Yuxiensis granularis gen. et sp. nov., a Novel Quellkörper-Bearing Fungal Taxon Added to Scortechiniaceae and Inclusion of Parasympodiellaceae in Coronophorales Based on Phylogenetic Evidence

An undetermined saprobic fungal taxon from Yunnan (China) is revealed as a new genus in Scortechiniaceae (Coronophorales). The novel taxon, Yuxiensis, is characterized by immersed to erumpent, semi-globose ascomata, which are not surrounded by any tomentum or conspicuous subiculum, a subcylindrical quellkörper in the centrum, clavate asci with long pedicels and allantoid hyaline ascospores with granular contents. Maximum likelihood and Bayesian posterior probability analyses based on LSU, ITS, tef1 and rpb2 sequence data depict a close phylogenetic relationship of the new genus to Pseudocatenomycopsis, hence, confirming its placement in Scortechiniaceae. Parasympodiellaceae, thus far belonging to Parasympodiellales, is transferred to Coronophorales based on multi-gene phylogenetic evidence. Additionally, the incertae sedis monotypic genus Arthrocristula is treated as a synonym of Parasympodiella, with Arthrocristula hyphenata recombined as Parasympodiella hyphenata comb. nov., as the type strain of Arthrocristula hyphenata clusters inside the Parasympodiellaceae clade along with other Parasympodiella taxa.


Specimen Collection and Morphological Studies
Dead twigs of an undetermined deciduous host were collected from Yuxi, Yunnan Province, China, during the dry season in May 2019. The samples were taken to the mycology laboratory at the Kunming Institute of Botany, in a plastic Ziploc bag and stored inside a paper envelope. External examinations were made using a Motic SMZ 168 Series stereo-microscope. Morphological characters were examined by hand sectioning of sporocarps and placed on water-mounted glass slides. Microscopic photography was conducted using a Nikon ECLIPSE 80i compound microscope (Nikon, Tokyo, Japan) fitted with a Canon EOS 600D camera. The structures which were observed and measured include diameter, height, shape, and color of ascomata; ascomatal wall's width and cell structure; quellkörper shape; shape, length, and width of asci and shape, size, and ornamentation of ascospores. The Tarosoft (R) Image Frame Work version 0.9.7. program was used for the measurements of photomicrograph structures. Images used for figures were processed with the Adobe Photoshop CS6 Extended version 13.0.1 software (Adobe Systems, San Jose, California). The holotype was deposited in the herbarium of Cryptogams Kunming Institute of Botany Academia Sinica (HKAS). Both Facesoffungi [20] and Index Fungorum [21] numbers were obtained.

DNA Extraction, PCR Amplification and Sequencing
No culture could be obtained for the collected sample despite several trials on various media, including malt extract agar, potato dextrose agar, corn meal agar, or water agar under different incubation conditions. Therefore, DNA was extracted directly from the fruiting bodies of the fungus as outlined by Wanasinghe et al. [22].
The PCR protocols were programmed as described in Wanasinghe et al. [28]. The PCR products were verified by staining with ethidium bromide on 1% agarose electrophoresis gels. They were then purified according to the company protocols and DNA sequencing was performed at Shanghai Sangon Biological Engineering Technology & Services Co. (Shanghai, P.R. China). Forward and reverse DNA sequence data were obtained and analyzed. Consensus sequences were generated using the SeqMan software (DNAStar, Inc., Madison, WI, USA). The newly acquired sequence data from the present study were deposited in GenBank for subsequent studies [29] (Table 1). Table 1. Taxa used in this study and their corresponding GenBank accession numbers. Generated sequence data for the new taxon are in bold.

Taxa Strains
GenBank Accession Numbers

Phylogenetic Analyses
Verified sequences were initially used for BLASTn analyses, following which closely related sequences were downloaded from GenBank based on BLAST similarities and relevant publications [2,8] (Table 1). Alignment of each locus was performed using MUSCLE in MEGA X (Molecular Evolutionary Genetics Analysis), using default conditions for gap openings and gap extension penalties. It was then improved whenever necessary in the BioEdit v.7.0.5.2 software [30].
Maximum likelihood (ML) and Bayesian posterior probability (BYPP) analyses were conducted using both individual and combined datasets. Prior to ML analysis, the sequence alignments were converted from FASTA into PHYLIP format using the ALTER (alignment transformation environment, http://www.singgroup.org/ALTER/, accessed on 30 August 2021) bioinformatics web tool [31]. They were then used to generate ML trees using RAxML-HPC2 on XSEDE (v.8.2.10) [32] with the GTRGAMMA substitution model and bootstrapping with 1000 replicates.
The BYPP analysis was generated using Markov Chain Monte Carlo sampling in MrBayes v3.1.2 [33,34]. MrModeltest v.2.3 [35] was used to estimate the best evolutionary model for each gene region under the Akaike Information Criterion (AIC) implemented in PAUP v.4.0b10 [36]. The best-fit model was determined as GTR+I+G for LSU, tef1, and rpb2 while GTR+G for ITS. Six simultaneous Markov chains were run for 4.5M generations with trees sampled every 100th generation. The first 20% of generated trees were the burn-in phase and discarded. The remaining 80% of trees were used to calculate posterior probabilities in the majority rule consensus tree. Phylograms were configured in FigTree v.1.4.0 [37] and modified in Microsoft PowerPoint (2013). The final alignments and phylogenetic tree were deposited in TreeBASE, submission ID: 28713 (http://www.treebase.org/ , accessed on 30 August 2021).

Taxonomy
In this section, the amended descriptions and notes for Coronophorales, Parasympodiellaceae, and Parasympodiella are given. Furthermore, descriptions, notes, and illustrations are given for the following taxa: Parasympodiella hyphenata comb. nov. and Yuxiensis granularis gen. et sp. nov. The families Bertiaceae, Ceratostomataceae, Chaetosphaerellaceae, Coronophoraceae, Nitschkiaceae, Parasympodiellaceae and Scortechiniaceae grouped together, each forming a monophyletic clade in the ML tree ( Figure 1). The tree topology resulting from the BYPP analysis mainly differed from the ML one with regard to the placement of Coronophora gregaria (ANM1555) (Coronophoraceae), which clustered within Parasympodiellaceae ( Figure A1).
Type family: Coronophoraceae Höhn. Notes: The phylogenetic analyses based on the combined LSU-ITS-tef1-rpb2 sequence data, conducted in the present study, supports the inclusion of Parasympodiellales in Coronophorales, as Parasympodiellaceae is sister to Scortechiniaceae with 96% ML and 1.00 BYPP statistical support within the order ( Figure 1). The description for the asexual morph of taxa in Coronophorales is therefore emended to include the morphological characteristics of Parasympodiellaceae. With the exact taxonomic placement being uncertain, the hyphomycetous taxon Arthrocristula has so far been maintained in Ascomycota genera incertae sedis [18,38]. In the present study, the type strain of A. hyphenata (CBS 583.82) was found to cluster within Parasympodiellaceae, indicating a close phylogenetic affinity to Parasympodiella (Figure 1). Morphologically, the conidiophores of Parasympodiella are unbranched or sparingly branched, thick-walled, and brown, becoming paler towards the conidiogenous regions. They are often geniculate, with terminal or intercalary sympodial conidiogenous cells, which are mostly thallic. The secession of conidia occurs schizolytically [13,14,16]. The conidiophores of Arthrocristula are, however, well-branched with no conspicuous geniculation and are initially hyaline, becoming thick-walled and brown upon maturity. The conidiogenous cells secede rhexolytically into arthroconidia, leaving the mature conidia with remnants of the separating cells at each end with small frills. The terminal cells of the conidiogenous hyphae remain as empty cells [17]. Considering these morphological differences, the two genera can be considered as distinct. However, given that Parasympodiella has a stylaspergillus-like synasexual morph, the fact that Arthrocristula also is another synasexual morph of Parasympodiella cannot be ruled out. Fungi have been reported to have two or more morphologically distinct asexual morphs [39][40][41]. Moreover, phylogeny supports the inclusion of Arthrocristula in Parasympodiella (Figure 1). Therefore, the former is synonymized under Parasympodiella in the present study. Associated with sclerotia of Sclerotinia minor in soil. Sexual morph: Undetermined. Asexual morph: Hyphomycetous. Mycelium in vitro comprises narrow, septate vegetative hyphae (2-3 µm diam.), hyaline when immature and brown at maturity. Conidiophores originating from vegetative hyphae, undifferentiated, initially hyaline, turning brown and thick-walled when mature, branched, narrow at first, widening and extending apically and laterally after that to give rise to fertile conidiogenous hyphae. Conidiogenous cells (80-100 × 4-5.5 µm), delimited by a basal septum, developing successively, generally unbranched, indeterminate, initially non-septate, hyaline, becoming randomly septate and pigmented in alternate cells on maturity, undergoing rhexolytic secession to give rise to arthroconidia. Conidia (5-12 × 6-7 µm), thallic-arthric, brown, aseptate to septate, cylindrical to oblong, formed in a chain with hyaline separating cells, with fragments of intervening cell walls remaining attached to both ends of conidium after secession, terminal cells of the chains remain empty. Chlamydospores in vitro (12-16 × 8-13 µm), thick-walled, brown, terminal and intercalary (adapted from Sigler et al. [17]).

Parasympodiellaceae
Notes: The type and single taxon of Arthrocristula, A. hyphenata is combined under Parasympodiella following the synonymy of Arthrocristula under Parasympodiella in the present study. Parasympodiella hyphenata comb. nov. has been reported from the sclerotium of Sclerotinia minor buried in the soil [17]. It is mainly characterized by branched conidiophores and alternatively pigmented conidiogenous cells, which produce conidia by seceding rhexolytically [17]. Saprobic on dead wood. Sexual morph: Ascomata immersed to erumpent, appearing superficial on worn off substrate, astromatic to stromatic, aggregated, black, non-ostiolate, semi-globose when fresh, collapsing when dry, without hair or bristles, subiculum inconspicuous, coriaceous. Ascomatal wall comprising 2-3 types of layers; outermost layer heavily pigmented, composed of thick-walled, very dark brown cells, inner layer comprising Saprobic on dead wood. Sexual morph: Ascomata immersed to erumpent, appearing superficial on worn off substrate, astromatic to stromatic, aggregated, black, non-ostiolate, semi-globose when fresh, collapsing when dry, without hair or bristles, subiculum inconspicuous, coriaceous. Ascomatal wall comprising 2-3 types of layers; outermost layer heavily pigmented, composed of thick-walled, very dark brown cells, inner layer comprising thick-walled cells of textura globulosa to textura angularis, innermost layer composed of flattened, thin-walled cells of textura prismatica toward the locule; Munk pores visible, few per cell. Hamathecium composed of large, subcylindrical quellkörper, attached to the roof of the centrum and extending until base of the ascoma; paraphyses indistinct. Asci 8-spored, unitunicate, clavate, long-pedicellate, rounded at apex, lacking an apical ring, thin-walled, evanescent. Ascospores irregularly arranged, cylindrical to allantoid, hyaline, aseptate, with granular contents, lacking mucilaginous sheath or appendage. Asexual morph: Undetermined.
Type: Yuxiensis granularis Bundhun, Wanas. & K.D. Hyde Notes: Yuxiensis is introduced in Scortechiniaceae as a new quellkörper-bearing taxon distinct from all other genera in the family, based on LSU, ITS, tef1 and rpb2 sequence data. The new genus has a close phylogenetic affinity to Pseudocatenomycopsis even though this relationship is not statistically significant (Figure 1). This low support may possibly be accounted for by insufficient taxon sampling. Nevertheless, the new taxon being introduced in the present study constantly clusters with Pseudocatenomycopsis in all phylogenies (single, not shown; and concatenated, Figure 1). The asexual morph for Yuxiensis could not be obtained in the present study. It thus cannot be morphologically compared with Pseudocatenomycopsis, which has been described in its asexual morph only. Pseudocatenomycopsis has been introduced from Zambia as a saprobe on the stem of Rothmannia engleriana (Rubiaceae) [11]. The new genus is also phylogenetically close to Euacanthe (Figure 1). It morphologically differs from Euacanthe in terms of ascomatal position and surface as well as ascospore ornamentation [12,43] Notes: In the multi-gene phylogeny, Yuxiensis granularis is more closely related to Pseudocatenomycopsis rothmanniae, followed by Euacanthe usambarensis (=Euacanthe foveolata [12]) ( Figure 1). The LSU sequence of Yuxiensis granularis is 95% similar to Pseudocatenomycopsis rothmanniae (GenBank KF777237; similarity = 869/910(95%), Gaps = 3/910(0%)). The ITS sequence of Yuxiensis granularis is 85% similar to Pseudocatenomycopsis rothmanniae (Gen-Bank KF777185; similarity = 470/552(85%), Gaps = 10/552(1%)). Pseudocatenomycopsis rothmanniae has only LSU and ITS sequence data deposited in GenBank, and hence the protein-coding genes, tef1 and rpb2 could not be compared. Morphological comparison between the two taxa is currently unfeasible since the single species of Pseudocatenomycopsis, P. rothmanniae, has been introduced in its asexual morph [11], while the asexual morph for Yuxiensis granularis could not be obtained.

Discussion
Sordariomycetes is a frequently-studied class, with several taxa having been recently introduced or revised [4,5,8,12,44]. The present study corroborates this fact, as supported, firstly, by establishing a new saprobic genus, Yuxiensis, in Scortechiniaceae based on a dual taxonomic approach. In addition to phylogeny, the familial placement of the new genus within Scortechiniaceae is morphologically confirmed by the presence of the quellkörper in

Discussion
Sordariomycetes is a frequently-studied class, with several taxa having been recently introduced or revised [4,5,8,12,44]. The present study corroborates this fact, as supported, firstly, by establishing a new saprobic genus, Yuxiensis, in Scortechiniaceae based on a dual taxonomic approach. In addition to phylogeny, the familial placement of the new genus within Scortechiniaceae is morphologically confirmed by the presence of the quellkörper in its centrum. Within Scortechiniaceae, Yuxiensis shares many overlapping characters with the other genera. For instance, like almost all the other genera in the family with a known sexual morph, Yuxiensis comprises ascomata which collapse upon drying, presence of munk pores in the ascomatal wall and inconspicuous paraphyses [4,8]. It also has 8-spored asci, similar to Biciliospora, Coronophorella, Euacanthe, Scortechinia, and Tympanopsis and long-pedicellate asci like Biciliospora, Neofracchiaea, Scortechinia, Scortechiniella, and Scortechiniellopsis [2,4,8]. It is equally similar to most taxa of Neocryptosphaerella and Pseudocryptosphaerella in that its ascomata are immersed to erumpent, appearing superficial when the substrate has worn away [2,12]. The new genus however, demarcates itself from the other genera in the family by several ways. Its ascomata are not seated on or surrounded by a well-developed, conspicuous subiculum unlike many taxa of Biciliospora, Coronophorella, Euacanthe, Neofracchiaea, Scortechinia, Scortechiniella, Scortechiniellopsis, or Tympanopsis [2,4,8]. The ascomata of Yuxiensis are also devoid of a tomentose or setose surface as compared to Euacanthe, Neofracchiaea, or some taxa of Neocryptosphaerella and Pseudocryptosphaerella [2,8]. Moreover, the ascospores of Yuxiensis do not have conspicuous guttules unlike those of Euacanthe, Neocryptosphaerella, and Pseudocryptosphaerella taxa and they lack appendage-like wall extensions on both ends, contrary to Biciliospora and Scortechiniella [2,45]. Since Yuxiensis is phylogenetically closely related to Pseudocatenomycopsis and Euacanthe (Figure 1), more details about their morpho-molecular comparisons have been given in the result parts 3.2.5 and 3.2.6 above. Based on all these morphological as well as phylogenetic differences, Yuxiensis is herein introduced as a new genus.
The inclusion of Parasympodiellales in Coronophorales in the present study also points toward the continuous amendment in the classification of Sordariomycetes. Herein, while phylogeny supports the addition of Parasympodiellaceae to Coronophorales, this inclusion is equally supported by the fact that taxa of Parasympodiellaceae have similar morphological characters with several taxa in other families of Coronophorales (Ceratostomataceae, Chaetosphaerellaceae, Scortechiniaceae ) in terms of unbranched or branched, generally brown and often septate conidiophores or conidia produced in chains [4,8,11]. Parasympodiellaceae distinguishes itself from the other families mainly by its sympodial and unbranched conidiogenous cells which undergo schizolytic or rhexolytic secession to form arthroconidia. Furthermore, the incertae sedis taxon Arthrocristula is synonymized under Parasympodiella, with Arthrocristula hyphenata recombined to Parasympodiella hyphenata and representing another synasexual morph of Parasympodiella. This arthrocristula-like synasexual morph of Parasympodiella is typically characterized by branched conidiophores and conidiogenous cells which secede rhexolytically to give rise to arthroconidia. It is also different from the stylaspergillus-like synasexual morph of Parasympodiella which is generally characterized by pale brown, phialidic conidiogenous cells originating from terminal or intercalary vesicle-like cells and filiform conidia which are produced in slimy masses [13].
An additional collection of Fracchiaea myricoides (HKAS 115760) was also made in the present study, and sequence data for the same have been used in the phylogenetic analyses and deposited in GenBank ( Table 1). The latter species was initially introduced as Coronophora myricoides based on LSU and ITS sequence data and the differences mentioned between this taxon and the type species of Coronophora, C. gregaria was mainly based on the shapes of the ascomata and ascospores [46]. Huang et al. [12] recently synonymized this species to Fracchiaea myricoides; our collection supports this synonymy and the inclusion of 'Coronophora myricoides' in Fracchiaea (Nitschkiaceae) (Figure 1).
Despite the advancement towards a natural classification of Sordariomycetes, uncertainties and confusions still prevail, as we note in the case of Parasympodiella longispora (CBS 544.84 and KACC 41225) (in Parasympodiellaceae clade, Figure 1). The latter is currently known as 'Bahusakala longispora' in Index Fungorum and MycoBank, with Parasympodiella longispora as an (obligate) synonym. The type strain of 'Bahusakala longispora', CBS 544.84, sequenced by Vu et al. [19], clusters in the Parasympodiellaceae clade ( Figure 1) with good statistical support. Bahusakala longispora was introduced by Tokumasu and Tubaki [47] as a taxon with conidiophores that are sympodial, rarely branched, erect in the lower part and become geniculate (zig-zag, as mentioned in the original description) in the upper part. Furthermore, the conidiogenous cells are hyaline, originating at regular intervals, while the arthroconidia, subhyaline to pale yellow. Chlamydospores are produced in vegetative hyphae. However, the species was later synonymized to Parasympodiella longispora since its morphology (based on its type) matched the description of Parasympodiella more accurately [48]. Bahusakala taxa have been reported to have conidiophores that branch at irregular intervals to produce brown conidiogenous hyphae at the terminal and intercalary positions. The conidia are usually brown to dark brown and originate from random disarticulation of the main conidiophore axes and conidiogenous hyphae [42,[48][49][50]. Based on the description of its type (and placement in the present phylogenetic tree), the species is better accommodated in Parasympodiella than Bahusakala. We may as well adopt a broader taxonomic perspective and decide that both Parasympodiella and Bahusakala are congeneric since, despite their morphological differences, the two genera are also characterized by similar features. Both are hyphomycetes with erect and septate conidiophores and produce arthroconidia which secede schizolytically [42,50]. However, no molecular data for the type species, B. olivaceonigra is yet available to enable any definite phylogenetic placement and eventually to confirm a taxonomic conclusion for Bahusakala.
We also take note that the GenBank accession numbers of the sequence data for Neocryptosphaerella globosa (GKM471N) that we use in our phylogeny have been assigned under different strain numbers, namely, LSU (GenBank FJ968977: strain GKM469N), tef1 (GenBank FJ969036: strain GKM471N), and rpb2 (GenBank FJ968935: strain GKM469N). In the original manuscript [2], these accession numbers are under the strain number Neocryptosphaerella globosa GKM471N. Therefore, we followed the original paper.
This remarkable finding of a new genus in a rarely collected order indicates how little we know of the fungal diversity of Yunnan and the broader region, including areas such as Thailand and Laos [51], with recent studies showing large numbers of novel taxa being discovered [52,53]. Further studies in other countries and habitats across this region will surely result in the discovery of numerous other taxa in Parasympodiella, Yuxiensis, and other poorly known taxa of Coronophorales [54].