Morphological and Phylogenetic Appraisal of Novel and Extant Taxa of Stictidaceae from Northern Thailand

Stictidaceae comprises taxa with diverse lifestyles. Many species in this family are drought resistant and important for studying fungal adaptation and evolution. Stictidaceae comprises 32 genera, but many of them have been neglected for decades due to the lack of field collections and molecular data. In this study, we introduce a new species Fitzroyomyces hyaloseptisporus and a new combination Fitzroyomyces pandanicola. We also provide additional morphological and molecular data for Ostropomyces pruinosellus and O. thailandicus based on new collections isolated from an unidentified woody dicotyledonous host in Chiang Rai, Thailand. Taxonomic conclusions are made with the aid of morphological evidence and phylogenetic analysis of combined LSU, ITS and mtSSU sequence data. Characteristics such as the shape and septation of ascospores and conidia as well as lifestyles among genera of Stictidaceae are discussed.


Introduction
Nannfeldt [1] established Ostropales to accommodate Ostropaceae, which was later synonymized under Stictidaceae [2]. The taxonomic position of Ostropales is within Ostropomycetidae (Lecanoromycetes and Ascomycota), which was assigned based on multigene phylogenetic analyses [3]. The outlines of Ostropales have been revised in several studies, and the number of families in the order has been subject to multiple changes over time [4][5][6][7]. Previously, 14 families had been included in Ostropales, namely, Coenogoniaceae, Gomphillaceae (syn. Solorinellaceae), Graphidaceae, Gyalectaceae, Odontotremataceae, Phaneromycetaceae, Phlyctidaceae, Trichotheliaceae (syn. Myeloconidaceae and Porinaceae), Protothelenellaceae (syn. Thrombiaceae), Sagiolechiaceae, Spirographaceae, Stictidaceae, Thelenellaceae and Thelotremataceae [2,[4][5][6][7][8][9]. However, multigene phylogenies have ultimately resulted in the transfer of most families into different orders, such as Baeomycetales, Thellenellales, Graphidales and Gyalectales. Presently, Stictidaceae is the only family assigned to Ostropales [10]. Morphology of ascospores, conidia and conidiogenesis of 32 genera in Stictidaceae. The orange area contains genera known only from sexual morphs; the area in purple comprises genera that were established based only on asexual morphs; and the area in blue encompasses genera that were described from both sexual and asexual morphs, while the area in green consists of a single genus, which reproduces via sticky propagules rather than sexual or asexual spores. The number of ascospores per ascus for each genus is noted in the outermost circle. The ascospores and conidia indicated with the red letters "T" were redrawn from the type species. The original references of these characters are cited for each genus.
Stictidaceae species commonly occur on bark, leaves, stems and wood of various plant hosts in terrestrial habitats [25] and have broad geographic distribution in Africa, Asia, Europe and North America [13]. This family contains a broad diversity of lifestyles, ranging from saprobes, pathogens and endophytes to lichens. Lichenicolous species have been recorded in Cryptodiscus, Nanostictis and Sphaeropezia [33,42,54], while Schizoxylon albescens, Stictis confusa and Stictis populorum show optionally lichenized lifestyles depending on the associated substrates [55,56]. Thus, Stictidaceae played an important Figure 1. Morphology of ascospores, conidia and conidiogenesis of 32 genera in Stictidaceae. The orange area contains genera known only from sexual morphs; the area in purple comprises genera that were established based only on asexual morphs; and the area in blue encompasses genera that were described from both sexual and asexual morphs, while the area in green consists of a single genus, which reproduces via sticky propagules rather than sexual or asexual spores. The number of ascospores per ascus for each genus is noted in the outermost circle. The ascospores and conidia indicated with the red letters "T" were redrawn from the type species. The original references of these characters are cited for each genus.
Stictidaceae species commonly occur on bark, leaves, stems and wood of various plant hosts in terrestrial habitats [25] and have broad geographic distribution in Africa, Asia, Europe and North America [13]. This family contains a broad diversity of lifestyles, ranging from saprobes, pathogens and endophytes to lichens. Lichenicolous species have been recorded in Cryptodiscus, Nanostictis and Sphaeropezia [33,42,54], while Schizoxylon albescens, Stictis confusa and Stictis populorum show optionally lichenized lifestyles depending on the associated substrates [55,56]. Thus, Stictidaceae played an important role in understanding how fungi can adapt to various successional substrates during habitat succession by switching lifestyle [55].
In this study, we aim to clarify the taxonomic placement of new species (Fitzroyomyces hyaloseptisporus sp. nov. and Fitzroyomyces pandanicola comb. nov) and two previously described species (Ostropomyces pruinosellus and O. thailandicus) by using morphological and multigene-based phylogenetic analyses. To provide an identification scheme for all genera of Stictidaceae, the diversity of ascospore/conidia morphology and the mode of conidial development is summarized. Descriptions of sexual and asexual morphs of Stictidaceae are refined to keep abreast of current literature. The lifestyles of genera in Stictidaceae are mapped on the phylogenetic tree in order to visualize these features in a phylogenetic context and to explore the transition of nutrition modes in Stictidaceae.

Sample Collections and Isolation
Specimens were collected from an unidentified woody dicotyledonous plant in northern Thailand. Pure cultures were obtained via single spore isolation as outlined in Senanayake et al. [57]. The obtained cultures were deposited in Mae Fah Luang University culture collection (MFLUCC), Chiang Rai, Thailand, and herbarium specimens were deposited in Mae Fah Luang University Herbarium (Herb. MFLU). The Faces of Fungi numbers were obtained following Jayasiri et al. [58], and species names were registered in Index Fungorum (2021).

Morphological Studies
Specimens were examined with a Motic SMZ 168 stereomicroscope. Hand sections of the ascomata and conidiomata were mounted in water for microscopic studies and photomicrography. A Congo red solution was used to observe asci and paraphyses. The key structures such as ascomata, exciple, paraphyses, asci, ascospores, conidiogenous cells and conidia were observed by using a Nikon ECLIPSE 80i compound microscope, and they were photographed with a DS-Ri2 camera attached to the compound microscope. The measurements were taken with the Tarosoft (R) Image Frame Work program, while images used for figures were processed with Adobe Photoshop CS3 (Version 15.0.0, Adobe ® , San Jose, CA, USA).Extended version 10.0 (Adobe Systems, San Jose, CA, USA).

DNA Extraction, PCR Amplification and Sequencing
Genomic DNA was extracted separately from both fresh fungal mycelia growing on potato dextrose agar (PDA) media and fruiting bodies using Biospin Fungus Genomic DNA Extraction Kit (BioFlux ® , Hangzhou, China) following the protocol of the manufacturer. Polymerase chain reaction (PCR) was carried out for three partial gene fragments including large subunit ribosomal rRNA (LSU), internal transcribed spacers (ITS) and mitochondrial small subunit ribosomal rRNA (mtSSU) with primers LR0R/LR5 [59], ITS5/ITS4 [60] and mrSSU1/mrSSU3R [61], respectively. Amplification reactions were performed in 25 µL of PCR mixtures containing 8.5 µL of ddH2O, 12.5 µL 2×PCR Master Mix (Bioteke Corporation, Beijing, China), 2 µL of DNA template and 1 µL of each primer. PCR amplifications of LSU and ITS genes were performed as described by Wanasinghe et al. [33], while mtSSU gene was amplified following Zoller et al. [61]. Amplified PCR products were sequenced in Tsingke (Kunming, China). Sequences generated in this study were deposited in GenBank, and accession numbers were obtained (see Table 1). The newly generated sequences are in bold font. The type strains are indicated with the symbol "T".

Sequence Alignment and Phylogenetic Analyses
Raw sequences generated in this study were assembled with Sequencing Project Management (SeqMan) [62]. Megablast search using the newly generated sequences as queries was performed to check for contamination and to reveal closely related taxa in the GenBank nucleotide database. The available taxa representing genera in Stictidaceae are listed in Table 1. Each gene matrix was independently aligned with MAFFT (http: //mafft.cbrc.jp/alignment/server/, accessed on 31 July 2021) [63]. Uninformative gaps and ambiguous regions were removed using Trimal available on the Phylemon 2.0 online platform [64]. Trimmed alignments were combined with Sequence Matrix v. 1.7.8 [65]. The combined alignment was used for maximum likelihood (ML) and Bayesian inference (BI) analyses.
Maximum likelihood analysis was performed using RAxML-HPC2 on XSEDE (8.2.10) in CIPRES Science Gateway V. 3.3 [66] by employing default parameters but with the following adjustments: Bootstrap iterations were set to 1000, and substitution model was set to GTR+GAMMA+I. The optimal nucleotide substitution models used for Bayesian analysis were independently selected for each locus under Akaike information criterion (AIC). Bayesian analysis was performed with MrBayes 3.2.7a in CIPRES Science Gateway v. 3.3 [67] in order to infer posterior probabilities (PP) [68,69] with Markov Chain Monte Carlo sampling (MCMC). Six simultaneous Markov chains were run for 2,000,000 generations, and trees were sampled every 1000 generations, resulting in 2000 trees. The first 25% of trees, representing the burn-in phase of the analyses, were discarded, while the remaining 75% of trees were used to calculate PP in the majority rule consensus tree.

Phylogenetic Analysis
Maximum likelihood phylogenetic analysis was conducted using combined LSU, ITS and mtSSU sequence data of 70 representative taxa of Stictidaceae, two species of Thelenellaceae and three species of Trapeliaceae. The tree is artificially rooted to Orceolina kerguelensis, Placopsis perrugosa and Trapelia placodioides in Trapeliaceae, following Baloch et al. [7]. Of the 32 genera within Stictidaceae, molecular data are available for 22, 17 of which contain the type species. Thus, the sequences from all 22 genera are used in the phylogenetic analysis herein. Alignment comprised 2140 characters, including gaps (LSU: 841; ITS: 569; mtSSU: 730), of which 1001 characters were constant, 203 variable characters were parsimony-uninformative and 936 (43%) characters were parsimony-informative. The ML analysis of the combined dataset yielded a best scoring tree with a final ML optimization likelihood value of −26,748.800570. The alignment had 1332 distinct alignment patterns, with 30.23% completely undetermined characters and gaps.

Taxonomy
dark olivaceous, doliiform, cylindrical, acicular, filiform, one-celled to multiseptated occasionally staurosporous.  Saprotrophic, lichenized or optionally lichenized on the wood, bark, stem and leaves of various plant hosts, or are lichenicolous on other microfungi or living as parasites and endophytes of living plants. Sexual morph: Ascomata immersed or semi-immersed to superficial, perithecioid or apothecioid. They are gregarious, opening by the entire pore or transverse slit. Discs vary in color ranging from white, grey and brown to dark and are usually pruinose. Exciple typically consists of interwoven hyphae, sometimes contains crystalline inclusions. Periphysoids present or not. Hymenium comprising asci and paraphyses, commonly enclosed in a thick gelatinous matrix. The subhymenium is hyaline or pigmented and composed of angular cells. The paraphyses is filiform, simple or branched and sometimes apically enlarged, circinate and adhering to form an epithecium.
Notes: In the ITS, LSU and mtSSU combined phylogenetic analysis, our strains MFLUCC 21-0113 and MFLU 21-0116 formed a clade with Ostropomyces thailandicus (MFLU 20-0539) with 100 MLBS/1.00 PP statistical support values ( Figure 2). The MFLU 21-0116 isolate shares similar morphological characteristics to the type specimen (MFLU 20-0539) in having irregular-shaped conidiomata, hyaline, cylindrical conidiogenous cells and catenulate conidia that easily break into small, numerous and ellipsoidal units ( Figure 4). Therefore, we introduced our isolate MFLU 21-0116 as a new collection of O. thailandicus. This species produces many crystal-like substances on PDA in vitro, which has rarely been reported in known Stictidaceae species. Additionally, based on the collections in this study and Thiyagaraja et al. [20], we found that O. thailandicus and O. pruinosellus likely occurred in close proximity to each other.
thailandicus. This species produces many crystal-like substances on PDA in vitro, which has rarely been reported in known Stictidaceae species. Additionally, based on the collections in this study and Thiyagaraja et al. [20], we found that O. thailandicus and O. pruinosellus likely occurred in close proximity to each other.

Discussion
It has been previously hypothesized that the common ancestor of Ostropales was lichenized and that the extant saprotrophic lineages are the result of multiple losses of

Discussion
It has been previously hypothesized that the common ancestor of Ostropales was lichenized and that the extant saprotrophic lineages are the result of multiple losses of lichenization [2,72]. Thiyagaraja et al. [20] also proposed that Stictidaceae was derived from loss of lichenization. In the phylogenetic analysis herein, we manually mapped lifestyles on the phylogenetic tree ( Figure 2). Information on species lifestyles was acquired from relevant references [2,26]. Taxa included in the phylogenetic analysis herein comprised five lifestyles: (1) saprotrophic, (2) lichenized, (3) optionally lichenized, (4) lichenicolous and (5) endophytic. The present phylogenetic tree shows that saprobes are broadly dispersed throughout the tree; lichenized species disperse in six clades, optionally lichenized in three, lichenicolous in two and endophytic in one. The high diversity of life modes is notable even within genera; for example, within Cryptodiscus three lifestyles are present, while there are two within Sphaeropezia and Stictis. Collective consideration of previous findings and those herein supports the conclusions of Thiyagaraja et al. [20] regarding the saprobic nature of the common ancestor of Stictidaceae and the convergent development of non-saprobic life modes in this family. One, however, cannot draw firm conclusions at this point as 10 genera lack molecular data; thus, their phylogenetic position within Stictidaceae is undetermined. Notably some of these genera are parasitic, and it would be of interest to see their distribution across the tree. Regardless, the observed notable plasticity of lifestyles might represent a strategy of fungi adapted to habitats, where succession commonly occurs [55] and a driving force facilitating speciation in Stictidaceae [20].
Of the ten genera that lack molecular data, Conotremopsis, Delpontia and Propoliopsis are monotypic. Conotremopsis is distinct in having superficial, elongated and keeve-like apothecia [34]. Delpontia has submuriform ascospores similar to those in Cryptodiscus and Topelia. Nonetheless, Delpontia can be easily distinguished from Topelia. The former forms wide-opening apothecia and has a non-lichenized lifestyle, while the latter has closed perithecioid apothecia and a lichenized lifestyle [13,43]. Cryptodiscus is diverse in terms of ascospore character morphology, apothecia nature and lifestyle [13,27,28], which can result in confusion as to its relationship with Delpontia. The other genera lacking molecular data are not monotypic. Dendroseptoria is distinct in its sporodochial fruiting bodies and staurosporous conidia [48], while Stictophacidium is set apart in having unicellular ascospores [13]. Biostictis, Karstenia, Lillicoa and Nanostictis have long clavate, cylindrical to filiform and multiseptate ascospores that resemble those of the majority of species in Stictis and Schizoxylon [13,35]. When taking shape and septation of ascospores into consideration, there is no clear-cut boundary among these eight genera. However, they can be distinguished based on other features. For instance, all members of Biostictis are reported as parasites of living leaves of plants [13]. Karstenia is unique in that it forms covering layers that consist of short-celled, vertically oriented hyphae ending in a fringe of hair-like projections [13]. All members of Lillicoa inhabit living leaves and seemingly are not lichenized. Conversely, all members of Nanostictis are obligately lichenicolous [13]. Hence, features such as lifestyle, substrate preference, nature of apothecia/conidiomata, structure of exciple and ascospore/conidia morphology are of taxonomic significance either independently or in combination for specific taxa. Molecular data are needed to examine the value of these diagnostic characters in intrageneric and intergeneric classification.
Sexual-asexual connections have been established for only a few genera of Stictidaceae including Stictis and Acarosporina using cultures and molecular evidence [28,52,53]. In Schizoxylon pseudocyanosporum, the asexual conidiomata were linked to its sexual apothecia solely on the basis of their close proximity to each other [13,53]. The connection, however, has not been confirmed by using pure cultures or molecular data. Similarly, this study and Thiyagaraja et al. [20] reported the intermingled occurrence of both sexual and asexual morphs together on the same substrate. Although it could be speculated that the two morphs belong to the same species, this is not very likely. Molecular data derived from pure cultures and subsequent sequence and phylogenetic analyses (see notes, Figure 2) collectively suggest that the two morphs are in fact separate species. In the phylogenetic analysis, the sexual morph grouped with Ostropomyces pruinosellus, while the asexual morph clustered with O. thailandicus. Moreover, the interspecific genetic diversity of all genetic markers used in this study between O. pruinosellus and O. thailandicus clearly indicates that the two are not conspecifics. Thus, establishing sexual-asexual links should not be based solely on the close proximity of the two morphs, as in the case of S. pseudocyanosporum, but supplemented with molecular data.
In this study, we have explored the taxonomic significance of characters for Stictidaceae. However, limited taxa sampling in the phylogenetic analyses hinders us from drawing a firm conclusion on this point. Additionally, taxonomic placements of pathogenic/parasitic species in Stictidaceae are as yet unclear due to the lack of molecular data. It would be interesting to observe if the addition of these taxa will confirm the saprobic nature of the ancestor of this family. Stictis contains many species, which were introduced solely on the basis of morphological observation, so gaps remain in terms of its natural classification using moleculebased phylogeny. Therefore, additional collections especially of early branching and undersampled species are urgently needed in the future to address the issues mentioned above.