Two Novel Genera, Neostemphylium and Scleromyces (Pleosporaceae) from Freshwater Sediments and Their Global Biogeography

Although the Pleosporaceae is one of the species-richest families in the Pleosporales, research into less-explored substrates can contribute to widening the knowledge of its diversity. In our ongoing survey on culturable Ascomycota from freshwater sediments in Spain, several pleosporacean specimens of taxonomic interest were isolated. Phylogenetic analyses based on five gene markers (ITS, LSU, gapdh, rbp2, and tef1) revealed that these fungi represent so far undescribed lineages, which are proposed as two novel genera in the family, i.e., Neostemphylium typified by Neostemphylium polymorphum sp. nov., and Scleromyces to accommodate Scleromyces submersus sp. nov. Neostemphylium is characterized by the production of phaeodictyospores from apically swollen and darkened conidiogenous cells, the presence of a synanamorph that consists of cylindrical and brown phragmoconidia growing terminally or laterally on hyphae, and by the ability to produce secondary conidia by a microconidiation cycle. Scleromyces is placed phylogenetically distant to any genera in the family and only produces sclerotium-like structures in vitro. The geographic distribution and ecology of N. polymorphum and Sc. submersus were inferred from metabarcoding data using the GlobalFungi database. The results suggest that N. polymorphum is a globally distributed fungus represented by environmental sequences originating primarily from soil samples collected in Australia, Europe, and the USA, whereas Sc. submersus is a less common species that has only been found associated with one environmental sequence from an Australian soil sample. The phylogenetic analyses of the environmental ITS1 and ITS2 sequences revealed at least four dark taxa that might be related to Neostemphylium and Scleromyces. The phylogeny presented here allows us to resolve the taxonomy of the genus Asteromyces as a member of the Pleosporaceae.


Introduction
The Pleosporaceae is one of the largest families within the order Pleosporales (Dothideomycetes) in terms of the number of species. It was introduced by Nitschke [1] and was considered for a long time a heterogeneous group of bitunicate ascomycetes with its genera distinguished primarily by their ascospore features (i.e., shape, color, septation, pigmentation, and presence or lack of mucilaginous sheaths) [2]. According to recent taxonomic revisions of the Dothideomycetes [3,4], based on morphological investigations and phylogenetic data, the Pleosporaceae is a well-delineated family that comprises 23 genera and more than 2000 species. Alternaria, Bipolaris, Curvularia, Exserohilum, Pyrenophora, and Stemphylium are the most species-rich genera in the family [3]. They show teleomorphs characterized by black ostiolate ascomata with thick-walled peridium, cellular pseudoparaphyses and bitunicate, fissitunicate, eight-spored asci, producing melanized, phragmosporous, or muriform ascospores [3]. More commonly, they present dematiaceous hyphomycetous anamorphs producing phragmo-or dyctioconidia from tretric (poroblastic) or blastic conidiogenous cells, although coelomycetous anamorphs with phialidic or anellydic conidia have also been described [3,5]. Although Pleospora was designed as the type genus in the family, with the advent of one fungus-one name initiative in the International Code of Nomenclature for algae, fungi, and plants (ICN; Melbourne Code) [6], the name Stemphylium was retained over Pleospora by the working group on Dothideomycetes of the International Commission on the Taxonomy of Fungi [7].
Members of the Pleosporaceae are widely distributed across the environment and have a wide range of lifestyles, i.e., saprophytic, endo-/epiphytic, and parasitic on various hosts in terrestrial and aquatic environments [8]. Among them, species of Alternaria, Bipolaris, Curvularia, or Stemphylium are important pathogenic fungi to plants of various crops, resulting in yield and economic losses [8][9][10]. However, they also include human and animal pathogens that cause infections with different clinical manifestations [11]. Several metagenomic studies reveal that pleosporalean fungi are well represented in aquatic environments [12][13][14]. Although some species have been found strictly adapted to aquatic ecosystems [15], many of them are commonly found in association with terrestrial plants and, therefore, they are not considered especially adapted to freshwater habitats [16]. Results of those studies also suggest that this group of fungi has been generally overlooked and undersampled in freshwater ecosystems, particularly in rivers, despite their relevant role in ecosystem functioning as saprophytes and parasites [12,17].
In our latest efforts to expand knowledge on the diversity of culturable Ascomycota from river sediments collected in Spain, several interesting specimens of dematiaceous filamentous fungi were isolated. A preliminary sequence analysis of the nuclear ribosomal operon (i.e., the 28S large ribosomal subunit--LSU, and the internal transcribed spacer--ITS, including the 5.8S rDNA gene) revealed that those specimens would belong to the Pleosporaceae, but they could not be identified at the genus level. The aim of the present study was, therefore, to resolve the taxonomy of the above-mentioned isolates based on morphological features and multi-locus phylogenetic analysis inferred with sequences of the nuclear markers mostly represented in the different members of the Pleosporaceae. These are the LSU and ITS regions of the rDNA, and partial fragments of the RNA polymerase II largest subunit (rpb2), the translation elongation factor 1-α (tef 1), and the glyceraldehyde-3-phosphate dehydrogenase (gapdh) genes [8,18,19]. Additionally, in order to elucidate the putative global geographic distribution of those isolates and to study their diversity hidden among environmental sequences, their ITS barcodes (i.e., full-length of ITS1 and ITS2 sequences) were blasted against the GlobalFungi database [20]. This is a recently created database, which currently includes accumulated data on fungal distribution and ecology generated from more than three hundred metagenomic studies published in the last decade (GlobalFungi database, accessed on 18 May 2022).

Sampling and Isolates
Sediment samples were collected in 2019 from natural areas of two Spanish provinces, Lleida and Madrid. Samples from Lleida were collected from the Segre River as it passes through Camarasa, an area characterized by a continental Mediterranean climate (https: //www.meteo.cat/wpweb/climatologia/el-clima-ahir/el-clima-de-catalunya/, accessed on 21 April 2022), with an average annual temperature of 13.5 • C, an average annual rainfall of 800 mm, an altitude of 800 m, and a vegetation dominated by holm oaks (Quercus ilex subsp. rotundifolia) (http://www.biodiver.bio.ub.es, accessed on 21 April 2022). Samples from Madrid were collected from two streams around Rascafría in the Guadarrama Natural Park. This area has a continental mountain climate with an average annual temperature of 11.8 • C, an average annual rainfall of 530 mm, an altitude of 1200 m, and a forest dominated by Cistus oromediterraneus, Juniperus communis, and Pinus sylvestris (https:// www.parquenacionalsierra guadarrama.es, accessed on 21 April 2022).
Sediments from the rivers or streams selected in the above-mentioned locations were collected randomly. Samples were obtained ca 10 cm below the surface layer from the riverbeds or edges using sterile 100 mL plastic containers, which were transported in a refrigerated container to the laboratory and processed immediately. Samples were vigorously shacked in the same containers; then, after 1 min at rest, the water was decanted and the sediment was poured into plastic trays onto several layers of sterile filter paper to remove excess water [21]. To achieve a greater fungal diversity in culture, three agar media were used: dichloran rose-bengal-chloramphenicol agar (DRBC; 2.5 g peptone, 5 g glucose, 0.5 g KH 2 PO 4 , 0.25 g MgSO 4 , 12.5 mg rose-bengal, 100 mg chloramphenicol, 1 mg dichloran, 10 g agar, 500 mL distilled water), DRBC supplemented with 0.01 g/L of benomyl, and potato dextrose agar (PDA; Pronadisa) supplemented with 2 g/L of chloramphenicol and 2 g/L of cycloheximide. Each sample was cultured in duplicate in each medium as follow: 0.5 g of sediment was mixed with melted medium at 45 • C in the same Petri dish and, once solidified, it was incubated at room temperature (22-25 • C) in the dark. Plates were examined weekly by stereomicroscope for 4-5 weeks. To obtain pure cultures, fragments of the colony or conidia of the fungi growing on primary cultures were transferred, using a sterile dissection needle, to plates containing PDA supplemented with chloramphenicol and incubated at 25 • C in darkness. These PDA cultures were used for a preliminary morphological identification and for extracting DNA of the fungi selected.
Living cultures of putative novel or rare fungi were preserved and deposited in the culture collection of the Faculty of Medicine in Reus (FMR, Spain) for further studies. Taxonomic information and nomenclature for the new species were deposited in MycoBank (https://www.mycobank.org/, accessed on 23 March 2022). Cultures from ex-type strains and holotypes, which consisted of dry colonies on the most appropriate media for their sporulation, were also deposited at the Westerdijk Fungal Biodiversity Institute in Utrecht (CBS, The Netherlands) (https://wi.knaw.nl/, accessed on 22 May 2022).
In addition, the ex-type and a reference strain of Asteromyces cruciatus were also examined in the current study, because a preliminary molecular comparison revealed this species as related to some of our isolates. According to Mycobank and the Index Fungorum database, A. cruciatus represents a monotypic genus with unclarified taxonomy.

Phenotypic Study
Microscopic characterization was carried out from the isolates growing on potato carrot agar (PCA; 20 g potato, 20 g carrot, 13 g agar, 1 L distilled water) after 7-14 d at 25 • C in darkness and mounted on slides with Shear's mounting solution (3 g potassium acetate, 60 mL glycerol, 90 mL ethanol 95%, and 150 mL distilled water) [22], using an Olympus BH-2 bright field microscope (Olympus Corporation, Tokyo, Japan). Size ranges of relevant structures in species descriptions were derived from at least 30 measurements. Micrographs were taken using a Zeiss Axio-Imager M1 light microscope (Zeiss, Oberkochen, Germany) with a DeltaPix Infinity × digital camera. Photoplates were assembled from separate photographs using PhotoShop CS6. Macroscopic characterization of the colonies was made on PDA, PCA and oatmeal agar (OA; 30 g oatmeal, 13 g agar, 1 L distilled water) after 7 days at 25 • C in darkness. Other culture media, such as OA and PCA with sterile plant debris (i.e., leaves and twigs of Dianthus caryophyllus), synthetic nutrient-poor agar (SNA; 1 g KH 2 PO 4 , 1 g KNO 3 , 0.5 g MgSO 4 × 7H 2 O, 0.5 g KCl, 0.2 g glucose, 0.2 g sucrose, 14 g agar, 1 L of distilled water), and V8 medium (16 g agar, 200 mL V8 juice, 1 L distilled water), were also used specifically for the FMR 18289 in order to stimulate its sporulation. For the same purpose, this isolate was submitted to the procedure described in Nishikawa and Nakashima [23] for Alternaria sporulation. Color notations in descriptions were according to Kornerup and Wanscher [24]. Growth rates were measured in duplicate on PDA after 7 d in darkness, at 5 • C intervals from 5 to 40 • C, and also at 37 • C.

DNA Extraction, Sequencing and Phylogenetic Analysis
Total genomic DNA was extracted through the modified protocol of Müller et al. [25] and quantified using Nanodrop 2000 (Thermo Scientific, Madrid, Spain). In order to reconstruct the phylogeny of the Pleosporaceae family, the loci amplified and sequenced were the ITS barcode and the D1/D2 domains of the LSU of the rDNA, as well as gene fragments of the rpb2, tef 1, and gapdh. Primers pairs for their amplification were ITS5/ITS4 [26], LR0R/LR5 [27], RPB2-5F2/fRPB2-7cR [28,29], EF1-728F/EF1-986R [30], and gpd1/gpd2 [31], respectively. Briefly, PCR conditions for ITS, LSU, gapdh, and tef 1 were set as follows: an initial denaturation at 95 • C for 5 min, followed by 35 cycles of 30 s at 95 • C, 45 s at 56 • C, and 1 min at 72 • C, and a final extension step at 72 • C for 10 min. PCR conditions for the rpb2 were an initial denaturation at 94 • C for 5 min, followed by 5 cycles of 45 s at 94 • C, 45 s at 60 • C, and 2 min at 72 • C, then 5 cycles of 45 s at 94 • C, 45 s with 58 • C, and 2 min at 72 • C, later 30 cycles of 45 s at 95 • C, 45 s with 54 • C, and 2 min at 72 • C, and a final extension step at 72 • C for 7 min. PCR products were purified and sequenced at Macrogen Corp. Europe (Madrid, Spain) with the same primers used for amplification. Consensus sequences were assembled using SeqMan v. 7.0.0 (DNAStar Lasergene, Madison, WI, USA).
A preliminary species identification of the sediment isolates was carried out by comparing their ITS region with those at the National Center for Biotechnology Information (NCBI) using the Basic Local Alignment Search Tool (BLAST; https://blast.ncbi.nlm. nih.gov/ Blast.cgi, accessed on 5 March 2021) and with the UNITE database (https://unite.ut.ee/, accessed on 5 March 2021). A maximum similarity level of ≥98% was used for species-level identification. Lower similarity values were considered as putative unknown fungi, and their taxonomic position was assessed from analyses of the loci mentioned above.
Sequences of related species and representatives of other genera belonging to the Pleosporaceae family were obtained from GenBank and are listed in Table 1. Individual and combined analyses using the LSU, ITS, gapdh, rpb2, and tef 1 sequences were carried out to assess the phylogenetic relationship of the unidentified isolates to the other taxa in the family. Datasets for each locus were aligned individually in MEGA (Molecular Evolutionary Genetics Analysis) software v.6.0 [46], using the ClustalW algorithm [47] and refined with MUSCLE [48] or manually adjusted, if necessary, on the same platform. Phylogenetic concordance of the five-locus datasets was tested individually in each single-locus phylogeny through visual comparison and using the Incongruence Length Difference (ILD) implemented in the Winclada program [49] in order to assess any incongruent results among nodes with high statistical support. Once their concordance was confirmed, individual alignments were concatenated into a single data matrix with SequenceMatrix [50]. The best substitution model for all gene matrices was estimated using MEGA software for Maximum Likelihood (ML) analysis, whereas for the Bayesian Inference (BI) analysis it was estimated using jModelTest v.2.1.3 following the Akaike criterion [51,52]. The phylogenetic reconstructions were performed with the combined genes using ML under RAxML-HPC2 on XSEDE v-8.2.12 [53] in CIPRES Science gateway portal [54] and BI with MrBayes v.3.2.6 [55].
For the ML analysis, phylogenetic support for internal branches was assessed by 1000 ML bootstrapped pseudoreplicates and bootstrap support (bs) ≥ 70 was considered significant [56]. The phylogenetic reconstruction by BI was carried out using 5 million Markov Chain Monte Carlo (MCMC) generations, with four runs (one cold chain and three heated chains), and samples were stored every 1000 generations. The 50% majority-rule consensus tree and posterior probability (pp) values were calculated after discarding the first 25% of samples. A pp value of ≥0.95 was considered significant [57]. The resulting trees were plotted using FigTree v.1.3.1 (http://tree.bio.ed.ac.uk/software/figtree/, accessed on 1 June 2022). The DNA sequences and alignments generated in this study were, respectively, deposited in GenBank (Table 1) and in Zenodo (https://doi.org/10.5281/zenodo.6973696, accessed on 9 August 2022).

Phylogeny and Geographic Distribution of Allied Environmental Sequences
In order to assess putative global geographic distribution and ecology of the novel fungi detected and their hidden diversity among environmental sequences, the full length of their ITS1 and ITS2 sequences were blasted against the GlobalFungi database [20]. At the time of accession (May 2022), this dataset contained 36,684 samples from 367 studies, 213, 747, 241 unique sequences for ITS1, and 582, 264, 149 for ITS2. Since GlobalFungi has separated ITS1 and ITS2 sequences, they were analyzed separately. In order to verify generic and species boundaries among downloaded ITS environmental sequences related to our fungi, we also included in the analyses ITS sequences of known species previously obtained from the GenBank and UNITE databases. Those known species were representatives of the well-delineated monophyletic genera (i.e., Alternaria, Asteromyces, Gibbago, Paradendryphyella, Pyrenophora, and Stemphylium), which were the closest taxa to our fungi in a full length ITS analysis carried out previously ( Figure S1 in Supplementary Material). ITS1 and ITS2 sequences of some of those fungi were also blasted against the GlobalFungi dataset and included in the respective analyses. In each case, we selected and downloaded environmental sequences that had a similarity of between 98 and 100% and a full-length coverage with the sediment isolates and with those related to the known species, apart from the ITS1 of the isolate FMR 18289, because its highest sequence similarity found in the database was lower, at 95%. Pleosporacean genus/species boundaries were inferred from ML trees of ITS1 and ITS2 sequences computed in RAxML. Virtual taxa, consisting of environmental sequences only, were defined as arbitrary phylotypes in the phylogenetic trees, following Réblová et al. [58,59]. Data on occurrence across environmental samples and metadata related to the particular samples (location, substrate, biome, or climatic data) were obtained for each taxon and are listed in Table S1 (Supplementary Material).

Results
Among pleosporacean fungi found in the freshwater sediments, we recovered five isolates (FMR 17886, FMR 17889, FMR 17893, FMR 17894, FMR 17895) exclusively from DRBC agar supplemented with benomyl. These isolates were identified initially as Stemphylium sp. because they showed similar morphological features but did not exactly fit into any of the known species described in that genus. Another interesting isolate (FMR 18289) was recovered from DRBC but could not be identified morphologically because it only produced sclerotium-like structures and failed to form fertile reproductive morphs (i.e., anamorph and/or teleomorph) despite the attempts to stimulate sporulation in various culture conditions.

Phylogeny
Molecular identification based on the BLAST query revealed that LSU sequences of the six unidentified isolates showed a high percentage of similarity with other members of the Pleosporaceae. Specifically, the stemphylium-like isolates showed a sequence identity of 99% with Stemphylium (S.) vesicarium (CBS 191.86) and Bipolaris (B.) microlaenae (CBS 280.91), while the sequence of the sclerotium-forming isolate was 99% similar to Pyrenophora (P.) seminiperda (CBS 127927) and 98% to Alternaria (A.) avenicola (CBS 121459). Similar values were obtained when sequences of species of other well-delineated genera in the Pleosporaceae were compared, which confirmed the low discriminatory power of this gene marker in the family. On the other hand, the genetic similarity was considerably lower when ITS sequences were compared with other members of the Pleosporaceae. The closest matches for the stemphylium-like isolates were Paradendryphiella (Pa.) salina (CBS 142.60 and CBS 141.60) with a similarity of 96%, followed by Pa. areniae (CBS 181.58) and S. vesicarium (CBS 191.86) with a 95%. BLAST results and the particular morphology of those isolates precluded them from being classified in the genus Stemphylium or in Paradendryphiella. The highest similarity for ITS sequence of the remaining isolate (FMR 18289) was 96% with A. avenicola (CBS 121459), followed by P. seminiperda (CBS 127927) with a similarity of 90%. BLAST searches using the remaining phylogenetic markers revealed even lower values of similarity (≤89.6%) with other members of the Pleosporaceae.
Since individual analyses with LSU, ITS, gapdh, rpb2, and tef 1 were visually similar and the ILD test did not show incongruences (p = 0.33), a multi-gene analysis was carried out with the five markers. The concatenated phylogeny encompassed 59 sequences that represented 17 genera in the Pleosporaceae with 3160 bp long (531 for ITS, 892 for LSU, 865 for rpb2, 624 for gapdh, 248 for tef 1), of which 1164 were variable sites (198 for ITS, 168 for LSU, 375 for rpb2, 281 for gapdh, 142 for tef 1) and 905 were phylogenetically informative sites (159 for ITS, 94 for LSU, 325 for rpb2, 231 for gapdh, 96 for tef 1). For the ML analyses, K2 + G + I was selected as the best fit model for ITS, LSU, and rpb2, the K2 + G for tef 1 and TN93 + G for gapdh. For the BI analyses, SYM + G + I was selected as the best fit model for ITS and rpb2, the K2 + G + I for LSU, the HKY + G + I for gapdh, and the K2 + G for tef 1. The RAxML tree (Figure 1) showed that FMR 17886, FMR 17889, FMR 17893, FMR 17894, and FMR 17895 clustered together in a monophyletic undescribed lineage, strongly supported (100 bs/1 pp), which was sister to a well-supported clade (83 bs/0.98 pp) that includes members of the genera Asteromyces, Paradendryphiella, and Stemphylium. Among these, Asteromyces currently has an uncertain taxonomic position among the Ascomycota. However, our analysis places its type species, A. cruciatus, into the Pleosporaceae. The undescribed lineage of the five sediment isolates represents a novel genus, which is proposed here as Neostemphylium (N.) and represented by the new species N. polymorphum (Figure 1). These isolates showed similar morphological features and had an intra-specific genetic variability ranging from 0.1 to 0.4% in the concatenated phylogenetic analysis. In the same phylogeny (Figure 1), the isolate FMR 18289 was located in a faraway single branch within a well-supported clade (99 bs/1 pp) together with other genera of the Pleosporaceae, representing a novel genus for the family. This is proposed below as Scleromyces (Sc.) and typified by the new species Sc. submersus. A detailed morphological characterization of the novel fungi is provided in the taxonomy section.
our analysis places its type species, A. cruciatus, into the Pleosporaceae. The undescribed lineage of the five sediment isolates represents a novel genus, which is proposed here as Neostemphylium (N.) and represented by the new species N. polymorphum (Figure 1). These isolates showed similar morphological features and had an intra-specific genetic variability ranging from 0.1 to 0.4% in the concatenated phylogenetic analysis. In the same phylogeny (Figure 1), the isolate FMR 18289 was located in a faraway single branch within a well-supported clade (99 bs/1 pp) together with other genera of the Pleosporaceae, representing a novel genus for the family. This is proposed below as Scleromyces (Sc.) and typified by the new species Sc. submersus. A detailed morphological characterization of the novel fungi is provided in the taxonomy section.

Biogeography and Ecology
A BLAST search in the GlobalFungi database revealed the presence of Neostemphylium and Scleromyces among environmental sequences from samples collected worldwide. When we compared the ITS1 sequences of Neostemphylium, this resulted in 469 unique environmental ITS1 sequences (similarity 98-100%), covering 739 samples. The ITS1 sequence of the Scleromyces isolate, as mentioned before, yield the highest similarity value, that of 95%, found in the database .At that value, we obtained 500 environmental sequences, which covered 291 samples. With so many environmental sequences related to our fungi, we were able to select for the analyses representatives from a variety of locations, substrates, and biomes (Table S1 in Supplementary Material), as well as other environmental sequences from different species of the genera Alternaria, Pyrenophora, and Stemphylium. The ITS1 phylogenetic analysis included 102 sequences, 225 characters, of which 156 were variable sites and 127 were phylogenetically informative sites. The ML tree was rooted in a branch leading to Comoclathris (Co.) typhicola (CBS 132.69) and Co. sedis (CBS 366.52) (Figure 2). The environmental ITS1 sequences selected clustered into eight phylotypes, two of which, with a total of 32 sequences, were related to the genus Neostemphylium, and another two phylotypes with three sequences were related to Scleromyces. phylium and Scleromyces among environmental sequences from samples collected worldwide. When we compared the ITS1 sequences of Neostemphylium, this resulted in 469 unique environmental ITS1 sequences (similarity 98-100%), covering 739 samples. The ITS1 sequence of the Scleromyces isolate, as mentioned before, yield the highest similarity value, that of 95%, found in the database .At that value, we obtained 500 environmental sequences, which covered 291 samples. With so many environmental sequences related to our fungi, we were able to select for the analyses representatives from a variety of locations, substrates, and biomes (Table S1 in Supplementary Material), as well as other environmental sequences from different species of the genera Alternaria, Pyrenophora, and Stemphylium. The ITS1 phylogenetic analysis included 102 sequences, 225 characters, of which 156 were variable sites and 127 were phylogenetically informative sites. The ML tree was rooted in a branch leading to Comoclathris (Co.) typhicola (CBS 132.69) and Co. sedis (CBS 366.52) (Figure 2). The environmental ITS1 sequences selected clustered into eight phylotypes, two of which, with a total of 32 sequences, were related to the genus Neostemphylium, and another two phylotypes with three sequences were related to Scleromyces. Most environmental sequences linked to Neostemphylium formed the phylotype representative of N. polymorphum, apart from four sequences that were designated as the phylotype ITS1-ENV1 and could represent a hypothetically distinct species from N. polymorphum. On the other hand, no environmental ITS1 sequences were matched to the novel species Sc. submersus. However, this species formed a divergent branch close to the phylo-types designated as ITS1-ENV2 and ITS1-ENV3, with one and two sequences, respectively, which might also represent two hidden taxa for the genus Scleromyces.
In contrast, more environmental ITS2 sequences were found related to the genus Scleromyces than to Neostemphylium. Namely, 26 sequences 98-100% similar were linked to Scleromyces and eight unique sequences to the latter genus, covering six and nine samples, respectively. The ITS2 dereplicated dataset had 92 sequences that were representative of members of the above-mentioned genera, with 166 characters, of which 90 were variable sites and 73 phylogenetically informative sites. The ML tree was rooted to Co. typhicola (CBS 132.69) and Co. sedis (CBS 366.52) (Figure 2). The environmental sequences were distributed into seven phylotypes. Specifically, eight sequences were linked to the phylotype of N. polymorphum and one to Sc. submersus, while the remaining sequences related to Scleromyces were designated as phylotype ITS2-ENV1 because they represented a distinct species from Sc. submersus. However, this Scleromyces phylotype does not correlate with any delineated in the ITS1 analysis, since ITS2-ENV1 includes environmental sequences from the USA, while ITS1-ENV2 and ITS1-ENV3 phylotypes both have sequences from Australian samples.
Biogeography and ecological parameters of the environmental sequences related to our novel fungi and inferred in ITS1 and ITS2 phylogenetic analyses are summarized in Table 2. Briefly, Oceania (mainly Australia) has the majority of samples containing ITS1 and ITS2 sequences linked to N. polymorphum, Sc. submersus, and the hidden phylotypes identified here. Nevertheless, many sequences linked to N. polymorphum were also found in samples from areas of Europe (France and Spain) and the USA, the most-sampled areas in GlobalFungi (33.55% and 23.74%, respectively). Interestingly, the Scleromyces phylotype ITS2-ENV1 is the only sequence sampled from aquatic environments collected in the USA (Table 2). Conversely, sequences linked to N. polymorphum, Sc. submersus, and to the other phylotypes related to the novel genera were sampled from soils or roots as the most frequently inhabited substrates in different biomes (grasslands, wetlands, croplands, woodlands, shrublands, or, rarely, forests) ( Table 2).
Habitat and geographical distribution: In addition to our freshwater sediment isolates from Spain, the environmental data suggest that members of Neostemphylium would primarily inhabit soil, but also air, rhizosphere soils, roots, and shoots from areas of Australia, Europe (France), and the USA (Figure 2).
Cardinal temperatures for growth: minimum 5 Distribution: Australia, France, Spain, and the USA (Figure 2, Table 2). Notes: The multi-gene phylogeny of the Pleosporaceae presented here shows that N. polymorphum is related to the genera Asteromyces, Paradendryphiella, and Stemphylium (Figure 1). However, it is not only placed distant from the clade representative of these three genera, but Neostemphylium also differs in several diagnostic morphological features. Although Neostemphylium and Stemphylium resemble each other in their anamorphs characterized by the formation of phaeodictyospores from apically swollen conidiogenous cells, the conidiophore branching pattern is more complex in Neostemphylium than in Stemphylium species. Conidiophores in the latter genus are commonly unbranched or rarely branched [9]. In addition, Neostemphylium produces a synanamorphic state characterized by blastic, brown phragmoconidia, sometimes branched, that are not reported in any species of Stemphylium. Paradendryphiella differs in the production of exclusively cylindrical to obclavate phragmoconidia with dark septa on narrow denticles, often aggregated at the apex of lateral or terminal conidiogenous cells [18], while the conidiogenous apparatus of Asteromyces is characterized by polyblastic swollen conidiogenous cells with long denticles in a radial arrangement, giving rise to one-celled dark brown conidia [60].
Gibbago is another pleosporacean genus, represented by G. trianthemae, which also resembles Neostemphylium in its conidiogenous cells and conidial morphology [38,61]. However, like in Stemphylium species, Gibbago produces mostly unbranched or rarely branched conidiophores, and no synanamorph or microconidiation cycle have been described in G. trianthemae. In addition, our phylogeny agrees with the Pleosporaceae phylogeny presented by Pem et al. [38], placing the genus Gibbago in a fully supported clade related to Exserohilum, which are both placed far from the new genus proposed here.
Scleromyces Torres-Garcia, Dania García and Gené, gen nov. MycoBank: MB 843291 Etymology: Name refers to the production of only sclerotium-like structures in in vitro conditions.
Habitat and geographical distribution: Aside from our freshwater sediment isolate from Spain, the environmental metadata suggest that members of Scleromyces would inhabit temperate climate areas (Australia and USA), colonizing primarily soils but also it can be found associated with plant material (roots and shoots) ( Figure 2, Table 2).

Discussion
The study of underexplored substrates can contribute to widening the knowledge of the Pleosporaceae diversity and, subsequently, to filling gaps in phylogenetic relationships among its taxa. In the present study, we describe two novel genera for the family, Neostemphylium and Scleromyces, sampled from Spanish freshwater sediments and cultured in vitro using the semi-selective medium DRBC. The efficacy of this medium to isolate pleosporacean fungi, such as Alternaria, Bipolaris, or Curvularia, was previously reported by Funnel-Harris et al. [64]. However, our study is the first to report that DRBC supplemented with benomyl can also be effective for culturing pleosporacean fungi of taxonomic interest, since all isolates of Neostemphylium were recovered from the latter medium. Anyway, its efficacy is well known to isolate ascomycetes from other groups, such as Microascales of clinical interest like Lomentospora or Scedosporium species [65,66].
The multi-locus phylogenetic analysis has been crucial for delimiting the novel fungi because of the resemblance of Neostemphylium to other genera, such as Stemphylium or Gibbago, and in the case of Scleromyces due to the absence of strictly sporulating structures. Neostemphylium shares with Stemphylium and Gibbago the production of phaeodictyospores from apically somewhat swollen and darkened conidiogenous cells [8,9]. However, it differs in the development of a synanamorph, which consists of blastic, cylindrical phragmoconidia, occasionally branched, that arise laterally or terminally on vegetative hyphae, and in the production of a microconidiation cycle not described in Gibbago or in Stemphylium. Similar structures to the N. polymorphum synanamorph have been described in the two species of the genus Berkeleiomyces (Be.), Be. bassicola and Be. rouxiae, although they were defined as septate chlamydospores [67]. Those fungi, however, belong to the microascaceous family Ceratocystidaceae and are phytopathogens to a wide range of plant hosts [67,68]. Conversely, the ability to produce a microconidiation cycle is known in the Pleosporaceae, since it has been reported in different species of Bipolaris and Curvularia [33,69], but the biological role of this state remains obscure. The microconidiation cycle observed in N. polymorphum is different from other genera because its mature secondary conidia resemble the primary ones ( Figure 3N-Q), that is, they become dark brown dictyoconidia. Secondary conidia described in Bipolaris and Curvularia are small, globose, and usually one-celled [33,70].
Despite being limited to form sclerotium-like structures, Scleromyces is phylogenetically distinct from other genera in the Pleosporaceae, at least from those with available DNA sequence data (Figure 1). According to Hongsanan et al. [3], cultures for some accepted genera in the family (i.e., Allonecte, Diademosa, Extrawettsteinina, Platysporoides, Pleoseptum, Prathoda, and Pseudoyuconia) are not available for comparison and lack DNA sequence data for confirming their classification. However, most of them were described as associated with plant material, producing only the teleomorph and placed in the family according to their morphological features [3,4]. As mentioned before, there are other pleosporalean fungi, such as A. slovaca [18,63] or P. pseudoerythrospila [42], which only produce sclerotia or chlamydospores and have been distinguished from other members of the genus exclusively by molecular data. Another example in the Pleosporales is the recently described monotypic genus Gambiomyces, which has been delimited according to the phylogeny of LSU, ITS, rpb2, and tef 1 and erected to accommodate the sterile fungus G. profunda, isolated from clinical specimens of a Gambian patient [71]. Examples from other fungal groups include the chaetothyrialean soil-inhabiting species Cyphellophora chlamydospora, which was described as producing only chlamydospores [72], and the species of the xylarialean endophytic genus Muscodor, which have been described as producing only sterile mycelia [73,74]. However, all of them have been clearly distinguished from their counterparts by their phylogeny, giving rise the possibility of naming relevant fungi like the species of Muscodor, which are important producers of volatile organic compounds with a wide range of potential applications in agriculture, medicine, and other sectors [74].
A huge number of unidentified environmental fungal sequences have been generated in the last decade by numerous metagenomic studies, with relevant information on ecology and distribution. One way to resolve their identification is currently to attempt to link them to sequences of known and well-established species [58,[75][76][77][78]. In this context, following the recent studies on the phylogeny and global distribution of Zanclospora and Codinaea [58,59], we traced the novel species N. polymorphum and Sc. submersus in the GlobalFungi database [20] to explore their putative geographical distribution as well as to detect hypothetical hidden taxa related. Our results revealed that N. polymorphum is a more common worldwide fungus than Sc. submersus, since its sequences can be linked to a large set of environmental ITS1 sequences from samples collected in Australia, Europe (France and Spain), and the USA (Figure 2). This distribution is not surprising since those are the most-sampled areas given in the GlobalFungi database (7.46%, 33.55%, and 23.74%, respectively, at the time of accession). On the other hand, Sc. submersus could be defined as a rarer or more geographically restricted fungus because only one ITS2 environmental sequence, originating from Australia, matched this species. Although N. polymorphum and Sc. submersus were discovered in freshwater sediment samples, we might assume that they are more likely terrestrial fungi given that the majority of environmental sequences they were linked to originated from soils of different terrestrial biomes in regions with humid and temperate climates ( Table 2). In fact, fungal communities that colonize terrestrial substrates like plants, soils, rocks, etc., will end up in the river sediments by lixiviation, where they can accumulate and survive under water conditions. That adds further support to river sediments as a suitable substrate for isolating a great diversity of fungi, including putative novel taxa.
Interestingly, our metabarcode analysis allowed us to detect four hidden phylotypes or "dark taxa", defined by Lücking et al. [78] as "new lineages known from sequence data only but for which no individual voucher specimens or cultures exist". Namely, one was related to the genus Neostemphylium (ITS1-ENV1) and three to Scleromyces (ITS1-ENV2, ITS1-ENV3, ITS2-ENV1) ( Figure 2). However, none of them were represented by fulllength ITS sequences as we did not find any correlation among the metadata from such phylotypes. That was in contrast to Réblová et al. [58], who obtained three whole ITS sequences among Zanclospora phylotypes, which were attributable to their geography and ecology data overlapping among various phylotypes found. We hope that all those "dark taxa" can be formally proposed soon following some of the options proposed by Lücking et al. [78] for naming fungi known only from environmental sequences.
Finally, our phylogenetic analysis has not only contributed to delineating two new genera but has also allowed us to confirm the taxonomic position of the monotypic genus Asteromyces [60] in the family Pleosporaceae, increasing to 26 the number of the accepted genera since its last review [3,4]. According to Mycobank, it was a genus classified in the family Dematiaceae (Helotiales) and in the Index Fungorum database as incertae sedis. We examined the morphology of the type species of the genus A. cruciatus and completed sequence data for its ex-type strain CBS 171.63 and for the reference strain CBS 536.92 (Table 1), confirming its particular features [60] and its relationship with other members of the family, such as Paradendryphyella and Stemphylium (Figure 1).

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jof8080868/s1. Figure S1: RAxML tree of the Pleosporaceae family from ITS, including the strains recovered from freshwater sediments studied in this work. The strains studied in this work are in bold and in red in the obtained tree. Determined by MEGA software v.6, the best nucleotide substitution model for ML analysis was K2 + G + I. The aligned data set was 531 bp long, with 217 variable sites and 198 phylogenetically informative. Branch lengths are proportional to phylogenetic distance. Bootstrap support values above 50% are indicated on the nodes. The tree is rooted to Neocamarosporium chichastianum CBS 137502 and Neocamarosporium goegapense CPC 23676. T = Ex-type strain. Table S1: Environmental and biogeographical information contained in all ITS1/ITS2 sequences downloaded from the GlobalFungi database included in our analysis (see Figure 2).