Four new genera and six new species of lyophylloid agarics (Agaricales, Basidiomycota) from three different continents

Four new genera encompassing six new species are described in the Lyophyllaceae (Agaricales): the monotypic Australocybe from Australia; the monotypic Phaeotephrocybe from Belize; the monotypic Nigrocarnea from Laos containing the novel conidia-producing species N. radicata and Praearthromyces containing two Asian taxa, the conidia-producing P. corneri known from Malaysia and Singapore and P. griseus from Thailand, which lacks conidial production. In addition, we describe Arthromyces glabriceps, which is the first species in the genus Arthromyces that lacks conidial production on the basidiome. Alternative options for a classification of these lyophylloid taxa are discussed and based on morphological, ecological and biogeographical considerations rejected.


Introduction
The genus Tephrocybe as introduced by Donk (1962), as a segregate from the genus Lyophyllum P. Karst., contains a number of small collybioid mushrooms, often with a hygrophanous pileus and grey colours (Donk 1962). Various molecular studies have shown the genus to be paraphyletic, which is consistent with the lack of diagnostic features (Hofstetter et al. 2002;Hofstetter et al. 2014;Bellanger et al. 2015). The most recent study which included several members of Tephrocybe found eight distinct evolutionary origins within the Lyophyllaceae (Bellanger et al. 2015). Several Tephrocybe species have been transferred back to Lyophyllum, or have been placed in new genera such as Myochromella, Sagaranella and Sphagnurus (Hofstetter et al. 2014). A previous study showed that the type of the genus, Tephrocybe rancida, is part of a termitomycetoid clade which, next to the termite-associated genus Termitomyces, also contains the insect-faecal associated genera Arthromyces and Blastosporella (van de Peppel et al. 2021). The study by van de Peppel et al. also showed that five different species, which were provisionally identified as Tephrocybe sp. (s.l.), are also part of this termitomycetoid clade. In the present study, we examined these five unnamed Section Editor: Zhu-Liang Yang * Lennart J. J. van de Peppel lennart.vandepeppel@wur.nl species and an additional specimen of an unnamed species of the genus Arthromyces. In order to test whether any of the currently accepted members of the genus Tephrocybe are part of the termitomycetoid clade, several additional fungarium specimens of known species were included in the analysis. A specimen from Australia (Tephrocybe sp. 1) is macroscopically similar to T. rancida but phylogenetically distinct, occupying an isolated branch in the termitomycetoid clade. Because of the strong phylogenetic evidence and to conserve monophyletic genera, we propose to place this species in the monotypic genus Australocybe. The sole member of this genus, A. olivacea, is described below.
The study by van de Peppel et al. (2021) recovered a subclade within the termitomycetoid clade containing T. rancida, Blastosporella zonata and Tephrocybe sp. 2 and 3. Fieldwork in Laos in connection with the production of the first booklet on edible, poisonous and medicinal fungi from this country (Laessøe et al. 2019) led to the recognition of Tephrocybe sp. 3, a species with internal conidia in the pileus context. Tephrocybe sp. 2 is a species from Belize, similar to T. rancida, which lacks the production of conidia on the basidiome. Due to the lack of shared characters among these two species and the genus Blastosporella, we propose to erect two monotypic genera; Phaeotephrocybe and Nigrocarnea. Tephrocybe sp. 3 is described below as Nigrocarnea radicata and Tephrocybe sp. 2 is described below as Phaeotephrocybe leucophylla. Baroni et al. (2007) reported a close morphological similarity of the palaeotropical/Asian Tricholoma furcatifolium to the Neotropical/American genus Arthromyces. Examination of Tephrocybe sp. 5 from the study by van de Peppel et al. (2021) showed this species to be very similar to T. furcatifolium. This species together with Tephrocybe sp. 4 make up a small clade sister to the Arthromyces-Termitomyces clade, and these two species thus are not related to the ectomycorrhizal genus Tricholoma. In addition, we examined the type collection of T. furcatifolium. A DNA sequence generated from one of the paratypes of T. furcatifolium shared 99% identity with Tephrocybe sp. 5, confirming that this species belongs to the Lyophyllaceae. The holotype of T. furcatifolium does not produce arthroconidia on the basidiome and therefore the holotype and paratype may represent different species. Therefore, we described Tephrocybe sp. 5 as a new species; Praearthromyces corneri. As both this species and Tephrocybe sp. 4 do not nest phylogenetically with any of the described genera within the termitomycetoid clade, we propose to place them in the new genus Praearthromyces and describe Tephrocybe sp. 4 as Praearthromyces griseus.
The genus Arthromyces currently encompasses two species. A. matolae was reported from Belize and A. claviformis was originally reported from Puerto Rico and the Dominican Republic (Baroni et al. 2007). Recently, A. claviformis has also been observed in Colombia near Cali (mushroomobserver. org, observations: 275090, 348275, 367362) and Cogua (observation 363150). A recent soil metabarcoding study identified OTU's as Arthromyces sp. from the Andean Yungas forest of Argentina (Eduardo et al. 2018). This indicates that the genus Arthromyces has a distribution range spanning the entire continent of South America and extending to the Caribbean islands and Central America. Furthermore, A. matolae has been collected as far north as Gainesville, Florida (van de Peppel et al. 2021). Although the production of dark arthroconidia on both the stipe and pileus of the basidiome is one of the key features of Arthromyces, we examined several specimens identified as Arthromyces sp., which lack conidia on the basidiome. Our molecular phylogenetic analysis shows that these specimens nest within the genus Arthromyces and is sister to A. matolae. These specimens are described as Arthromyces glabriceps below.

Morphological analysis
Colour codes used in species descriptions refer to Kornerup and Wanscher (1978). Presence of hymenial cystidia is examined in longitudinal sections of lamellae and additionally for cheilocystidia in whole mounts of lamellar edges of at least 1 mm in length. All measurements of microscopic structures were made under 1000× magnifications.

DNA extraction and PCR
For DNA extraction, small fragments of dried fungarium specimens were used. DNA was extracted following a modified cetyltrimethylammonium bromide (CTAB) protocol (Nieuwenhuis et al. 2019). For known members of the genus Tephrocybe, the nuclear ribosomal region containing the internal transcribed spacer 1 (ITS1), 5.8S and the internal transcribed spacer 2 (ITS2) was amplified using a conventional polymerase chain reaction (PCR). For the five unnamed species of Tephrocybe and the unnamed species of Arthromyces, a partial sequence of the nuclear ribosomal large subunit (LSU, 28S) and the sequence encoding for elongation factor 1-alpha (EF1α) were also amplified using PCR. The following primer pairs were used: the fungal-specific forward primer ITS1f and the universal reverse primer ITS4 for the ITS marker (White et al. 1990;Gardes and Bruns 1993), LROR and LR5 for the LSU (Vilgalys and Hester 1990) and EF595F and EF1160R for EF1α (Kauserud and Schumacher 2001). After initial amplification of the holotype of T. furcatifolium failed, a PCR was conducted targeting ITS1 and ITS2 separately using the primer pairs ITS1f/ITS2 and ITS3/ITS4 (White et al. 1990;Gardes and Bruns 1993).
All PCR programs were conducted as described previously (van de Peppel et al. 2018). Sequencing was performed by Eurofins genomics (Ebersberg, Germany) using both the forward and reverse primer. For LSU only, the forward primer (LROR) was used as the forward read was of sufficient quality for subsequent analysis. Inspection of the chromatograms and assembly of the forward and reverse sequences was performed using the CLC Genomics Workbench 8. All new sequences were deposited in Genbank (Supplementary Table 1).

Phylogenetic analysis
To test whether any of the current members of the genus Tephrocybe are part of the termitomycetoid clade, ITS sequences obtained from various species of Tephrocybe (Supplementary Table 1) were compared using NCBI BLAST, to a dataset consisting of all currently known members of the termitomycetoid clade and a representative number of species from the Lyophyllaceae. Since the phylogenetic relationships within the termitomycetoid clade already have been estimated and are well supported, based on a phylogenetic reconstruction on a dataset of 1131 orthologous genes (van de Peppel et al. 2021), a three-gene phylogeny would not add any additional phylogenetic resolution. We therefore limited our phylogenetic reconstruction of the termitomycetoid clade to specimens of the genus Arthromyces, since A. glabriceps was not included in the study by van de Peppel et al. Sequences for each individual marker (ITS, LSU and EF1α) were aligned using MAFFT v7.409 online alignment software, using default settings (Katoh et al. 2019). The three resulting alignments were trimmed and concatenated in Geneious 10.0.9 (Kearse et al. 2012). A maximum-likelihood tree was reconstructed using the webserver of IQ-tree with default settings and 1000 ultrafast bootstrap replicates (Trifinopoulos et al. 2016).

Morphology
All taxa described here share the following microscopic features: basidiospore walls are inamyloid but evenly cyanophilic, basidia and basidioles possess siderophilic and cyanophilic bodies (Baroni 1981) typical of all members of the Lyophyllaceae.

Phylogenetic analysis
The only known species of Tephrocybe which had a close match to a member of the termitomycetoid clade was T. oldae. The ITS sequence of this species was very similar (99.72%) to that of our T. rancida specimens (OM905964 and OM905967). The study by van de Peppel et al. (2021) found significant support for all branches within the termitomycetoid clade, resulting in a fully resolved tree (Fig. 1A). Our phylogenetic reconstruction on Arthromyces specimens shows significant bootstrap support for A. glabriceps being sister to a clade containing A. matolae and A. claviformis (Fig. 1B).

Taxonomy
Australocybe olivacea T.J. Baroni Diagnosis: Tephrocyboid in stature with the habit of T. rancida but differing significantly by the dark fuscous brown colour of the pileus, olive-grey lamellae, pale golden-olive coloured pseudorhiza, presence of cheilocystidia and pleurocystidia with dark golden-brown resinous pigments inside of and also extruded from these cystidia and distinct molecular profile of its combined ITS, LSU and EF1a sequences.
Cheilocystidia abundant (lamella edge sterile), lageniform, ventricose-rostrate, rostrum sometimes capitate, also some subglobose to ovate and pedicellate, 30-45 × 10-16 μm, mostly with dark golden brown resinous coagulated pigments in the apical regions of the cystidia, sometimes completely filled with resinous pigments (somewhat like chrysocystidia); this golden resinous pigment can be extruded and coating the apex of the cystidia. Pleurocystidia scattered but fairly common, as cheilocystidia in shapes and frequently filled with golden brown resinous pigment. Lamella trama pale to moderately brown from parietal pigments in the hyphal walls, many hyphae also finely to moderately encrusted with brown pigments, hyphae parallel, cylindrical or inflated, 2.5-10 μm in diam. with abundant internal cyanophilic bodies. Pileipellis a cutis of repent dark brown strongly encrusted hyphae, hyphae 3-9 μm in diam, producing scattered or clustered short   Discussion: Australocybe is the earliest diverging lineage of the termitomycetoid clade and the only taxon with clamp connections in combination with cheilo-and pleurocystidia with resinous contents. Tephrocybe and Blastosporella also produce clamp connections (Table 1), while all the other taxa in this clade leading up to Termitomyces, lack clamp connections on their hyphae. The presence of resin producing hymenophoral cystidia are also distinctive for Australocybe, and found in species of Arthromyces so far, although refractive resinous particles can be found on basidia and basidioles in Nigrocarnea radicata as well. Presumably these resin encrustations would be a deterrent to hymenophore grazing fungivores, thus allowing for prolonged basidiospore production.
It is not clear at the present time if this old world taxon is associated with insects, but the presence of a welldeveloped pseudorhiza suggests this possibility since other members of the clade, i.e. Blastosporella, Arthromyces and Termitomyces, have evolved this symbiotic (or distinctive) relationship. Future investigations are needed to determine if Australocybe, Nigrocarnea, Praearthromyces, and Phaeotephrocybe all share some type of insect association. Diagnosis: Tephrocyboid in stature and because of the long pseudorhiza similar to the habit of T. rancida, differing however by the striking black, pulverulent pileus flesh consisting of chains of smooth, dark-coloured arthroconidia and the absence of clamp connections in all tissues, and by the distinct molecular profile of its combined ITS, LSU and EF1a sequences.
Basidiospores 4.5-8 × 3.0-5.0 μm, av. 5.4 × 3.7 μm, Qav. 1.48 (N = 50), not abundant and difficult to find in mounts, ellipsoid, smooth. Basidia 15-23 × 6-9 μm (N = 16), clavate, 4-sterigmate. Basidia and basidioles moderate to densely Discussion: Molecular data shows that N. radicata is closely related to Blastosporella zonata (Fig. 1). Morphologically, they look somewhat similar as they both produce conidia in/ on the pileus and have a rooting stipe. However, they can be distinguished by the type and position of conidia as N. radicata produces chains of arthroconidia in the pileus context and B. zonata produces ornamented blastoconidia in small spore heads on the surface of the pileipellis. These two species have no overlap in distribution as N. radicata occurs in the paleotropics while B. zonata occurs in the neotropics. An ITS sequence generated from an additional collection UDB034575/HNL501934 confirms that this collection belongs to the same species. These collections were made some 10 km apart in the same area, forest type and altitude and with similar macroscopic and microscopic features.
Because of the reported insect-faecal association in the closely related B. zonata (van de Peppel et al. 2021), we attempted to dig down to find the nutrient substrate to which the pseudorhiza was attached, but it was not possible to follow the very thin strand deeper than about 10 cm in hard soil. The type locality was very rich in ectomycorrhizal fungi and also in species of Termitomyces. Diagnosis: Tephrocyboid in stature and somewhat similar in appearance to Tephrocybe rancida but distinguished from this grey coloured species by the brown colours of the pileus and stipe and the white gills. It also differs from Blastosporella zonata, Nigrocarnea radicata and species of Arthromyces by the absence of dark blasto-or arthroconidia in/on the pileus and from species of Praearthromyces, and all species and genera discussed here, by its distinct molecular profile of combined ITS, LSU, and EF1a sequences.
Discussion: Phaeotephrocybe leucophylla is a neotropical species that is similar in general appearance compared to the European/North American Tephrocybe rancida but differs most notably in colour of the pileus, stipe and lamellae. Phaeotephrocybe leucophylla, does not produce conidia on its basidiomata, but is phylogenetically related to the arthroconidia producing Nigrocarnea from the paleotropics an d t he b la stospo re co nidium pr oduc ing ge nus

B l a s t o s p o r e l l a f r o m t h e n e o t r o p i c s ( F i g . 1 A ) .
Phaeotephrocybe leucophylla has no clamp connections, a feature shared with its phylogenetic sister taxon Nigrocarnea radicata, and all taxa (Praearthromyces, Arthromyces, Termitomyces) in the evolutionary lineage leading to the termite farmed agarics. Clamp connections are found in the three early branching genera in the termitomycetoid clade, Australocybe, Tephrocybe and Blastosporella.  Diagnosis: Species with tephrocyboid or collybioid or mycenoid habit with narrow, densely crowded lamellae, with thin, deeply rooting pseudorhiza, no clamp connections, cheilocystidia variously shaped and mostly lacking internal pigments or resinous materials, and phylogenetically distinct by molecular profile of its combined ITS, LSU and EF1a sequences. Diagnosis: Distinguished by densely caespitose collybioid basidiomata with long connate pseudorhizae, densely crowded, narrow and frequently forked lamellae, and concentrically white and grey zonate pileus with darker zones producing profuse dry arthroconidia, and by the distinct molecular profile of combined ITS, LSU, and EF1a sequences.
Pileus 10-25 mm wide, applanate to concave or slightly infundibuliform, disc depressed or subumbilicate, margin undulate, smooth, surface strongly velutinous but densely pruinose when older and staining fingers when handled with olive grey dusty conidia, concentrically zoned in white, dark and light grey, the darker zones covered with chains of arthroconidia. Context dull white or pale grey, soft. Lamellae densely crowded, narrow (less than 1 mm), often forked near attachment but also in the middle or near the pileus edge, pale grey or grey, edges white and fimbriate, finely eroded. Stipe 5-13 × 1-3 mm, cylindrical or compressed and twisted, gradually tapering to a long pseudorhiza, pale to dark grey, pale greyish over the pseudorhiza, surface strongly pruinose on above ground parts.
Discussion: Species in Praearthromyces closely resemble Arthromyces by having a pseudorhiza and in the case of P. corneri by producing chains of dark arthroconidia on the basidiome. However, phylogenetic reconstruction shows a sister relationship of Praearthromyces to the Arthromyces-Termitomyces clade (Fig. 1) (van de Peppel et al. 2021). Placing P. corneri in the genus Arthromyces would make the latter genus paraphyletic or would require Arthromyces to be synonymized with Termitomyces, which is not warranted because of the lack of a termite association (see also discussion at the end). The genus can be distinguished from Australocybe, Blastosporella and Tephrocybe by the lack of clamp connections, a feature that is shared with Arthromyces, Phaeotephrocybe, Nigrocarnea and Termitomyces. Praearthromyces and Arthromyces can be distinguished by their geographic origin, Arthromyces from the new world and Praearthromyces from the old world, or by DNA sequence data.
The study by van de Peppel et al. (2021) showed that this taxon is part of a two species clade, sister to the Arthromyces-Termitomyces clade (Fig. 1). Corner (1994) described Tricholoma furcatifolium as being characterised by very crowded, narrow, regularly furcate lamellae (Corner 1994). The description does not refer to the production of dark arthroconidia on the basidiome. Unfortunately, we were unable to obtain a sequence from the holotype. However, another collection of T. furcatifolium, a paratype (Singapore, Gardens Jungle, 22 November 1940 made by Corner and illustrated by a painting in the original description) has abundant arthroconidia on the pileus surface and we managed to obtain an ITS sequence from this collection; this sequence was 99% identical to the sequence we obtained from a more recent collection by Dirk Stubbe, DS06 151, made in the Pasoh National Forest Reserve in Malaysia. We are unable to resolve whether the collection designated as the holotype in the protologue by Corner, Singapore, Corner s.n., 30 August 1939, and this second collection, a paratype with arthroconidium producing basidiomata, made by Corner, represent different species or morphological variation within this species. Stubbe's image below and the hand painted icons of Corner's arthroconidium producing paratype look very much alike. All the microscopic features of the basidiomata are also alike in Corner's arthroconidium producing paratype from Singapore and Stubbe's collection from the Pasoh National Forest in Malaysia. As we could not determine whether Corner's holotype and paratype collections represent the same or different species, we chose to describe our collection as the new species P. corneri. Diagnosis: With a mycenoid or somewhat tephrocyboid habit similar to Tephrocybe rancida but differing by having cheilocystidia and no clamp connections, and is phylogenetically distinct from Tephrocybe and other morphologically similar taxa by the molecular profile of its combined ITS, LSU and EF1a sequences.
Discussion: Praearthromyces griseus is distinguished from P. corneri by its more delicate mycenoid habit, lack of zonate pileus surface, the lack of arthroconidia, by the unforked lamellae, and by not being caespitose. Praearthromyce griseus is currently only known from a single collection made in Thailand. Etymology: glabri-ceps Latin, for the glabrous pileus. Diagnosis: Similar in morphology to Arthromyces matolae, but differing mainly by the lack of arthroconidia production on the pileus surface.
Discussion: Arthromyces glabriceps differs from A. matolae and A. claviformis in lacking the production of conidia on the pileus, and in its known distribution; A. glabriceps is apparently restricted to the Guiana Shield, known only from the western Pakaraima Mountains of Guyana, and from Brazil. Van de Peppel et al. (2021) reported the presence of conidiophores in laboratory cultures of A. matolae. Since a culture for A. glabriceps was not available, we could not determine the presence of conidiophores in A. glabriceps. We have no reports for A. glabriceps growing on insect-faecal pellets in the same fashion as has been reported for A. matolae and A. claviformis (van de Peppel et al. 2021). However, a collection from Pará, Brazil, was found to be growing on faecal pellets. This collection could not be unambiguously identified as A. glabriceps and DNA sequence data is also lacking.
Distribution and habitat: In Dicymbe corymbosa monodominant forest or adjacent mixed forest; despite

Discussion
A recent phylogenomic study, based on 1131 core nuclear genes, included five of the six newly described species (as Tephrocybe sp1-5.) (van de Peppel et al. 2021), while the sister group relation between Arthromyces matolae and A. glabriceps was based on EF1α and the ITS and LSU of the rDNA produced in this study. The study by van de Peppel et al. already showed that geography and morphological features such as conidial production, clamp connections and insect-faecal associations are not consistent with the clades recovered from their phylogeny (see also Table 1). The use of these characters as sole foundation for a generic classification that is consistent with the phylogeny is therefore hardly possible. Translating the phylogeny into a new generic classification, primarily based on the genetic data with morphological and biogeographical criteria as additional data, allows a number of possibilities, ranging from one genus for the complete clade to five genera. In order to evaluate these various options, we used the six criteria proposed by Vellinga et al. (2015). These six criteria are (1) monophyly, (2) sufficient phylogenetic coverage, (3) sufficient branch support, (4) discussion of alternative options, (5) no single marker phylogeny  (6) include all supporting evidence and background information. To those criteria we add one additional consideration, viz., the desirability that the approach taken for this clade within the Lyophyllaceae is consistent with earlier treatments of that family (Bellanger et al. 2015;Hofstetter et al. 2014).
Criteria (1), (2), (3), (5) and (6) have been amply discussed in Van de Peppel et al. (2021) and all background information on molecular data is available in the public domain. We refer to that study for the information and concentrate on criterion (4), the evaluation of alternative classifications.
Three main taxonomic proposals are consistent with the phylogeny: (A) put all species in one genus, which should be called Termitomyces; (B) recognise eight genera, four of which are newly described, as proposed in our paper; (C) recognise three genera, viz., Australocybe (for A. olivacea, as this species cannot be placed in another genus without making that genus paraphyletic, except when placed in Termitomyces, our option (A)), Tephrocybe (including Blastosporella, Phaeotephrocybe and Nigrocarnea-as the group is monophyletic), and Termitomyces (including Arthromyces and Praearthromyces). Further options, e.g., combining Termitomyces and Arthromyces, while recognising Praearthromyces as a separate genus; or combining Phaeotephrocybe and Nigrocarnea, while treating Tephrocybe and Blastosporella as autonomous species would also be consistent with the phylogeny but would result in the description of three new genera, hardly different from our proposal for four new genera.
The most radical proposal would be to recognise one genus that is then called Termitomyces. For mycologists from the temperate region, it would certainly be curious that a species that is never associated with termites (and its distribution area largely excludes the distribution area of termites) has to be placed in Termitomyces. Further disadvantages of that proposal would be that the genus has no unique morphological characters (the rooting stipe equally occurs in some more distantly related Lyophyllaceae), and, more importantly, is extremely heterogeneous in its ecological lifestyles. Lifestyles and lifestyle transitions have been considered as major arguments for a taxonomy of the early-branching Lyophyllaceae, as proposed by Hofstetter et al. (2014) and Bellanger et al. (2015), and where generic distinctions like Sphagnurus (parasitic on Sphagnum) and Sagaranella (nitrophilic, known as ammonia fungi) are partly justified by the ecological transition. It is also noteworthy that many genera in these early-branching Lyophyllaceae are very species-poor, often containing one to three species. Previous studies have found Tephrocybe (a species-rich genus in the pre-molecular era) to be paraphyletic with up to eight independent origins within the Lyophyllaceae, with one species (T. rancida) in the termitomycetoid clade (Hofstetter et al. 2014;Bellanger et al. 2015) and both studies have contributed to the dismemberment of this paraphyletic taxon by describing new species rather than by including members of other genera in a very largely circumscribed Tephrocybe. Recognising one very broadly circumscribed Termitomyces next to a large number of species-poor genera in the early-branching Lyophyllaceae would therefore create an unbalanced familial taxonomy. We consider these arguments sufficient to reject that first, radical alternative. Rejecting this option would automatically imply the recognition of a new genus Australocybe. While A. olivacea, is macro-morphologically similar to T. rancida, phylogenetic analysis shows that it does not nest in the same clade as T. rancida but occupies an isolated branch (van de Peppel et al. 2021).
The third option, which would necessitate the erection of one new genus only (Australocybe) as the other new species are accommodated in the existing genera Tephrocybe and Termitomyces, would run into similar problems as our radical alternative. Putting the species of Arthromyces and Praearthromyces in Termitomyces would imply recognition of a genus Termitomyces with non-termite associated species that occur in biogeographical regions where the fungusgrowing termites do not occur (both species of Arthromyces occur in the Neotropics). Although Arthromyces species are known to grow on insect faeces and so may be insect-associated, it is highly unlikely that these insects are termites (van de Peppel et al. 2021). As mentioned previously, we consider that the (unique) termite association of Termitomyces necessitates maintaining the current definition of the genus. As noted above, only transferring the palaeotropical Arthromyces to Termitomyces would still necessitate the erection of the genus P r a e a r t h r o m y c e s . C o m b i n i n g A r t h r o m y c es a n d Praearthromyces in one genus, which could be justified based on morphology, would not be allowed because of paraphyly.
Putting Phaeotephrocybe, Blastosporella and Nigrocarnea in Tephrocybe seems at odds with the earlier attempts by Hofstetter et al. (2014) and Bellanger et al. (2015) to create a monophyletic genus that is morphologically recognisable and that is characterised by a saprotrophic lifestyle. Both N. radicata and Ph. leucophylla share the tephrocyboid habit with T. rancida and Ph. leucophylla also shares the feature of not producing conidia on the fruitbodies with T. rancida. Both N. radicata and B. zonata produce conidia, while only the latter produces ornamented blastoconidia and all other conidia producing taxa in this clade, including Nigrocarnea and Arthromyces, form chains of arthroconidia. The genera Phaeotephrocybe and Nigrocarnea can be separated from the genera Blastosporella and Tephrocybe by the lack of clamp connections. Nigrocarnea can be distinguished from all other conidia-producing genera as it is the only species to produce the conidia inside the pileus trama. We consider this solution as a possible retrograde step as the variation in traits (both absence and presence of conidia on fruitbodies, clamp connections, insect-association) within Tephrocybe would leave no synapomorphies to separate Tephrocybe from the other related genera. Placing N. radicata and Ph. leucophylla into a single genus does unlikely seem a useful option as these species do not share many morphological characters (Table 1). Biogeographical considerations equally argue against combining both into one genus, as Nigrocarnea is only known from the paleotropics in Southeast Asia, whi le Phaeotephrocybe is only known from the neotropics in Central America. Furthermore, branch lengths suggest significant genetic divergence between the two. The lack of clamp connections in these genera sets them apart from Blastosporella and Tephrocybe, both of which have clamp connections, a feature that is shared with the genus Australocybe. The genus Phaeotephrocybe does not produce any type of conidia and is therefore easily distinguished from Nigrocarnea which produces dark-pigmented arthroconidia in the pileus context of the basidiomata. This trait of conidial production in the pileus trama morphologically distinguishes Nigrocarnea from Phaeotephrocybe and all other taxa in this clade.
With the disadvantages associated with the radical option (one genus Termitomyces) and the option to recognise three genera (Australocybe, Tephrocybe, and Termitomyces), it seems unavoidable to describe new, currently monophyletic genera that are now morphologically, biogeographically and ecologically, in terms of life style, rather well separated. While it could be stated that erection of these small new genera is an expression of what has been called the excessive multiplication of genera (Romagnesi 1977), we think that the specific evolutionary history of the Lyophyllaceae makes such action inevitable.
The addition of four new genera to the Lyophyllaceae and especially the termitomycetoid clade shows that even more ecological transitions may have occurred within this group than previously thought (Hofstetter et al. 2014). Insect associations have only been demonstrated in the genera Arthromyces, Blastosporella and Termitomyces; however, the presence of a pseudorhiza and other shared characters with Termitomyces suggest that other species also could be insectassociated. The diversity and widespread geographic distribution of these genera suggests that more species await discovery.
Data availability The DNA sequences generated during the current study are available in NCBI Genbank.

Declarations
Ethics approval and consent to participate Not applicable.

Competing interests The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.