Revision of Cerinomyces (Dacrymycetes, Basidiomycota) with notes on morphologically and historically related taxa

Cerinomyces (Dacrymycetes, Basidiomycota) is a genus traditionally defined by corticioid basidiocarps, in contrast to the rest of the class, which is characterized by gelatinous ones. In the traditional circumscription the genus is polyphyletic, and the monotypic family Cerinomycetaceae is paraphyletic. Aiming for a more concise delimitation, we revise Cerinomyces s.l. with a novel phylogeny based on sequences of nrDNA (SSU, ITS, LSU) and protein-coding genes (RPB1, RPB2, TEF1-α). We establish that monophyletic Cerinomyces s.s. is best characterized not by the corticioid morphology, but by a combination of traits: hyphal clamps, predominantly aseptate thin-walled basidiospores, and low content of carotenoid pigments. In our updated definition, Cerinomyces s.s. encompasses five well-supported phylogenetic clades divided into two morphological groups: (i-iii) taxa with arid corticioid basidiocarps, including the generic type C. pallidus; and (iv-v) newly introduced members with gelatinous basidiocarps, like Dacrymyces enatus and D. tortus. The remaining corticioid species of Cerinomyces s.l. are morphologically distinct and belong to the Dacrymycetaceae: our analysis places the carotenoid-rich Cerinomyces canadensis close to Femsjonia, and we transfer the clamps-lacking C. grandinioides group to Dacrymyces. In addition, we address genera related to Cerinomyces s.l. historically and morphologically, such as Ceracea, Dacryonaema and Unilacryma. Overall, we describe twenty-four new species and propose nine new combinations in both Cerinomycetaceae and Dacrymycetaceae.


INTRODUCTION
The Dacrymycetes is a phylogenetically and morphologically well-established class of basidia-bearing fungi, inhabiting dead wood and causing brown rot (Oberwinkler 2014, Floudas et al. 2015, Nagy et al. 2016). The class is relatively small, and includes two orders (Dacrymycetales and Unilacrymales), four families (Cerinomycetaceae Jülich, Dacrymycetaceae J. Schröt., Dacryonaemataceae J.C. Zamora & Ekman, and Unilacrymaceae Shirouzu, Tokum. & Oberw.), more than 10 generally accepted genera, and ca. 400 published species names. Many of these names are taxonomic synonyms, and the true number of described species is at least 120. Members of the class can be readily distinguished from other basidiomycetes by their bisterigmate Y-shaped basidia, with the single exception of Unilacryma unispora, which has unisterigmate basidia (Wells 1994, Shirouzu et al. 2013. On an ultrastructural level, dacrymycetes are characterized by dolipore septa with imperforate parenthesomes, or rarely with a single pore (Maekawa 1987, Oberwinkler 1994, Shirouzu et al. 2013. In terms of  1900190919231928193619491961196419821986Brasfield (1938 mentioned that "Ceracea auct. not Cragin" is a synonym to Arrhytidia. Later he changed his mind and purported that Ceracea is a correct name for species with resupinate basidiocarps without a distinct "root" in substrate. Since the type of Ce. vernicosa was assumed to be lost, he intended to typify Ceracea by Ce. lagerheimii (Brasfield 1940). Martin (1940) accepted Brasfield's concept of the genus and described a new species Ce. canadensis. Luckily, the original material of Ce. vernicosa was found, and Martin (1949) revealed it to represent an anamorphic ascomycete fungus, possibly a parasite of polypores. Thus, the name became unavailable for corticioid dacrymycetes. To encompass the latter, Martin introduced the genus Cerinomyces consisting of C. pallidus as the type, and new combinations C. crustulinus and C. canadensis. The last species of Ceracea, Ce. aurea and Ce. brasiliensis, appeared in a post-mortem publication of .
In the following years, the taxonomic position of Cerinomyces was a matter of discordance. Donk (1956Donk ( , 1957, Eriksson (1958) and Eriksson & Ryvarden (1973) suggested the genus is affined with corticioid fungi, while others treated it as a member of dacrymycetes (Martin 1957, Kennedy 1958a, Parmasto 1961. Within the most comprehensive revision of dacrymycetes to date, McNabb (1964) accepted the latter opinion, followed by the rest of the community. For example, Donk (1966) had already cited Cerinomyces in the Dacrymycetales. In the same work, McNabb described C. grandinioides and combined C. lagerheimiia species that differs from the rest of the genus by its absence of clamps.
By the end of the twentieth century, the genus was expanded with a number of new taxa. Cerinomyces ceraceus Ginns joined a morphogroup of clampless Cerinomyces species. Boidin and Gilles introduced C. albosporus, which was later accompanied with macromorphologically similar C. aculeatus N. Maek., C. curvisporus N. Maek. & M. Zang, and C. fasciculatus Gilb. & Hemmes. A few more taxa with less certain relations were described as C. crustulinus var. latisporus B. de Vries (the first strictly gelatinous taxon in the genus), C. pengii Liu & Fan, and C. megalosporus Duhem.
Shirouzu and co-authors pioneered the phylogenetic studies of the class, and in an ongoing series of works demonstrated that most genera in the class are polyphyletic, and Cerinomyces is not an exception (Shirouzu et al. 2007(Shirouzu et al. , 2009(Shirouzu et al. , 2013(Shirouzu et al. , 2016(Shirouzu et al. , 2020. They showed that species with corticioid arid basidiocarps, threeseptate basidiospores and clampless septa, like C. lagerheimii, C. ceraceus, and C. grandinioides, belong to the core Dacrymycetaceae as a sister clade to some typical Dacrymyces and Guepiniopsis Pat. species. On the other hand, species like Dacrymyces punctiformis Neuhoff with pustulate gelatinous basidiocarps, aseptate basidiospores and clamped hyphae clustered in the Cerinomycetaceae (Shirouzu et al. 2009(Shirouzu et al. , 2013. Aside from these shifts, the Cerinomycetaceae appeared to be a well-supported clade in most phylogenetic analyses. In a recent work,  confirmed the earlier results and clarified a mean stem age for the Cerinomycetaceae, estimated at 197 million years ago. They also pointed out that the studied species did not have conspicuous carotenoid contents, and young spores in the genus are binucleate, in contrast to uninucleate ones in the rest of dacrymycetes (see also Yen 1947). Finally, a continuing production of dacrymycete genomes including species of Cerinomyces s.l. in the 1 000 Fungal Genomes Project promises further refinement of the group phylogeny (Grigoriev et al. 2014).
The aims of this paper are: (i) to assess all existing species of Cerinomyces s.l. and describe new ones; (ii) to revise the Cerinomycetaceae as a family consisting of Cerinomyces s.s.; (iii) to find practical nomenclatural solutions for the species excluded from Cerinomyces s.s.; and (iv) to establish connections between phylogenetic groups and morphological characters whenever possible.  Supplementary Table S1). Herbaria acronyms follow Index Herbariorum (http://sweetgum. nybg.org/science/ih/). Collector's numbers are shown without a collector's name abbreviation. Studied type specimens are accompanied with exclamation mark (!), and these specimens are not duplicated under "Specimens examined". Descriptions are based primarily on sequenced specimens marked with an asterisk (*), and specimens without sequences are incorporated only in the absence of sequenced ones or when their morphology agrees well with the adopted species concept. Detailed information on specimens, high-resolution macro photographs, and scanned notes and labels are available under CC BY 4.0 license via the PlutoF platform (https://plutof.ut.ee, Abarenkov et al. 2010).

Morphological study
Microscopic studies were performed with Leica DMLB, Leica DM1000 LED and Nikon Eclipse 80i microscopes. For slide preparation small parts of basidiocarps were moisturized with tap water, then cut with a razor blade and placed for a short time in a small drop of water; excess of water was removed with filter paper before further dying. The routine mountant used for measurements and drawings was Cotton Blue (CB): 0.1 mg aniline blue (Merck 1275) dissolved in 60 g of pure lactic acid. After applying a cover-glass, the slide was heated without reaching the boiling point, then the preparation squashed by tapping on the cover-glass, and excess of CB removed with filter paper. In cases when CB was not suitable for measurements, 1 % KOH with addition of water solution of Congo Red was used instead. Whenever spore measurements in KOH are reported, it is mentioned. Illustrations were produced from microscopic slides using either a drawing tube at ×1 000 magnification (×2 000 for spore drawings) or rarely from integrated camera  Table 1. Specimens and corresponding sequence accession numbers used in this study. Accession numbers of newly generated sequences are indicated in bold.

Species
Specimen     Shirouzu et al. (2007) TNS-F-61296 Japan -LC585245 LC003899 --this study, Shirouzu et al. (2016) TNS-F-61306 Japan -LC585246 LC003909 --this study, Shirouzu et al. (2016) TNS-F-61316 Japan -LC585247 LC003919 --this study, Shirouzu et al. (2016) TNS-F-61317 Japan -LC585248 LC003920 --this study, Shirouzu et al. (2016) TNS       photos; later vectorized with Wacom DTK-2700 in CorelDRAW 2017. Measurements were done using ×1 000 magnification, oil immersion, and phase contrast illumination; eyepiece scale bar with 1 μm grid was used, and dimensions were estimated with a subjectively defined accuracy of 0.1 μm; when working with Nikon hardware, spores were photographed and measured in NIS-Elements BR < v. 5.20.00. Spore statistics were produced in LibreOffice Calc v. 6.0.6.2. When preparing summaries, individual measured spores were omitted only when considered immature or overgrown; not more than five spores from original measurements were excluded per taxon. The following abbreviations are used in descriptions: L for mean spore length, W for mean spore width, Q for L/W ratio, Q' for variation of length to width ratio of individual basidiospores. To show variation in basidiospore dimensions, 5 % of measurements from each end of the range are excluded and given in parentheses. In case of identical values, parts in parentheses are omitted. For the types and representative specimens at least 30 randomly selected mature basidiospores and well-developed basidia were measured when possible; a total number of measured structures against a number of studied specimens is shown as "n = 30/1". By default, spore walls thickness measurements were obtained from outer walls, not septal, which are often thicker. Following Ingold (1983), we distinguish parts of two-lobed apiculus (or hilar  REVISION OF CERINOMYCES WITH NOTES ON RELATED TAXA www.studiesinmycology.org apparatus) in basidiospores of dacrymycetes as: (A) hilum itself the point of attachment to basidia; and (B) more prominent hilar appendix (Fig. 3). Herewith, length of apiculus is not added to the spore length. Terms "basidia" and "sterigmata" are used for parts below and above the bifurcation point, respectively. The widths of basidia were measured immediately below the bases of sterigmata. Sterile elements in hymenium are referred here as "hyphidia", instead of traditional "dikaryophyses" (see Discussion). Raw morphometric data for the studied specimens are provided in the Supplementary Table S2. Taxonomic novelties were deposited in MycoBank (Crous et al. 2004).

DNA extraction, PCR, and sequencing
Protocols for most of UPS and TNS herbaria specimens follow respectively  and Shirouzu et al. (2016); most other materials were processed as indicated below. DNA from the fresh specimens was routinely extracted in 100 μL of "10× reaction buffer B" (0.8 M Tris-HCl pH 8.  extracted and stored at -80°C. For older specimens and types the High Pure PCR template preparation kit (Roche Applied Science, Penzberg, Germany) was used following the protocol of manufacturer. Polymerase chain reaction (PCR) was performed with the following primers; forward and reverse ones are separated by slash. The whole ITS region with a part of LSU: ITSOF (ACTTGGTCATTTAGAGGAAGT, Tedersoo et al. 2008

Phylogenetic analysis
General sequence management and contig assembly were done in Geneious v. 7.0.6 and 9.  (Ronquist et al. 2012) as implemented at CIPRES portal (Miller et al. 2010). Default priors were used, nucleotide substitution models were estimated with model jumping method (nst = mixed) with gamma-distributed rate variation across sites and proportion of invariable sites not estimated. Analyses were carried out in four parallel runs with four MCMC chains each, for 10 M generations, sampling trees every 5 000 generations, with temperature constant 0.1. A burn-in was set to a fraction 25 %. The analyses were automatically stopped if the average standard deviation of split frequencies dropped below 0.01. Effective sample sizes (ESS) were assumed sufficient with values reaching above 200, and potential scale reduction factor (PSRF) approximating to 1. Tracer v. 1.7.1 (Rambaut et al. 2018) and RWTY (Warren et al. 2017) were used to observe convergence of model parameters and tree topologies. The consensus tree was built using 50 % majority rule.

RESULTS
A dataset of SSU, 5.8S, LSU, TEF1-α, RPB1, and RPB2 genes was used to infer the class phylogeny and resolve the positions of Cerinomyces s.l. clades. Our analysis confirmed the earlier reported family arrangement of the Dacrymycetes, with the robustly supported Cerinomycetaceae as a sister clade to the Dacrymycetaceae (Fig. 4). From a morphological perspective, the combination of characters that unites the Cerinomycetaceae when compared to other families is: (i) presence of simple clamps on all hyphal septa; (ii) curved-cylindrical thin-walled basidiospores that only rarely and tardily develop up to three transverse septa; (iii) low amount of carotenoid pigments in hyphae and basidiospores; and (iv) corticioid, resupinate, pustulate, pulvinate and only slightly cerebriform basidiocarps. In addition, young basidiospores of the family members appear to be binucleate, while in the rest of the class uninucleate state is usual. The character was observed in fresh material of ten Cerinomyces taxa and reported in descriptions.
In contrast to the traditional view, corticioid morphotype alone does not define the family: while most corticioid dacrymycetes do belong to the Cerinomycetaceae, a number of Cerinomyces s.l. are found in the Dacrymycetaceae. In agreement with earlier studies, one of the excluded groupsa clade containing D. grandinioidesis resolved as a sister to Dacrymyces stillatus, D. chrysospermus and Guepiniopsis buccina. Here we formally transfer all taxa related to D. grandinioides to Dacrymyces. Two other corticioid species, "Cerinomyces" canadensis and D. corticioides, are recovered in proximity to Femsjonia peziziformis. We refrain from nomenclatural rearrangements of these until a dedicated study of Femsjonia is undertaken.
Based on the same dataset, we empirically designated five clades within the Cerinomycetaceae to highlight connections between phylogenetic and morphological groups. We treat all clades as part of Cerinomyces: in our opinion, division of the family into several genera is impractical. Even though morphology is generally uniform within the clades, it is still not possible to identify characters that would unambiguously define every putative genus. In addition, establishing generic boundaries is not justified given the low support values in some of the deeper nodes in the family phylogeny.
Two main types of basidiocarps, arid corticioid and gelatinous pustulate, are found in the Cerinomycetaceae. As shown in the Fig. 4, corticioid species are scattered across three groups (labelled here as C. pallidus, C. albosporus and C. borealis clades), while the gelatinous ones form two (C. enatus and C. tortus clades). To increase the resolution within these clades, we utilized ITS and TEF1-α sequences. ITS was too variable for family-wide alignments, so we divided the data into clade-or even subclade sets, presented in separate trees (Figs 5-7). SSU and LSU were also incorporated in the concatenated sets, but predictably showed little parsimony-informative signals at the species level. For the species delimitation, we primarily used ITS as a barcoding marker available for most of the taxa. In total, the genus Cerinomyces includes 29 species, of which 20 are newly described here. In addition, we propose three new combinations, designate four informal taxa and exclude seven species from the genus. The clades are detailed below.
The C. pallidus clade encompasses the generic type C. pallidus and its nine corticioid relatives. These inhabit angiosperm and gymnosperm wood mostly in the temperate zones of the Northern and Southern hemispheres. Basidiocarps are pale ochraceous, varying from arachnoid to crustose with a thin cottony subiculum. Cerinomyces pallidus and C. paulistanus develop hyphal pegs, and several related species demonstrate microscopic peg-like hyphal constructions (like Fig. 39 C), but never as abundant and regular as in the C. albosporus clade members. Microscopic features often blend between species, requiring comparison of multiple characters for non-molecular identification. Because of the differences in ITS, we prepared separate phylogenies for the C. atrans and C. volaticus subclades ( Fig. 5 A, B). Basidiocarps in the first subclade are more likely to become crustose and darken to different extents. This character was observed in all members, namely in C. atrans, C. concretus, C. pallidus, C. paulistanus, C. ramosissimus, and C. verecundus. Species in the second subclade (C. fugax, C. nepalensis, C. tristis, and C. volaticus) usually remain arachnoid or solid, but still cottony, and light-coloured.
The C. albosporus clade (Fig. 5 C), sister to the C. pallidus clade, includes four corticioid species confined to climates ranging from humid temperate to tropical. Three of these species (C. aculeatus, C. albosporus, and C. brevisetus) have abundant, regularly distributed hyphal pegs and larger microstructures than in other arid Cerinomyces. The fourth species, C. inermis, is distinguished by the absence of hyphal pegs and basidiospores that conform better to some members of the C. borealis and C. pallidus clades. In the absence of sequence data, our judgement from morphology indicates that further peg-bearing species are also allied with the clade: C. fasciculatus collected from Hawaiʻi and C. curvisporus from Southwest China, known for its strongly bent basidiospores.
The C. borealis clade includes three corticioid species growing on coniferous wood. Two closely related species from temperate Europe and North America, C. borealis and C. favonius, have the narrowest basidiospores in the genus. The third species, C. pinguis, was collected from mountains of Nepal, and possesses much larger basidiospores. Macromorphologically the group is difficult to distinguish from other arid Cerinomyces members, though it develops more delicate, often arachnoid basidiocarps without pegs. Clade K from Shirouzu et al. (2016Shirouzu et al. ( , 2020 corresponds to a part of the C. borealis clade and represents Japanese environmental samples and strains. Their sequences differ from the specimen-based ones, suggesting undescribed diversity in the clade (Fig. 7 A).   The C. enatus clade members ( Fig. 6) look like gelatinous Dacrymyces species with pustulate basidiocarps, though they lack the hallmark bright yellow tints of Dacrymycetaceae, being instead pale yellow, ochraceous, dark brown and reddish brown. Microscopically, the clade is characterized by heavily gelatinized hyphae and the presence of dendroid hyphidia in the hymenium of all the species. The clade consists of C. cokeri, C. creber, C. enatus, C. enterolaxus, and C. neuhoffii, all of which grow in moderate to highly humid conditions in the biogeographic Northern temperate zone. We found that widely recognized C. enatus (≡ D. enatus [Berk. & M.A. Curtis] Massee) occurs only in North America and East Asia. For European material formerly identified as D. enatus, we introduce a new species, C. aeneus, that belongs to the C. tortus clade. Cerinomyces crustulinus, whose name was massively misapplied to corticioid members, is likely to be related to the C. enatus clade. However, in the absence of fresh collections and sequence data, its position is difficult to resolve with confidence.
The C. tortus clade (Fig. 7 B) comprises four species with gelatinous basidiocarps, including C. tortus, one of the earliest described Dacrymyces species (as D. tortus [Willd.] Fr). The distribution and morphological characters of the clade often overlap with the C. enatus clade, but in the phylogeny their relation is not well supported (Fig. 4). The clade members develop mostly pustulate basidiocarps, with the exception of C. aeneus, that also can demonstrate coalescing to resupinate morphology. Cerinomyces tortus itself has a number of morphologically similar relatives, from which we formally describe only C. lipoferus. At least two more Nordic taxa are represented by scarce specimens that are not suitable as types, and whose intraspecific variation is poorly known. The main character that helps to effectively distinguish C. tortus, C. hesperidis and C. lipoferus from the other gelatinous species is the lack of finely branched hyphidia. In addition, C. lipoferus demonstrates a high amount of lipid droplets in hyphae, which is unusual for the family, and scattered 3-septate basidiospores, unique among the gelatinous species studied and very infrequent in the genus. Morphology-based identification in C. enatus and C. tortus clades is possible, though differentiating European species can be problematic if basidiocarps are young or weathered. Identification by ITS marker can also be difficult due to high levels of intragenomic polymorphism hampering Sanger sequencing and contig assembly.
All other known corticioid species of Cerinomyces s.l. belong to the Dacrymycetaceae and are considerably different from the Cerinomycetaceae. They have more robust, richly coloured corticioid basidiocarps with hymenial surfaces that become waxyor firmly-gelatinous if moisturized, thick light-coloured subiculum, and fimbriate margins. In these groups we describe four new species, propose six combinations and one informal taxon.
The C. canadensis lineage consists of a single species collected in East Asia and North America. It has corticioid pale to dark orange basidiocarps of various shapes, sometimes effused over several centimeters. The species sometimes possesses socalled pseudoclamps that are rare among dacrymycetes (see a note under the species and Fig. 23 D). "Cerinomyces" canadensis together with D. corticioides are related to Femsjonia (Fig. 4), which is well reflected in their morphological similarity to Femsjonia species.
The D. corticioides lineage encompasses one species with circular, separate, later coalescing basidiocarps of yellow to orange colour, hyphae with clamp connections and tardily septate basidiospores. Here we show that D. corticioides, traditionally recognized as a North American species, has an amphi-Pacific distribution in the Northern Hemisphere and is identical to C. altaicus described from the Russian Far East. Femsjonia uniseptata with brightly yellow firm-gelatinous basidiocarps appears to belong to D. corticioides. European D. confluens is also very similar morphologically to D. corticioides, but due to the lack of sequence data, we treat it separately.
The seven species of the D. grandinioides clade tend to have relatively thick basidiocarps, a waxy-gelatinous yellow hymenial layer, clampless hyphae and three-septate, usually thick-walled basidiospores. Members of the clade occur in Africa and in the Americas, and are accordingly divided into two subclades (Fig. 8). On the grounds of morphology, we associate two nonsequenced species with this group: D. pulchrus and D. lagerheimii. We suppose "Cerinomyces bambusicola" nom. prov. belongs here as well because of its colour and prominently odontioid basidiocarps (Gminder 2016). In this group, microscopical characters vary even within a single specimen, which makes species delimitation particularly difficult.

Taxonomy Key and identification tables
The key covers all dacrymycetes with corticioid basidiocarps regardless of colour, as well as small gelatinous taxa without long stalks, of bleak yellow, ochraceous, light to dark brown and reddish brown colour. Some of them can be confused with members of the Cerinomycetaceae: certain Dacrymyces species and members of the families Unilacrymaceae and Dacryonaemaceae, excluding Dacryonaema rufum that has unique, easily recognizable synnematous basidiocarps (Nannfeldt 1947. We render difficult groups as separate character tables outside of the main key: clamped, arid, pale-coloured corticioids from the C. pallidus, C. borealis and C. albosporus clades (Table 3), Cerinomyces species with gelatinous basidiocarps from the C. enatus and C. tortus clades (Table 4), and clampless, slightly gelatinous, yellow corticioid species with three-septate basidiospores -Dacrymyces species from the D. grandinioides clade ( Basidiospores cylindrical to slightly curved-cylindrical (Q 2.2), basidia < 60 μm in length. Clamps present on all septa...………………………………………………………6
A monotypic family in the class Dacrymycetes, consists of the genus Cerinomyces.
Description: Basidiocarps either arid and corticioid or gelatinous and pustulate, pulvinate, and resupinate; colour from white to ochraceous, pale yellow, brown, and dark brown. Basidiocarps monomitic, consist of hymenium and supporting structure (subiculum), some species have a distinct margin. Hymenium amphigenous, hymenial surface either smooth or in some corticioid species bears hyphal pegs. Hyphae transparent, smooth, hyphal width and wall thickness can vary inside a single basidiocarp, wall gelatinization occurs in many species, gelatinous layer sometimes roughened. Clamps present on each septum except occasional clampless secondary septa inside basidia. Clamps mostly simple, loop-like clamps can occur in subicular areas. Hymenium includes basidia and, in some species, hyphidia with thickened base and thinner cylindrical apical part, which is either branched or simple. Hyphidia of the second type tend to organize in hyphal pegs. Mature basidia long-clavate, with two subulate sterigmata. Basidia occasionally bear one or rarely three sterigmata. Basidiospores cylindrical to curvedcylindrical, aseptate or rarely with up to three transverse septa in few species, walls thin and smooth, contents usually hyaline, germination with germ tubes or subglobose to cylindrical conidia.
Habitat: decayed, usually decorticated wood of gymnosperm and angiosperm trees and shrubs. According to our observations, cause brown rot.
Hymenium composed of basidia and hyphidia. Hyphidia with thickwalled clavate base up to 11 μm in diam and simple cylindrical thinwalled apical part; occasional or in groups developing into hyphal pegs. Basidia clavate, 12-46 × 3-7 μm, with sterigmata up to 29(-66) μm in length ( Notes: The species was synonymized with C. albosporus shortly after description (Maekawa & Zang 1997). The two species are related, but clearly different in the ITS sequences, basidiocarp appearance, spore size, and distribution. Maekawa (1987) provided a detailed study on the cultural characteristics of the species, showing that young basidiospores are binucleate, and when deposited on MEA, aseptate basidiospores develop one to four transverse septa before or rarely after germination. He also described curved-cylindrical conidia 3.6 -5.6 × 1.2-1.6 μm produced from hyphae in monokaryotic culture and pictured basidiospores with conidiogenous scars. Shirouzu et al. (2009) reported subglobose conidia born from basidiospores, measuring 6 × 3 μm.
Notes: The species has the most robust basidiocarps and the largest basidiospores among the peg-bearing Cerinomyces   Etymology: atrans (Lat.)darkening; due to a brown colour of well-developed basidiocarp areas.
Description: Basidiocarps arachnoid, arid and light ochraceous, in older areas smooth, solid, crustose, slightly waxy when wet, dark ochraceous to brown, especially in prominent or bruised areas; with thin subiculum and indistinct arachnoid margins.
Habitat and distribution: Angiosperm (Alnus and unident.) and possibly gymnosperm wood; Western North America. Notes: Older, prominent, and damaged areas of the studied basidiocarps have dark brown tints, but we suppose this character can be absent in young fungi. Compared to C. pallidus, another North American species inhabiting mostly angiosperm wood, C. atrans does not produce hyphal pegs visible to the naked eye, but only microscopic aggregations of hyphidia. When dealing with occurrences on gymnosperm wood, a great attention is needed to distinguish between C. atrans-C. pallidus and exclusively conifers-inhabiting North American species. However, basidiocarps of the latter group do not turn into darkened crust so easily, instead remaining arachnoid or firm-cottony.
An earlier published ITS sequence of a culture TUFC30545 aligns well with our sequences, though we did not examine its voucher material UBC6108. Etymology: borealis (Lat.)northern.
Habitat and distribution: Gymnosperm wood (Picea, Pinus, and unident.); Northern Europe.  Notes: Cerinomyces borealis is a relatively widespread species in Northern Europe. In Nordic herbaria, the majority of the specimens labelled as C. crustulinus belong to C. borealise.g., GB-0071198, cited as such in "The Corticiaceae of North Europe" (Eriksson & Ryvarden 1973). The species can be distinguished from similarly looking C. volaticus by longer, thinner basidiospores (mostly 3.0 μm in width) and the more delicate, arachnoid basidiocarps. Intermediate forms with large basidiospores also occur, but even then, higher Q values of C. borealis still hold.

Cerinomyces brevisetus
Habitat and distribution: Unidentified wood; South America (known only from the type locality).
Notes: Cerinomyces brevisetus has the shortest pegs in the genus, which helps to differentiate it from the closest relatives in  the C. albosporus group, as well as from the similarly looking C. pallidus and C. paulistanus. In addition, all other members of the C. albosporus clade have longer basidiospores, while the C. pallidus clade is characterized by the shorter ones.
Habitat and distribution: Gymnosperm wood (Abies, Picea, Tsuga, and unident.); East Asia and North America.  Notes: The species does not belong to Cerinomyces s.s., but we postpone raising a new synonym (see Discussion). The type material could not be located in TRTC hebarium (S. Margaritescu, 15 Sep. 2016, pers. comm.), but morphology of the studied specimens agrees well with the original concept of Ceracea canadensis. It can be confused with Dacrymyces corticioides, though the latter species possesses larger microstructures, basidiospores with up to three septa and basidiocarps of well-defined circular form. Maekawa (1987) reported ovoid to straight cylindrical conidia born from hyphae in a monokaryotic strain of C. canadensis (CCCM 0194), but we did not check identity of this strain.  Under pseudoclamp we understand a clamp which development was arrested before the terminal part of a clamp fused with the parental hypha (upper Fig. 23 C, three rightmost pictures on Fig. 23 D, Cl emençon et al. 2004). The character is rather rare in the species but was found in all studied specimens. Sequences from TAAM007082 and TAAM061880 were not used in the analyses, but they are similar to the ITS of Spirin 8468. Description: Basidiocarps gelatinous, of irregular shapes, pustulate or cerebriform with thin resupinate margins, up to 15 mm in the longest dimension, pale yellow when fresh, light brown at margins to brown at middle, dark brown when dried, often erumpent through bark. Hyphae clamped, in subiculum 1.5-2.5 μm in diam, walls 1.5-3.0 μm in width together with gelatinous layer; small knots of ramificated hyphae rarely occur; subhymenial hyphae of the same diam, walls 0.2-0.3(-0.5) μm in width; margins covered with simple cylindrical or slightly clavate anastomosing hyphae 1.5-2.5 μm in diam, walls 0.2-0.5 μm in  Notes: Cerinomyces cokeri can be confused with C. enatus, both occuring on coniferous wood with bark. Cerinomyces enatus is generally darker coloured, becoming almost black in dry conditions, while C. cokeri remains dark brown. Also, we did not find in C. cokeri swollen hyphal compartments and such thick gelatinous layer on hyphae as in C. enatus. Etymology: concretus (Lat.)grown together, condensed. Reference to a dense hymenium becoming crustose.
Habitat and distribution: Gymnosperm wood (Cupressus, Juniperus, and unident.); Western and South-Western Europe. Notes: Cerinomyces creber is similar to C. neuhoffii in the branched hyphidia and tendency to develop dense coalescing basidiocarp groups, but the former species has smaller basidiospores, slightly smaller basidia, and often more lightcoloured basidiocarps that coalesce more readily. Also, it www.studiesinmycology.org seems that C. creber is confined to Cupressaceae wood (according to the substrates we were able to identify), while all studied specimens of C. neuhoffii were found on Pinus wood. Cerinomyces aff. crustulinus 1 resembles C. creber by macroand micromorphology but represents a distinct taxon with even paler basidiocarps growing on angiosperm wood. Sequence of UPS:F-979574 was produced during the revision of this study; it was not used in the analyses but matches other included sequences.
Notes: No dendroid hyphidia were observed in contrast to an earlier report (McNabb 1964). We assume hyphidia were unevenly distributed in basidiocarps and destroyed with past preparations. Among the species described in this revision, the closest morphological relative to C. crustulinus is resupinate C. aeneus that also inhabits angiosperm wood in Europe, but differs in darker cerebriform-resupinate basidiocarps with reddish tint and abundant branched hyphidia. The extremely scanty type material and absence of fresh collections that fully agree with our observations prevent resolution of the exact position of C. crustulinus (but see C. aff. crustulinus 1 below). Roberts (2006) reported the species from Jamaica, and while it does resemble C. crustulinus, the robust gelatinous discoid to coalescing brown basidiocarps with light margins, abundant branched hyphidia, larger spores (11.4-) 11.8 -13.8(-14.5) × 4.0-5.2(-5.7) μm place it closer to either C. enatus or even members of the Dacrymycetaceae.
Habitat and distribution: Angiosperm wood; Europe.
Notes: The taxon is the closest match to C. crustulinus except for difference in branched hyphidia (but see commentaries on C. crustulinus above about previous report of hyphidia in this species). In absence of additional material, we are inclined to keep this specimen and the type material of C. crustulinus as two separate taxa. Habitat and distribution: Gymnosperm wood (Pinus); Southeast Asia (known only from the type locality).
Habitat and distribution: Angiosperm (Alnus, Castanopsis, Clethra, Quercus, Rhododendron, and unident.) and gymnosperm (Pinus) wood; East Asia, North America.  Notes: Cerinomyces enatus is similar to C. cokeri but demonstrates darker colouration, swollen cells and thicker gelatinous layer on internal hyphae. In all studied specimens septate basidiospores are extremely rare. The species shows high variability in morphological characters, nrDNA sequences and substrate preferences. In all publications of Shirouzu et al. Japanese materials were cited as C. canadensis or C. pallidus. For American specimens CFMR:HHB-671 and CFMR:HHB-7334 we were able to sequence only 5.8S and part of ITS2. The sequences were too short to use in the phylogeny reconstruction but sufficient to identify them as C. enatus. Interestingly, Seifert  (1983) studied a culture UBC 6124 from CFMR:HHB-671 (D. punctiformis in his paper), and found in a wood block decay test that it was almost unable to degrade Pinus wooda substrate, from which it was originally collected.
The studied types of C. enatus from K and FH are contaminated by mycoparasites, contain only few, collapsed basidiospores, and show a highly agglutinated hymenial layer, which altogether deny proper measurements. Nevertheless, characteristic macromorphology, presence of branched hyphidia, thick-walled subicular hyphae with occasional swollen compartments and simple marginal hyphae help to connect type material to the other collections cited here.
Dacrymyces gangliformis has 0-1-septate basidiospores that are slightly wider than typically in C. enatus -9.2-12 × 4.5-6 μm. The species-defining "ganglia" were found to be altered swollen basidiospores, germinating with hyphae. The type specimen is scanty, poorly preserved and contaminated with anamorphic fungi. Considering similarities in microstructures and basidiocarp, we believe the specimen belongs to C. enatus. Dacrymyces fuscominus type is also in a rather bad condition, but observed morphology allows us to treat the species as a synonym of C. enatus, which has already been suggested by Kennedy (1958b Etymology: έντερα (Gr.)intestines; laxus (Lat.)loose; in reference to wide and loosely arranged subicular hyphae.

Notes:
We have not seen the specimens, but according to the described morphology, C. fasciculatus is related to the C. albosporus clade. The types could not be traced in BPI (several loan requests) or ARIZ (A.E. Arnold, 7 Jul. 2019, pers. comm.).   Notes: Cerinomyces favonius can be separated from other corticioid North American species by its narrower basidiospores and often by thinner basidiocarps. Though, specimen GB-0071216, that was identified as C. favonius primarily by the spores, has relatively robust basidiocarps and cylindrical hyphidia with thickened bases that arrange into microscopic hyphal pegs, which is more characteristic to the C. pallidus clade members. The closest phylogenetic relative to C. favonius, European C. borealis, features slightly larger basidiospores and basidia. Etymology: fugax (Lat.)shy, timid; reference to subtle basidiocarps.
Habitat and distribution: Gymnosperm wood (Pinus, Pseudotsuga); North America. Notes: Cerinomyces fugax together with C. favonius and C. tristis form a difficult morphogroup of conifers-dwelling North American species. However, basidiospores of C. fugax are wider than of C. favonius and longer than of C. tristis. Micromorphologically, C. fugax is almost indistinguishable from its closest phylogenetic neighbor -European C. volaticus. Among the specimens that we were unable to sequence but tentatively identified as C. fugax, one has larger basidiospores 9.0-10.9 × 2.9-4. Etymology: hesperidis (Gr., latinized)of evening, western.

Notes: Cerinomyces hesperidis differs from European gelatinous
Cerinomyces species by basidiocarps that are more regularly shaped, often with a central depression or of cupulate form, rooted in substrate with a short stalk, without a tendency to coalesce, resembling some collections of C. tortus. Cited here nonsequenced specimens fit well in the concept of C. hesperidis, but differ from the type by having thick-walled, pronouncedly clavate terminal hyphae covering basidiocarp stems. Etymology: inermis (Lat.)unarmored, referring to the absence of hyphal pegs.
Habitat and distribution: Gymnosperm wood; New Zealand (known only from the type locality).
Habitat and distribution: Gymnosperm wood (Picea, Pinus, and unident.), Northern Europe. Notes: The species resembles C. tortus in rarely branched hyphidia but differs in high number of lipid bodies in hyphae, presence of 1-or rarely 3-septate basidiospores, and paler fruitbodies.  referred to the species as to Dacrymyces tortus s. lat. 1. Sequences of ENZ20001 were produced during the revision of this study; they were not used in the analyses but match other included sequences.
Similarly large amount of lipid drops and matching spore measurements were observed in a specimen from a montane habitat (China, Jilin, Huang Song Pu, Chang Bai Shan Forest Reserve, 1 200 m a.s.l., on Larix olgensis, 11 Sep. 1983, L. Ryvarden 21458 [K(M): 34585]). We do not assign it to C. lipoferus due to the geographical distance between occurrences. Etymology: Reference to the type collection locality.
Habitat and distribution: Unidentified, probably gymnosperm wood; South Asia (known only from the type locality).
Notes: The species has the shortest basidiospores in the genus. In our opinion, pronounced apicular part of the basidiospores (and large space between hilum and its appendix) can indicate their immaturity, so the reported measurements may not represent the full range for the trait. Type locality of C. pinguis is also in Nepal, but this species has substantially larger microstructures than C. nepalensis. The type specimen was first reported as C. pallidus by Hjortstam & Ryvarden (1984).
Etymology: In honour of the German mycologist Walther Neuhoff.
Habitat and distribution: Gymnosperm wood (Pinus and unident.), Europe. Notes: Cerinomyces neuhoffii and C. tortus are among the most common gelatinous Cerinomyces species. The former one is widespread in Europe, while the latter occurs mostly in its northern part. Basidiocarps of C. neuhoffii are generally smaller, darker, with a softer consistency, and often organized in large, closely crowded groups. On microscopic level, C. tortus very rarely forms dendroid hyphidia, in contrast to the finely branched, abundant structures of C. neuhoffii. Also, C. neuhoffii has shorter and stouter basidia, and at least some of them tend to be slightly bent or asymmetric. Cerinomyces creber is highly similar to C. neuhoffii in hyphidia and gregarious coalescing basidiocarps, but it is often more lightcoloured, has smaller basidiospores, slightly smaller basidia, and seems to be associated with Cupressaceae. Two undescribed taxa designated below as C. aff. tortus 1 and 2 both have branched hyphidia, but their basidiospores are on average shorter than of C. neuhoffii. We also found that basidiospores of C. neuhoffii specimens from Southern and Eastern Europe are slightly wider compared to the Northern collections.  cited the species as Dacrymyces tortus s. lat. 4. Neuhoff (1934) actually distinguished two putative species (D. punctiformis and D. romelii) in what had been previously considered as Dacrymyces tortus s.l. Although the original descriptions may partially cover the morphological characteristics of C. neuhoffii, they also fit within the variation of C. tortus. The type specimens of both D. punctiformis and D. romelii turned out to be conspecific to C. tortus (see note under that species), so in absence of available names we describe C. neuhoffii as a new species, while honouring Neuhoff's clue that C. tortus has to be split.
Habitat and distribution: Angiosperm (Malus, Quercus, and unident.), very rarely gymnosperm (Pinus) wood; North America. Notes: Cerinomyces pallidus can be identified by association with angiosperm substrates and by the presence of pegs in mature basidiocarps. It shares these features with C. paulistanus, but the latter species occurs only in South America and slightly differs in morphology. In some collections, pegs can be very rare or even absent, leading to confusion with another North American angiosperm-inhabiting species, C. atrans, that has similar micromorphology, but darker basidiocarps. Both species can occasionally appear on gymnosperm wood, and then it is difficult to tell them apart from strictly conifers-dwelling taxa.
In the protologue of C. pallidus Martin assigned a "type" and, perhaps already after publication, distributed its duplicates to a number of herbaria. Maekawa (1987) cited a holotype deposited in TRTC, but the specimen could not be traced (S. Margaritescu, 14 Sep. 2016, pers. comm.). Worth noting, that under the same number 4673, Martin put another gathering from 28 July of 1948 (ISC-V-0045115), also marked as a type. Besides of the same locality (Iowa City) and substrate (oak), it is unknown how exactly the specimen is connected to the original type material from 1939. Martin (1949) reported ellipsoid conidia produced at the tops of its hyphal pegs, but we did not see any anamorphic structures in C. pallidus.
A non-sequenced specimen UBC:F873 (Canada, British Columbia) deserves a specific mention having unusually short, ellipsoid basidiospores (6.5-8 × 3-4 μm, L = 7.4 μm, W = 3.6 μm, Q = 2.1, Q' = 1.6-2.6, n = 16/1), and can represent either an immature C. pallidus or a potential new species. Sequence of ARIZ-M-AN09245 was produced during the revision of this study; it was not used in the analyses but matches other included sequences. Etymology: Reference to the type collection locality.
Habitat and distribution: Unidentified, probably angiosperm wood; South America. Notes: Cerinomyces paulistanus is a close relative to C. pallidus in phylogeny, being also highly similar morphologically, but having slightly shorter basidia, swellings in subicular hyphae, and occurring in South instead of North America. They are the only two species in Cerinomyces outside of the C. albosporus clade that develop macroscopic hyphal pegs, though we noted that in C. paulistanus many of otherwise well-developed basidiocarps have no such pegs. Etymology: pinguis (Lat.)fat; reference to the shapes of some basidia.
Habitat and distribution: Unidentified, probably gymnosperm wood; South Asia (known only from the type locality).

Cerinomyces ramosissimus
Notes: Basidiocarps of C. tortus are similar to the ones of C. neuhoffii, but generally larger, more light-coloured, occasionally yellowish when fresh or dull orange when dry. In addition, C. tortus has a northern distribution and does not produce such easily coalescing groups as C. neuhoffii or C. creber. The branched hyphidia of C. tortus are rare and more robust than of any other gelatinous Cerinomyces species. Cerinomyces lipoferus is different from C. tortus by its in average wider basidiospores, as well as by longer basidia and much higher content of lipid droplets in hyphae. In C. lipoferus, single-septate basidiospores are rare but consistently present, whereas in C. tortus they occur only in exceptional cases (e.g., in the specimen Miettinen 21768). Other affined taxa, C. aff. tortus 1 and 2, possess well-branched hyphidia. In Zamora & Ekman (2020) C. tortus sensu typi was referred to as Dacrymyces tortus s. lat. 3.
When describing Tremella torta, Willdenow (1788) mentioned that it is a rare fungus occurring in hedges near Berlin, but he did not elaborate much on morphology. Fries (1828) transferred the species to Dacrymyces and specified Pinus as the only substrate. He also synonymized Dacrymyces lacrymalis sensu Sommerfeld with D. tortus. It caused some confusion, since Sommerfeld's concept of D. lacrymalis is a mix of two species: one growing on gymnosperm wood and another one on angiosperm (Sommerfelt 1826). Fries' herbarium in UPS contains two specimens that both support the current concept of C. tortus as a conifers-dwelling species. We could not prove that this material was used for the sanctioning description, and therefore it is unavailable for lectotypification (ICN 2018, Art. 9.4).
Later Bourdot (1932) described Dacrymyces deliquescens var. castaneus characterized by discoid, short stalked, brown with orange tint basidiocarps, absence of hyphidia, and occasionally 1-septate basidiospores. We think the variety represents a well-developed C. tortus s.l. (For information on D. castaneus Rabenh. see the next section.) Neuhoff (1934) added superfluous D. punctiformis and D. romellii. Excellent condition of D. punctiformis holotype allowed us to sequence ITS region that, together with the typical morphology, confirmed the type identity to C. tortus. We have checked D. romellii isotype and found that it is very scanty, contains no spores, and is massively contaminated by intrahymenial Tremella species. Curiously enough, Neuhoff in pers. comm. to Kennedy (1957) said that D. romellii is a "luxuriant state" of D. punctiformis, which is opposite to the state of the types. Basidiocarps and hyphal structure of D. romellii are similar to C. tortus, though some microstructural differences are present, such as hyphal knots in subiculum and more thick-walled hyphidia. We consider it a synonym of C. tortus, assuming morphological peculiarities could be caused by the parasite and the old specimen age, respectively.
A possibly related to C. tristis specimen demonstrates some of the lowest Q values in Cerinomyces, having basidiospores (5-) 5.7-7(-7.2) × (2.9-)3.1-3.9(-4) μm, L = 6.3 μm, W = 3.5 μm, Q = 1.8, basidia 11-19 × 3.5-4 μm, and sterigmata up to 13 μm in length (Canada, British Columbia, Vancouver island, China beach, on Picea sitchensis, 20 Sep. 1967, B. Eriksson, J. Eriksson 8401 & J. Ginns [GB-0071226]). Hilar appendix in spores of this specimen is often distanced from the hilum itself. This pattern was also observed in the holotype of C. nepalensis, potentially indicating spore immaturity. For the same Canadian specimen, authors of "The Corticiaceae of North Europe" reported 3-to 4-sterigmate basidia (Eriksson & Ryvarden 1973). We could not confirm any of these, but 3-sterigmate basidia do occasionally occur in other species of Cerinomyces (own observations). Interestingly, the specimen NY:ex TRTS20941 bears a name "Ceracea temagamensis Jacks."the author correctly assumed it belongs to a new species, but never published the result. Sequence of O. Miettinen 19013 was produced during the revision of this study; it was not used in the analyses but matches other included sequences. Etymology: verecundus (Lat.)modest, shy; as a reference to bleak basidiocarps.
Habitat and distribution: Gymnosperm wood (Pinus); New Zealand (known only from the type locality).
Habitat and distribution: Gymnosperm wood (Taxodium); North America. Notes: Compared to the allied North American species, D. burdsallii has smaller microstructures and much less abundant pegs than D. grandiianother species that also inhabits gymnosperm wood. At the same time, D. ceraceus and D. sobrius apparently prefer angiosperm substrates. Aside the intron TTGAGAG present in the end 5.8S, D. burdsallii is different from relatives by changes in the ITS, SSU, LSU and RPB2 regions. A non-sequenced specimen CFMR:HHB-6685 possesses swollen up to 6.5 μm cells in subhymenial hyphae, but never as wide as in D. ceraceus. ( Description: Basidiocarps appear as cottony white subicular patches, later covered with solid, ochraceous to yellowish brown hymenium, arid to waxy, with rare scattered hyphal pegs up to 200 μm in height. Margin wide, fimbriate, white. Hyphae without clamps, subicular hyphae 2-4 μm in diam, walls 0.5-1.0 μm in width; subhymenial hyphae 2-3.5 μm in diam, walls 0.3-0.7 μm in width. Swollen cells up to 20 μm, with thickened walls, occasional to abundant, occurring primarily in subhymenium. Marginal hyphae have cylindrical to clavate ends often with few constrictions, 3-5 μm in diam, with walls 0.5-1.2 μm in width. Margin contains crystal aggregations, marginal and hymenial hyphae often covered with crystal shielding. Hymenium densely Habitat and distribution: Angiosperm wood (Magnolia); North America (known only from the type locality).

Dacrymyces ceraceus
Notes: Dacrymyces ceraceus is the most similar species to D. sobrius, while related D. burdsallii and D. grandii differ in growing on coniferous substrates. Marginal cells, often with moniliform constrictions that were not found in other species, can help in identification. Dacrymyces ceraceus is conventionally characterized by the presence of wide swollen compartments in subhymenial hyphae. However, abundance of swollen cells varies highly in preparations even from the same basidiocarp, and sometimes they can be found in other species (D. burdsallii, D. grandinioides, and D. sobrius) so this character should be used with caution. Another character that was utilized in the key by Ginns (1982), thickness of hyphal wall and its level of gelatinization in KOH, have doubtful value as an identification cue. It is often difficult to tell apart the wall itself and gelatinous layer, while the extent of gelatinization varies, depending on time of exposure to water, age of specimens, and nature conditions before collection. At this point, the most reliable way to identify the species is sequencing of ITS marker. As for culture characteristics, Ginns (1982) reported obovoid to cylindrical, 3-4.5 × 2 μm conidia produced in monokaryotic culture.
The ITS region of CFMR:HHB-6817 (USA, Florida, Levy co., State Route 24, near Otter creek, on Taxodium distichum, 27 Jul. 1972, H.H. Burdsall) aligns well with the D. ceraceus ex-type sequence but differs in a few base pairs. Considering low sequence quality, we cannot conclude whether it is an indication of a separate species or intraspecific variation. Notably, the specimen also has slightly moniliform terminal cells. Notes: The species is similar to D. lagerheimii but has smaller basidiospores and much more delicate basidiocarps. Basidiocarp in the specimen BPI726063 is partially covered with visible by the naked eye large crystal agglomerations of unknown origin. Another South American species that does not demonstrate hyphal pegs, D. pulchrus, has substantially longer basidiospores with up to five septa.
Habitat and distribution: Gymnosperm wood (Abies, Pinus and unident.); Europe.   Notes: Three-septate basidiospores are extremely rare in specimens from both Finland and Germany; perhaps, septation occurs only immediately before germination.
Habitat and distribution: Gymnosperm (Abies, Picea, Pinus, and unident.) and perhaps angiosperm wood; East Asia, North America.  Notes: Dacrymyces corticioides can be identified by conspicuous circular coalescing basidiocarps with white margins that are fimbriate in dry state. Basidiospores are mostly aseptate but become 3-septate at maturity. Majority of the specimens were reported from gymnosperm wood with a single possible exception of TAAM126607. Sequences from TAAM102301, TAAM126607, and TAAM150056 were not used in the analyses, but they are similar to the other included sequences.
Notes: The species has averagely the largest basidiospores and basidia among its North American relatives, as well as the most abundant and prominent hyphal pegs. Description: Basidiocarps appear as small white subicular patches, later coalescing. Well-developed basidiocarps can cover up to 10 cm in the longest dimension, with thick, solid, waxy-gelatinous, yellow to orange hymenial surface, thick cottony subiculum, white fimbriate margins, and abundant lightcoloured hyphal pegs or denticles up to 300 μm in height, oriented downwards. Hyphae without clamps, subicular hyphae 2.5-4.5 μm in diam, with some compartments and hyphal junctions swollen up to 5 μm in diam, walls 0.5- Notes: Dacrymyces grandinioides and D. venustus are morphologically cryptic African species. However, their ITS sequences are distinct enough for easy delimitation: for example, D. grandinioides lacks insertion AACGTA in the end of 5.8S region, which is characteristic to the other species. The type specimen of D. grandinioides has smaller basidiospores and more subtle basidiocarps compared to the recent collections, which are also of much deeper orange colour. In the type we recorded cylindrical to curved-cylindrical conidia 3.7-4.1 × 1.3-1.8 μm but were unable to confirm their origin. Habitat and distribution: Angiosperm wood (Chusquea and unident.); South America (known only from the type locality). In the absence of sequenced material, and considering the high morphological variability in the group, we limit the scope of the species exclusively to the Ecuadorian collections.
Edward A. Burt labelled but never published "Ceracea triseptata Burt, n. sp.", that was intended to be different from Ce. lagerheimii by 3-septate basidiosporesthe protologue of Ce. lagerheimii mentions only 1-septate sporesand crystalline matter in basidiocarps (Trinidad, 1912-1913. It is not known if the specimen stands for a distinct taxon, but small spores and presence of crystals place it closer to D. cereus. Ginns (1982) suggested to separate Cerinomyces lagerheimii from C. ceraceus and C. grandinioides on the basis of its smooth surface, gelatinization of hyphae in 2 % KOH, and predominantly 1-septate basidiospores. While the absence of pegs is indeed a key feature, the latter two characters can vary depending on the condition and treatment of a specimen, or development stage, respectively.
Habitat and distribution: Unidentified, probably angiosperm wood; South America (known only from the type locality).
Notes: The species is morphologically similar to D. grandinioides, and the easiest way to distinguish them is to compare ITS regions. The type was collected from deeply charred wood. Notes: The taxon is almost indistinguishable from D. venustus and D. grandinioides. Considering this and also scanty molecular data available for analysis, we do not erect a formal species. The studied specimen also includes Asterostroma species (Russulales). Notes on taxa historically or morphologically related to Cerinomyces s.l.
The species probably belongs to the Cerinomycetaceae, but its precise position cannot be resolved, since the isotypes are extremely scanty and fresh collections are absent. It has a unique substrate among all dacrymycetes, growing on pine cones. The variety used to be associated with D. tortus (Kennedy 1958b), but according to the protologue and revision of , it possesses at least 3-septate basidiospores and thus belongs to the Dacrymycetaceae, probably as a synonym to D. adpressus. The variety has prominent branched hyphidia like many taxa in the Cerinomycetaceae, but considering regularly 3-septate thickwalled basidiospores, it probably belongs to the Dacrymycetaceae as a relative of D. paraphysatus. It was first published in a not effective way as D. enatus var. brunnescens (Kennedy 1957).
number of septa in basidiospores, Ster.
One of the isotypes (H6014412) shows a large number of basidiospores distinctly smaller to what was reported by , which can be a sign of spore immaturity. Here we report the data anyway, to claim for attention on this character, which suitability may need to be re-evaluated with further material. Table 5. Species with clampless septa and brightly coloured corticioid basidiocarps, waxy-gelatinous when wet (the Dacrymyces grandinioides group). Ster.

DISCUSSION
Previous phylogenetic studies have settled the boundaries of Cerinomycetaceae, separated it from neighbouring families, and shown that Cerinomyces s.l. is polyphyletic (Shirouzu et al. 2013. In this work we took the next logical step: a formal reassessment of the family, using both molecular and morphological data on all known taxa. The number of new taxa was unexpectedly high considering how rarely the genus occurs in nature, and how rarely it is collected. Many new species, particularly arid Cerinomyces, were recovered by extensive herbarium samplingthis, once again, points to the importance of natural history repositories. Only gelatinous Cerinomyces species are sufficiently represented with the recent specimens, while the rest of the genus is covered primarily by material from 1955-1985 collected by experts of "holobasidial" corticioid fungi. In many fungal groups, specimens of this age are considered to be unfit for sequencing, but Sanger sequencing of certain DNA regions (e.g., parts of the nrDNA, such as ITS) of dacrymycetes has a relatively high rate of success. Such DNA preservation may be facilitated by the ability of dacrymycetes to survive dry periods in a desiccated state and then restore sporulation after humidity increases (Shirouzu et al. 2013). We suppose that gelatinization of basidiocarps, a landmark character in the class, is connected to this ability, and gelatinous matter serves as a water storage and protection for hyphal structures. Concluding from the phylogenies, this character was independently lost several times and substituted by corticioid basidiocarps of different degrees of aridness. The ecological background of this change is yet to be understood, but we hypothesize that the transitions can be linked with adjustments to micro-niches. For example, we observe that the gelatinous Cerinomycetaceae species grow mostly on fine, hard branches, while corticioids prefer larger, more decayed trunks. In this context, basidiocarps appearing on small, often suspended debris are at risk of rapid drying that interrupts sporulation, and therefore they depend on gelatinization for revival. Meanwhile, species preferring stably wet habitats like half-decomposed logs do not need to prepare for several sporulation episodes. In this case the corticioid morphotype is beneficial, providing a larger hymenial surface and relief from production of voluminous gelatinous matrices.
Compared to the last revision (McNabb 1964), the main changes made to Cerinomyces s.s. are: the exclusion of the brightly pigmented femsjonia-like corticioids and species lacking hyphal clamps, and the incorporation of dacrymyces-like members with gelatinous pale ochraceous and brownish basidiocarps. Following this pattern, a specimen can be rather reliably identified as Cerinomyces without a microscope just by the visibly low carotenoid content, regardless of whether it is gelatinous or corticioid. When working with gelatinous species, one should also consider Unilacrymaceae and Dacryonaemataceae, but basidiocarps in these families are generally much smaller than in Cerinomyces. There are a few species in the Dacrymycetaceae with brownish basidiocarps (e.g., Calocera fusca Lloyd, Guepiniopsis fulva P. Delivorias), but they have unusual shapes for the Cerinomycetaceae. Lack of colour or semitransparency are not rare in gelatinous dacrymycetes, but this normally appears when the basidiocarps have developed in darkness (Bulat 1954, Vail & Lilly 1968. This is held to be a stable character in only a few species (e.g., Ditiola haasii Oberw., Dacrymyces cylindricus Shirouzu), which differ from Cerinomyces by basidiocarp morphologies. Discoloured morphs in Cerinomyces were seen in C. cokeri (Fig. 11 C) and C. tortus, but can be expected in other species growing in the shade or during the darkest part of the year.
Introducing new taxa to Cerinomyces s.s. is relatively straightforward now that the family Cerinomycetaceae consists of one genus only. At the same time, transfer of brightly coloured corticioids to the Dacrymycetaceae raises the question of how to choose generic names for them. Genus division in this family is clearly a difficult task, complicated by the poly-and paraphyly of the current genera, numerous morphological homoplasies, and lack of molecular data. These uncertainties explain the minimalistic approach to nomenclature we have adopted here. First, for the Dacrymyces corticioides and "Cerinomyces" canadensis lineages we decided to postpone new combinations until the upcoming studies of Femsjonia. Second, we included the D. grandinioides clade members to the genus Dacrymyces, judging from their close position to Dacrymyces s.s. in the core Dacrymycetaceae. The alternative solution of establishing a new genus is impractical from our viewpoint. Considering the diversity of undescribed taxa in the vicinity of the D. grandinioides clade, premature generic splitting may only multiply the problems in this part of the family. Moreover, lack of a common stark trait in Dacrymyces should not be a sole reason for splitting. Morphological plasticity even within the well-defined and monophyletic genera of dacrymycetes seems to be the norm rather than the exception. Unilacryma has members with 1-and 2-sterigmate basidia, Dacryonaema includes taxa with synnematous and pustulate basidiocarps, and Cerinomyces now unites arid and gelatinous species. In the same way, Dacrymyces can be potentially redefined as both monophyletic and morphologically inclusive genus.
For the convenience of identification, we provided detailed information on all dacrymycetous corticioids from the two families. In certain morphogroups many of the reviewed traits are not diagnostic alone, and instead a combination of characters is needed for identification. This prevented us from extending the dichotomic key to the species level in all cases. As a substitution, identification tables (Tables 3-5) should be sufficient for most identification purposes. However, there are a few morphological species groups that entail difficulties in identification of young or weathered basidiocarps. These include North American corticioids C. atrans, C. favonius, C. fugax, and C. tristis; European gelatinous C. creber, REVISION OF CERINOMYCES WITH NOTES ON RELATED TAXA www.studiesinmycology.org C. lipoferus, C. neuhoffii, and C. tortus together with undescribed taxa; and possible, yet uncovered, complexes of European angiosperm-dwelling members related to C. crustulinus or C. aeneus. The D. grandinioides clade is particularly difficult, and in many cases morphological identification within the group seems impossible. While the South American peg-lacking species D. cereus, D. lagerheimii and D. pulchrus can be distinguished using spore size and septation, identification of their peg-bearing relatives is challenging. For both the African complex of D. grandinioides, D. venustus, and D. aff. venustus 1; and North American complex of D. burdsallii, D. grandii, D. ceraceus, and D. sobrius, we detected wide variation and overlap in trait measurements. More material is needed to bring higher morphological resolution to the group and provide new ways for identification. For example, many specimens identified in the D. grandinioides group demonstrate abundant conidia and conidiogenous structures of largely unexplored diagnostic value. Anamorphic development has a high importance in some species in dacrymycetes, up to apparent loss of teleomorphic stages (members of Dacryoscyphus), hence this area should be assessed in taxonomic studies in more detail.
We argue that biogeography is a working determination criterion even for species based on a single specimen or with exceedingly restricted known localities. Species ranges in the studied groups are likely to be confined to continents and biogeographical zones, judging from the overall low number of cosmopolite species in dacrymycetes. Only three species in this study have a multicontinental distribution, found in both East Asia and North America ("Cerinomyces" canadensis, C. enatus, D. corticioides). Notably, there are no findings of Cerinomyces from the African region, except for C. albosporus found at the island of R eunion, and no gelatinous species are recorded in South America. Conversely, corticioid species from the D. grandinioides clade were found only in Africa and the Americas, which is unlikely to be caused by sampling density alone.
Tree host specialization in Cerinomyces s.l. is generally limited to angiosperm or gymnosperm wood; association with a certain tree species cannot be confidently established. The majority of Cerinomyces s.s. prefer coniferous wood, except for C. aeneus, C. crustulinus, and C. pallidus that occur only on angiosperm trees; C. albosporus on Asteraceae shrubs; C. enatus on both angio-and gymnosperms; and a few more species with dubious preferences or unidentified hosts. Shirouzu et al. (2014) provided additional insight on host specificity with their wood block decay test, which confirmed the ability of C. enatus to decompose both Pinus densiflora and Fagus crenata (data under C. canadensis in Shirouzu et al. 2014, but also mind our note under C. enatus). At the same time, members of the D. grandinioides clade seem to occur on a high variety of substrates including shrubs, bamboos, charred wood, etc. Hence, it is difficult to judge host preferences, particularly when only few specimens are available.
Spore measurements have always been a principal component of dacrymycete identification. However, not all of the spore characteristics are equally useful. The maximal number of internal septa in basidiospores should be used with caution: this number fluctuates within one specimen and changes during the basidiocarp lifetime. It is very common that in fresh specimens basidiospores are aseptate while still attached, but after discharge they can massively generate 1-3 septa before germination (Maekawa 1987). This quality led to errors in protologues of C. albosporus, Dacrymyces confluens, and D. lagerheimii for which only one septum was reported, instead of the actual maximal three. In case of uncertainty, we strongly recommend checking the number of septa with germinating spore prints on agar. If this material is not viable, one can look for discharged basidiospores in substrate scrapes close to the basidiocarp.
On a methodological note, we observed that microscopic slide mountants affect the measurements and shapes of microstructures. In this work we used Cotton Blue (CB) in lactic acid, providing sufficient contrast and stability of microstructures. Notably, use of KOH-based solutions in concentrations conventional for studies of aphyllophoroid fungi can lead to unreproducible results, as gelatinous matter that covers hyphae can thicken or even dissolve depending on the concentration of KOH and exposure time. Spore walls easily swell, most dramatically in thick-walled basidiospores of the Dacrymycetaceae, adding several micrometers to spore dimensions. On the other hand, preparations with KOH (with or without a stain) help to reveal the organization of densely interwoven hyphae in cases where CB fails. Therefore, for measurements in dacrymycetes we recommend using CB as a primary mountant and a weak solution of KOH (e.g., 1 %) where necessary.
Aiming to reduce number of superfluous terms describing fungal morphology, we propose to replace the word "dikaryophysis" (sterile element in hymenium of dacrymycetes) with the more general term "hyphidium". The term "dikaryoparaphysis" was introduced by Lowy (1954) as a name for sterile hymenial hyphae in all basidiomycetesin such a manner the author wanted to counterpose it to ascomycetous monokaryotic "paraphysis". Kennedy (1957) started a tradition to use it in a simplified form: "dikaryophysis". The term has never been widely used and apparently survived only in dacrymycete-related works, adding unnecessary confusion. Here we report hyphidia only when they are dendroid (the C. tortus, C. enatus clades), simple with a thickened base and long thin apical part (the C. pallidus clade), or simple cylindrical of even width (the D. corticioides lineage). We would not use "hyphidia" to describe occasional subhymenium extensions or young cylindrical probasidia that can be found in most dacrymycetes.
Further sequencing of more gene regions in the leastrepresented groups of the Cerinomycetaceae will allow for better-supported phylogenies. Building a robust multigene phylogeny for the Dacrymycetaceae will help to refine nomenclature of former Cerinomyces species if they need to be treated as separate genera. Continuing collection efforts and herbarium searches in unidentified Dacrymyces will certainly uncover more new species. This applies not only to understudied regions, but also to Europe, where knowledge on diversity of gelatinous species is still incomplete. Higher sampling density and established distributions will facilitate recognition of endangered species as potential red lists candidates. Ecological roles and requirements of dacrymycetes are poorly understood and deserve be studied more extensively. For example, several dacrymycetes including C. tortus are abundant in the pine wood-decay communities in the Northern boreal zone, but their habitat specialization strategies, if any, are mostly unknown. Environmental DNA obtained from wood will be invaluable in expanding both the taxonomical and ecological knowledge of this group. Finally, genomic and enzymatic studies in Cerinomycetaceae and minor families of