The Hidden Diversity of Diatrypaceous Fungi in China

In this study, we investigated the diversity of diatrypaceous fungi from six regions in China based on morpho-molecular analyses of combined ITS and tub2 gene regions. We accept 23 genera in Diatrypaceae with 18 genera involved in the phylogram, and the other five genera are lacking living materials with sequences data. Eleven species included in four genera (viz. Allocryptovalsa, Diatrype, Diatrypella, and Eutypella) have been isolated from seven host species, of which nine novel species (viz. Allocryptovalsa castaneae, A. castaneicola, Diatrype betulae, D. castaneicola, D. quercicola, Diatrypella betulae, Da. betulicola, Da. hubeiensis, and Da. shennongensis), a known species of Diatrypella favacea, and a new record of Eutypella citricola from the host genus Morus are included. Current results show the high diversity of Diatrypaceae which are wood-inhabiting fungi in China.

During the investigation of forest pathogens in China, 86 diatrypaceous specimens associated with various disease symptoms were collected from Beijing City, Xinjiang Uygur Autonomous Region, and four other provinces in China viz. Hubei, Hebei, Jiangsu, and Yunnan. The objectives were to supplement a multi-gene DNA dataset of Diatrypaceae including ITS and tub2, improve the phylogenetic systematics of this family, and provide a theoretical basis for the identification of diseases and pathogens.

Isolates
Symptomatic branches or twigs were collected from seven tree hosts (Betula albosinensis, B. davurica, B. platyphylla, Castanea mollissima, Juglans regia, Morus alba, and Quercus mongolica) from Beijing City, Xinjiang Uygur Autonomous Region, and four other provinces in China viz. Hubei, Hebei, Jiangsu, and Yunnan. Eighty-six fresh specimens of Diatrypaceae were put into envelopes with records of their altitude, collector, collecting time, host, longitude, and latitude. A total of 21 representative isolates were obtained by removing the ascospores or conidial mass from fresh specimens on the surface of 1.8% potato dextrose agar (PDA) and incubating at 25 • C for 24 h. Single germinating spore was transferred onto a fresh PDA plate. Specimens and isolates were deposited in the Beijing Forestry University (BJFU) and the Beijing Museum of Natural History (BJM). Strains of the new species are maintained in the China Forestry Culture Collection Centre (CFCC).

Morphological Analysis
Species identification was based on morphological features of fruiting bodies and micromorphology supplemented by cultural characteristics. Macro-morphological observations including structure and size of stromata, ectostromatic disc, and ostioles were determined using a Leica stereomicroscope (M205 FA) (Leica Microsystems, Wetzlar, Germany). Micro-morphological photographs were captured using a Nikon Eclipse 80i microscope (Nikon Corporation, Tokyo, Japan), including conidiophores, asci, and conidia/ascospores. Adobe Bridge CS v. 6 and Adobe Photoshop CS v. 5 were used for manual editing. At least 10 conidiomata/ascomata, 10 asci, and 30 conidia/ascospores were randomly selected for measurement to calculate the mean width/length and respective standard deviations (SD). Cultural characteristics of strains incubated in the dark at 25 • C were recorded. Colony morphology was described using the color charts of Rayner (1970). Nomenclatural novelties were deposited in the MycoBank ( 1 Crous et al., 2004).

DNA Extraction, PCR Amplification, and Sequencing
Fungal mycelium grown on the cellophane on PDA was scraped for the extraction of genomic DNA following the modified CTAB method (Doyle and Doyle, 1990). Two loci were amplified, including the internal transcribed spacer (ITS) region and partial beta-tubulin (tub2) using the primer pairs ITS1/ITS4 (White et al., 1990) and T1/Bt2b (Glass and Donaldson, 1995;O'Donnell and Cigelnik, 1997), respectively. The additional combination of Bt2a and Bt2b (Glass and Donaldson, 1995) was used in case of amplification failure of the primer T1 and Bt2b. The polymerase chain reaction (PCR) assay was conducted as described in Fan et al. (2020). PCR amplification products were estimated via electrophoresis in 2% agarose gels. DNA sequencing was performed using an ABI PRISM R 3730XL DNA Analyzer with a BigDye Terminater Kit v. 3.1 (Invitrogen, United States) at the Shanghai Invitrogen Biological Technology Company Limited (Beijing, China).

DNA Sequence Analysis
The initial identities of our strains sequenced were obtained by morphological observations and nucleotide BLAST search. To clarify the phylogenetic position, the alignment based on a combined matrix using ITS and tub2 sequences was performed to compare with other available species in Diatrypaceae. Reference sequences were selected based on ex-type or exepitype sequences available from relevant recently published literature (de Almeida et al., 2016;Senwanna et al., 2017;Shang et al., 2017Shang et al., , 2018Hyde et al., 2019Hyde et al., , 2020aPhookamsak et al., 2019;Dayarathne et al., 2020a,b;Konta et al., 2020;Supplementary Table 2). Xylaria hypoxylon (CBS 122620) was selected as the outgroup. For each gene, sequences were aligned using MAFFT v. 7 (Katoh and Standley, 2013) and manually improved where necessary using MEGA v. 6 (Tamura et al., 2013). Ambiguously aligned sequences were excluded from the analysis. Alignments were used to infer a preliminary phylogenetic relationship for our sequences based on Maximum Parsimony (MP) with PAUP v. 4.0b10 (Swofford, 2003), Maximum Likelihood (ML) with PhyML v. 3.0 (Guindon et al., 2010), and Bayesian Inference (BI) analyses with MrBayes v. 3.1.2 (Ronquist and Huelsenbeck, 2003).
Maximum parsimony analysis was performed using a heuristic search option of 1,000 random-addition sequences. The tree bisection and reconnection (TBR) was selected as option to the branch swapping algorithm (Swofford, 2003). The branches of zero length were collapsed, and all equally parsimonious trees were saved. Clade stability was assessed with a bootstrap analysis of 1,000 replicates (Hillis and Bull, 1993). Tree length (TL), consistency index (CI), retention index (RI), and rescaled consistency (RC) were calculated (Swofford, 2003). ML analysis including 1,000 bootstrap replicates (Hillis and Bull, 1993) was conducted with a general time reversible (GTR) model of site substitution, including gamma-distributed rate heterogeneity and a proportion of invariant sites (Guindon et al., 2010). The nucleotide model of evolution for each of the data partitions were estimated by MrModeltest v. 2.3 (Posada and Crandall, 1998) before the Bayesian analysis. BI analysis was performed using a Markov Chain Monte Carlo (MCMC) algorithm with Bayesian posterior probabilities (Rannala and Yang, 1996). Two MCMC chains were run for 1,000,000 generations with a sampling frequency at every 100th generation. The first 25% of trees were discarded as the burn-in phase of each analysis, and the posterior probabilities (BPP) were calculated to assess the remaining trees (Rannala and Yang, 1996). The branch support from MP and ML analyses were evaluated with a bootstrapping (BS) method of 1,000 replicates (Hillis and Bull, 1993). The resulting trees were plotted in Figtree v. 1.4.4 and edited in Adobe Illustrator CS6 v. 16.0.0. All sequences from this study were deposited in GenBank (Supplementary Table 2). The multi-gene sequence alignment files were submitted to TreeBASE ( 2 accession number: S27126).

Phylogenetic Analyses
The phylogenetic analysis combined ITS and tub2 contained 146 ingroup strains with 1,175 characters including gaps (713 for ITS and 462 for tub2), of which 471 were constant, 191 variable characters were parsimony-uninformative, and 513 characters were variable and parsimony-informative. The MP analysis resulted 500 parsimonious trees, and the first tree (TL = 3,637, CI = 0.362, RI = 0.771, RC = 0.279) was presented in Figure 1. For BI analyses, the best-fit model of nucleotide evolution was deduced on the AIC (ITS: GTR + I + G; tub2: HKY + I + G). Tree topologies of ML and BI analyses did not significantly differ from the MP. Topology of the phylogenetic analyses were similar to the relevant recently published literature (Senwanna et al., 2017;Shang et al., 2017Shang et al., , 2018Hyde et al., 2019Hyde et al., , 2020aPhookamsak et al., 2019;Dayarathne et al., 2020a;Konta et al., 2020).
Some confused taxa were excluded in the current phylogram after the primary analyses.  (CBS 198.49,R191,and DL26C), which was recognized as known species.
Clade 12 (Allocryptovalsa/Eutypella sensu lato): This clade comprises Allocryptovalsa and part Eutypella species clustered with strong support values (MP/ML/BI = 87/94/1) in Figure 1. Two isolates (CFCC 52433 and CFCC 52434) grouped together with Eutypella citricola (HVVIT07 and HVGRF01) with strong support (MP/ML/BI = 100/100/1). The isolates CFCC 52432 formed a separate branch separated from Eutypella citricola and  1 | Phylogram of Diatrypaceae based on combined ITS and tub2 sequence data. The MP and ML bootstrap support values above 70% are shown at the first and second positions, respectively. Thickened branches represent posterior probabilities above 0.95 from the BI. Ex-type strains are in bold, type species are denoted with the superscript "TS" and the disputable type species are denoted with the superscript "TSQ." Strains from the current study are in blue.
Culture characteristics: Cultures are initially white with irregular margin, becoming dark green at the margin and stopping growing with 7 cm in diam. after 2 weeks, comprising dense, irregular, flat mycelium.
Known Notes: Three new strains isolated from branches of Castanea mollissima and Juglans regia, show high support value (MP/ML/BI = 99/100/1) with the closely clustered isolates in Allocryptovalsa (Figure 1; Clade 12: Allocryptovalsa/Eutypella sensu lato). Moreover, this species has different morphological characters. Morphological comparison of members of Allocryptovalsa is provided in Supplementary Table 3.
MycoBank MB 837787. Etymology: Named after the host genus from which it was collected, Castanea.
Culture characteristics: Colonies are initially white, uniform, becoming dark after 2 weeks.
Known host and distribution: Known only on Castanea mollissima in Hebei Province, China.
Moreover, Allocryptovalsa castaneicola can differ from Cryptovalsa species by having a yellow rather than white powdery entostroma appeared on the ascomatal outer surface (Dayarathne et al., 2020b). Also, phylogenetic analyses show affinities of this fungus with strains from Eutypella spp. Therefore, the assignment of the strains to the genus Eutypella sensu lato (Figure 1; Clade 12) may require future reconsideration.
Notes: The genus Diatrype was established by Fries (1849) with Diatrype disciformis as the generic type, which have often been regarded as saprobes on decaying wood and have a strong ability to resist harsh conditions (Senanayake et al., 2015). Due to the taxonomic confusion, Diatrype may require a thorough revision together with the entire family in the future. Diatrype betulae H.Y. Zhu & X.L. Fan sp. nov. Figure 4.
MycoBank MB 837784. Etymology: Named after the host genus from which it was collected, Betula.
Culture characteristics: Cultures are white, uniform, dense, slow growing, reaching 4 cm after 2 weeks, not produced pigmentation on PDA media.
Culture characteristics: Colonies are white, dense, not produced pigmentation on PDA media. Pycnidia distributed irregularly on colony surface with yellow cream conidial drops exuding from the ostioles.
Known host and distribution: Known only on Castanea mollissima in Hebei Province, China.
MycoBank MB 837786. Etymology: Named after the host genus from which it was collected, Quercus.
Diagnosis: Phylogenetically, Diatrype quercicola formed a separate clade. However its asci are polysporous and differ from the common 8-ascospores asci in Diatrype.
Culture characteristics: Colonies are white, irregular, reaching 9 cm after 7 days, not produced pigmentation on PDA media.
Known  (Vasilyeva and Stephenson, 2005). Nevertheless, based on phylogeny analyses, this taxon appears best placed in Diatrype. Therefore, the assignment of this species to Diatrype may require reconsideration due to the taxonomic confusion around Diatrypaceae.
Notes: Diatrypella was introduced by Nitschke (1867) Glawe and Rogers (1984) believed Diatrypella was well distinguished genus as its well-delveloped stromata and the single host affiliation (Da. verruciformis on Alnus and Da. favacea on Betula). Further studies are needed to clarify them. Vasilyeva and Stephenson (2005) pointed out that Diatrypella morphologically resembled Cryptovalsa. Diatrypella and Cryptovalsa were mentioned as the polysporous complement of Diatrype and Eutypa (Vasilyeva and Stephenson, 2005). Nevertheless, it is still difficult to determine the differences between Diatrypella and Cryptovalsa based on morphological characters (Acero et al., 2004;Vasilyeva and Stephenson, 2005). Therefore, multilocus phylogeny including more representative taxa are needed to clarify the relationship among species in Diatrypella (Mehrabi et al., 2015).
Newly generated nine isolates show affinities to Diatrypella favacea clade based on phylogenetic analyses. Therefore, we prefer the assignment of the strains to the genus Diatrypella (Figure 1; Clade 06) preliminarily and tentatively, as it may require future reconsideration after the typification work on type species of Diatrypella. Diatrypella betulae H.Y. Zhu & X.L. Fan sp. nov. Figure 7.
MycoBank MB 837778. Etymology: Named after the host genus from which it was collected, Betula.
Diagnosis: Phylogenetically sister to Diatrypella shennongensis, differ by the number of perithecia.
Culture characteristics: Cultures are flat, reaching 9 cm diam. after 7-10 days. Colonies white, rough on surface, not produced pigmentation on PDA media.
Culture characteristics: Cultures are fluffy, reaching 9 cm after 7 days, becoming pale yellow at the margin after 2 weeks. Colonies dense with aerial mycelium at the center, sparse at the margin.
Culture characteristics: Cultures are white, uniform, attaining 9 cm in 7 days. Colonies sparse at the center, medium dense at the margin, rough on the surface, not produced pigmentation on PDA media.
Known host and distribution: Known on various hosts with worldwide distribution 3 . Notes: Diatrypella favacea was reported to be restricted to Betula spp. in previous study (Saccardo, 1882;Glawe and Rogers, 1984), but then it got involved in the problematic species concept and delimitation with Da. verruciformis (Farr et al., 1989). Diatrypella favacea and Da. pulvinata clustered in a single clade until Hyde et al. (2020b) reported Da. yunnanensis. The strain CFCC 52409 clusters with Diatrypella favacea (CBS 198.49,DL26C,and R191) in a separate lineage. Morphologically, our strain is similar to those previously reported in terms of the size of asci (64-124 × 9.5-12 vs. 70-90 × 8-12 µm) and ascospores (4-5.5 vs. 6-8 µm) (Vasilyeva and Stephenson, 2005). The current definition of Diatrypella favacea seems to be difficult due to the lack of type material with available living culture or DNA sequence data. Thus, the current 3 https://nt.ars-grin.gov/fungaldatabases identification is preliminary and awaits further studies of typification. Diatrypella hubeiensis H.Y. Zhu & X.L. Fan sp. nov. Figure 10.
MycoBank MB 837781. Etymology: Named after the location where it was collected, Hubei Province.
Culture characteristics: Cultures are white, fluffy, fast growing, attaining 9 cm in 7 days. Colonies dense, slightly raised with aerial mycelium, not produced pigmentation on PDA media.
Known host and distribution: Known only on Betula davurica in Hubei Province, China.
MycoBank MB 837780.   Etymology: Named after the location where it was collected, Shennong Ding.
Culture characteristics: Cultures are white, dense, uniform, fluffy, growing up to 4 cm in diam. After 3 days, and reaching 9 cm within 10 days. Colonies do not produced pigmentation on PDA media.
Known host and distribution: Known only on Betula albosinensis in Hubei Province, China.
Notes: Diatrypella shennongensis can be distinguished from its closest relative, Da. betulae, by its number of perithecia (less than 10 vs. more than 10) in one stroma and base number difference (67/665 in ITS and 13/416 in tub2). In addition, the multigene phylogenetic analyses support this species as a new species with high statistical support (MP/ML/BI = 100/100/1).

Other Genera Included in Diatrypaceae
In here, we follow Hyde et al. (2020a); Wijayawardene et al. (2020) to list the genera in Diatrypaceae. A taxonomic key to distinguish 23 genera of Diatrypaceae is provided. Known distribution: Thailand (Li et al., 2016;Konta et al., 2020).
Notes: Allodiatrype was established by Konta et al. (2020) and typified by A. arengae. In the meanwhile, A. elaeidicola, A. elaeidis, and A. thailandica were also accommodated to this genus (Konta et al., 2020). The genus shares most similarities with Diatrype, whereas they can be distinguished by the shape and size of stromata (Konta et al., 2020).
Notes: Dothideovalsa was introduced to accommodate D. diantherae, D. eutypoides, D. tucumanensis, and D. turnerae. At present, only four species epithets of Dothideovalsa were listed in Index Fungorum (2021). However, there was no available sequence data in GenBank. Therefore, this genus is still doubtful and needs further studies.
Notes: Halocryptovalsa was established by Dayarathne et al. (2020b) to accommodate species resembling Cryptovalsa from marine environments namely Cryptovalsa avicenniae and a new species Halocryptovalsa salicorniae. The genus is characterized by poorly developed stromata and poly-spored asci, with a J-, cylindrical, conspicuous apical or subapical ring. Ascospores are hyaline or yellow-brown to brown, allantoid, with small, fat globules at the end (Dayarathne et al., 2020b Known distribution: Thailand (Dayarathne et al., , 2020b. Notes: Halodiatrype was established by Dayarathne et al. (2016) to accommodate H. avicenniae and H. salinicola isolated from mangroves. The characteristic of this genus is having ascomata lacking stromatal tissues, 8-spored, cylindrical to clavate, pedicellate asci, oblong to allantoid or sub-inequilateral, larger ascospores with septa, and libertella-like asexual morphs, which can easily identify Halodiatrype from other genera in Diatrypaceae  Known distribution: Congo, Mexico, Philippines (Ellis and Everhart, 1896;Rehm, 1913).
Notes: Monosporascus was introduced by Pollack and Uecker (1974) with M. cannonballus as the type species. The genus is characterized by pyriform asci and the formation of one (rarely two) single large, sphaerical ascospores (Pollack and Uecker, 1974 Known distribution: Australia, India, and Thailand (Hyde and Jones, 1992;Pande, 2008;Dayarathne et al., 2020b).
Notes: Quaternaria was introduced by Tulasne and Tulasne (1863) and was typified by Q. persoonii. Clements and Shear (1931) lectotypified the illegitimate name Q. quaternata to Q. persoonii and considered Quaternaria as a synonym of Eutypella (Tulasne and Tulasne, 1863). Based on molecular phylogeny and the discussion of Gams (1994), Quaternaria was considered to be an independent genus by Acero et al. (2004). The genus is characterized by stromata was cryptosphaeroid in appearance and developed within the bark parenchyma.
Notes: Rostronitschkia was introduced to accommodate the type species R. nervincola. At present, only the type species was listed in Index Fungorum (2021). However, there were no available sequence data and strains. Therefore, this genus is still doubtful and needs to further study.
Key to genera of Diatrypaceae  Konta et al., 2020). The current study revised the Diatrypaceae and accepted 23 genera in this family (Supplementary Table 1). However, there only exist 18 genera in current phylogenetic analyses due to availability of molecular data (Figure 1). In China, 11 genera and 62 species belong in Diatrypaceae have been recorded (Supplementary Table 5). However, 40 species (66.67%) do not have molecular data until now.
Birch is of high economic, medicinal, and ornamental value. Six species of Diatrypaceae were recorded from Betula spp. with DNA sequences in the current study, including Diatrype betulae, Diatrypella betulae, Da. betulicola, Da. favacea, Da. hubeiensis, and Da. shennongensis. All species of Diatrypella 2 clade were isolated from Betula spp., except for Da. pulvinata and Da. yunnanensis isolated from an unidentified plant. It showed that many Diatrypella species may have obvious host specificity. The other four species (viz. Cryptosphaeria venusta, Diatrype platystoma, D. undulata, and Eutypella halseyana) of Diatrypaceae were recorded from Betula spp. in China but no materials with DNA sequences, including Cryptosphaeria venusta, Diatrype platystoma, D. undulata, and Eutypella halseyana (Teng, 1996;Vasilyeva and Ma, 2014). A morphological key was provided to separate them in the current study.
The Allocryptovalsa species clustered within the same clade as Eutypella species in our phylogenetic analyses (Figure 1; Clade 12: Allocryptovalsa/Eutypella sensu lato). Allocryptovalsa was introduced by Senwanna et al. (2017) which resembles Cryptovalsa in morphology by having polysporous asci different from the 8-spored asci of Eutypella. The number of ascospores per ascus (eight spores vs. multiple spores) has been used traditionally to differentiate the genera of Diatrypaceae (Diatrype vs. Diatrypella and Cryptovalsa vs. Eutypella). However, the recent studies indicated that the polysporous ascus feature maybe not significant in Diatrypaceae (Acero et al., 2004;Trouillas et al., 2011;Chacón et al., 2013;Liu et al., 2015). A thorough revision is needed to resolve the problematic situation of Diatrypaceae, which includes a mass of misidentified genus/species, as a result of the unstable phylogenetic frame with type materials. Therefore, it seems to better if the future work could treate the strains from clade 12 into one genus Allocryptovalsa.
The strains of Diatrype formed a clade with high support values (Figure 1; Clade 09: Diatrype sensu stricto). However, in this clade, some strains of Diatrypella (Da. quercina, Da. iranensis, and Da. macrospora) were placed between Diatrype species. Acero et al. (2004) reported that Cryptosphaeria, Diatrype, Diatrypella, Eutypa, and Eutypella were polyphyletic and confused probably due to lack of tub2 gene sequences or misidentified species. However, these five genera are still polyphyletic within the family from previous studies (de Almeida et al., 2016;Shang et al., 2017;Mehrabi et al., 2019;Dayarathne et al., 2020a,b;Konta et al., 2020) based on the ITS and tub2 sequences data. Allodiatrype and Halodiatrype reported their morphological resemblance to Diatrype Konta et al., 2020). However, asci of Halodiatrype lack an apical ring, the asci of Allodiatrype and Diatrype have J-, cylindrical, conspicuous apical ring Maharachchikumbura et al., 2016;Konta et al., 2020). Additionally, Diatrype can be differentiated from Allodiatrype by the color of ascospores. The ascospores of Diatrype are hyaline becoming yellowish ascospores at maturity, whereas the ascospores of Allodiatrype are hyaline (Maharachchikumbura et al., 2016;Konta et al., 2020). Though Diatrype and Diatrypella are not polyphyletic any more, some strains of Diatrype and Diatrypella species still need further study.
Eutypa guttulata is closely related to the genus Halodiatrype and Pedumispora from the phylogenetic analyses (Figure 1) as a basal branch. Eutypa guttulata can be distinguished by fusiform ascospores, whereas Pedumispora has allantoid ascospores (Hyde and Jones, 1992). It is also different from Halodiatrype because of lacking an apical ring . Diatrype brunneospora is closed to Eutypa guttulata (Figure 1), which is morphologically similar to members of Eutypa spp. (Trouillas et al., 2010b). Therefore, the assignment of this isolate to the genus Diatrype may require reconsideration in the future. Eutypa flavovirens is closely related to the genus Cryptosphaeria with no support (Figure 1). Diatrypella can be distinguished by hyaline to subhyaline, rarely pale olivaceous ascospores, whereas Diatrype palmicola has pale yellowish to pale brown ascospores at maturity (Liu et al., 2015). For those species, the assignment of these isolates still remain unclear, which may require reconsideration in the future.
Eutypa microasca appeared in a strongly supported clade along with two Peroneutypa species with fusiform asci in Figure 1 (Clade 19: Peroneutypa). Carmarán et al. (2006) suggested that the morphology of the ascus could explain the phylogenetic relationships within Diatrypaceae better than stromata, although our study indicates that it is not entirely supported in Peroneutypa. The phylogenetic signal of the ascus shape and the phylogenetic placement of Eutypa microasca should be further tested in future study.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.