, there were no nucleotide differences in ITS, LSU and SSU gene regions. Therefore, we recognize these three isolates belong to one species (Jeewon and Hyde, 2016), which we introduce as a new species herein.

Notes
Acrocalymma yuxiense, collected from dried leaves of Quercus glauca in Yunnan, is in an independent lineage with weak support and is phylogenetically distinct from other extant species of Acrocalymma (Figure 2). This new species differs from other taxa in Acrocalymma in having conidia with 3 vertical eusepta while other species produce aseptate conidia.

Culture Characteristics
Colonies grew on PDA at 20 • C in the dark reached 2 cm diam., within 14 days, dense, circular, slightly raised, surface smooth, entire margin, white in surface view and off-white to gray in reverse.

DISCUSSION
The mountainous region of Yunnan Province, China is an important global biodiversity hotspot for studying the evolution of plants, animals, and fungi (Feng and Yang, 2018). The mountains of Southwest China; Eastern Himalaya-Nepal-India and Indo-Burma-India-Myanmar are included in the world's 35 biodiversity hotspots, and these three hotspot regions intersect in Yunnan Province (Myers et al., 2000;Mittermeier et al., 2005). As a result of the diverse landscape and climatic conditions within Yunnan Province, fungi in Yunnan Province have higher rates of endemism; however, they also share evolutionary connections with species from other regions of the world (Feng and Yang, 2018). Among them, wood-decaying Basidiomycota in tropical China are well studied, and many new species have been documented (Dai et al., 2003(Dai et al., , 2004(Dai et al., , 2011Cui et al., 2009;Wang et al., 2011;Yuan and Dai, 2012). This has facilitated better understanding of species diversity and the systematics of woody-based basidomycetous groups, such as Polyporales. However, woody-based microfungi such as Dothideomycetes are relatively neglected compared to the level of research conducted on Basidiomycetes . In the last few years, there has been increasing attention on woody-based microfungal occurrences, and more studies are reporting on the microfungal diversity, especially in Dothideomycetes, of Yunnan Province Huang et al., 2018;Luo et al., 2018;Wanasinghe et al., 2018aWanasinghe et al., , 2020Wanasinghe et al., , 2021Rathnayaka et al., 2019;Qiao et al., 2020;Thiyagaraja et al., 2020b;Yasanthika et al., 2020).
Neoheleiosa has a close phylogenetic affinity to Heleiosa with 100% ML and 1.00 BI support values (Figure 1). Kohlmeyer et al. (1996) introduced Heleiosa as a monotypic genus to accommodate Heleiosa barbatula, which is characterized by cylindrical asci with a short pedicel, 1-septate, ellipsoid ascospores with appendages. This fungus was found on senescent leaves of Juncus in salt marshes on the Atlantic coast of the United States (North Carolina, Virginia). Ascomata, asci and ascospore shapes of both genera were shown to be similar. However, the ascospores of Neoheleiosa do not have any appendages (Figures 4m-q), whereas Heleiosa have 10 or more short and curved hair-like subapical appendages at each end. Ascospores with a gelatinous sheath or appendages may help fungi to attach to plant substrates in aquatic or marine habitats (Shearer, 1993;Hyde and Goh, 2003;Jones, 2006;Devadatha et al., 2018;Hashimoto et al., 2018). The subapical appendages of Heleiosa ascospores could be an adaptation for its marine-based habitat and loss of appendaged ascospores in Neoheleiosa is potentially advantageous to adaptation to a nonaquatic habitat.
The surface ornamentation of spores, such as the presence of or absence of appendages, is often used in ascomycetous taxonomy to delineate species or genera (Jeewon et al., 2003;Liu et al., 2014;Phookamsak et al., 2014;Wijayawardene et al., 2016;Paz et al., 2017;Réblová et al., 2015Réblová et al., , 2018Pem et al., 2019b). Abbott and Currah (1997) and Villegas et al. (2005) reported spore ornamentation as a useful character in differentiating various genera within Helvellaceae and Gomphales. Jeewon et al. (2002) clearly demonstrated that species bearing appendages can form distinct phylogenetic lineages and discussed how this character can be a significant phylogenetic marker at the intergeneric level. Considering the similarities shown between our newly proposed genus and Heleiosa, we based our placement of these specimens in a new genus due to the lack of any appendages on the ascospores, a significant variation in morphology compared to the spores of Heleiosa. The ascospores of Neoheleiosa are ornamented with longitudinal streaks (Figure 4q), whereas Heleiosa have smooth-walled ascospores. Furthermore, the arrangement of peridium cells has been reported to be useful to demarcate different genera (Hyde et al., 2013;Tian et al., 2015;Ekanayaka et al., 2017;Paz et al., 2017;Senanayake et al., 2018). In our study, we observed that the peridium cells of Heleiosa are textura angularis, while Neoheleiosa have cells of textura globulosa. It could be possible that with wider collections in the future, species from these two genera will constitute distinct lineages that will further substantiate their generic placement, a scenario which has been reported in other studies Wanasinghe et al., 2018b;Jaklitsch and Voglmayr, 2020).
Acrocalymmaceae includes a single genus with eight species viz. Acrocalymma aquaticum, A. bipolare, A. cycadis, A. fici, A. medicaginis, A. pterocarpi, A. vagum and A. walker. These are reported from terrestrial habitats, excluding Acrocalymma aquatica and A. bipolare (Zhang et al., 2012;Dong et al., 2020). The majority of these species are saprobic on various host substrates (Hongsanan et al., 2020a,b;Tennakoon et al., 2021). Only Acrocalymma medicaginis and A. vagum are reported as pathogens Trakunyingcharoen et al., 2014). Moreover, Acrocalymma medicaginis is known as the causal agent of root and crown rot of Medicago sativa (Alcorn and Irwin, 1987). Even though there are only eight described species, more than 200 ITS sequences have been deposited from these species in GenBank. Endophytic strains of Acrocalymma vagum have the highest number of reported sequences, followed by A. medicaginis. Thambugala et al. (2015) introduced Magnabotrioscoma to accommodate M. uniseriata (≡Misturatosphaeria uniseriate), which is morphologically distinct from the remaining genera in Floricolaceae. Magnabotrioscoma species have been reported as woody-based saprobes on Clematis vitalba, Malus halliana, Ribes sanguineum, Robinia pseudoacacia, Salix sp., and Vaccinium myrtillus from Belgium, China, Germany, Norway and the United Kingdom (Jaklitsch et al., 2016;Hyde et al., 2017;Phukhamsakda et al., 2020). In this study, we classified another species in this genus, Magnibotryascoma kunmingensis, growing on the decaying woody litter of Acer cappadocicum var. sinicum and Machilus yunnanensis. The sexual morph of the genus is characterized by erumpent to superficial ascomata lacking a subiculum and fusiform to elliptical and guttulate ascospores (Mugambi and Huhndorf, 2009;Thambugala et al., 2015;Phukhamsakda et al., 2020), and the asexual morph has pycnidial conidiomata featuring aseptate and brown conidia Jaklitsch et al., 2016). One interesting finding in this genus is the close connection between uncultured fungal strains (i.e., JF945459, KC978028, KX515906, LS994072) and endophytic strains (i.e., EF419935, EF419972, EF419996) with our new strains in BLAST search results. These strains are not linked to morphological details, and it is therefore difficult to provide further insights into their morpho-phylo relationships. This study reveals that there are potentially many new species awaiting discovery in this region .

DATA AVAILABILITY STATEMENT
The datasets generated for this study can be found in the NCBI GenBank, MycoBank and TreeBASE.

AUTHOR CONTRIBUTIONS
PM, DW, and RJ designed the study. DW did the sample collection. PM and DW were involved in the phylogenetic analyses. SL and J-CX contributed for the research funds. All authors contributed to the writing, preparation, and submission of the manuscript.