Mycosphaerangium and Neomelanconium (Cenangiaceae) are closest relatives: phylogenetic relationships, morphology and a new species

Based on molecular phylogenetic analyses of a multigene matrix of partial nuSSU-ITS-LSU rDNA, RPB1, RPB2 and TEF1 sequences and by morphological evidence, the genus Mycosphaerangium is shown to be the closest relative of Neomelanconium, and confirmed to be a member of the Cenangiaceae (Leotiomycetes). While Mycosphaerangium and Neomelanconium share many traits like similar conidia, conidiogenesis, asci and ascospores, their apothecia differ particularly in excipular features and are therefore recognized as distinct genera. Mycosphaerangium tiliae, described from North America, is excluded from the genus but shown to represent the sexual morph of the European Neomelanconium gelatosporum, and it is therefore synonymized with the latter. Based on morphology, Neomelanconium deightonii is assumed to be congeneric with Neomelanconium gelatosporum, and it is lectotypified. Dermatea tetraspora and Phaeangium magnisporum, the basionyms of Mycosphaerangium tetrasporum and M. magnisporum, respectively, are lectotypified as well, and for M. tetrasporum, the asexual morph is recorded for the first time. Mycosphaerangium quercinum sp. nov. is described as a new species from various Quercus hosts in Europe, where it is shown to be widely distributed. It morphologically and ecologically closely resembles the North American M. tetrasporum, but differs in paraphysis and ascospore morphology and by croziers at its ascus base. The three accepted species of Mycosphaerangium and the two of Neomelanconium are described and illustrated. Mycosphaerangium magnisporum, M. quercinum and M. tetrasporum are recorded to be constantly associated with species of Coryneum, indicating a fungicolous habit, but no evidence for fungal associations has been found in Neomelanconium deightonii and N. gelatosporum.


Introduction
The anamorph genus Melanconium is morphologically characterized by unicellular ellipsoid brown conidia produced in acervuli (Sutton 1964). Melanconis, a diaporthalean genus, was commonly accepted to be its asexual morph (Barr 1978). However, as outlined in Rossman et al. (2015) and Jaklitsch and Voglmayr (2020), the generic concept of Melanconium and the true identity of its generic type, M. atrum, are obscure, and therefore the well-defined Melanconis was protected over Melanconium (Turland et al. 2018, Appendix III). Through its history, numerous species have been described within Melanconium, making it a heterogeneous assemblage. Therefore, many Melanconium species have been subsequently transferred to other genera such as Arthrinium, Greeneria and Harknessia (Rossman et al. 2015). Likewise, also the large genus Melanconis has been shown to be polyphyletic, and it has been restricted to eight closely related species (Jaklitsch and Voglmayr 2020).
Melanconium gelatosporum was described by Zimmermann (1913) from Tilia, for which Petrak (1940) Section Editor: Roland Kirschner established the new genus Neomelanconium. Subsequently, Petrak (1954) and recently Wijayawardene et al. (2016) added two additional species, N. deightonii and N. spartii, respectively. In the original description, Zimmermann (1913) supposed diaporthalean affinities, but Petrak (1940Petrak ( , 1954 considered it to be the asexual morph of a member of Massariaceae (Dothideomycetes). Until recently, the systematic affiliation of the type species remained unresolved and was based solely on morphological observations. Nowadays, the genus is classified as Pezizomycotina incertae sedis in Index Fungorum (accessed 14 Sep. 2020). However, Crous et al. (2019) published sequence data for N. gelatosporum which revealed a phylogenetic affiliation with Cenangiaceae (Leotiomycetes). Also, they decided to place N. spartii in a new genus Pseudomelanconium based on morphological differences, albeit no sequence data were available for the latter. In lack of type studies and sequence data, the status of the third species, N. deightonii, remained unclear. No sexual morph has been published for any Neomelanconium species.
The sexual morph genus Mycosphaerangium was established by Verkley (1999) for three species originally described in the genus Sphaerangium (Seaver 1951), with M. tetrasporum as type species. This is a later homonym of the moss genus Sphaerangium Schimp. The three species share many traits: erumpent apothecia, dark brown to black hymenium, a lighter brown coloured excipulum, thin-walled asci with a broadly rounded apex without a visible apical apparatus, and large, subglobose to ellipsoid, dark brown ascospores surrounded by a distinct refractive, hyaline gel sheath (Verkley 1999). Due to the peculiar morphology and the lack of fresh material, Verkley (1999) classified Mycosphaerangium as Helotiales incertae sedis, which was subsequently commonly followed by other authors (e.g. Jaklitsch et al. 2016). No sequence data have yet become available for the genus, and its systematic affinities within Leotiomycetes are currently unclear.
Several collections from corticated Quercus twigs made by the first author revealed an apparently undescribed coelomycete resembling the genus Neomelanconium. Concurrently, the third author made several holomorphic collections of a discomycete on Quercus in Spain which showed close affinities to the North American Mycosphaerangium tetrasporum. A morphological comparison of the asexual morphs of all these collections revealed that they fully matched, which was further confirmed by sequence data from cultures obtained from sexual and asexual morphs. This led us to initiate a detailed study on the genera Neomelanconium and Mycosphaerangium. Fresh material of Neomelanconium gelatosporum was collected for morphological studies, pure culture isolation and sequencing. Type material of all three Mycosphaerangium and of the two Neomelanconium species was reviewed. In addition, multigene phylogenetic analyses were performed to solve the placement of these genera.

Field survey, sample collection and sample sources
During the period 2015-2020, recently shed or still attached dead, corticated twigs of Quercus spp. and Tilia spp. were collected, examined for the presence of fungi of interest and air dried for subsequent morphological analysis, pure culture isolation and molecular studies. Material was collected in Austria, Greece, Italy and Spain mostly in light, mixed deciduous forests, occasionally on solitary trees in alleys or parks. Typical habitats are shown in Fig. 1. After processing, specimens were deposited in the fungaria of the University of Vienna (WU) and of the Consejería de Medio Ambiente (Junta de Andalucía), Sevilla (JA-CUSSTA).
In addition to fresh collections, specimens of Mycosphaerangium and Neomelanconium spp. were received on loan from herbaria FH, GAM, GZU, NY, and W and morphologically investigated. Details of the specimens used for morphological investigations are listed in the Taxonomy section; herbarium acronyms are according to Thiers (2020), and citation of exsiccatae follows Triebel and Scholz (2020).

Morphological observations
Stereomicroscopy photographs were captured with a Nikon SMZ 1500 stereomicroscope equipped with a Nikon DS-U2 digital camera or with a Sony Cyber-shot DSC-HX100V camera mounted in one of the oculars of a binocular lens, using a LED lamp for lighting. For certain images of ascomata, the stacking software Zerene Stacker version 1.04 (Zerene Systems LLC, Richland, WA, USA) was used.
For light microscopy, fresh collections were rehydrated and examined according to "vital taxonomy" in the sense of Baral (1992). Hand sections of ascomata and conidiomata were made using a razor blade and the following reagents were used as mounting media: tap water, 3% KOH, Congo red, Lugol's reagent (IKI, with 3% KJ and 1% iodine) and Indian ink. Slides were examined and photographed using a Zeiss Axio Imager.A1 (Zeiss, Jena, Germany) microscope equipped with a Zeiss Axiocam 506 colour digital camera or a Canon EOS 40D camera installed on an OPTIKA B-350 microscope. Measurements were done with the NIS-Elements D v.3.0 or Zeiss ZEN Blue Edition software packages. Measurements are reported as maxima and minima in parentheses and the range representing the mean plus and minus the standard deviation of a number of measurements given in parentheses; in addition, means of measurements (Me) and of l/w ratios (Qe) are given for ascospores and conidia.

Pure culture isolation
Mature conidiomata or apothecia were horizontally or vertically cut using a sterile razor blade and the surrounding host tissue was removed. Subsequently, the sections were transferred to a sterile drop of water on a microscope slide, torn apart with a forceps to release the conidia or the ascospores, and the suspension was pipetted on 2% corn meal agar (CMA, Sigma-Aldrich) supplemented with 2% w/v dextrose (CMD) or 2% malt extract agar (MEA) plates supplemented with 200 mg/l penicillin G and streptomycin sulphate (Sigma-Aldrich, St. Louis, MO) and incubated at room temperature. As the conidia and ascospores germinated slowly, the isolation plates were regularly checked for contaminations, which were removed. Germinated conidia or ascospores were then transferred to 2% CMD or 2% MEA plates, which were sealed with laboratory film and incubated at 22°C. Cultures were deposited at the Westerdijk Fungal Biodiversity Centre, Utrecht, The Netherlands (CBS culture collection).

DNA extraction, PCR and sequencing
Growth of liquid cultures and extraction of genomic DNA was done according to Voglmayr and Jaklitsch (2011), using the DNeasy Plant Mini Kit (QIAgen GmbH, Hilden, Germany). Six loci (ITS, LSU, and SSU rDNA, RPB1, RPB2, TEF1) were amplified. For details on the primers and annealing temperatures used for PCR and sequencing, see Table 1. PCR products were Fig. 1 Typical habitats of Mycosphaerangium quercinum (a-c) and Neomelanconium gelatosporum (d) purified using an enzymaticPCR cleanup (Werle et al. 1994) as described in Voglmayr andJaklitsch (2008). DNA was cyclesequenced using the ABI PRISM Big Dye Terminator Cycle Sequencing Ready Reaction Kit v. 3.1 (Applied Biosystems, Warrington) and the PCR primers; in addition, for some loci, internal primers were used (see Table 1). Sequencing was performed on an automated DNA sequencer (ABI 3730xl Genetic Analyzer, Applied Biosystems).

Phylogenetic analyses
For phylogenetic analyses, a matrix of aligned nucleotide sequences from six different phylogenetic markers (ITS, LSU, and SSU rDNA, RPB1, RPB2, TEF1) was produced. Sequences obtained in the present study were aligned to selected GenBank sequences of Cenangiaceae, Rutstroemiaceae and Sclerotiniaceae from Pärtel et al. (2017) and Johnston et al. (2019) and complemented with a few additions from GenBank. Five species of Chlorociboria (Chlorociboriaceae) were selected as outgroup according to Johnston et al. (2019). Sequences were aligned with the server versions of MAFFT (Katoh et al. 2002;http://www.ebi.ac.uk/Tools/mafft (Madeira et al. 2019), https://mafft.cbrc.jp/alignment/server (Katoh et al. 2019)) and subsequently checked, trimmed and refined using BioEdit version v. 7.0.9.0 (Hall 1999), removing excessive leading and trailing gap regions and adjusting the proteincoding gene alignments according to the correct amino acid reading frame. The combined sequence matrix contained 6677 nucleotide positions (555 from ITS, 1299 from LSU, 1633 from SSU, 1104 from RPB1, 1076 from RPB2, 1010 from TEF1). GenBank accession numbers of the sequences included in the phylogenetic analyses are given in Table 2; the species classification follows Galán et al. (2015), Johnston et al. (2019) and Perić et al. (2019).
Maximum likelihood (ML) analyses were performed with RAxML (Stamatakis 2006) as implemented in raxmlGUI 1.3 (Silvestro and Michalak 2012), using the ML + rapid bootstrap setting and the GTRGAMMA substitution model with 1000 bootstrap replicates. The matrix was partitioned for the individual gene regions, with separate substitution model parameters implemented.
Maximum parsimony (MP) analyses were performed with PAUP v. 4.0a167 (Swofford 2002), using 1000 replicates of heuristic search with random addition of sequences and subsequent TBR branch swapping (MULTREES option in effect, steepest descent option not in effect). All molecular characters Table 1 Primers used to amplify and sequence the nuclear internal transcribed spacer (ITS), large subunit (LSU) and small subunit (SSU) rDNA regions, the RNA polymerase II largest (RPB1) and second largest (RPB2) subunit genes and the translation elongation factor 1-α (TEF1) gene  Carbone and Kohn (1999) 1 Internal primers used only for sequencing Table 2 Isolates and accession numbers used in the phylogenetic analyses. Taxa with quoted genus names are generically misplaced (polyphyletic) but not yet formally recombined. Isolates/sequences in bold were isolated/sequenced in the present study Taxon were unordered and given equal weight; analyses were performed with gaps treated as missing data; the COLLAPSE command was set to NO. Bootstrap analysis with 1000 replicates was performed in the same way, but using 10 rounds of random sequence addition and subsequent TBR branch swapping during each bootstrap replicate. Bootstrap support below 70% was considered low, between 70 and 90% moderate, above 90% high and 100% maximum.

Culture characteristics
Culture images of Mycosphaerangium quercinum and Neomelanconium gelatosporum grown on 2% CMD and 2% MEA are shown in Fig. 2. Culture descriptions are given under the respective species.

Molecular phylogeny
Of the 6677 nucleotide positions, 1882 were parsimony informative (238 from ITS, 198 from LSU, 140 from SSU, 526 from RPB1, 493 from RPB2 and 287 from TEF1). The parsimony analyses revealed 1 MP tree 10,574 steps long (not shown). The best tree revealed by RAxML (− ln = 54,940.906) is shown as phylogram in Fig. 3. Except for a few deeper nodes, topologies of the ML tree were largely compatible with the MP tree. Tree topologies of our phylogenetic analyses agree well with those of Johnston et al. (2019). In the phylogenetic analyses, the Cenangiaceae clade and the Rutstroemiaceae-Sclerotiniaceae clade receive maximum support. Within the latter, the Sclerotiniaceae are resolved as a highly supported clade, but they are embedded within a paraphyletic Rutstroemiaceae, which consists of three highly supported subclades. Within Cenangiaceae, Mycosphaerangium quercinum and Neomelanconium gelatosporum are sister species with maximum support, and the sister group relationship of the Mycosphaerangium-Neomelanconium clade with Trochila craterium is moderately supported. As previously reported (e.g. Pärtel et al. 2017, Johnston et al. 2019, several genera like Cenangium, Lambertella, Moellerodiscus and Trochila are polyphyletic, indicating that the morphological characters currently used for generic circumscription do not reflect phylogenetic relationships, necessitating additional morphological and molecular investigations to reach more appropriate generic concepts within Helotiales. Sexual morph apothecial, erumpent through bark, solitary or in groups, sessile to subsessile; basal stroma absent. Recpetacle paler than the disc, beige, orange to dark brown, with raised, rim-like margins, curved inward when dry. Disc circular, elliptic to irregularly polyangular, concave to plane, dark brown to black, rough. Paraphyses embedded in hymenial gel, longer than the asci, filiform, septate, simple or branched, sometimes anastomosing, with hyaline to brownish, slightly to distinctly swollen or irregularly sinuousknobby tips, refractive vacuolar bodies (VBs) absent. Asci cylindrical-clavate, thin-walled, unitunicate, with broadly rounded to subtruncate apex without a distinct apical apparatus and opening by an apical rupture, with 4 uniseriately arranged ascospores. Ascospores dark brown, aseptate, subglobose or ellipsoid, with a thick, dark brown, minutely to distinctly warted wall, surrounded by a prominent hyaline gelatinous sheath rapidly expanding in water.
Asexual morph acervular, usually conspicuous due to black, glossy, effused conidial masses, associated with conidiomata or stromata of Coryneum spp. Conidiomata subperidermal, not to slightly elevating the bark, solitary, scattered, lenticular, with thin, whitish pseudoparenchymatous base. Conidiophores emerging from a pseudoparenchymatous base, reduced to the conidiogenous cells. Conidiogenous cells percurrent, smooth, cylindrical to filiform, hyaline, commonly turning yellowish-brown with age, with a single, often indistinct annellation. Conidia dark brown, aseptate, subglobose, broadly ellipsoid, ovoid to pyriform, apex broadly rounded, base truncate, thick-walled, guttulate, usually with one large and numerous small guttules, with smooth to finely verruculose wall, surrounded by a prominent hyaline gelatinous sheath rapidly expanding in water.
Habitat: on dead corticated branches of Betula nigra, associated with effete conidiomata of Coryneum betulinum.
Distribution: only known from GA, southeastern USA. Notes: Mycosphaerangium magnisporum is so far only known from two sites in GA, USA (see https://mycoportal. org/). The type collection, investigated and described by Edith K. Cash at the USDA, Beltsville, USA, was distributed to several herbaria in the USA, of which we here select the copy NY 02975298 as lectotype. In addition, a specimen from the same locality but with an earlier collection date was received from GAM; although it is marked as co-type, it is not mentioned in the protologue and therefore does not qualify for a type. The specimens seen contain only few ascomata and are in rather poor condition. Most apothecia seem to have fallen off which is already mentioned in the original description (Cash 1940). To save material, microscopy of the specimens was kept to a minimum, and some characters like excipulum morphology could not be documented. No conidiomata were seen on the specimens. Like in M. quercinum and M. tetrasporum, the apothecia of M. magnisporum are associated with a Coryneum species; it could be identified as C. betulinum (Fig. 4q). Despite the lack of sequence data and conidiomata, morphology and ecology leave no doubt about its affiliation with Mycosphaerangium. It differs from the other species by distinctly longer, ellipsoid ascospores with darker brown ends and a different host.
MycoBank: MB 836870. Etymology: referring to its Quercus hosts. Diagnosis: Mycosphaerangium quercinum is distinguished from the other species of the genus mainly by irregularly sinuous-knobby paraphysis tips encrusted by amorphous, olive to reddish brown plaques.
Culture characteristics: Colonies on CMD 50-mm diam after 3 months at 22°C, first hyaline, partly or entirely turning brownish or ochre, covered by floccose tufts or a dense white mat of aerial hyphae, with irregular wavy margin; reverse ochraceous with a caramel to dark brown centre. Conidia not observed in culture.
Habitat: on dead corticated branches of Quercus spp., constantly associated with effete conidiomata of Coryneum spp.
Distribution: apparently widely distributed in south-central and Southern Europe; confirmed from eastern Austria, Greece (Crete), Italy and Spain.
Holotype Notes: Mycosphaerangium quercinum resembles the North American M. tetrasporum, which, however, mainly differs by distinctly swollen, clavate, red brown paraphysis tips, lack of croziers at the ascus base and distinctly verruculose ascospore walls. It differs from M. magnisporum by distinctly shorter, subglobose to broadly ellipsoid, concolorous ascospores and a different host. In all specimens investigated, conidiomata as well as ascomata of M. quercinum are associated with effete conidiomata of Coryneum spp., indicating that it is fungicolous. Although M. quercinum seems to be widely distributed in Southern Europe, ascomata are so far only known from Spain.
Notes: Mycosphaerangium tetrasporum resembles the European M. quercinum, which, however, differs by irregularly sinuous-knobby hyaline paraphysis tips encrusted by amorphous brown plaques, the presence of croziers at the ascus base, and by finely verruculose spores appearing smooth in profile view. It differs from M. magnisporum by distinctly shorter, subglobose to broadly ellipsoid, concolorous ascospores and a different host. In all specimens investigated, conidiomata as well as ascomata of M. tetrasporum are associated with effete conidiomata or stromata of Coryneum species. To our knowledge, besides the type specimen from Newfield, NY (USA), only a recent Canadian collection is known (see above), which unfortunately is scant and in poor condition, and therefore no DNA data could be generated.
Notes: No sexual morph has been known for Neomelanconium until this study, and the connection with the mycosphaerangium-like sexual morph is confirmed by morphological data. Neomelanconium differs from the closely related Mycosphaerangium mainly by eight-spored asci and marginal, inversely stellate excipular teeth with a textura prismatica-porrecta covered by an external layer of white crystals. In addition, unlike in Mycosphaerangium, there is no evidence for association with fungi in Neomelanconium.
Morphology of the second species, N. deightonii, fits the generic type, and it is therefore retained within the genus, albeit no sequence data are available to verify its phylogenetic affiliation with Neomelanconium. Based on the lack of a gel sheath around the conidia, Crous et al. (2019) transferred the third species of the genus, N. spartii, to Pseudomelanconium.
Sexual morph: not known.
Distribution: known only from the type locality in Sierra Leone (Africa).
Notes: The conidiomata, conidiogenous cells and conidia of Neomelanconium deightonii resemble those of the generic type, N. gelatosporum. As already mentioned by Petrak (1954), N. deightonii differs from N. gelatosporum mainly by irregularly plurilocular conidiomata, which led him to propose the subgenus Neomelanconiopsis for it. Although no sequence data are available, for the time being, we decide to maintain N. deightonii within Neomelanconium based on morphological resemblance.
There is some confusion in the literature about the gel sheath surrounding the conidia. While Petrak (1954) described a rapidly dissolving gel sheath 1.5-2.5 μm wide, Sutton (1980) reported the absence of a gelatinous sheath. Re-investigations of the original collections showed that within conidiomata, the conidia were embedded in an extensive amorphous gel matrix (Fig. 9h) that apparently originated from dissolved conidial gel sheaths, as remnants were seen on some conidiophores (Fig. 9j). With the evidence at hand, we consider Petrak's (1954) observations to be correct; a gel sheath surrounding the conidia may be only observed in fresh material, becoming confluent and agglutinated to an amorphous matrix after prolonged storage.
Habitat: on dead corticated branches of Tilia spp.
Distribution: E u r o p e ( k n o w n f r o m A u s t r i a , Czech Republic, Germany) and North America (eastern USA).
Typification: Czech Republic, Morava, Lednice (formerly Eisgrub), Unterwald, on dead stems and branches of Tilia sp., 15 Jan. 1913, H. Zimmermann (W 1978   By their dark brown verruculose ascospores and conidia surrounded by a massive gel sheath, the lack of a visible apical apparatus and, at least in M. quercinum, of vacuolar bodies (VBs), the genera Mycosphaerangium and Neomelanconium deviate significantly from other Cenangiaceae, widening its morphological diversity. Based on the deviating ascal and ascospore features, Verkley (1999) even speculated about a closer affinity of Mycosphaerangium with Pezizales, but he could not determine the dehiscence mechanisms in the old herbarium material he had at hand for investigation. However, the fresh collections of the newly described M. quercinum indicate dehiscence by a rupture of the ascus apex.
Remarkably, Mycosphaerangium and Neomelanconium show some similarities to Cenangium ferruginosum, which also has erumpent apothecia, inamyloid asci, and ascospores with a gel sheath. In addition, the VBs that are characteristic of many members of the family are absent in Cenangium and in Mycosphaerangium. However, in the phylogenetic analyses, Cenangium is not closely related to Mycosphaerangium and Neomelanconium, indicating an independent evolutionary origin of these similarities.
Ecologically, Mycosphaerangium and Neomelanconium fit well the Cenangiaceae by growth on corticated branches still attached to the trees, which requires adaptations to drought. Colonization of aerial host parts (twigs or leaves) is a common ecological character of Cenangiaceae, which is connected with longevity and drought resistance of their apothecia (Pärtel et al. 2017). Therefore, morphological characters like rupture of the ascus apex and thick-walled, dark brown ascospores and conidia surrounded by a thick gel sheath may represent an adaptation to the dry conditions in their habitats.