Phylogeny of Crepidotus applanatus Look-Alikes Reveals a Convergent Morphology Evolution and a New Species C. pini

Crepidotus applanatus is known as a common wood inhabiting fungus widely distributed throughout the Northern Hemisphere. There have been contrasting opinions about the delimitation and taxonomic treatment of the similar species C. malachius. Our phylogeny did not support the close relationship of these two morphologically similar species and the grouping of collections labelled by both names within each phylogenetic species reflects unreliable species delimitations in the traditional literatures. Both species inhabit the wood of deciduous trees, and our morphological analysis identified the size of basidiospores as a significant difference between them. The collections from Pinus sylvestris are recognised as a new species, C. pini sp. nov., and its morphological identification requires a combination of both basidiospore and cheilocystidia characters.


Introduction
Fungi of the genus Crepidotus (Fr.) Staude (Crepidotaceae, Agaricales, Agaricomycotina) have small to medium-sized pileate basidiomata with a typically reduced or absent stipe, a lamellate hymenophore, and a brown spore print. They grow on various dead plant substrates-mainly on wood, less on soil, and exceptionally on fruiting bodies of other fungi. They are circumpolarly widespread and relatively common from tropical to hemiboreal regions [1][2][3][4][5]. Kirk et al. [6] reported about 200 species described worldwide before 2008 and 55 new species have been published more recently (https://www.mycobank.org, accessed 28 April 2022).
The recent phylogenetic studies have confirmed that all four species form independent species lineages, but the morphological concept of C. applanatus and C. malachius is ambiguous (the clade of C. malachius contained specimens identified as C. applanatus) [12,13]. Some older morphology-based studies recognized these two species [2,4], while others synonymized them [1,14] or did not mention the name C. malachius at all [3,15,16].

Phylogenetic Analyses
The ITS sequences with 93% similarity or higher to any of the studied taxa and the LSU with 97.5% similarity were retrieved from GenBank (Supplementary Table S2). The ITS alignment is also supplemented by sequences published by Jančovičová et al. [20] and by sequences of selected European species with non-globose basidiospores retrieved from Unite (https://unite.ut.ee, accessed on 14 April 2022) and GenBank (www.ncbi.nlm.nih. gov/genbank, accessed on 14 April 2022). The LSU alignment is supported by sequences from Kumar et al. [21]. Raw sequence files were edited in Geneious version R10 [22]. Intraindividual polymorphic sites having more than one signal were marked with NC-IUPAC ambiguity codes. Several important samples and lineages retrieved from GenBank were represented only by one sequence, therefore the ITS and LSU alignments were analysed separately. The final alignments were aligned by MAFFT version 7 using the strategy E-INS-i [23] and were manually improved in Geneious version R10 [20]. The hypervariable regions were removed by gBlocks [24] using settings for less stringent selection. The final datasets were analysed with two different methods: Maximum Likelihood (ML) and Bayesian inference (BI). Maximum likelihood was computed at an IQ-tree web server [25] under GTR+I settings with 1000 ultrafast bootstrap replicates. The aligned fasta files were converted to the nexus format using Mesquite 3.61 [26] and further analysed using MrBayes 3.2.6. [27] on XSEDE at the Cipres Science Gateway [28]. Bayesian runs were computed independently twice with four MCMC chains for 10 million generations until the standard deviation of split frequencies fell below the 0.01 threshold. The convergence of runs was visually assessed using the trace function in Tracer 1.6 [29]. Three samples of the genus Simocybe were selected as an outgroup. The results were further edited with TreeGraph 2 [30].

Morphological Analyses
Our morphological observations were focused on the collections corresponding to the morphological profile of C. applanatus and C. malachius, which was defined by the lack of orange tint on the lamellae and pilei, not the hymeniderm pileipellis structure and without the predominant narrowly utriform cheilocystidia [10] (Supplementary Table S1).
The macromorphological characters were described based on fresh materials. The width of the pileus denotes a dimension from right to left; the pileus diameter was measured as the shortest distance between the attachment to the substrate and the pileus margin. References to colours follow the findings of Kornerup & Wanscher [31]. The decay stages of wood were classified according to Ripková et al. [32] and Olsson & Jonsson [33].
The micromorphological structures were examined from dried herbarium specimens under an Olympus BX41 light microscope using an oil immersion lens at a magnification of 1000×. All tissues were examined in a combination of 3% aqueous solution of KOH and aqueous Congo red. The colour of the structures was only observed in 3% aqueous solution of KOH. The drawings were made with an Olympus U-DA drawing attachment at a projection scale of 2000×. An Olympus Camedia C-5060 digital camera was used to take microscopic photos. The micromorphology was based on 20 (basidia, basidiola, and tramal cells) and 30 (basidiospores, cheilocystidia, terminal, and subterminal cells of pileipellis) measurements per specimen. Except for tramal cells, the range of microscopic characters is given as the minimum, maximum (in parentheses), average ± standard deviation, and average values. For tramal cells, only the range of minimum and maximum sizes is given. The statistical analyses were computed in R 3.6.2 (R Core Team, 2020, Vienna, Austria) and the results were visualized as box plots using the ggplot2 package [34]. The graphical outputs were further processed in CorelDRAW X5 (Corel Corporation, Ottawa, ON, Canada). The frequency of occurrence of the micromorphological structures (cheilocystidia, terminal cells of pileipellis) is related to the following values: 100% = always, 50.1-99% = mostly, 25.1-50% = often, 5.1-25% = sometimes, 0.1-5% = rarely, 0% = never. The descriptions of micromorphological characters were based only on sequenced specimens (Supplementary Table S1). The morphological terminology is adopted from Berger et al. [35] and Vellinga [36].
Abbreviations: L = number of lamellae reaching the stipe/the (closest) point of attachment to the substrate; l = number of lamellulae between each pair of lamellae; and Q = ratio of length and width of basidiospores, cheilocystidia, and terminal cells of pileipellis.

Phylogenetic Analyses
The topologies of ITS and LSU trees ( Figure 1) were similar but not consistent in all parts. All the included European members of the genus with globose to subglobose basidiospores were strongly supported by ITS. Three collections collected on pine stumps and with similar morphology as C. applanatus were sister to C. stenocystis and they are part of a clade that is sister to typical C. applanatus collections from the wood of deciduous trees. These pine inhabiting collections are described as a new species, C. pini. In the ITS tree, the material that is identified as C. malachius was clustered with C. crocophyllus and C. ehrendorferi (BI = 0.95, ML = 66). In the LSU tree, C. malachius was only clustered with C. crocophyllus and C. globisporus (BI = 1, ML = 33), and C. ehrendorferi was placed in a separate clade nested in a large lineage mainly with species without globose spores. In conclusion, C. applanatus and C. malachius were placed in unrelated phylogenetic lineages and the C. applanatus lineage contained additional look-alike species C. pini. conclusion, C. applanatus and C. malachius were placed in unrelated phylogenetic lineages and the C. applanatus lineage contained additional look-alike species C. pini.  Sequences retrieved from GenBank are provided with their original identifications, accession numbers, and codes of countries of the origin (https://www.iban.com/country-codes, accessed on 14 April 2022) and US and Canada state/province codes (https://www.fs.fed.us/database/feis/format.html, accessed on 14 April 2022). Sequences generated in this study are provided by herbarium numbers instead of Genbank accession numbers. Ex-type sequences are labeled with "T!". ML bootstrap support values greater than 50% and BI posterior probabilities greater or equal than 0.95 are indicated at the nodes.

Morphological Analyses
After an initial morphological analysis, the characters of basidiospores and cheilocystidia were selected as having potential distinguishing values and were supported by additional observations. The basidiospore width and cheilocystidia length of 30 elements per collection were measured (Supplementary Tables S3 and S4).
There were significant differences in the basidiospore width between C. applanatus and C. malachius, with no overlapping average values or medians ( Figure 2). For C. applanatus, 75% of the basidiospore width variability was below 5.5 µm, and for C. malachius, 75% of this character variability was above 6 µm. The new species, C. pini, had an intermediate size and its basidiospore width did not overlap with the median and average values of either of the previous species, but its interquartile range overlapped towards its 25th percentile with C. applanatus and towards its 75th percentile with C. malachius. sequences are labeled with "T!". ML bootstrap support values greater than 50% and BI posterior probabilities greater or equal than 0.95 are indicated at the nodes.

Morphological Analyses
After an initial morphological analysis, the characters of basidiospores and cheilocystidia were selected as having potential distinguishing values and were supported by additional observations. The basidiospore width and cheilocystidia length of 30 elements per collection were measured (Supplementary Tables S3 and S4).
There were significant differences in the basidiospore width between C. applanatus and C. malachius, with no overlapping average values or medians ( Figure 2). For C. applanatus, 75% of the basidiospore width variability was below 5.5 μm, and for C. malachius, 75% of this character variability was above 6 μm. The new species, C. pini, had an intermediate size and its basidiospore width did not overlap with the median and average values of either of the previous species, but its interquartile range overlapped towards its 25th percentile with C. applanatus and towards its 75th percentile with C. malachius. The interquartile values of cheilocystidia length overlapped all three species ( Figure  3). There was no significant difference between C. applanatus and C. malachius. Crepidotus pini had average values of all three sequenced collections that were higher than 40 μm and the highest value among the collections of both former species was 39.6 μm. Considering a relatively high range of cheilocystidia length values of all species within the individual collections, this character needs to be used with caution. The interquartile values of cheilocystidia length overlapped all three species (Figure 3). There was no significant difference between C. applanatus and C. malachius. Crepidotus pini had average values of all three sequenced collections that were higher than 40 µm and the highest value among the collections of both former species was 39.6 µm. Considering a relatively high range of cheilocystidia length values of all species within the individual collections, this character needs to be used with caution. In conclusion, C. applanatus and C. malachius can be identified by basidiospore width, but the combination of the former character and cheilocystidia length is necessary to identify C. pini. Another feature to support the identification of C. pini can be the cheilocystidia length/width ratio. The cheilocystidia width of all species is similar, but in C. pini they are mostly narrowly clavate, and this type is less frequent in the other two species (Table 1). Crepidotus pini also differs from C. applanatus and C. malachius by incrustation of hyphae in pileipellis and by consistent presence of yellow tints on the pileus surface (never being white when young). Table 1. Basidiospore and cheilocystidia characteristics measured on Crepidotus applanatus (n = 300), C. malachius (n = 360) and C. pini (n = 90). The size of basidiospores and cheilocystidia, the length/width ratio of basidiospores (Q), and the frequency of occurrence of cheilocystidia types (mostly = 50.1-99%, often = 25.1-50%, sometimes = 5.1-25%, rarely = 0.1-5%) was calculated based on the sequenced specimens (Supplementary Tables S1-S4); n = total number of measurements (30 measurements per specimen); the measurements are presented as averages ± standard deviation, values in parenthesis = minimum and maximum.

Characters
Species Measurements basidiospore size (μm) C. applanatus  In conclusion, C. applanatus and C. malachius can be identified by basidiospore width, but the combination of the former character and cheilocystidia length is necessary to identify C. pini. Another feature to support the identification of C. pini can be the cheilocystidia length/width ratio. The cheilocystidia width of all species is similar, but in C. pini they are mostly narrowly clavate, and this type is less frequent in the other two species (Table 1). Crepidotus pini also differs from C. applanatus and C. malachius by incrustation of hyphae in pileipellis and by consistent presence of yellow tints on the pileus surface (never being white when young).

Taxonomy
Crepidotus applanatus (Pers.) P. Kumm., Der Führer in die Pilzkunde: 74, 1871. (Figure 4 and Figure 5A). Ecology: Most of our collections of C. applanatus were from the wood of fallen trunks, fewer were from fallen branches, and one collection was from a standing trunk. The most common host tree was Fagus sylvatica, and less common was Acer pseudoplatanus, Quercus sp., and Salix sp.; that it grew on Picea abies is questionable. The fallen trunks and branches were usually 20 to 40 cm thick, in full contact with the soil and basidiomata grew mainly on their lateral side; basidiomata on a standing trunk grew to a height of about 150 cm. The substrates of the 3rd to 5th decay stages were often overgrown with mosses. The findings were from the months of July, September, and October. Most of our collections were from old-growth forests protected as nature reserves in Slovakia, few collections were from managed forests. Of the habitats, the species was most common in beech and fir-beech forests (Fagus sylvatica dominates the tree layer, along with admixed Abies alba, Acer pseudoplatanus, Picea abies, and others); one collection was from oak-hornbeam woods (tree layer is predominantly occupied by Carpinus betulus, Quercus petraea, and Q. robur); another one was from a former pasture having gradually been covered with wood trees of Betula pendula, Populus tremula, and Salix caprea; and one Norwegian collection was from a boreal Picea abies forest. The elevation of the stands was 60-900 m a.s.l.      Figure 5B and Figure 6).  Holotype: ALT 5730 (from New England U.S.A.); isotype: FH-258654. Description: Basidiomata pileate, sessile (laterally or dorsally attached to the substrate) or with a rudimentary lateral stipe, gregarious in sparse to dense groups, often imbricate. Pileus rounded flabelliform, flabelliform, rarely reniform; convex, plano-convex to applanate; flat, sometimes umbonate or depressed near the point of attachment; 8-75 mm wide, 10-50 mm in diameter; hygrophanous; margin usually involute, becoming inflexed, rarely straight; entire, in larger basidiomata often undulate, translucently striate (up to oneeighth to one-third of the pileus diameter); surface of young and fresh basidiomata white (watery white when soaked), when mature becomes yellowish gray (4B2-putty) or ivory (4B3), when dried sometimes more yellow (4A3-cream, 4A4-light yellow); sericeous or velutinous, smooth; at the point of attachment white mycelial tomentum extending up to oneeighth to one-quarter of the pileus diameter. Stipe (if present) cylindrical, 1-3 × 1-2 mm, whitish, pubescent. Lamellae l = 7-15; L = 14-28, 1-7 mm wide, ventricose, adnexed to decurrent; when young whitish to orange-white (5A2), when mature more gray and orange (5B2-orange-gray, 5B3-pale orange, 5C3-brownish orange) to golden blond (5C4) and up to brown (5D6-oak brown or 5E6-mustard); lamellae edges concolorous when young, paler (whitish) than the lamellae sides when mature; entire or irregularly serrulate. Spore print brown (5E6-mustard brown, 6E6-cacao brown). Context up to 2 mm thick, whitish to ivory (4B3), smell and taste indistinct.
Ecology: Most of our collections of C. malachius were from the wood of fallen trunks, fewer from stumps, and one collection was from a standing trunk and one from a fallen branch. The most common host tree was Fagus sylvatica, less common were Acer campestre, Fraxinus excelsior, and Quercus sp. The stumps and trunks were about 20 to 50 cm thick, and the branches were about 10 cm thick; fallen substrates were usually in full contact with the soil and basidiomata grew mainly on their lateral sides; basidiomata on a broken (still standing) trunk grew to a height of about 90 cm. The substrates of the 3rd to 5th decay stages were often overgrown with mosses. The findings were from the months of June to October. Most of our collections were from old-growth forests protected as nature reserves in Slovakia with few collections from managed forests. Of the habitats, the species was most common in beech and fir-beech forests (Fagus sylvatica dominates the tree layer, along with admixed Abies alba, Acer pseudoplatanus, Fraxinus excelsior, Picea abies, and others); few collections were from oak-hornbeam woods (tree layer is predominantly occupied by Carpinus betulus, Quercus petraea, and Q. robur); the collections from the Czech Republic    Figure 5C, Figure 7 and Figure 8).   Description: Basidiomata pileate, sessile (even without a rudimentary stipe on the youngest basidiomata; laterally or dorsally attached to the substrate); growing in tight groups of more than five to nearly a hundred basidiomata (usually on one place on the stump), imbricate; or growing in more sparse groups of fewer than five basidiomata, sometimes single (often scattered in several places on the stump). Pileus mostly rounded flabelliform, sometimes flabelliform; convex, plano-convex to applanate; flat or umbonate, sometimes depressed near the point of attachment; 7-50 mm wide, 6-30 mm in diameter; hygrophanous; margin usually involute, becoming inflexed to reflexed, somewhat exceeding the lamellae; entire, in some (mostly in larger) basidiomata crenulated, crenate or undulate, not or only when very moist translucently striate; surface of young basidiomata grayish-yellow (4B4-champagne or 4B5-corn), with a distinctly lighter margin (4A4-light yellow), just a short time pruinose, soon glabrous, smooth, when drying with no or slightly lighter colour changes (4A4, 4B4, 4B5), mature and fresh basidiomata more brownish (5D5-clay, 5D6-oak brown, 5E6-mustard brown) to dark brown when very old (7F8), of the same colour or lighter when dried (4B4, 4B5); sometimes with fine sun brown (6D5) scales on the clay (5D5) background in older basidiomata; whitish mycelial tomentum at the point of attachment present on both sides of the pileus, strongly contrasting with the colour of pileus and lamellae, extending up to one-fifth to one-quarter of the pileus diameter. Lamellae l = (3)5-7; L = 9-22, 1-5 mm wide, ventricose, adnexed to decurrent; when young yellowish-white to grayish-yellow (4A2-yellowish white, 4A3-cream, 4B3-ivory, 4B4-champagne, 4B5-corn), when mature blond (5C4-golden blond to 5D4-dark blond), the oldest coloured up to brown (5E6-mustard brown or 5E7-linoleum brown) to dark brown (7F8); lamellae edges concolorous or paler (whitish) than lamellae sides; entire or irregularly serrulate. Spore print brown (6E5 to 6E6-cacao brown, 6F5 to 6F6-dark brown). Context up to 2 mm thick, grayish yellow (4B4-champagne) to dark blond (5D4), smell indistinct, taste indistinct or slightly like mushrooms or sour.
Ecology: Our collections of C. pini were from 80 to 100 years old stumps of Pinus sylvestris in the 3rd to 5th decay stages. The stumps of felled trees were 30 to 50 cm thick and the basidiomata were formed at their base, on strongly decayed wood, and almost in contact with soil. Some stumps were overgrown with mosses (Hypnum cupressiforme, Aulacomnium androgynum) and some were also colonized by other basidiomycetes (Mycena sp., Artomyces sp.), ascomycetes (Ascocoryne sp.), or myxomycetes. The stumps were located in well-lit places, namely along the forest road, at the forest edge, or where the clearing was created by the felling of trees. The findings were from the months of August to October. We have found C. pini at three localities in the Záhorská nížina Lowland (north-western Slovakia), in forests with cultivated Scots pine (Pinus sylvestris), and at an elevation of 190 to 220 m a.s.l. The forests were between 80 and 100 (110)

Discussion
Our analyses confirmed that C. applanatus and C. malachius are morphologically distinct and phylogenetically unrelated taxa. These two species show similar field appearance and basidiospores shape, but they are not closely related. For example, C. malachius is more closely related to C. crocophyllus, the species with brown squamulose to fibrillose pileus. Members of serie Applanatus, which is defined by a white or whitish and hygrophanous pileus surface as well as globose to subglobose basidiospores, are not monophyletic and represent a morphotype that evolved convergently.
Our phylogenetic study showed little variation in ITS or LSU regions among samples of C. applanatus that is distributed throughout the Northern Hemisphere, but there is a geographic clustering observed within C. malachius (Figure 1). This variation needs further analyses of additional loci, preferably including protein coding genes. If there are geographic variants or closely related taxa diversified by geographical distance, they may correspond to some varieties of either species. However, previously described varieties may also represent completely different taxa, for example C. malachius var. trichifer was synonymised with C. stenocystis [10].
The new species C. pini seems to have strict preference to conifers and our material of C. applanatus and C. malachius is from the wood of deciduous trees. There are only two other European species supposedly preferring substrates of conifers, C. kubickae and C. stenocystis [43,44]. Only C. stenocystis has a similar basidiospore shape to our new species, but it differs by its larger spores and base-inflated (mostly utriform) cheilocystidia and terminal cells in pileipellis [12]. Previous reports of C. applanatus from coniferous substrates, e.g., [3], may represent C. pini and need a revision. A search of publicly available sequences did not result in a positive match with our new species, and we believe that the diversity of recently described Crepidotus species is covered by our sequence sampling and is proof that C. pini has not been described recently. Among older species described outside Europe, C. avellaneus Hesler & A.H. Sm. and C. campylus Hesler & A.H. Sm. from the USA are similar, but they both grow on deciduous substrates and the first differs by its considerably small (up to 15 mm), white basidiomata and the second has cylindrical and capitate cheilocystidia [2].
The first phylogenetic studies with sufficient sampling across the genus Crepidotus were based on the LSU nr DNA region [45]. Subsequent Crepidotus studies used this genetic marker to assess the species delimitation and to estimate the interspecies relationships, e.g., [21,42,46]. With the general implementation of ITS as a universal fungal barcode [47,48], there is an increasing number of Crepidotus sequences available from mycological studies that are not aimed at the genus taxonomy. For this reason, some recent studies analyzed the ITS region in addition to the LSU or were used as the only genetic marker [12,20,[49][50][51]. Since sequence data for many important Crepidotus collections are available only for one of these two regions of ribosomal DNA, we did not combine them in a single phylogenetic analysis.

Conclusions
This study solved the long-lasting question about the identity and delimitation of common globose-spored Crepidotus species, C. applanatus and C. malachius. Our phylogenetic study supported two species and placed them in two well-supported lineages within the genus Crepidotus. Their field aspect and microscopy are similar except for the spore size that we propose as a distinguishing character. Collections from coniferous substrate initially named as C. applanatus represent a third species, C. pini, which may also be identified based on the yellowish colours on its pileus surfaces when young, but a combination of spore and cheilocystidia characters are needed to recognise it from other similar species.
We highlighted that the majority of Crepidotus studies used either ITS or LSU nr DNA regions to support their species hypotheses. This makes the available sequence data inconsistent and did not allow us to combine both regions in a single two-loci tree. We applied a coalescent approach and analysed the ITS and LSU datasets separately, which also strengthened our argumentation that similar morphotypes represented by C. applanatus and C. malachius evolved convergently (which was supported by both analyses). The grouping of species with globose spores was not consistent between the trees. While ITS analysis supported the monophyly of the group, an LSU tree placed a small group of C. ehrendorferi and related globose-spored species close to species with non-globose spores. This clearly demonstrated that building a multi-locus tree including low copy genes is necessary to resolve phylogenetic relationships within the genus Crepidotus.
Crepidotus pini seems to be specialized to coniferous substrates and we did not find any apparent differences in the ecology of C. applanatus and C. malachius. The challenging question is if ecological conditions had driven the convergent evolution of this morphotype and if there is some niche differences and specialization between the two species dwelling deciduous substrates.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jof8050489/s1, Table S1: Details on the specimens (arranged alphabetically and within Slovakia from west to east) examined in this study. Specimens marked with asterisk symbol have been sequenced; Table S2: The list of ITS and LSU nrDNA sequences used in the phylogenetic analysis. Taxon names (in alphabetical order) are presented in their original form (C. = Crepidotus). Sequences prepared in this study are in bold; Table S3: Estimated ranges of basidiospore dimensions based on 30 measurements represented as average ± standard deviation, in parenthesis are minimum and maximum values, the average is underlined, Q = the length/width ratio. Specimens marked with asterisk have been sequenced; Table S4: Estimated ranges of cheilocystidia dimensions based on 30 measurements represented as average ± standard deviation, the average is underlined. Morphological groups of cheilocystidia: cla-clavate, cyl-cylindrical, utr-utriform, ven-ventricose, for-forked, lob-lobate, pyr-pyriform, glo-globose (undifferentiated here whether it is a broadly or narrowly type of the cheilocystidia). Specimens marked with asterisk have been sequenced.