Taxonomy and phylogeny of the basidiomycetous hyphomycete genus Hormomyces

Abstract

The taxonomy of the genus Hormomyces, typified by Hormomyces aurantiacus, which based on circumstantial evidence was long assumed to be the hyphomycetous asexual morph of Tremella mesenterica (Tremellales, Tremellomycetes) or occasionally Dacrymyces (Dacrymycetales, Dacrymycetes), is revised. Phylogenies based on the three nuc rDNA markers [internal transcribed spacers (ITS), 28S large ribosomal subunit nrDNA (28S) and 18S small ribosomal subunit nrDNA (18S)], based on cultures from Canada and the United States, suggest that the genus is synonymous with Tulasnella (Cantharellales, Agaricomycetes) rather than Tremella or Dacrymyces. Morphological studies of 38 fungarium specimens of Hormomyces, including the type specimens of H. callorioides, H. fragiformis, H. paridiphilus and H. peniophorae and examination of the protologues of H. abieticola, H. aurantiacus and H. pezizoideus suggest that H. callorioides and H. fragiformis are conspecific with H. aurantiacus while the remaining species are unlikely to be related to Tulasnella. The conidial chains produced by H. aurantiacus are similar to monilioid cells of asexual morphs of Tulasnella species formerly referred to the genus Epulorhiza. The new combination Tulasnella aurantiaca is proposed and the species is redescribed, illustrated and compared with similar fungi. The ecological niche of T. aurantiaca and its possible relationship to orchid root endophytes is discussed. A key to asexual genera with similar conidium ontogeny to T. aurantiaca is provided.

Citation: Mack J, Assabgui RA, Seifert KA (2021). Taxonomy and phylogeny of the basidiomycetous hyphomycete genus Hormomyces. Fungal Systematics and Evolution 7: 177–196. doi: 10.3114/fuse.2021.07.09

INTRODUCTION

The genus Hormomyces was introduced by Bonorden (1851) for a single hyphomycetous species, H. aurantiacus, collected on old oak (Quercus sp.) wood in Germany. The protologue described and illustrated orange, gelatinous sporodochia with branched chains of hyaline, globose conidia; no measurements for microscopic structures were reported (Fig. 1). The location of Bonorden’s herbarium is uncertain (Stafleu & Cowan 1976) and no type specimen is known to exist for this species. Hormomyces is rarely mentioned in modern taxonomic literature (Seifert et al. 2011). The only modern study is by Tubaki (1976), who reported on its blastic, acropetal conidium ontogeny and the characters of H. aurantiacus in axenic culture based on a strain isolated from a fallen twig of Tsuga canadensis collected in New York State, United States. His concept was generally accepted. The colonies of H. aurantiacus are conspicuous and distinctive but they may be disregarded as uninteresting “jelly fungi” by those focusing on microfungi, or discarded as a “trivial asexual morph or anamorph” by those interested in macrobasidiomycetes. Despite this lack of academic attention, there are frequent reports by field mycologists, mostly accurately reflecting the Tubaki (1976) concept, e.g. Mushroom Observer includes 42 (Wilson et al. 2020) and MyCoPortal 180 records (MyCoPortal 2019).

Fig. 1.

A copy of the lectotype figure for Hormomyces aurantiacus, a reproduction of fig. 234, taf. XI from Bonorden (1851).

Seven species of Hormomyces were described: H. abieticola, H. aurantiacus, H. callorioides, H. fragiformis, H. paridiphylus, H. pezizoideus and H. peniophorae (supplementary table S1). Two were first described in Hypsilophora prior to being transferred to Hormomyces by Saccardo (1888), namely Hy. fragiformis and Hy. callorioides. Saccardo (1888), who mostly compiled descriptions from other mycologists and often did not examine material himself, considered the main distinction among these species to be sporodochial colour: orange for H. aurantiacus, purple for H. fragiformis and pink for H. callorioides. Lloyd (1916) questioned the value of colour to distinguish these species, suggesting that they might be conspecific without formally synonymizing them. McNabb (1969) synonymized H. callorioides under H. fragiformis, re-emphasizing their exclusion from Hypsilophora, which he considered the correct generic name for the asexual morph of the fungus now known as Erythricium salmonicolor (Corticiales). McNabb did not propose a synonymy of H. fragiformis with H. aurantiacus, and the former name has often been used for specimens collected in North America (MyCoPortal 2019), irrespective of sporodochial colour. Tubaki (1976) accepted only H. aurantiacus in the genus, although he did not critically revise the other named species.

In their compilation of hyphomycete genera, Seifert et al. (2011) listed two other genera as synonyms of Hormomyces: Sphaerocolla (following the opinions of von Höhnel 1917; Donk 1962) and Hormisciopsis. Sphaerocolla Karsten (1892) comprises S. aurantiaca from Finland from living Betula bark, reported as producing large, effuse, orange sporodochia up to 10 cm long, with branched chains of globose conidia 3–9 μm diam, reminiscent of H. fragiformis or H. aurantiacus. Hormisciopsis Sumstine (1914) was proposed for Ho. gelatinosa, producing red, gelatinous sporodochia on wood collected in Pennsylvania, with branching chains of globose to ellipsoidal conidia 6–10 × 5–6 μm. The description corresponds well with H. fragiformis and H. aurantiacus, and the illustration is very similar to those of H. aurantiacus provided by Lloyd (1916) and H. fragiformis by Patouillard (1900). Although Sumstine (1914) compared Hormisciopsis gelatinosa to Hormiscium, he was apparently unaware of Hormomyces.

The taxonomic relationships of Hormomyces were a matter of debate, although it has usually been regarded as a basidiomycete. Patouillard (1900) suggested that Hormomyces is the budding state of a Dacrymyces. Lloyd (1916) disagreed but did not provide an alternative classification. That same year, Saccardo suggested that H. aurantiacus might be the asexual morph of Tremella mesenterica (Saccardo 1916). Bresadola (1932) reiterated this putative connection in a short note, reporting that the fungus he examined had conidia up to 3 μm long but he did not illustrate it; unfortunately, no specimens were mentioned. Donk (1962) also suggested a relationship between Tr. mesenterica and H. aurantiacus, noting that immature Tr. mesenterica also produced conidia. McNabb (1969) suggested that this connection was “currently accepted in Europe.” Later, however, Pippola & Kotiranta (2008) described the asexual morph of Tr. mesenterica as producing chains of ellipsoidal conidia (1.8–)2.2–4.5(–5.9) × (1.6–)1.8–3.8(–4.2) μm, originating from clamped conidiogenous cells, which matches the dimensions given by Bresadola (1932). The characters diverge significantly from H. aurantiacus as described by Bonorden (1851), i.e. with globose conidia and no clamp connections. Tubaki (1976) also disagreed with this supposed connection, remarking that many asexual morphs of Tremella species are yeasts, while his culture of H. aurantiacus was filamentous. Thus, the purported asexual-sexual morph connection was a speculation that was never established experimentally and was contradicted by subsequent observations. Given their different ecological niches, H. aurantiacus on rotten wood and Tr. mesenterica as a mycoparasite usually on rather solid wood, it is unlikely that the two were seen in close proximity.

In 2015, we isolated a culture from a freshly collected specimen of H. aurantiacus and our initial 18S nrDNA sequences obtained from that culture suggested an affinity of Hormomyces with Tulasnella rather than Tremella or Dacrymcyes. The genus Tulasnella contains about one hundred sexual species, most of which occur on wood, and most known asexual species are usually isolated from orchid roots or liverworts. The basidiomes are usually thin and waxy and the basidia have four basally swollen sterigmata delimitated from the probasidium by a septum. So far, no asexual species of Tulasnella have been recorded from wood. The morphology and ontogeny of the conidial chains of H. aurantiacus are similar to the monilioid hyphae of the orchid mycorrhizal Tulasnella asexual morphs formerly attributed to Epulorhiza (Roberts 1994), and together with its dikaryotic hyphae, support a relationship with Tulasnellaceae. In this paper, our goal is to clarify the family and genus level classifications of this fungus using Bayesian Inference (BI) and Maximum Likelihood (ML) analyses of nuc rDNA regions and genes, and to evaluate the morphological characters of H. aurantiacus and the other described species of the genus based on studies of types and supplementary specimens and cultures. Should Hormomyces be considered distinct from or a synonym of Tulasnella? And if H. aurantiacus is classified in Tulasnella, might it be the asexual morph of a known sexual species, or identical with one of the other asexual species previously attributed to Epulorhiza? Because H. aurantiacus is apparently lignicolous, understanding its relationship with other Tulasnella species and mycorrhizal species formerly classified in Epulorhiza may provide insight into ecological patterns in this family.

MATERIALS AND METHODS

Cultures, specimens and morphological examination

Fresh specimens of H. aurantiacus were collected in 2014 and 2015 near Ottawa, Ontario, Canada, with additional specimens and one culture provided by colleagues; all material studied is listed in the Specimens Examined paragraphs of the Taxonomy section. Cultures were obtained by squashing a small fragment of a sporodochium and making a slurry that was further diluted with sterile dH₂O. Ten 20 μL drops of the dilution were pipetted onto malt extract agar (MEA, recipe from Raper and Thom 1949), using BD Bacto Malt Extract, and 15 g/L BD Bacto Agar, with the addition of 0.1 g/L ZnSO₄.7H₂O and 0.005 g/L CuSO₄.5H₂O, and examined 24 h later for germinating conidia, which were then transferred individually to new MEA plates to reduce the risk of contamination. Thereafter, we did not attempt to make single spore isolations because the conidia produced in vivo remain in strongly coherent chains and can be separated only with difficulty. No evidence of contamination of the cultures by other fungi or bacteria was evident under the dissecting microscope or by light microscopy of mounts from the original isolates.

Culture characters were observed by growing all strains on MEA and oatmeal agar (OA; Gams et al. 1987).

For each strain and medium, one plate was incubated under nrUV light for 16 h and in darkness for 8 h per day at ambient RT (varying between 20–25 °C), and another in darkness in a 25 °C incubator. Specimens were inoculated at a single, central point in a 90 mm polystyrene Petri dish. All cultures were then examined and measured every 7 d for 4 wk, with coloration noted on days 14 and 28. To determine cardinal temperatures, two strains were grown in darkness at temperatures ranging from 5–40 °C, at increments of 5 °C (except 37 °C was used instead of 35 °C), and measured every 7 d for 4 wk. Culture photographs were taken with an Olympus Tough TG-5 camera (Olympus, Tokyo) with black velvet as a background.

Herbarium specimens, including available types, were borrowed from BPI (26 specimens), DAOM (7), K(M) (5) and HKAS (1) (herbarium acronyms follow Thiers 2019).

Microscopic observations of living specimens were made using tissue mounted in water. Herbarium specimens were rehydrated with a small drop of dH₂O for 10 min, then small fragments of sporodochia were removed using jewellers’ forceps and heated in 85 % lactic acid for five minutes on an electric hot plate. Herbarium specimens and cultures were examined using an Olympus SZX12 dissecting microscope and BX50 compound microscope (Olympus, Tokyo), and photographed using Infinity 2 or Infinity X USB microscope cameras, using Infinity Capture software (Lumenera, Ottawa). Specimen and colony colours were described using the alphanumeric codes and names (with initial capitals) used by Kornerup & Wanscher (1976). Photographic plates were assembled using Photoshop v. 5.5 (Adobe Systems, San Jose, CA).

To demonstrate dikaryotic nuclei, hyphae from a 9-d-old culture of DAOMC 251988 grown on MEA were mounted in 10 μL of SYTO 13 green fluorescent nucleic acid stain (Invitrogen, Carlsbad, California) on a microscope slide. After the placement of the cover glass, the sample was left in darkness for 30 min at 37 °C, then examined with a Nikon ECLIPSE E800 fluorescence microscope (Nikon, Tokyo) using the UV2 setting. Pictures were taken with a Nikon DS-Ri2 camera (Nikon, Tokyo).

For each specimen, 50 conidia were measured, and mean and standard errors are provided. Q values for conidia were calculated as length divided by width. To assess the possible significance of conidial dimensions as a diagnostic character for species identification, means (with upper and lower limits represented by standard error, and outliers in brackets) were compared for all specimens. All calculations were made with Excel 2016 (Microsoft, Redmond).

DNA extraction, sequencing and phylogenetic analysis

Genomic DNA was extracted from pure cultures of H. aurantiacus using the DNeasy UltraClean Microbial Kit (Qiagen, Hilden), following the manufacturer’s protocol. Permission was not obtained to attempt extraction of genomic DNA from specimens loaned by fungaria. For preliminary phylogenetic placement of an early isolate of H. aurantiacus (DAOMC 252084), a partial sequence of the 18S nrRNA gene was generated using primers NS1 and NS4, and an amplification profile of 95 °C for 10 min for the initial denaturing of the DNA template, then 40 cycles with denaturation at 95 °C for 60 s, annealing at 56 °C for 45 s and extension at 72 °C for 90 s, with a final extension at 72 °C for 10 min. Because of the relationship with Tulasnellaceae (Cantherellales) suggested by these results, we tried a Tulasnella-specific ITS4 primer designed by Taylor & McCormick (2008) in combination with ITS5; this resulted in sporadic and low yield amplification for our strains of H. aurantiacus. Amplification of the 28S nrRNA gene with LROR and either LR8 or LR5 invariably failed or yielded multiple PCR bands. Amplification was eventually achieved using V9G as the forward primer and LR3 as the reverse primer. See Results for primer references. The PCR profile had an initial denaturation at 94 °C for 90 s, five cycles of denaturation at 94 °C for 45 s, annealing at 56 °C for 45 s and extension at 72 °C for 60 s. Annealing temperatures were then decreased by 1 °C every five cycles until reaching 51 °C, which was used for another 30 cycles, with a final extension of 10 min, for a total of 55 cycles.

Our sequences of H. aurantiacus were edited and trimmed using Geneious v. 11.1.5 (Biomatter, Auckland). BLAST analyses of all three nrDNA markers indicated a relationship to reference sequences in Tulasnellaceae, which guided our sampling for subsequent phylogenetic analyses. Our sequences were aligned using MUSCLE v. 3.8.425 (Edgar 2004) in datasets containing sequences of Cantharellales downloaded from GenBank (Table 1 for ITS, Table 2 for 28S), sampled to include all available taxa of Tulasnellaceae and a few similar, mostly ITS, environmental sequences and unidentified species uncovered during BLAST searches, along with selected representatives of other genera in the order. We also used an ITS and partial 28S sequence of H. aurantiacus (NBRC 30400, the culture referred to by Tubaki 1976) published in the NBRC culture collection database (NITE Biological Resource Center 2019), which were not deposited in GenBank. Reference sequences of Craterellus tubaeformis were used as the outgroup for the ITS analysis, which examined relationships among species of Tulasnella, Epulorhiza and H. aurantiacus. A reference sequence of Ustilago maydis was used as the outgroup for the LSU analysis, which focused on the relationship between Hormomyces and Tulasnella in the Cantharellales, and tested prior hypotheses of relationships between Hormomyces and Tremella or Dacrymyces.

Table 1.

ITS sequences retrieved from GenBank or newly generated during this study.

Species	Strain no.	Origin	Host	GenBank accession no.	Reference
Ceratobasidium albasitensis	EaB-T2	–	–	AJ427398	Gonzalez, unpublished
C. angustisporum	CBS 568.83	Eyre Peninsula, SA, Australia	Pterostylis mutica endophyte	AJ427403	Gonzalez, unpublished
C. cereale	C13	–	Triticum aestivum	AJ302009	Gonzalez et al. (2002)
C. cornigerum	C6	–	Erigeron canadensis	AJ301902	Gonzalez et al. (2002)
C. ramicola	CBS 758.79T	Florida, USA	Pittosporum, leaf	AJ427404	Gonzalez, unpublished
C. stevensii	CBS 477.82	Kentucky, USA	Malus domestica, twig	AJ427405	Gonzalez, unpublished
Craterellus tubaeformis	S9	–	–	MH394713	Jensen-Vargas & Marizzi (2018)
	1D3	Kunigami, Okinawa, Japan	–	AB973729	Matsuoka, unpublished
Tulasnella albida	KC 110	–	–	AY373294	McCormick et al. (2004)
T. amonilioides	–	–	–	JF907600	Almeida et al. (2014)
T. anaticula	13o004	Mt. Hambeak, South Korea	Platanthera chlorantha	KT164598	Direct submission
T. asymetrica	MA FF305808 Clone C002	Australia	Thelymitra epipactoides	KC152347	Cruz et al. (2014)
	MAFF 305808 Clone C005	Australia	Thelymitra epipactoides	KC152348	Cruz et al. (2014)
	MAFF305808 Clone C009	Australia	Thelymitra epipactoides	KC152349	Cruz et al. (2014)
T. aurantiaca	DAOMC 251988	Pennsylvania, USA	Rotten wood	MK626686	This study
	DAOM 970795
	DAOMC 251989	Tennessee, USA	Rotten wood and Crepidotus spp.	MK626533	This study
PBM4158
	DAOMC 252083	Victoriaville, Quebec, Canada	Fomitopsis betulina	MK626687	This study
DAOM 970821
	DAOMC 252084	Ottawa, Ontario, Canada	Rotten Populus wood		This study
DAOM 970822	MK626567
	DAOMC 252086	Montreal, Quebec, Canada	Rotten wood	MK593626	This study
DAOM 970819
	DAOMC 252085	Ottawa, Ontario, Canada	Rotten wood	MK626568	This study
DAOM 970820
	NBRC 30400	New York, USA	Twig of Tsuga canadensis	Sequence not in Genbank	NITE Biological Resource Center, 2019
T. australiensis oreophilus	CLM 031	New York, USA	Arthrochilus	KF476602	Arifin et al. (2020)
	CLM 1945T	Nowra, NSW, Australia	Cryptostylis erecta root	MT003730	Arifin et al. (2020)
	CLM 2004	Northern Sydney NSW, Australia	Cryptostylis erecta root	MT003715	Arifin et al. (2020)
T. calospora	CBS 326.47	–	–	AY373298	McCormick et al. (2004)
	Ch5-3	–	–	HM450045	Idris (2010)
T. concentrica	CLM 2071	Morton NP, NSW, Australia	Cryptostylis leptochila root	MT036547	Arifin et al. (2020)
	CLM 2098T	Nowra, NSW, Australia	Cryptostylis erecta root	MT003744	Arifin et al. (2020)
	CLM2198	Bunyip SP, Vic, Australia	Cryptostylis leptochila root	MT036533	Arifin et al. (2020)
T. cumulopuntioides	MAFF 245682	Tsukuba City, Ibaraki, Japan	Spiranthes sinensis	LC175323	Fujimori et al. (2019)
T. danica	KC 388	–	–	AY373297	McCormick et al. (2004)
T. deliquescens	MAFF 244717	–	Spiranthes sinensis	LC175329	Fujimori et al. (2019)
T. densa	CLM 2110	Bulahdelah, NSW, Australia	Cryptostylis hunteriana	MT036526	Arifin et al. (2020)
	CLM 2111	Bulahdelah, NSW, Australia	Cryptostylis hunteriana	MT036525	Arifin et al. (2020)
	CLM 2117T	Bulahdelah, NSW, Australia	Cryptostylis hunteriana	MT036520	Arifin et al. (2020)
T. dentritica	MAFF 244709T	–	Spiranthes sinenis	LC175308	Fujimori et al. (2019)
T. eichleriana	KC 852	–	–	AY373292	McCormick et al. (2004)
	K(M) 143600	England, UK	Wood of decorticated sapling	KC152381	Cruz et al. (2014)
T. ellipsoidea	MAFF 245686	Tsukuba City, Ibaraki, Japan	Spiranthes sinensis	LC175315	Fujimori et al. (2019)
T. epiphytica	AERO_3.2	–	–	JF907598	Almeida et al. (2014)
T. irregularis	CBS 574.83T	NT, Australia	Dendrobium dicupum root	MH861654	Vu et al. (2019)
T. occidentalis	CLM 1938T	Boyanup, WA, Australia	Cryptostylis ovata root	MT008096	Arifin et al. (2020)
	CLM 1942	Boyanup, WA, Australia	Cryptostylis ovata root	MT008092	Arifin et al. (2020)
	CLM 1943	Boyanup, WA, Australia	Cryptostylis ovata root	MT008091	Arifin et al. (2020)
T. prima	CLM 159T	Blue Mountains, NSW, Australia	Chiloglottis trilabra	KF476556	Linde et al. (2013)
	CLM 377	Kanangra Boyd NP, NSW, Australia	Chiloglottis aff. jeanesii	KF476544	Linde et al. (2013)
	5O5.III.3	–	Chiloglotis dyphilla	HM196792	Roche et al. (2010)
	SRBG01.II.3	Australian National Botanical Garden, Acton, ACT, Australia	Chiloglottis trapeziformis	HM196793	Roche et al. (2010)
T. pruinosa	DAOM 17641	Ontario, Canada	Sporophore on Populus sp.	AY373295	McCormick et al. (2004)
T. punctata	CLM 2012	Northern Sydney, NSW, Australia	Cryptostylis subulata root	MT008124	Arifin et al. (2020)
	CLM 2017T	Northern Sydney, NSW, Australia	Cryptostylis subulata root	MT008122	Arifin et al. (2020)
	CLM 2018	Northern Sydney, NSW, Australia	Cryptostylis subulata root	MT008121	Arifin et al. (2020)
T. rosea	CLM 1770	WA, Australia	Spiculaea ciliata root	MN947568	Arifin et al. (2020)
	CLM 1773T	WA, Australia	Spiculaea ciliata root	MN947569	Arifin et al. (2020)
	CLM 1774	WA, Australia	Spiculaea ciliata root	MN947570	Arifin et al. (2020)
T. secunda	CLM 274	Talbot, WA, Australia	Paracaleana triens	KF476580	Linde et al. (2013)
	CLM 222	Talbot, WA, Australia	Paracaleana minor	KF476568	Linde et al. (2013)
	CLM 009T	Talbot, WA, Australia	Drakaea elastica	KF476575	Linde et al. (2013)
Tulasnella sp. ECU 5	DC 225	Ecuador	Branch	KC152397	Cruz et al. (2014)
	DC 225	Ecuador	Branch	KC152398	Cruz et al. (2014)
Tulasnella sp. ECU 6	DC 185	Ecuador	Branch	KC152401	Cruz et al. (2014)
	DC 262	Ecuador	Branch	KC152409	Cruz et al. (2014)
Tulasnella sp.	DC 294 C009	Germany	Rotten wood	KC152387	Cruz et al. (2014)
	DC 294 C016	Germany	Rotten wood	KC152394	Cruz et al. (2014)
	CLM 084	Australia	Arthrochilus oreophilus	KF476594	Linde et al. (2013)
	CLM 085	Australia	Arthrochilus oreophilus	KF476595	Linde et al. (2013)
T. sphagneti	12033.1T	Kosciuszko NP” NSW, Australia	Chiloglottis aff. valida	KY095117	Linde et al. (2017)
	13102.1	Kosciuszko NP” NSW, Australia	Chiloglottis turfosa	KY445924	Linde et al. (2017)
	13065.2	Kosciuszko NP” NSW, Australia	Chiloglottis sp.	KY445925	Linde et al. (2017)
T. tomaculum	KC 429	–	–	AY373292	McCormick et al. (2004)
	K(M)123675	England, UK	–	KC152380	Cruz et al. (2014)
T. violea	DC 177	Ecuador	Decaying wood	KC152414	Cruz et al. (2014)
	DC 292	Germany	Decaying wood	KC152412	Cruz et al. (2014)
	DC 292	Germany	Decaying wood	KC152435	Cruz et al. (2014)
	DC 293	Germany	Decaying wood	KC152437	Cruz et al. (2014)
	KC 851	–	–	AY373293	McCormick et al. (2004)
	K(M) 164256	England, UK	Underside of Fagus sylvatica log	KC152411	Cruz et al. (2014)
T. warcupii	CLM 007	Atherton, Tablelands, QLD, Australia	Arthrochilus oreophilus	KF476600	Linde et al. (2013)
	CLM 022	Australia	Arthrochilus oreophilus	KF476601	Linde et al. (2013)
	CLM 028	Australia	Arthrochilus oreophilus	KF476599	Linde et al. (2013)

^TIndicates type specimens.

Table 2.

Gene sequences (28S) retrieved from GenBank, and newly generated in this study.

Species	Strain no.	Origin	Host	GenBank accession no.	Reference
Agaricus bisporus	CBS 151.46	–	–	MH867670	Vu et al. (2019)
Botryobasidium botryosum	AFTOL-ID 604	–	–	DQ089013	Nilsson, unpublished
Bo. isabellinum	GEL2109	–	–	AF393047	Nilsson, unpublished
Bo. subcoronatum	GEL 1286	–	–	AF287850	Binder & Hibbett (2002)
Burgella lutea	Etayo 27623T	Bolivia	Corticolous lichens	KC336075	Diederich et al. (2014)
Bu. flavoparmeliae	JL192-01	Oklahoma, USA	Flavoparmelia baltimorensis	DQ915469	Lawrey et al. (2007)
Burgoa moriformis	VCH 33T	Inisherk, Crom, Fermanagh, Ireland	Salix bark	DQ915477	Lawrey et al. (2007)
Cantharellus addaiensis	BB 96.010	Zambia	–	KM484680	Shao et al. (2014)
Ca. altipes	BB 07.019T	USA	–	KF294627	Buyck et al. (2014)
Ca. amethysteus	993/estades	–	–	MG450679	Buyck et al. (2018)
Ca. cibarius	CC15SWE	Sweden	Betula	JX030441	Foltz et al. (2013)
Ca. ferruginasecens	GE sn	France	–	KM484681	Shao et al. (2014)
Ca. formosus	BB 13.163	USA	–	KM484683	Shao et al. (2014)
Ca. lateritius	JJ NC-Canth 2	USA	–	KM484686	Shao et al. (2014)
Ca. longisporus	ER 107	Madagascar	–	KM484688	Shao et al. (2014)
Ca. minor	BB 07.057	USA	–	KF294632	Buyck et al. (2014)
Ca. texensis	BB O7.018T	USA	–	KF294626	Buyck et al. (2014)
Ceratobasidium bulbillifaciens	Eichler-Cezanne 8193	Germany	Bark of Acer platanoides	KC336073	Diederich et al. (2014)
C. bulbilllifaciens	Eichler-Cezanne 8067	–	Bark of Fraxinus	KC336071	Diederich et al. (2014)
C. globisporum	CBS 569.83T	Queensland Australia	–	MH873365	Vu et al. (2019)
C. pseudocornigerum	CBS 568.83T	Australia	–	MH873364	Vu et al. (2019)
C. ramicola	Java 11	Java	Theobroma cacao	HQ424243	Samuels et al. (2012)
C. theobromae	South Sulawesi 1	South Sulawesi	Petiole of Theobroma cacao	KU319575	Samuels et al. (2012)
	South Sulawesi 6	South Sulawesi	Petiole of Theobroma cacao	KU319577	Samuels et al. (2012)
	South Sulawesi 11	South Sulawesi	Theobroma cacao	HQ424241	Samuels et al. (2012)
	South Sulawesi 10	South Sulawesi	Theobroma cacao	HQ424242	Samuels et al. (2012)
Ceratorhiza oryzae-sativae	CBS 439.80	Japan	Oryza sativa	MH873047	Vu et al. (2019)
Clavulina amazonensis	AMV1973	Colombia	Pseudomonotes tropenbosii	KT724123	Vasco-Palacios (2016)
Cl. cinerea	KHL 11694	Lammi, Finland	–	AM259211	Nilson et al. (2006)
Cl. cf. cristata	BB 12.083	Italy	–	KM484694	Shao et al. (2014)
Cl. purpurascens	ZP-3065	China	Soil	MK564124	Wu et al. (2019)
Clavucilium delectabile	KHL 11147	Norway	–	AY586688	Larsson et al. (2004)
Craterellus cinereofimbratus	JOH4	Columbia	Pseudomonotes tropenbosii	KT724159	Vasco-palacios (2016)
Cr. lutescens	BB 13.048	Canada		KM484696	Shao et al. (2014)
Cr. tubaeformis	BB 1324	USA	–	KM484697	Shao et al. (2014)
	BB 07.293	Slovakia	–	KF294640	Buyck et al. (2014)
Dacrymyces chrysospermus	FPL11353	–	–	AF287855	Hibbett et al. (2000)
D. stillatus	CBS 195.48	France	–	MH867857	Vu et al. (2019)
Gloeotulasnella cystidiophora	KW 2871	–	–	AY585831	Shefferson et al. (2005)
Haplotrichum conspersum	AFTOL ID 1766	–	–	DQ521414	Matheny et al. unpublished
Hydnum elatum	FRI62309	Kampung Jelawat-Tasik Bera Pahang Malaysia	–	KU612691	Feng et al. (2016)
H. ellipsosporum	FD3281	Switzerland	–	KX086217	Beenken, unpublished
H. magnorufescens	161209	Slovenia	–	KU612669	Feng et al. (2106)
H. rufescens	BB 07.340	Slovakia	–	KM484698	Shao et al. (2014)
H. versterholtii	HKAS92342	Yulong snow mountain, Yunnan, China	–	KU612646	Feng et al. (2016)
Minimedusa obcoronata	CBS 120605	Thailand	Eucalyptus camaldulensis	GQ303309	Cheewangkoon et al. (2009)
Multiclavula mucida	DSH96-056	–	–	AF287875	Hibbett et al. (2000)
M. mucida	TUB 011734	–	–	EU909345	Krause et al. (2011)
M. vernalis	GB-BN-1	Sweden	–	AM259214	Nilsson et al. (2006)
Rhizoctonia floccosa	CBS 337.36T	Indonesia	–	MH867319	Vu et al. (2019)
R. quercus	CBS 313.35T	Italy	Root of quercus pedunculata	MH867202	Vu et al. (2019)
R. repens	CBS 298.32	Netherlands	Orchis morio	MH866781	Vu et al. (2019)
Schizophyllum commune	MUT 4875	Mediterranean Sea	Flabellia petiolata	MF115832	Poli et al. (2018)
Sistotrema adnatum	FCUG 700	–	–	DQ898699	Moncalvo et al. (2006)
S. biggisae	FCUG 782	–	–	DQ898697	Moncalvo et al. (2006)
S. coronilla	AFTOL-ID 618	–	–	DQ457641	Moncalvo et al. (2006)
S. hypogaeum	CBS 394.63T	Australia	Soil	MH869926	Vu et al. (2019)
S. oblongisporum	FCUG 2117	–	–	DQ898703	Moncalvo et al. (2006)
S. sernanderi	CBS 926.70	–	–	AF518650	Hibbett & Binder (2002)
Sistotremella brinkmanii	CBS 186.39	Michigan, USA	–	MH867474	Vu et al. (2019)
Sistotremella perpusilla	CBS 126048	North Carolina, USA	Abies	MH875516	Vu et al. (2019)
Thanatephorus cucumeris	AFTOL-ID 2022	–	–	DQ917658	Matheny et al. unpubl.
	CBS 340.51	England, UK	–	MH868410	Vu et al. (2019)
Tremella macrobasidiata	AM453	Portugal	Lecanora chlorotera	KT334595	Zamora et al. (2016)
Tr. mesenterica	AM30	Wedin, Sweden	–	JN043569	Millanes et al. (2011)
	CBS:6973T	Vancouver, British-columbia, Canada	Alnus rubra	KY109900	Vu et al. (2016)
T. anaticula	UAMH 5428	Alberta, Canada	Roots of Calypso bulbosa	AY243520	Taylor et al. (2003)
T. asymmetrica	MAFF P305806		Thelymitra luteocilium	DQ388046	Suarez et al. (2006)
T. aurantiaca	DAOMC 251988	Pennsylvania, USA	Rotten wood	MK627511	This study
DAOM 970795
	DAOMC 251989	Tennessee, USA	Rotten wood and Crepidotus spp.	MK627512	This study
PBM 4158
	DAOMC 252083	Victoriaville, Quebec, Canada	Fomitopsis betulina	MK627513	This study
DAOM 970821
	DAOMC 252084	Ottawa, Ontario, Canada	Rotten Populus wood	MK627514	This study
DAOM 970822
	DAOMC 252086	Montreal, Quebec, Canada	Rotten wood	MK627515	This study
DAOM 970819
	DAOMC 252085	Ottawa, Ontario, Canada	Rotten wood	MK627516	This study
DAOM 970820
T. calospora	SPRR.R2	India	Paphiopedilum druryi	MN271391	Parthibhan & Rammasubu (2020)
T. eremophila	13062 MD	–	Euphorbia officinarum	KJ701189	Crous et al. (2015)
T. irregularis	CBS 574.83	NT, Australia	Dendrobium dicuphum	AY243519	Taylor et al. (2003)
T. phuhinrongklaensis	SDBR-CMU-CR41T	–	–	MF427703	Rachanarin et al. (2018)
	SDBR-CMU-CR42	–	–	MF427704	Rachanarin et al. (2018)
	SDBR-CMU-CR43	–	–	MF427705	Rachanarin et al. (2018)
	SDBR-CMU-CR44	–	–	MF427706	Rachanarin et al. (2018)
T. pruinosa	DAOM 17641	Richmond Hill, Ontario, Canada	Populus sp.	AF518662	Hibbett & Binder (2002)
Tulasnella sp.	GEL 4461	–	–	AJ406436	Langer (2001)
	GEL 4745	–	–	AJ406436	Langer (2001)
	GEL 5130	–	–	DQ898731	Moncalvo et al. (2006)
T. violea	DAOM 222001	–	–	AY293216	Binder et al. (2005)
	AFTOL-ID 1879	–	–	DQ520097	Garnica & Weiß, unpublished
Ustilago maydis	CBS 358.32	–	–	MH866814	Vu et al. (2019)

^TIndicates type specimens or ex-type strains.

Independent phylogenetic analyses were conducted using the Maximum Likelihood (ML) and the Bayesian Inference (BI) algorithms for both markers. The ML analysis was done with PHYML v. 3.0 using the GTR + G + I model as the most suitable model for both 28S and ITS (Guindon et al. 2010). For BI, JModelTest v. 0.1.1 (Darriba et al. 2012) was used to determine the most suitable model, GTR + G for both ITS and 28S. The BI analyses were run using MrBayes v. 3.2 (Ronquist et al. 2012), with four simultaneous Markov chains run until the average standard deviation of split frequency reached < 0.01. Convergence was assessed when the standard deviation of split frequency reached < 0.01. Sampling frequency was 1 in 500 generations with the first 25 % of the trees discarded as burnin. Trees were visualized using FigTree v. 1.4.3 (Rambaut 2016) and modified using Adobe Illustrator 10 (Adobe, San Jose) and PowerPoint 2016 (Microsoft, Redmond). The alignments and phylogenetic trees were deposited in TreeBASE (Treebase.org Study ID: 25624 and 27256). Proposed new names and typifications were deposited in MycoBank (MB) and the MycoBank typification (MBT) database (Westerdijk Fungal Biodiversity Institute, Utrecht).

RESULTS

Phylogenetic analysis

An initial 1 045 bp sequence of the 5’ end of the 18S nrRNA gene of DAOMC 252084 (deposited as GenBank MN719097) indicated a relationship of H. aurantiacus with Tulasnella, with a 99.2 % similarity to an AFTOL generated sequence of Tulasnella violea (AY707097), based on a query coverage of 99 %, with sequence similarities of about 80–99 % with other Tulasnellaceae, mostly with query coverages of about 55 %. The distance tree accompanying the BLAST search was consistent with the close placement of our sequence in this family. Because the taxon sampling for 18S sequences of Tulasnellaceae is so sparse in GenBank, with only three named species among the 32 reference sequences available, we did not pursue further 18S analyses.

Our initial attempts to amplify ITS and 28S separately from cultures of H. aurantiaca using universal primers, or a putative Tulasnella-specific ITS primer (Taylor & McCormick 2008), were unsuccessful, with only faint or multiple bands visualized. Diverse techniques such as annealing point gradients from 60–55 °C, a touchdown PCR of 40 cycles each at 57 and 55 °C, and altering primer and DNA concentrations were all unsuccessful in amplifying the DNA. This was solved using the forward primer V9G, slightly upstream from the ITS locus at the end of the 18S (de Hoog & Gerrits van den Ende 1998) along with the reverse primer LR3, slightly downstream from the LROR primer on the 28S (Raja et al. 2017). Sequences of this fragment showed that H. aurantiacus DAOMC 251989 has mismatched binding sites in the ITS4, ITS5, ITS1F and LROR priming regions. For subsequent amplification and sequencing, we designed the new primers ITS4_hormo, ITS5_hormo and LROR_hormo (Table 3), correcting the mismatches identified by our first sequence. This enabled the routine amplification and sequencing of both markers.

Table 3.

Primer names and sequences used for this study, including the new primers ITS5_hormo, ITS4_hormo and LROR_hormo and the Tulasnella-specific ITS4 primer, ITS4_tul, with melting temperature (T_m) in degrees Celsius. Point mutations are in bold, nucleotide additions are in bold and underlined and nucleotide deletions are shown as a blank space with a bold underline.

	Sequence (5’ to 3’)	Tm (°C)	Reference
V9G	TTACGTCCCTGCCCTTTGTA	56	de Hoog & Gerrits van den Ende (1998)
ITS5	GGAAGTAAAAGTCGTAACAAGG	51	White et al. (1991)
ITS5_hormo	GGAAGTACAAGTCGTAACAAGG	53	This study
ITS4	TCCTCCGCTTATTGATATGC	52	White et al. (1991)
ITS4_tul	CCGCCAGATTCACACATTGA	55	Taylor & McCormick (2008)
ITS4_hormo	TCCTCCGCTGAATAATATGC	52.1	This study
LROR	ACCCGCTGAACTTAAGC	52	Moncalvo et al. (2000)
LROR_hormo	ACCCGCTTGA_TTTAAGC	50	This study
LR3	CCGTGTTTCAAGACGGG	53	Moncalvo et al. (2000)
LR3R	GTCTTGAAACACGGACC	50	Moncalvo et al. (2000)
LR5	TCCTGAGGGAAACTTCG	51	Moncalvo et al. (2000)
LR8	CACCTTGGAGACCTGCT	54	Hopple & Vilgalys (1999)

The sequences of our six strains of Hormomyces were very similar. For the ITS (Fig. 2), four of the strains had identical sequences; DAOMC 251989 had a single nucleotide insertion, and DAOMC 252084 differed from the other strains by 3–4 nucleotides. For the 28S (Fig. 3), the differences between each strain were no more than two nucleotides, resulting in similarities > 99.5 %.

Fig. 2.

Phylogenetic tree for species of Tulasnella and Ceratobasidium for the ITS region, based on Bayesian inference analysis using MrBayes. The values above the branches are Bayesian posterior probabilities/maximum likelihood bootstrap values. Bootstrap values of ≥ 70 % and Bayesian posterior probabilities of ≥ 0.70 are shown with bootstrap values of 100 % and Bayesian posterior probability of 1.00 replaced by an asterisk (*). ^Tindicates type specimens.

Fig. 3.

Phylogenetic tree for species of Cantharellales including species of Tulasnella for the 28S gene, based on Bayesian inference analysis using MrBayes. The values above the branches are Bayesian posterior probabilities/maximum likelihood bootstrap values. Bootstrap values of ≥ 70 % and Bayesian posterior probabilities of ≥ 0.70 are shown with bootstrap values of 100 % and Bayesian posterior probability of 1.00 replaced by an asterisk (*).^T indicates type specimens.

All analyses of both genes confirmed the phylogenetic relationship with Tulasnellaceae, Cantharellales, suggested by the initial 18S sequence, remote from the previously proposed relatives of H. aurantiacus in Tremellales or Dacrymycetales. This directed our sampling of taxa to test whether Hormomyces was nested within or distinct from existing genera of Tulasnellaceae. All BI and ML analyses resulted in all strains of H. aurantiacus forming a monophyletic clade nested within Tulasnella and separate from other genera of Cantherellales. In the ITS analysis, H. aurantiacus belongs with high support to a clade including various Tulasnella species including T. violea, T. eichleriana and several orchid root endophytes. Unfortunately our BI analysis of the ITS loci tends to result in polytomy for this clade, although each species remains phylogenetically coherent and distinct. The exact sister group of T. aurantiaca in this clade is unclear. A phylogenetic analyses of the ITS locus by Arifin et al. (2020), which used sequences of H. aurantiacus obtained during our study, also suggested the placement of H. aurantiaca in the same clade within Tulasnella, but sister to two unidentified Tulasnella species from Ecuador.

Our 28S analysis also nests H. aurantiacus within Tulasnella, suggesting it is sister to Tulasnella sp. GEL4461, with a difference of 47 bp (94.6 % similarity). Based on a MegaBLAST search of GenBank, the closest identified match for the LSU locus was Tulasnella obscura (AJ406435, identities = 485/517(94 %), 7 gaps (1 %) with a 57 % query cover). Other matches were less than 85 % similar and with an E value below 1e-100.

Based on our analyses, Tulasnella appears to be monophyletic (with the exception of Tulasnella eremophila), forming two distinct clades with high statistical support in the ITS phylogeny, and possibly four distinct clades in the 28S phylogeny. Because only six species of Tulasnellaceae, including our own, have sequences for both ITS and 28S, a concatenated analysis was not attempted. All of the asexual species previously referred as Epulorhiza currently sequenced using the ITS and LSU loci are easily separated from H. aurantiacus, suggesting that H. aurantiacus is a distinct species of Tulasnella.

Comparison of herbarium specimens of H. aurantiacus, H. fragiformis and H. callorioides

All of the examined specimens identified as H. aurantiacus, H. fragiformis or H. callorioides had similar micromorphology and no distinct separation was detected based on an extensive analysis of conidial sizes (Supplementary Fig. S1). Measurements overlapped with the lectotype specimen of H. fragiformis [(7–)7.5–9.5(–11) × 6–7.5(–8.5) μm (mean 8.66 ± 0.1 × 6.75 ± 0.1, Q 1.29, n = 50)] and the holotype of H. callorioides [(7–)7.5–9.5(–11) × (5–)5.5–6.5(–7) μm (mean 8.45 ± 0.1 × 6.21 ± 0.1, Q 1.37, n = 50)], with the average conidial size for all examined specimens being 8.43 ± 0.1 × 6.37 μm ± 0.1 (Supplementary Fig. S2).

Although the protologue of H. callorioides described its sporodochia as pink (Kalchbrenner & Cooke 1880), our examination of the holotype showed dried sporodochia that were terracotta to dark brown. The original pigments may have degraded over the years or darkened with preservation. The colours observed on the lectotype of H. fragiformis designed by McNabb (1969) and holotype of H. callorioides, both in dry specimens and after rehydration, were within the same range as that we saw in our own specimens and those preserved in BPI and DAOM. We found no evidence to support a hypothesis that sporodochial colour can be used to distinguish these three putative species when the specimens were examined side by side. Similarly, there were no other morphological or microscopic characters that allowed the distinction of the several specimens identified as H. fragiformis or H. aurantiacus, or the holotype of H. callorioides, despite the disjunct geographical locations among the three species.

Taxonomy

The genus Tulasnella was proposed for protection against Epulorhiza and Hormomyces by Stalpers et al. (2021), in accordance of the Shenzhen code (Turland et al. 2018). The remaining names in Epulorhiza were transferred to Tulasnella there. Here, the generic description of Tulasnella is emended to include the asexual morph characters for taxa described in Hormomyces and Epulorhiza.

Tulasnella J. Schröt., Kryptogamen-Flora von Schlesien 3.1(25–32): 397. 1888. nom. cons. prop.

Synonyms: Hormomyces Bonord., Handb. Allgem. mykol. (Stuttgart): 150. 1851 (asexual synonym).

Prototremella Pat., J. Bot., Paris 2: 269. 1888, fide Donk 1966.

Pachysterigma Johan-Olsen ex Bref., Unters. Gesammtgeb. Mykol. (Liepzig) 8: 5. 1888, fide Donk 1966.

Muciporus Juel, Bih. K. svenska VetenskAkad. Handl., Afd.: 23. 1897, fide Donk 1966.

Gloeotulasnella Höhn. & Litsch., Wiesner Festschrift (Wien): 57. 1908, fide Donk 1966.

Hormisciopsis Sumst., Mycologia 6: 32. 1914 (asexual synonym).

Epulorhiza R.T. Moore, Mycotaxon 29: 94. 1987 (asexual synonym).

Typification: Tulasnella lilacina J. Schröt. 1888.

Basidiomes (when present) crust-like, often on rotten wood, leaves or litter, often pinkish or purple. Hymenium composed of hyphae with or without clamps depending on the species, subhymenial structures often absent. Basidia with four sterigmata that are strongly swollen at the base, each separated by a septum from the clavate basal cell, which often collapses. Basidiospores variably shaped, from globose to helicoid; secondary conidia occasionally produced on germinating primary spores (Ingold 1984). Septa with central dolipores with continuous parenthesomes. Asexual morphs (when present) forming mycorrhizae with orchid roots or liverworts, or conspicuous orange to red pustulate, gelatinous, sporodochia on rotten wood. Stroma absent. Conidia in blastic, acropetal chains, which are dichotomously branched in some species, and arise directly from submerged hyphae or in sporodochia. Conidia aseptate, smooth or with minute pits, globose, subglobose, ellipsoidal to barrel-shaped, aseptate, hyaline, sometime with orange droplets. Somatic hyphae dikaryotic (Currah et al. 1990), lacking clamp connections when associated with asexual structures. Sclerotia occasionally produced.

Notes: The sexual morphs of Tulasnella are well-characterized by the production of thin, resupinate basidiomes on decaying substrates. They have unique “tulasnelloid” basidia with a basally swollen sterigma separated by a septum from the club-shaped basal cell, which usually collapses (Roberts 1994). Ultrastructure of dolipore septa is often used as a character to support class-level classification of Basidiomycota (Celio et al. 2006). Comparatively few exemplars are studied for each class and the character is rarely used for genus or species level classifications, which tend to be supported by DNA sequencing (e.g. Almeida et al. 2014, Linde et al. 2017, Arifin et al. 2020). We did not examine septal ultrastructure in this study. Most known asexual morphs of Tulasnella are characterized by the production of acropetal chains of conidia, and for most of the known asexual species, a relationship with orchids. Tulasnella aurantiaca is the exception because it grows on wood.

Apart from the species formerly included in Epulorhiza, a few other asexual states are described for other species of Tulasnellaceae but lack the characteristic monilioid conidial chains of T. aurantiaca. Tulasnella valentini, which grows on rotten wood, is reported to produce single apiculate, obclavate or fusiform conidia terminating irregularly lanceolate conidiogenous cells (Van de Put & Antonissen 1996). This is an unusual character state for a member of Tulasnellaceae and this asexual-sexual connection needs to be confirmed experimentally to eliminate the possibility that the observed conidia might have been those of a mycoparasite; for that reason, we did not include these deviating character states in the generic diagnosis above. Stilbotulasnella conidiophora, described from palm fronds but to date unsequenced, can be distinguished by its synnematous conidiomata and ellipsoidal ameroconidia produced in slimy masses from percurrently proliferating conidiogenous cells (Bandoni & Oberwinkler 1982).

Seifert et al. (2011) synonymized the monotypic Hormisciopsis with Hormomyces based on the protologue, and we include it as a synonym of Tulasnella above. We did not examine specimens of Hormisciopsis gelatinosa, but the protologue (Sumstine 1914) provides observations that are identical to what we have seen in specimens of T. aurantiaca. Seifert et al. (2011) followed von Höhnel (1917) and Donk (1962) in synonymizing Sphaerocolla with Hormomyces. However, re-examination of the protologue and slides from the holotype of S. aurantiaca (on Betula, Mustiala, June; H herb. Karsten PAK 3341; Supplementary Fig. S3), the type species of Sphaerocolla, suggests this was an error. Although the micromorphology is similar, the conidia of S. aurantiaca are slightly oblate rather than slightly ellipsoidal, and the fungus is reminiscent of the poorly-documented yeast O. margaritiferum, which occurs on slime fluxes on trees (Kurtzman 2011), a similar ecological niche to that reported for S. aurantiaca on living Betula trees. A notable difference is that O. margaritiferum produces endospores, which we did not observe in the material of S. aurantiaca, but these are apparently produced only in culture (Smith 1997) and no cultures of S. aurantiaca were isolated. Therefore, S. aurantiaca should be considered distinct from T. aurantiaca and may be conspecific with O. margaritiferum.

Based on our nuc rDNA phylogenies, Hormomyces is congeneric with Tulasnella, but appears to be distinct from all sequenced species of the latter genus. Therefore, a new combination is proposed:

Tulasnella aurantiaca (Bonord.) J. Mack & Seifert, comb. nov. MycoBank MB832426. Fig. 4.

Fig. 4.

Tulasnella aurantiaca. A, B. Appearance of fresh sporodochia on rotten wood. C. Detail of dried sporodochia. D. Detail of rehydrated sporodochia. E–G: 28-d-old culture, E on CMA with near-UV, F on OA with near-UV, G on OA in darkness. H, K. Living conidial chains in water. I, J. Conidial chains from culture on OA for 27 d under near-UV. L. Conidial chains in lactic acid. Scale bars: C–D = 500 μm, H, K–L = 10 μm, I, J = 20 μm.

Synonyms: Hormomyces aurantiacus Bonord., Handb. Allgem. mykol. (Stuttgart): 150. 1851.

Typification: fig. 234, taf. XI, in Bonorden 1851, Handb. Allgem. mykol. (Stuttgart): (lectotype, proposed here MBT388543, reproduced here as Fig. 1).

Hypsilophora callorioides Kalchbr. & Cooke, Grevillea 9: 18. 1880.

Hormomyces callorioides (Kalchbr. & Cooke) Sacc., Syll. fung. (Abellini) 6: 813. 1888.

Hypsilophora fragiformis Cooke, in Farlow, Appalachia 3: 247. 1884

Hormomyces fragiformis (Cooke) Sacc., Syll. fung. (Abellini) 6: 182. 1888.

Hormisciopsis gelatinosa Sumst., Mycologia 6: 32. 1914.

Basidiomes unknown. Sporodochia effuse, often pustulate, confluent in masses up to 5 cm long, occasionally more, or rarely solitary and <1 cm long, gelatinous or cartilaginous when fresh, waxy when dried, colours variable, from deep orange to garnet red when fresh, ranging from blonde to brown with various shades of orange and red when dry. Hyphae immersed, septate, branched, 2–4 μm wide, dikaryotic, binucleate (Supplementary Fig. S4), clamp connections absent, no stroma formed. Conidiophores arising from hyphal cells, clamp connection absent. Conidia blastic, in monilioid, branched acropetal chains, branching bifurcate, pattern variable, occurring mainly near the base of the chains, with usually fewer than five conidia between bifurcations, the terminal chains typically longer, often with up to 15 conidia and occasionally more, chains not readily separating into individual conidia. Conidia hyaline, often with conspicuous orange guttules when fresh, aseptate, smooth, subglobose to ellipsoidal or occasionally globose, often truncate, variable in length and width, (4.5–)7.5–9.5(–13) × (4–)5.5–7(–8.5) μm (mean 8.4 ± 0.1 × 6.4 ± 0.1, Q 1.31), thick-walled with walls ~1 μm thick.

Colony diam after 14 d, on MEA with near-UV (spectral range 300–400 nm) 30–50 mm on MEA in darkness 30–50 mm on CMA with near-UV 25–45 mm, on CMA in darkness 30–50 mm, on OA with near-UV 55–65 mm, on OA in darkness 40–55 mm. Colonies flat, often immersed, filamentous, often circular or irregular with undulate margins. Hyphae 2–5 μm wide. Sporodochia produced after 1–2 wk abundantly on OA, and in some strains on MEA, sterile on CMA, concolorous with mycelium, or more vibrantly coloured, especially when exposed to light: on CMA with near-UV after one month Salmon (6A4) to Pinkish White (9A2) in four strains and Cognac (6E7) in one; on CMA in darkness white, Blonde (5C4), Pinkish White (9A2) or Cognac (6E7); on MEA with near-UV white in four strains and Tangerine (6B7) in one; on MEA in darkness white in four strains and Yellowish White (4A2) in one; on OA under near-UV Greyish Red (7B4) to Pastel Red (8–9A5); on OA in darkness Pale Yellow (3A3) to Orange White (5A2) or Pale Orange (6A3). Conidia similar to those in vivo but slightly longer and narrower, (6.5–)7.5–10(–12.5) × (4–)5–6.5(–8) μm (mean 9 ± 0.5 × 5.8 ± 0.3, Q 1.57± 0.1).

Cardinal temperatures: Optimum 25 °C, minimum <5 °C, maximum 30–37 °C. Growth does not resume in cultures left for 1 mo at 37 °C, when moved to an incubator at 20 °C.

Distribution: Widely distributed in eastern North America, known from Ontario and Québec south to North Carolina and Tennessee and westward to Ohio. Tulasnella aurantiaca is known from western North America (one specimen collected in Arizona), Europe (Austria, Germany) and South Africa (Western Cape).

Habitat: Lignicolous and apparently saprobic, reported on rotten wood of conifers (Thuja) and angiosperms such as species of Populus, Platanus, Liquidambar and Vaccinum. It sometimes overgrows other fungi such as Fomitopsis betulina and Crepidotus spp.

Specimens examined (*indicates specimens that were cultured): As Hormomyces aurantiacus: Austria, Salzburg, Salzburg, on bark of rotten wood, 1880, E.A. Rau (BPI 726623). Canada, Ontario, Ottawa, Portobello Park, on rotten wood, probably Populus sp., Aug. 2014, J. Mack (DAOM 970822); ibid., Jul. 2015 (*DAOMC 252084); Ottawa, on rotten wood, probably Populus sp., 15 Jul. 2017, J. Mack (DAOM 970820, *DAOMC 252085); Quebec, Montreal, on rotten wood, 25 Oct. 2018, G. Cartier, isol. J. Mack (DAOM 970818, *DAOMC 252086); Victoriaville, on Piptoporus betulinus, 13 Jun. 2016, B. Fortier, isol. J. Mack (DAOM 970821, *DAOMC 252083). South Africa, on bark of unidentified tree, undated, V. Duthie Augusta (BPI 702911). USA, Arizona: Portal, Greenhouse Canyon, on rotten wood of Platanus sp., 26 Aug. 1956, J.L. Lowe & R.L. Gilbertson (BPI 726600); Kentucky, Crittenden, on rotten wood, 11 Jul. 1910, C.G. Lloyd (BPI 702908); Maryland, Great Falls, on rotten wood, 10 Oct. 1936, J.A. Stevenson (BPI 726621); Great Falls, on rotten log, 17 Oct. 1936, J.A. Stevenson (BPI 726619); Patuxent Wildlife Refuge, on wood of Liquidambar styraciflua, Jul. 1952, F. Berry (BPI 726599); Sligo, on rotten wood and polypores, 16 Oct. 1936, V.K. Charles & E.E. Dick (BPI 726619); New York, Adirondack Mountains, Paul Smiths, on rotten wood, 21 Sep. 1927, H.M. Fitzpatrick (BPI 702905); Greenport, on bark of Vaccinium sp., 3 May 1921, L. Roy (BPI 702904); North Carolina, Macon County, Highland, Mirror Lake Rd, on a fallen branch, 27 Aug. 1989, R.J. Bandoni no. 8495 (DAOM 970797, as Hormomyces sp.). Pennsylvania, Meadville, on rotten wood, 1922, E.C. Smith (BPI 702909); no location, on rotten wood, 2018, D. Newman, isol. J. Mack (*DAOMC 251988); Tennessee, Great Smoky Mountains National Park, Cosby, on rotten wood and Crepidotus spp., 27 May 2018, B.P Matheny, isol. B.P. Matheny (received as a culture, *DAOMC 251989); West Virginia, Fairmont, on rotten wood, no date, A. Boutlou (BPI 702903). As Hormomyces fragiformis: Canada, Ontario, Coopers Falls, on rotten log, 16 Sep. 1952, R.F. Cain (DAOM 82339); South of Pottageville, on hardwood plank, 6 Jul. 1954, R.F. Cain (DAOM 52048); New Durham, on decaying log, 11 Nov. 1930, R.F. Cain (DAOM 81729). USA, New Hampshire, Shelburne, on dead wood and polypore, Jun. 1883, M.C. Cooke K(M) 257483 (holotype); Maryland, Grand Falls, on rotten wood, 11 Oct. 1936, J.A. Stevenson (BPI 726622); Missouri, Perryville, Jun. 1883, C.H. Demetrio (BPI 726620); North Carolina: Asheville, on rotten wood, 1918, C.G. Lloyd (BPI 726625), Winston-Salem, on bark of rotten wood, 4 Jul. 1936, P.O. Schallert (BPI 726626); Ohio, Cincinnati, on bark, 3 Oct. 1920, C.G. Lloyd (BPI 702914); Chapel Hill, 14 Jan. 1924, on rotting deciduous wood, J.N. Couch (UNC 7236, in DAOM); Vermont, Middlebury, on dead wood, 20 Aug. 1896, E.A. Burt (BPI 702913); Virginia, W slope of Mt. Elliot, Augusta Co., on rotten wood, 17–21 Jul. 1936, J.A. Stevenson (BPI 726617); Rapidan River, Shenandoah National Park, rotten wood, 24 Sep. 1936, J.A. Stevenson (BPI 726616); West Virginia, Fairmont, on bark of rotten wood, no date, A. Boutlou (BPI 626624). As Hormomyces callorioides: South Africa, Somerset West, on rotting wood, no date, MacOwan (holotype K(M) 257481).

Notes: Tulasnella aurantiaca is characterized by orange to red sporodochia growing on wood or rarely, on other fungi, and the production of ellipsoidal conidia generally shorter than 10 μm. No sexual morph is known. While most species of Tulasnella occur on rotten wood (Roberts 1994), until now their asexual morphs have only been reported as symbionts of orchids or liverworts. Most of these symbiotic Tulasnella spp. have conidia > 10 μm long. The conidia of T. aurantiaca are most similar to those of T. epiphytica, which are also < 10 μm long and occur in branched, monilioid chains; those of T. epiphytica, however, have pitted walls (Pereira et al. 2003). Like T. aurantiaca, T. calendulina also produces orange colonies in culture, but its monilioid cells are larger (Zelmer & Currah 1995). Monilioid chains have been reported from basidiomes of T. violea, but the component cells are also >10 μm long (Roberts 1994). Sporodochial asexual morphs resembling T. aurantiaca include Oosporidium margaritiferum (Eurotiomycetes, Ascomycota), which also produces gelatinous sporodochia with long, branched chains of globose conidia, but differs by yeast-like growth in culture and its habit on living woody plants (Kurtzman 2011). Calloria fusarioides (Dermateaceae, Helotiales, Ascomycota; asexual morph formerly known as Cylindrocolla urticae), also produces gelatinous, orange sporodochia with long, branched acropetal chains but its conidia are cylindrical (Seifert et al. 2011). Heteromycophaga glandulosa (tentatively Tremellomycetes, cf. Weiß et al. 2014), a parasite of basidiomes of Exidia glandulosa, produces clavate conidia attached to conidiogenous cells by a clamp connection (Roberts 1997); it is unclear from the protologue whether the conidia are single or formed in chains.

Tulasnella aurantiaca grows easily on standard mycological media and grows in vitro between 5 and 30 °C, suggesting that it may be well-adapted to grow in different biomes. Most of the specimens examined were collected between May to October, suggesting that T. aurantiaca has a long sporulating season, which generally correlates with its optimum growth temperature of 25 °C in temperate climates. However, this species can also occur during the winter, as an examined specimen was collected in January.

After studying the available types of all described Hormomyces species and many other specimens, only H. fragiformis and H. callorioides were appropriately placed in Tubaki’s (1976) concept of Hormomyces. After careful studies of sporodochial coloration and conidial dimensions (Supplementary Fig. S1), we tentatively consider these two taxa to be conspecific with H. aurantiacus in agreement with Lloyd (1916). As far as we have been able to determine by microscopy, the herbarium specimens from Europe and South Africa are identical to the North America material. The colour variations we observed by growing cultures on different media in darkness or light, rewetting of sporodochia on ca. 30 herbarium specimens, and evaluation of approximately 35–40 photographs of both dried and fresh specimens on Mushroom Observer (Wilson et al. 2020), suggests that the colour distinctions used by previous authors, who examined only a few specimens, may not be diagnostic. However, the taxonomic significance of colour differences should be re-evaluated if additional evidence suggests the existence of cryptic, phylogenetic species within T. aurantiaca.

The cultures we examined and sequenced were all from North America and form a phylogenetically coherent clade. Bonorden’s H. aurantiacus was described from Europe. Unfortunately, despite the many collecting excursions by the senior author in the Netherlands in the mid-1980s, two visits to South Africa in 1996 and 2006, and requests to both professional and amateur colleagues to watch for this fungus over the past five years, we were unable to obtain fresh specimens or cultures of T. aurantiaca-like asexual morphs from Europe or Africa. Whether phylogenetically distinct species might occur in other parts of the world remains unknown, and we chose to exercise caution and not to propose any of the currently available material to epitypify T. aurantiaca. In any case, Bonorden’s illustration (reproduced here as Fig. 1), must serve as the lectotype (Art. 9.12, Turland et al. 2018) as proposed in the nomenclator above. If future studies of cryptic species support geographic separation, then H. fragiformis may be the appropriate name for material from North America, and H. callorioides for specimens from South Africa.

As reviewed in the Introduction, T. aurantiaca often was considered the asexual morph of Tremella mesenterica (McNabb 1969). Based on our phylogenetic results, the description of the asexual morph of Tr. mesenterica by Pipolla & Kotiranta (2008) and the yeast-like rather than filamentous growth of the latter in cultures (Fenwick 1995), this speculative connection is clearly untrue. The sexual morph of T. aurantiaca, if extant, remains unknown, but would be expected to have typical Tulasnella basidiomes and basidia.

Excluded species

Only H. aurantiacus, H. fragiformis and H. callorioides, discussed above, conform with the Tubaki (1976) concept of Hormomyces, now merged with Tulasnella. The remaining described species are discussed here.

Hormomyces abietinus P. Karst., Hedwigia 29: 271. 1890.

Synonym: ? Dacrymyces deliquescens Nees, 1816, fide Kennedy 1958.

We did not examine the type specimen. Kennedy (1958) considered H. abietinus a synonym of Dacrymyces deliquescens. McNabb (1973) synonymized D. deliquescens with D. stillatus and described the asexual morph of D. stillatus, which produces gelatinous, orange sporodochia with chains of arthroconidia 8–16 × 2.5–5.5 μm, similar to the conidial dimensions of H. abietinus reported by Karsten (1890). Neither Kennedy (1958) nor McNabb (1973) examined the holotype of H. abietinus and our request to examine the specimen in H remained unanswered; the synonymy remains tentative for this reason.

Hormomyces paridiphilus M. Zang & S.L. Wang, Acta bot. Yunn. 19: 324. 1997.

This species was reported as a inhabitant of tubers of Paris polyphylla var. yunnanensis (Melanthiaceae) in China (Zang & Wang 1997). The holotype consists of fragments of uncertain composition overgrown by a dry, greyish oidiodendron-like hyphomycete with chains of light brown, subglobose, finely ornamented conidia (2–)2.5–3(–3.5) × (1.5–)2–2.5 μm (mean 2.78 ± 0.4 × 2.35 ± 0.1, Q 1.18, n = 50). Globose chlamydospores-like structures (7–)9–13(–15) × (5–)6–10 μm (mean 10.1 ± 0.6 × 7.89 ± 0.4, Q 1.3, n = 15) composed of oblong, pale brown cells, 2–4(–6.5) × (1–)1.5–3.5 μm (mean 3.03 ± 0.1 × 2.13 ± 0.1, Q 1.45) were also observed, growing singly and terminally on hyphae. It seems unlikely that the authors would have misinterpreted the Oidiodendron as a Hormomyces, and dimensions of the observed conidia do not match the protologue; it seems more likely that the original specimen was overgrown by this mould later on. The species will have to be re-collected to re-evaluate its taxonomy. Until then, the taxon should be considered a nomen dubium.

Specimen examined: China, Yunnan, on tubers of Paris polyphylla var. yunnanensis, 10 Sep. 1995, S.L. Wang (holotype HKAS 30237).

Hormomyces peniophorae P. Roberts, Mycotaxon 63: 214, 1997.

The holotype consisted of twigs covered with basidiomes of Peniophora lycii, with inconspicuous gelatinous sporodochia ~1 mm diam forming pale spots on the hymenium. Conidia, conidiophores and haustoria consistent with those described by Roberts (1997) were observed. However, because the conidia are not formed in chains, there is no morphological reason to include this fungus in Hormomyces, and the presence of “tremellaceous” haustoria described and mycoparasitic ecology are more suggestive of Tremellales than Cantharellales. Because the haustoria are terminal, Roberts (1997) suggested a possible relationship of his species with the genus Sirotrema. However, the species of Sirotrema parasitize hosts in Rhytismataceae, and its species have clamp connections and yeast-like asexual morphs (Bandoni 1986). Hormomyces peniophorae should be recollected, cultured and sequenced before changes are proposed to its classification.

Specimens examined: UK, England, Devon, Scadson woods, Torquay, as a mycoparasite of Peniophora lycii growing on Rubus idaeus, 21 Jan. 1996, P. Roberts (holotype K(M):337706); Slapton woods, mycoparasite of Peniophora lycii growing on Ulmus, 10 Dec. 1994, P. Roberts (K(M):33198).

Hormomyces pezizoideus Speg., Boln Acad. nac. Cienc. Córdoba 2: 467. 1889.

The holotype specimen (LPS 28379) could not be provided on loan, but with the kindness of the curators, we were able to examine Spegazzini’s pencil drawing and a macro photograph of the specimen. These suggest a fungus with reddish sporodochia and branched chains of cuboidal or globose conidia ~1.5–2 μm diam, growing on a bamboo. These limited observations do not allow accurate identification of the species, but the small conidial size and bamboo habitat suggest it is not congeneric with T. aurantiaca.

DISCUSSION

Our taxonomic and phylogenetic revision of the hyphomycete genus Hormomyces provides the evidence for the formal synonymy with Tulasnella (Cantharellales) proposed by Stalpers et al. (2021), and the transfer of the type species H. aurantiacus to the latter genus. We consider two described species to be synonyms of T. aurantiacus based on type studies, and one species described in Hormisciopis is considered a synonym based on its protologue. Tulasnella aurantiaca is a relatively common fungus in temperate North America, conspicuous because of its gelatinous orange to reddish sporodochia and abundant globose to ellipsoidal conidia in branched, acropetal chains. While this combination of characters is diagnostic for T. aurantiaca, other sporodochial fungi with either Basidiomycetous or Ascomycetous affinities could be confused with T. aurantiaca. A key to similar genera is provided below. The species grows on rotten wood and occasionally overgrows wood-decaying fungi such as Fomitopsis betulina and Crepidotus sp. We did not observe microscopic structures that might indicate mycoparasitism, such as haustoria, in any specimens, suggesting that T. aurantiaca sometimes opportunistically overgrows fleshy basidiomes without parasitizing them. Hormomyces species are sometimes recorded from unusual substrates. One specimen of a Hormomyces sp. was reported from the palm Rhopalostylis sapida (McKenzie et al. 2004). A single strain identified as H. aurantiacus was isolated from the surface of a hibernating bat (Myotis septentrionalis) collected in a cave in New Brunswick, Canada (Vanderwolf et al. 2013). We did not examine either of these collections.

As described above, sporodochial colour is variable in this species. Our interpretation is that differences in colours of sporodochia constitute infraspecific variation related to age or environmental factors, and there are presently no correlations noted with other characters. In several other fungi, pigmentation is influenced by abiotic factors such as light, (e.g. Yu & Fischer 2018). Other sporodochial hyphomycetes, such as Clonostachys rosea, are well-known to produce conidiomata with variable colouration, ranging from yellow to orange, pink to red (Schroers et al. 1999), with a striking bluish green colour defining C. rosea f. catenulata (Schroers 2001). Variation and overlapping characters in basidiome colour or micromorphology are also problems for the identification of basidiomes of Tulasnella species in the absence of DNA sequencing (Cruz et al. 2014).

When this species was known as H. aurantiacus, it was frequently considered the asexual morph of Tremella mesenterica. However, all nuc rDNA analyses confidently place this species in Tulasnella among sequences identified as T. violea. The latter species is often considered a synonym of T. lilacina, the type species of Tulasnella, but this synonymy has not yet been evaluated by DNA sequencing. Sequences attributed to T. violea appear to be monophyletic in the ITS phylogenetic tree, with the exception of a single strain that may be a misidentified strain of T. eichleiriana. In the 28S tree, the two GenBank sequences labelled as T. violea are polyphyletic but the strains sampled are different from those in the ITS trees and it is unclear whether one or both are misidentified. Based on our ITS phylogenetic analysis, Tulasnella appears to be divided into two clades with T. aurantiaca in the same clade, with high support, as all the sequences labelled as T. violea. If the synonymy of T. violea and T. lilacina is eventually confirmed, and the concept of the species is clarified, it is probable that Hormomyces would remain a synonym of Tulasnella.

Until very recently, only nuc rDNA sequences were available for Tulasnella species. Sequences for the nuc glutamate synthase gene (C4102), and mito ATP synthase (C14436) were introduced by Arifin et al. (2020) for species from Australia belonging to a clade within Tulasnella that includes orchid root endophytes and possibly T. aurantiaca. Only a few RNA polymerase subunit I (RPB1) and RNA polymerase subunit II (RPB2) sequences are available (Matheny et al. 2006, Moncalvo et al. 2006). A phylogenetically robust generic concept and any subdivision of Tulasnella into segregate genera will require the generation of addition sequences covering the entire family, possibly using C4102 and C14436. Even if such studies justify splitting Tulasnella into segregate genera, it is probable that Hormomyces would stay nested within any narrower concept of the genus. Understanding the phylogenetic structure and species concepts in the core clade of Tulasnella around the type species clearly requires increased sampling and critical study. The inclusion of strictly asexual species in such analyses should provide additional phylogenetic and morphological characters and perhaps some clarity as future revisions proceed. The eventual taxonomic fate of Epulorhiza is less certain. The type species, E. repens (= Tulasnella calospora), and the other sequenced species of Epulorhiza are in the second ITS clade within Tulasnella, but this grouping is less evident in the 28S phylogeny. Therefore, it is possible that Epulorhiza could be re-instated if Tulasnella were divided.

Tulasnella species form mycorrhizae with several types of plants, especially several genera of orchids (Currah et al. 1997, Dearnaley et al. 2012, Almeida et al. 2014, Linde et al. 2017, Oberwinkler et al. 2017, Fujimori et al. 2019, Arifin et al. 2020), and liverworts (Preußing et al. 2009). Using mito 28S, Almeida et al. (2014) showed that the species formerly included in the asexual genus Epulorhiza, namely T. amonilioides, T. epiphytica, T. albertaensis and T. anaticula, do not form their own distinct clade, but are distributed across Tulasnella. In their phylogenetic analysis, strains identified as the type species of Epulorhiza, E. repens or its sexual morph T. calospora, are polyphyletic, and most of the clades represent unnamed endophytes isolated from orchid roots. We wonder whether T. aurantiacus may also be able to infect orchid roots because closely related species such as T. prima, T. secunda, T. warcupii, T. australiensis, T. rosea and several others occur in that niche (Linde et al. 2017, Arifin et al. 2020). Cantharellales are known for accelerated evolution of their nuc rDNA loci (Moncalvo et al. 2006) and the mutations in T. aurantiaca at the binding sites for the universal nuc rDNA primers, ITS4, ITS5 and LROR, may explain why it has not been detected in environmental samples. Different mutations in ITS primer binding sites of Tulasnella species were reported by Cruz et al. (2011), who designed specific primers different than those we designed for our study. Our modified primers for T. aurantiacus (Table 3) may be interesting to try for DNA surveys of orchid mycorrhizae or for the detection of related cryptic species.

Key to genera of hyphomycetes similar to Tulasnella aurantiaca (A = Ascomycota, B = Basidiomycota, U = unknown).

• 1. Clamp connections present on hyphae .............................................................................................................. Heteromycophaga (B)

• 1’. Clamp connections absent on hyphae ................................................................................................................................................ 2

• 2. No sporodochia produced ................................................................................................................................................................... 3

• 2’. Sporodochia produced ........................................................................................................................................................................ 5

• 3. Stipe dematiaceous ................................................................................................................................................... Phaeomonilia (A)

• 3’. Stipe hyaline ........................................................................................................................................................................................ 4

• 4. Producing large, inflated conidiogenous cells, often on Eucalyptus ............................................................................ Quambalaria (B)

• 4’. Conidia arising from simple hyphae conidia in long branched chains, only known from cultures derived from ascospores........................................................................................................................................................................................ Chaenothecopsis haematopus (A)

• 5. Conidia not in chains .......................................................................................................................................................................... 6

• 5’. Conidia in chains that may be branched ............................................................................................................................................. 8

• 6. Sporodochia cup-shaped, conidiogenous cells not in chains ........................................................................................... Ditangium (B)

• 6’. Sporodochia not cup-shaped, conidiogenous cells in chains ............................................................................................................... 7

• 7. Conidiogenous cells clavate .......................................................................................................................................... Algonquinia (U)

• 7’. Conidiogenous cells cuneiform ........................................................................................................................... Catenocuneiphora (U)

• 8. Sporodochia gelatinous or slimy ......................................................................................................................................................... 9

• 8’. Sporodochia dry ................................................................................................................................................................................ 14

• 9. Conidia 1-septate .............................................................................................................................................. Dacrymyces stillatus (B)

• 9’. Conidia aseptate ............................................................................................................................................................................... 10

• 10. On orchid roots, liverworts or slime fluxes on trees ......................................................................................................................... 11

• 10’. In beetle galleries, or on rotten wood or mushrooms ..................................................................................................................... 12

• 11. Endophytic on orchid roots, or liverworts ............................................................................... Tulasnella (excluding T. aurantiaca)(B)

• 11’. On slime fluxes on living trees .................................................................................................................................. Oosporidium (A)

• 12. In beetle galleries ........................................................................................................................................................... Raffaelea (A)

• 12’. On rotten wood, litter or rotten mushrooms .................................................................................................................................. 13

• 13. Conidia cylindrical .............................................................................................................................................................. Calloria (A)

• 13’. Conidia mostly ellipsoidal ............................................................................................................................ Tulasnella aurantiaca (B)

• 14. Conidia yellow and with a very thick cell wall....................................................................................................... Sphaerosporium (A)

• 14’. Conidia not as above ....................................................................................................................................................................... 15

• 15. Parasitic on fresh fruits ...................................................................................................................................................... Monilia (A)

• 15’. Not parasitic of fresh fruits ............................................................................................................................................................. 16

• 16. Conidia cuneiform ................................................................. Hyaloscypha (asexual morphs formerly included in Pseudaegerita)(A)

• 16’. Conidia ellipsoidal to fusiform, not cuneiform ............................................................................................................................... 17

• 17. Conidia hyaline.............................................................................................................................................................. Cylindrium (A)

• 17’. Conidia pigmented .................................................................................................................................................... Hoornsmania (U)

Supplementary Material: http://fuse-journal.org/

Figure S1.

Mean conidial dimensions (with error bars representing standard error) for all herbarium specimens and cultures examined, with the holotype of H. fragiforme represented by orange bars and the holotype of H. callorioides represented by green bars.

Figure S2.

Lectotype of H. fragiformis (A, C, E) and holotype of H. callorioides (B, D, F). A, B. Rehydrated sporodochia. C–F. Conidial chains. Scale bars: A, B = 500 μm. C–F = 10 μm.

Figure S3.

Conidia and conidial chains. A. Oosporidium sp. (DAOM 970823) identified using DNA sequencing. B. Holotype of Sphaerocolla aurantiaca (H). Both have similar conidial morphology and dimensions, suggesting that S. aurantiaca may be conspecific with Oosporidium margaritiferum. Scale bar = 10 μm.

Figure S4.

Nuclear staining of hyphae of DAOMC 251988, showing dikaryotic, binucleate hyphae, A, using near-UV light showing the stained nuclei and B with regular light. Scale bar = 20 μm.

Table S1.

Species, geographical location, host and herbaria for known type specimens of Hormomyces species.