Morpho-Molecular Evidence Reveals Four Novel Species of Gymnopus (Agaricales, Omphalotaceae) from China

Nine collections of gymnopoid fungi were studied based on morpho-molecular characteristics. The macromorphology was made according to the photograph of fresh basidiomata and field notes, while the micromorphology was examined via an optical microscope. Simultaneously, the phylogenetic analyses were performed by maximum likelihood and Bayesian inference methods based on a combined dataset of nrITS1-nr5.8S-nrITS2-nrLSU sequences. Integrated analysis of these results was therefore, G. efibulatus belonging to sect. Androsacei, G. iodes and G. sinopolyphyllus belonging to sect. Impudicae and G. strigosipes belonging to sect. Levipedes are proposed as new to science. The detailed descriptions, colour photos of basidiomata and line-drawings of microscopic structures are provided. The comparisons with closely related species and a key to known species of Gymnopus s. str. reported with morpho-molecular evidence in China is also given.


Introduction
Gymnopus (Pers.) Roussel is a group of white-spored macrofungi with collybioid, rarely tricholomatoid, marasmioid or omphalinoid habit and distributed worldwide . Its habit, causing the similarity in macromorphology, is the main reason for the confusion in taxonomy in the past. When revisiting the history of the genus, it is clear that the materials from Holarctic regions, comprised of Palaearctic and Nearctic subregions, were well studied [1][2][3][4][5][6][7][8]23,24]. Morphological studies on samples from Europe and northern Africa, a part of the Palaearctic subregion, and a few samples from North America, belonging to the Nearctic subregion, dominated the major revision of the generic concept [1,2,[23][24][25]. However, the highly phenotypic plasticity that causes the close species often forms a complex, such as the G. dryophilus complex, making classification in the species level hard if only depends on morphology [26,27]. The taxonomic study was improved following the use of the polyphasic method. Studies on mating systems and isozymes were used to distinguish the G. dryophilus complex, confirming the morphological results [26,27]. The molecular-phylogenetic analysis plays a critical role in species recognition and generic revision as a tool for evolutionary inference. Based on the morphological study and multigene phylogeny, the concepts of G. dryophilus (Bull.) Murrill and G. ocior (Pers.) Antonín and Noordel. were clarified [4]. Additionally, several results inferred from this method inspired mycologists to reconsider the inter-and infrageneric taxonomic position [7,28,29].

Morphological Studies
Macromorphological features were described based on photographs of fresh basidiomata and field notes. The colours terminology and code follow Kornerup and Wanscher [36]. Two types of lamellae were counted, of which the number of full-length lamellae is represented by 'L', and the number of lamellulae tiers is represented by 'l'.
Tiny tissue was cut from the dried basidiomata using a sharp blade and then mounted in 5% KOH on a glass slide for microscopic observation via a light microscope (ZEISS Axioscope 5). When needed, the Congo Red solution was used to highlight the outline of microscopic structures, and Melzer's reagent was used to test amyloid or dextrinoid reactions. In microscopic description, 'n' refers to the number of measured elements. Measurements of basidiospores are given as (a)b-c(d), of which 'b-c' refers to the minimum of 90% of the measured values. The main body excluding sterigmata or excrescences of basidia, basidioles and cystidia were measured.

DNA Extraction, Amplification and Sequencing
The dried tissue was used for genomic DNA extraction using NuClean Plant Genomic DNA Kit (Cowin Biotech Co., Ltd., Taizhou, China). The nuclear ribosomal internal transcribed spacer (nrITS) and nuclear ribosomal large subunit (nrLSU) sequences were amplified using primer pairs of ITS5/ITS4 and LR0R/LR5, respectively [37][38][39]. The polymerase chain reaction (PCR) programs were followed according to Li et al. [18], and the PCR products were sent to Sangon Biotech (Changchun, Jilin, China) for sequencing.

Phylogenetic Analyses
Sequences generated in this study were combined with those obtained from Gen-Bank by Basic Local Alignment Search Tool (BLAST) and then added to the matrix used by Li et al. [18], which covered all the sections of Gymnopus s. str. Based on two overlapping reads, targeted sequences were assembled and trimmed via BioEdit v. 7.0.9 [40]. Quality control, such as degenerate bases checking, was completed before depositing the sequences to be used in GenBank (see Table 1) [41]. The ML (Maximum Likelihood) and BI (Bayesian Inference) analyses follow Li et al. [18]. The Bootstrap Proportions ≥70% for ML analysis (ML-BP) and Posterior Probability ≥0.95 for Bayesian Inference analysis (BI-PP) were considered a significant value. The alignment file of the combined dataset used for phylogenetic analyses and both the two phylograms generated from two methods was deposited in Treebase (https://www.treebase.org/treebase-web/home.html (accessed on 18 March 2022). Newly generated sequences are highlighted in bold and sequences derived from type specimen are marked with an asterisk (*). The haplotypes were deduced from forward and reverse sequences.

Results
A combined dataset of two markers, including 1712 bases, comprising 64 nrITS sequences and 56 nrLSU sequences, was used to calculate ML and BI analyses. Amongst the dataset, 1237 were constant sites, 128 were variable and parsimony-uninformative sites, and 347 (≈20.27%) were parsimony-informative sites. According to the BIC criterion, the GTR+F+I+G4, K2P, HKY+F+G4 and HKY+F+I+G4 models were selected as the best-fit model for the nrITS1 region, 5.8S marker, nrITS2 region and nrLSU gene, respectively. Only the tree with a better topology generated by the ML method was shown ( Figure 1).
In the newly produced phylogram, the clade Paragymnopus helps to separate the Gymnopus s. str., forming a monophyletic clade with high support (BI-PP/ML-BP = 1.00/100%). Sequences of the nine studied samples were restricted in the Gymnopus s. str. clade forming four distinct lineages. Amongst them, the lineage of the collections HGASMF01-11995 and HGASMF01-7052 is highly supported (BI-PP/ML-BP = 1.00/98%) and nested in an unsolved clade representing G. sect. Androsacei. Moreover, the lineage of the collections HGASMF01-10068, HGASMF01-10069 and HMJAU 60388 is highly supported (BI-PP/ML-BP = 1.00/100%) within a well-supported clade, namely, G. sect. Impudicae (BI-PP/ML-BP = 1.00/96%). Furthermore, the lineage of the collections HMAS 295796 and HMAS 295797 belongs to the clade of G. sect. Levipedes, and both are strongly supported by the two methods analyses (BI-PP/ML-BP = 1.00/100%). Besides, collections HMJAU 60386-60387 are clustered in the clade of G. sect. Impudicae, and their sequences only differ by having two degenerate bases but including the same nucleotide in the corresponding site.
pudicae (BI-PP/ML-BP = 1.00/96%). Furthermore, the lineage of the collectio 295796 and HMAS 295797 belongs to the clade of G. sect. Levipedes, and both a supported by the two methods analyses (BI-PP/ML-BP = 1.00/100%). Besides, HMJAU 60386-60387 are clustered in the clade of G. sect. Impudicae, and their only differ by having two degenerate bases but including the same nucleotide responding site.    Diagnosis: This species is characterized by its brownish grey to dark-brown pileus, whitish to yellowish white rhizomorphs and ellipsoid to oblong basidiospores 7-9 × (3.5-)4-4.5(- 5.5) µm in size.
Basidiospores: [n = 60] 7-9 × Notes: Morphologically, the pale-coloured stipe and rhizomorphs of G. efibulatus is strongly reminiscent of G. pallipes J.P. Li and Chun Y. Deng and G. cremeostipitatus Antonín, Ryoo and Ka in the field. However, both G. pallipes from China and G. cremeostipitatus from South Korea, with smaller basidiospores (6.06 × 3.24 µm, 7.1 × 3 µm resp.) and the latter one with a pubescent stipe and scattered-to-frequent caulocystidia helps distinguish them from the new species [11,17]. Additionally, the absence of clamp connections, which is not quite common in the genus, strongly supports that they are not conspecific [11,17]. What is noteworthy is a description of Marasmius aurantiobasalis Desjardin and E. Horak, a member of Marasmius sect. Androsacei Kühner (≡ Gymnopus sect. Androsacei), based on individuals from Indonesia shares indistinguishable morphological features with the new species [49]. After checking the original description, Ma. aurantiobasalis from New Zealand differs by having smaller basidiospores (6-7.5 × 3-3.5 µm) and a pileipellis of subhymeniform that is made up of clavate to irregular, sometimes lobed, cells with dense diverticula [50].
Phylogenetic analyses agree with the morphological study, which showed the new species is close to G. spongiosus. Horak are close to the new species. However, G. agricola, from the USA, can be distinguished by its estriate pileus margin, a cartilaginous and non-strigose stipe [5]; G. hybridus, from France, and G. sepiiconicus, from South Solomons, are characterized by the non-strigose stipe and the absence of caulocystidia [2,14]; G. vitellinipes, from Indonesia, has larger basidiospores (8.3-9.3 × 4-4.4 µm) and a poorly developed Dryophilastructure in the pileipellis [15]; G. loiseleurietorum, from Austria, differs by the absence of true cheilocystidia and hyphae turn green in KOH [2]; G. spongiosus, from the USA, has smaller basidiospores (8.4 × 3.6 µm) that often turn olive green in alkali [20].
Phylogenetic analyses agree with the morphological study, which showed the new species is close to G. spongiosus. When revisiting these four gymnopoid fungi in morphology, it is not hard to find that they share the striate pileus and the distant to subdistant lamellae [6,12]. Similarly, G. densilamellatus, G. polyphyllus and G. sinopolyphyllus, forming an independent clade, share the very close to crowded lamellae [6,12]. Besides, G. sect. Androsacei is still an unsolved clade thus far, phylogenetically. Formally, Li et al. discussed the sectional circumscription and noted the broom cells were absent or weakly present in the pileipellis of several taxa [17]. Furthermore, the pileipellis of G. efibulatus also lack the bloom cells but subglobose cells were observed. Accordingly, this section is worthy of further exploration in the future based on morphology inferred from more materials and multilocus phylogenetic analyses.