Integrating Different Lines of Evidence to Establish a Novel Ascomycete Genus and Family (Anastomitrabeculia, Anastomitrabeculiaceae) in Pleosporales

A novel genus, Anastomitrabeculia, is introduced herein for a distinct species, Anastomitrabeculia didymospora, collected as a saprobe on dead bamboo culms from a freshwater stream in Thailand. Anastomitrabeculia is distinct in its trabeculate pseudoparaphyses and ascospores with longitudinally striate wall ornamentation. A new family, Anastomitrabeculiaceae, is introduced to accommodate Anastomitrabeculia. Anastomitrabeculiaceae forms an independent lineage basal to Halojulellaceae in Pleosporales and it is closely related to Neohendersoniaceae based on phylogenetic analyses of a combined LSU, SSU and TEF1α dataset. In addition, divergence time estimates provide further support for the establishment of Anastomitrabeculiaceae. The family diverged around 84 million years ago (MYA) during the Cretaceous period, which supports the establishment of the new family. The crown and stem age of Anastomitrabeculiaceae was also compared to morphologically similar pleosporalean families.

Several pleosporalean taxa are pathogens associated with a broad range of hosts including bamboo. Bamboo (Poaceae) comprises over 115 genera with around 1500 species [16][17][18], can be found in diverse climates [17], and are widely distributed in various forest types in Thailand [18,19]. It has been estimated that around 1100 fungal species belonging to over 200 genera have been described or recorded worldwide on bamboo and most of these bamboo-associated fungi are ascomycetes [20,21].

Sample Collection, Isolation and Identification
Dead bamboo culms were collected from a freshwater stream from Krabi province, Thailand, in 2015. The samples were incubated in plastic boxes with sterile and moist tissue at 25-30 • C for 3 days. Pure fungal colonies were obtained using single-spore isolation [28]. Germinating spores were transferred aseptically to potato dextrose agar (PDA) and malt extract agar (MEA) (Difco™). The cultures were incubated at 25 • C with frequent observations. Fungal characters were observed using a stereo microscope (Zeiss SteREO Discovery v. 8) fitted with an Axio Cam ERc5S and a Leica DM2500 compound microscope attached with a Leica MC190 HD camera. All microscopic measurements were carried out using Tarosoft (R) Image Frame Work program and the images were processed with Adobe Photoshop CS6 version 13.0 software (Adobe Systems, San Jose, CA, USA). The type specimens were deposited in the Mae Fah Luang University (MFLU) Herbarium, Chiang Rai, Thailand, and pure cultures were deposited at the Mae Fah Luang University Culture Collection (MFLUCC). The new taxon was linked with Facesoffungi numbers (FoF) [29] and Index Fungorum (Index Fungorum 2020, http://www.indexfungorum.org/, accessed on 2 December 2020) and established based on guidelines recommended by Jeewon and Hyde [30].

DNA Extraction, PCR Amplification and DNA Sequencing
DNA extraction, PCR amplification, DNA sequencing and phylogenetic analysis were carried out as detailed in Dissanayake et al. [31]. Total genomic DNA was extracted from fresh mycelium with a Biospin Fungus Genomic DNA Extraction Kit (BioFlux ® ) (Hangzhou, P.R. China) following the manufacturer's protocol. The nuclear ribosomal large subunit 28S rRNA gene (LSU) [32], the nuclear ribosomal small subunit 18S rRNA gene (SSU) [33] and the translation elongation factor 1-alpha gene (TEF1α) [34] were amplified using primers (LSU: LROR/LR5, SSU: NS1/NS4 and TEF1α: 983F/2218R). Polymerase chain reaction (PCR) was performed using PCR mixtures containing 5-10 ng DNA, 1X PCR buffer, 0.8 units Taq polymerase, 0.3 µM of each primer, 0.2 mM dNTP and 1.5 mM MgCl 2 . PCR conditions were set at an initial denaturation for 3 min at 94 • C, followed by 40 cycles of 45 s of denaturation at 94 • C, annealing for 50 s at 56 • C for LSU, SSU and 52 • C for TEF1α and extension for 1 min at 72 • C, with a final extension of 10 min at 72 • C. All the PCR products were visualised on 1% Agarose gels with added 6 µL of 4S green dyes, per each 100 mL. Successful PCR products were purified and sequencing was performed by Shanghai Sangon Biological Engineering Technology & Services Co. (Shanghai, P.R. China). All sequences generated in this study were submitted to GenBank (Table 1) and the ITS region of Anastomitrabeculia didymospora was deposited with the accession number MW413900 (MFLUCC 16-0412) and MW413897 (MFLUCC 16-0417).

Phylogenetic Analysis
The sequence data were assembled using BioEdit v. 7.2.5 [35] and subjected to a BLAST search (https://blast.ncbi.nlm.nih.gov/Blast.cgi) to find the closest matches with taxa in Pleosporales. Reference sequence data of this order and some representatives of other orders of Dothideomycetes were downloaded from previously published studies [1,6,[36][37][38][39]. The sequences were automatically aligned using default settings in MAFFT v. 7 (http://mafft.cbrc.jp/alignment/server/) [40]. A combined dataset of three gene regions (LSU, SSU and TEF1α) was prepared and manually adjusted using BioEdit and AliView [41]. Phylogenetic analyses of the combined dataset were performed using maximum likelihood, maximum parsimony and Bayesian inference method. Maximum likelihood analyses (ML), including 1000 bootstrap pseudoreplicates, were performed at the CIPRES web portal [42] using RAxML v. 8.2.12 [43]. Maximum parsimony analysis was conducted using PAUP v.4.0b 10 [44] with the heuristic search option and number of replications 1000 each. The Tree Length (TL), Consistency Indices (CI), Retention Indices (RI), Rescaled Consistency Indices (RC) and Homoplasy Index (HI) were documented.
The best model for different genes partition was determined in JModelTest version 2.1.10 [45] for posterior probability (PP). The general time reversible (GTR) model with a discrete gamma distribution plus invariant site (GTR+I+G) substitution model was used for the combined dataset. Posterior probabilities [46] were estimated by Markov Chain Monte Carlo sampling (MCMC) in MrBayes v. 3.2.6 [47]. Four simultaneous Markov chains were run for 10 million generations and trees were sampled every 1000th generation, thus resulting in 10,000 trees. The suitable burn-in phase was determined by inspecting traces in Tracer version 1.7 [48]. The first 10% of generated trees representing the burn-in phase of the analyses were discarded, while the remaining trees were used to calculate posterior probabilities (PP) in the majority rule consensus tree. The phylograms were visualized with FigTree v1.4.0 program [49] and edited using Adobe Illustrator CS6 v15.0 (Adobe Systems, USA).

Fossil Calibration and Divergence Time Estimates
Divergence times were estimated with BEAST 2.6.2 [50] based on the methodology described in Phukhamsakda et al. [4]. The aligned sequence dataset (LSU, SSU and TEF1α) used for the phylogenetic analyses were loaded into BEAUTI 2.6.2 to prepare the XML file. Nucleotide substitution models were determined using JModelTest version 2.1.10. The GTR+I+G nucleotide substitution model was applied to LSU and TEF1α partitions. The symmetrical (SYM) model with a discrete gamma distribution plus invariant site (SYM+I+G) substitution model was applied to the SSU partition. The data partitions were set with unlinked substitution, linked clock model and linked tree. An uncorrelated relaxed clock model with lognormal distribution was used. The Yule speciation process, which assumes a constant rate of speciation divergence, was used as the tree prior [51]. The analysis was performed in BEAST 2.6.2 for 100 million generations, sampling every 1000 generations. The effective sample size (ESS) was analysed with Tracer version 1.7 to check that the values were greater than 200, as recommended by Drummond et al. [52]. The first 20% trees were discarded as the burn-in phase and the remaining trees were combined in LogCombiner 2.6.2. The maximum clade credibility was calculated in TreeAnnotator v 2.6.2. The phylograms were visualized with FigTree v.1.4.0 program.

Fossil Calibration and Divergence Time Estimates
The topology of the maximum clade credibility (MCC) tree ( Figure 2) was congruent with the tree obtained from the Bayesian inference analysis and the maximum likelihood analysis. The divergence times of the dating analysis are listed in Table 2. The crown age of Dothideomycetes is estimated at 263 MYA during the Permian period based on the MCC tree. The split of Arthoniomycetes and Dothideomycetes occurred around 323 MYA during the Carboniferous period. The crown age of Pleosporales is estimated at 206 MYA, and Hysteriales diverged from Pleosporales approximately 236 MYA during the Triassic period. The crown age of Anastomitrabeculiaceae is estimated at around 2.6 MYA, and it diverged from Halojulellaceae at around 84 (52-116) MYA. Anastomitrabeculiaceae formed an independent lineage with close relationship to Halojulellaceae with strong posterior probability in the MCC tree (0.99 BYPP). The divergence time of Anastomitrabeculiaceae was compared to Pleosporalean families with trabeculate pseudoparaphyses, cylindrical asci and ascospores with a sheath ( Table 3). The divergence time of Anastomitrabeculiaceae was also compared to Didymosphaeriaceae as they are morphologically similar by having trabeculate pseudoparaphyses and cylindrical asci.
Note: Anastomitrabeculiaceae is introduced to include Anastomitrabeculia, which is reported as a saprobe on bamboo culms. Anastomitrabeculiaceae is characterised by semiimmersed, coriaceous or carbonaceous ascomata with septate, trabeculate pseudoparaphyses and hyaline ascospores with longitudinally striate wall ornamentation, surrounded by mucilaginous sheath. Anastomitrabeculiaceae formed a well-supported independent lineage closely related to Halojulellaceae, but Halojulellaceae differs by its cellular pseudoparaphyses and golden-brown ascospores.
Anastomitrabeculia Bhunjun, Phukhams. and K.D. Hyde, gen. nov. Index Fungorum number: IF556560, Facesoffungi number: FoF 09522. Etymology: Referring to the trabeculate pseudoparaphyses anastomosing between the asci and at the apex. Colonies on natural substrate umbonate at the centre, circular, black shiny dots are visible on the host surface. Ascomata on surface of the host, immersed under a clypeus, gregarious, uniloculate, subglobose, carbonaceous. Ostiole orange pigment near ostiole. Peridium comprising multilayers of brown to hyaline cells of textura angularis, inner layers composed of thin, hyaline cells. Asci 8-spored, bitunicate, fissitunicate, broad cylindrical to cylindrical-clavate, with a bulbous pedicellate, rounded at the apex, with an ocular chamber. Ascospores biseriate, broadly fusiform, tapering towards the ends, hyaline, with guttules in each cell, constricted at the septa, with longitudinally striate wall ornamentation, surrounded by mucilaginous sheath.
Note: Anastomitrabeculia is established as a monotypic genus. It is characterised by the presence of carbonaceous ascomata, with orange pigment near ostiole and ascospores with longitudinally striate wall ornamentation. Anastomitrabeculia is morphologically similar to members of Pleosporales in having perithecioid ascomata, bitunicate asci and hyaline ascospores.

Discussion
In this study, we introduce a new species, genus and family for a collection of Pleosporales found on bamboo. The introduction of new taxa, even at the family level, is not surprising, considering that about 93% of fungi remain unknown to science despite ca. 2000 species described every year [59,60]. Pleosporalean species can occur in terrestrial, marine and freshwater habitats [7][8][9]. Several studies have reported new pleosporalean taxa from freshwater or marine habitats or from bambusicolous hosts [1,3]. Pleosporales have unique characters such as perithecioid ascomata typically with a papilla and bitunicate, generally fissitunicate asci, bearing mostly septate ascospores of different colours and shapes, with or without a gelatinous sheath [7]. The morphology of Anastomitrabecu-

Discussion
In this study, we introduce a new species, genus and family for a collection of Pleosporales found on bamboo. The introduction of new taxa, even at the family level, is not surprising, considering that about 93% of fungi remain unknown to science despite ca. 2000 species described every year [59,60]. Pleosporalean species can occur in terrestrial, marine and freshwater habitats [7][8][9]. Several studies have reported new pleosporalean taxa from freshwater or marine habitats or from bambusicolous hosts [1,3]. Pleosporales have unique characters such as perithecioid ascomata typically with a papilla and bitunicate, generally fissitunicate asci, bearing mostly septate ascospores of different colours and shapes, with or without a gelatinous sheath [7]. The morphology of Anastomitrabeculiaceae is similar to members of the Pleosporales based on the presence of pseudoparaphyses, perithecioid ascomata, bitunicate asci and hyaline ascospores. Anastomitrabeculiaceae is characterised by semi-immersed to superficial ascomata, trabeculate pseudoparaphyses, cylindrical asci and ascospores with longitudinally striate wall ornamentation, surrounded by mucilaginous sheath. The newly discovered species formed a well-supported independent lineage basal to the Halojulellaceae based on phylogenetic analyses of the combined dataset (0.99 PP/65% MLBT). Halojulellaceae differs by its cellular pseudoparaphyses and golden brown ascospores [2]. The new taxon is also phylogenetically closely related to Neohendersoniaceae, which differs by its cellular pseudoparaphyses and smooth-walled ascospore [61]. A novel genus Anastomitrabeculia is therefore introduced to accommodate one new species, Anastomitrabeculia didymospora. A new family, Anastomitrabeculiaceae, is also introduced to accommodate this independent lineage.
Divergence time estimate has been widely used as supporting evidence to clarify taxonomic status of extant or novel families in fungal taxonomy [4,6,23,24,26,27,67]. In this study, the MCC tree was congruent with the topology of the trees generated from Bayesian inference analysis and maximum likelihood analyses. The divergence time estimates for the crown age of Dothideomycetes (263 MYA), the split of Dothideomycetes and Arthoniomycetes (323 MYA), the crown age of Pleosporales (206 MYA) and the divergence of Hysteriales from Pleosporales (236 MYA) are similar to previous studies [4,7,11]. Hyde et al. [27] recommended that the divergence times of families should be between 50 and 150 MYA. The stem age is usually preferred to the crown age in taxa ranking as it is not affected by the sample size of the clade [27]. Based on the MCC tree, Anastomitrabeculiaceae and Halojulellaceae share the stem age of 84 MYA which supports the establishment of Anastomitrabeculiaceae.
The divergence time of Anastomitrabeculiaceae was also compared to Pleosporalean families with trabeculate pseudoparaphyses, cylindrical asci and ascospores with a sheath (Table 3). Cyclothyriellaceae has an estimated crown age of 66 MYA and it diverged at 95 MYA. Fuscostagonosporaceae has a crown age of approximately 26 MYA and it diverged around 63 MYA. Bambusicolaceae, which was also isolated from dead bamboo culms, has a crown age of 29 MYA and a stem age of about 57 MYA. The stem age of Anastomitrabeculiaceae lies within the range of divergence times of those with similar morphology, but the crown age of Anastomitrabeculiaceae (2.6 MYA) is much earlier compared to these families. Bambusicolaceae was introduced by Hyde at al. [2] to include three bambusicolous taxa, and it currently has 15 species [7]. Fuscostagonosporaceae was introduced by Hyde at al. [66] to accommodate one bambusicolous taxon and it currently has four species [7]. Ariyawansa et al. [64] introduced the pleosporalean family, Caryosporaceae, which is morphologically similar to Astrosphaeriellaceae and Trematosphaeriaceae [7]. Based on Liu et al. [11], the stem age of Caryosporaceae (85 MYA) is similar to Trematosphaeriaceae (88 MYA) compared to Astrosphaeriellaceae (113 MYA), but the crown age of Caryosporaceae (2 MYA) is much earlier compared to Astrosphaeriellaceae (55 MYA) and Trematosphaeriaceae (65 MYA). Astrosphaeriellaceae currently has 111 species, and Trematosphaeriaceae has 103 species, whereas Caryosporaceae has ten species [7]. Compared to their morphologically similar families, the early crown of Anastomitrabeculiaceae and Caryosporaceae could be due to their smaller sample size. Therefore, further collections are needed for an accurate estimation of the crown age as it is affected by the sample size of the clade [27]. This could also be due to rapid speciation of pleosporalean fungal species given their high adaptation capabilities.
The estimated crown age of Pleosporales (206 MYA) lies within the early Triassic period. The origin of monocotyledons is estimated within the late Cretaceous period (around 145 MYA) [68]. This period is associated with the diversification of pleosporalean families, which continued during the early Cretaceous period when there was a major diversification and radiation of angiosperms, which favoured further diversification of Pleosporalean families to adapt to various hosts [69].
Hosts and their symbionts can speciate in parallel, which relates to a high level of congruence between the phylogeny of the hosts and their symbionts [70,71]. Therefore, studies focusing on divergence time is important for a better understanding of host-pathogen interaction as well as co-evolutionary interactions [72]. This study uses a polyphasic approach based on morphology, multi-locus phylogenetic analyses and divergence time estimates. By implementing a polyphasic approach, we provide strong evidence for introducing the new family based on congruent results supporting the establishment of a new family.