Brahmaculus gen. nov. (Leotiomycetes, Chlorociboriaceae)

Abstract A second genus in Chlorociboriaceae is described here as Brahmaculusgen. nov. Macroscopically distinctive, all species have bright yellow apothecia with several apothecial cups held on short branches at the tip of a long stipe. The genus is widely distributed across the Southern Hemisphere; the four new species described here include two from Chile (B. magellanicussp. nov., B. osornoensissp. nov.) and one each from New Zealand (B. moonlighticussp. nov.) and Australia (B. packhamiaesp. nov.). They differ from species referred to Chlorociboria, the only other genus in Chlorociboriaceae, in their terrestrial habitat and ascomata that are noticeably more hairy than the known Chlorociboria species, most of which have apothecia with short, macroscopically indistinct hair-like elements. Based on our analyses, Chlorociboria as accepted here is paraphyletic. Additional study is needed to clarify where alternative, monophyletic generic limits should be drawn and how these genera may be recognised morphologically. Also described here are three new Chlorociboria spp. from New Zealand (C. metrosiderisp. nov., C. solandrisp. nov., C. subtilissp. nov.), distinctive in developing on dead leaves rather than wood and in two of them not forming the green pigmentation characteristic of most Chlorociboria species. New Zealand specimens previously incorrectly identified as Chlorociboria argentinensis are provided with a new name, C. novae-zelandiaesp. nov.


Introduction
The modern-day distribution of Nothofagaceae forests of the Southern Hemisphere and their associated fungi are often explained in terms of vicariance in relation to the breakup of Gondwana (e.g. Horak 1983). This explanation has been challenged in recent years (May 2017), with their distribution now thought to be due to a complex mix of ancient vicariant and geologically more recent long distance dispersal events, with evidence from Nothofagaceae phylogeny (e.g. Knapp et al. 2005) along with the phylogeny of some of their specialised fungal associates (e.g. Peterson et al. 2010b). The importance of these forests to the vegetation of southern South America and New Zealand has meant they have been amongst the most intensively studied mycologically in these regions (McKenzie et al. 2000;Johnston et al. 2012, Gamundí et al. 2017Romano et al. 2017a). Despite this, much of the fungal diversity in these forests remains undiscovered (e.g. Johnston et al. 2012;Romano et al. 2017b).
An example of this undiscovered diversity comes from recent collections of a beautiful, small terrestrial fungus from Nothofagaceae forests in South America, New Zealand and Australia that could not be matched to any known genus. Microscopically they had a clear affinity to Leotiomycetes. The unique apothecia are morphologically complex with a branched stipe and each branch ending in one or more cups, the hymenial surface in these cups forming a complex pattern comprising separate regions with asci and paraphyses, and with hair-like elements. Preliminary sequencing of ribosomal genes of both Australasian and South American specimens showed that these fungi are phylogenetically closely related and that they are also related to the Leotiomycetes genus Chlorociboria.
Here we describe four species in the newly erected genus Brahmaculus based on a combination of unique morphological and molecular characters. We incorporate Brahmaculus DNA sequences into a broad multigene Leotiomycetes phylogeny to show that these fungi represent a second genus in Chlorociboriaceae. Including Brahmaculus in the phylogeny makes Chlorociboria paraphyletic but the morphological and ecological differences between Chlorociboria and Brahmaculus species means that it is not sensible to treat them as a single genus. More intensive genetic sampling of additional Chlorociboria species will be needed to better resolve phylogenetic relationships within Chlorociboriaceae and to clearly define the phylogenetic and morphological limits of the genus Chlorociboria.
It is surprising that specimens of the morphologically spectacular Brahmaculus have not been collected more often in the Nothofagaceae forests of the Southern Hemisphere. Although clearly widespread geographically, these fungi presumably fruit rarely.

Samples
Specimens were collected during surveys of fungal diversity in Southern Hemisphere forests. Brief notes on macroscopic appearance were prepared and then the specimens dried and stored in the New Zealand Fungarium (PDD), National Herbarium of Victoria (MEL), Museo Nacional de Historia Natural (SGO) and the Florida Museum of Natural History (FLAS).

Morphology and culturing
Microscopic examinations were made from dried material routinely rehydrated and in 3% KOH and mounted in Melzer's Reagent, or where indicated, rehydrated and mounted in water. Vertical sections about 10 µm thick were prepared from apothecia rehydrated in 3% KOH using a freezing microtome and mounted in lactic acid. Where available, living cultures were grown from germinated ascospores and are stored in the ICMP culture collection, Manaaki Whenua-Landcare Research, Auckland.

Phylogenetic analyses
Two phylogenetic analyses were carried out. In the first, LSU, ITS, MCM7, RPB1 and RPB2 sequences from Brahmaculus specimens from South America and New Zealand, together with a set of Chlorociboria and Cyttaria specimens with multi-gene data available (Table 1), were incorporated into the alignments from Johnston et al. (2019 -data available from https://doi.org/10.7931/T5YV-BE95). Cyttaria was added because the analysis presented by Peterson and Pfister (2010a) suggested a relationship to Chlorociboriaceae and additional genes had recently become available for Cyttaria nigra. The expanded dataset was reanalysed using the same methods as Johnston et al. (2019). Briefly, genes were aligned using MAFFT (Katoh and Standley 2013), a maximum likelihood (ML) analysis of the concatenated alignments was run using IQ-TREE (Nguyen et al. 2015;Chernomor et al. 2016), using models selected by ModelFinder (Kalyaanamoorthy et al. 2017) for each partitioned gene, and ultrafast bootstrap (BS) analysis with 1000 replicates estimated branch support in the ML tree (Hoang et al. 2018). Xylaria hypoxylon and Neurospora crassa were used as outgroups.
The second analysis used ITS sequences only, treating all four Brahmaculus species, together with all Chlorociboria species with ITS sequences available, using Cenangiaceae  as the outgroup. The methods were the same as those used for the multi-gene analysis, except with the TIM2+F+I+G4 model, selected using ModelFinder. Alignments and partitions for each of the analyses are provided through the Landcare Research -Manaaki Whenua Datastore, https://doi.org/10.7931/2xet-fc88.

Phylogenetic analyses
Helotiales form a strongly supported monophyletic clade, and most families accepted within this order also form strongly supported clades (Fig. 1). The family-level clades of those families clustered in collapsed clades in Fig. 1 have 100% bootstrap support. Chlorociboriaceae and Cyttariaceae are strongly supported as monophyletic but their relationship to each other, and to other basal family-level clades within the Helotiales, is poorly resolved. A fully expanded version of Fig. 1 is available as a nexus file from the Landcare Research -Manaaki Whenua Datastore, https://doi.org/10.7931/2xet-fc88.
In both the multi-gene and ITS analyses, Brahmaculus forms a monophyletic clade within Chlorociboriaceae, but Chlorociboria as accepted here is paraphyletic (Figs 1, 2). The Brahmaculus species form a well-supported clade sister to a well-supported clade comprised of Chlorociboria aeruginella and C. halonata (from Northern Europe and New Zealand, respectively). The clade comprised of Brahmaculus plus these two species of Chlorociboria is sister to a clade containing the bulk of sequenced species of Chlorociboria, including the type species of the genus, C. aeruginosa.
At the species level, the ITS analysis supports the molecular phylogenetic distinctiveness of the novel species of Brahmaculus and Chlorociboria accepted here (Fig. 2).
Etymology. From Hindu mythology, named after Brahma, the four-headed creator god, reflecting the multiple heads of the apothecia, and the masculine diminutive -culus.
Diagnosis. Phylogenetically Chlorociboriaceae, distinguished from Chlorociboria by its terrestrial habitat, and apothecium with stipe branched near apex, each branch with an apothecial cup.
Description. Apothecia stipitate, yellow rhizomorphs at base of stipe, the stipe branched apically several times, each branch holding an apothecial cup. Receptacle and stipe densely covered with short hairs. Hairs more or less straight, cylindric, thin walled, with a few septa, pale brown intracellular pigment, externally densely encrusted with yellowish material, encrusting material dissolving in KOH + Melzer's reagent. The hymenium within each apothecial cup is typically divided into smaller segments, with areas comprising asci and paraphyses separated by clumps of hair-like elements. Excipulum comprises cylindric cells arranged more or less parallel to the surface, cells mostly long-cylindric, but  sometimes with outermost 1-2 layers of cells short and broad-cylindric, cell walls slightly thickened, hyaline, cells near base of hairs with pale brown vacuolar pigment. Asci with wall thickened at apex, amyloid pore extending through the wall, flaring near the inside and especially toward outside of the wall, 8-spored, with croziers. Paraphyses simple or tapering to apex, of similar length as asci. Ascospores oblong-elliptic, 0-septate, hyaline.  Notes. The four species described below are phylogenetically distinct but remarkably similar morphologically. There appear to be small differences in size and colour of the apothecia and shape of the paraphyses and hairs but having only a single specimen available for each species makes the significance of these differences difficult to assess. The rhizomorphs at the base of the stipe appear to be associated with tree roots. Based on the collecting sites, in South America and New Zealand the roots are likely to be Nothofagaceae, in Australia they may also be Nothofagaceae but Eucalyptus species were also growing in the vicinity. Observations from the South American specimens showed a loose weft of mycelium around the Nothofagaceae roots but there was no clear evidence of a mantle or ectomycorrhizal association. It is possible that these fungi are root endophytes, or perhaps parasites of Nothofagaceae-associated ectomycorrhizal fungi. Etymology. Refers to the Magellanic forests of the type locality. Diagnosis. Phylogenetically distinct from other known Brahmaculus spp., apothecia 3-8 × 1-2.5 mm, paraphyses undifferentiated to rounded apex, ascospores 5.5-9 × 1.5-2 µm (average 7.3 × 1.7 µm).

Brahmaculus magellanicus
Description. Apothecia 3-8 mm high, stipe 0.4-0.6 mm wide, cap 1-2.5 mm wide, the more or less globose cap comprising several closely packed apothecial cups, these arising from short, branches at the top of the stipe, hymenium pale yellow, hymenial areas broken into smaller segments by groups of bright yellow, hair-like elements amongst the fertile parts of the hymenium. Receptacle densely covered with stiff, bright yellow hairs, stipe with shorter hairs. Hairs 45-70 × 2.5-4.5 µm, straight, cylindric, tapering gradually in apical half toward small, rounded apex, thin-walled, sparsely septate, pale brown vacuolar pigment, densely encrusted with coarse, bright yellow crystals that dissolve in KOH + Melzers. Ectal excipulum comprising narrow-cylindric cells 8-20 × 2.5-3 µm oriented at low angle to receptacle surface, wall slightly thickened, mostly hyaline except cells at the base of hairs have pale brown vacuolar pigment. Medullary excipulum similar in structure but cells wider, 4.5-8 µm diam. Paraphyses 1.5-2.5 µm diam., undifferentiated at rounded apex, about same length as asci. Asci 40-55 × 4-5 µm, cylindric, apex rounded, wall thickened, amyloid pore extends through the wall, diffuse and flaring slightly towards the outside of the wall, crozier at base, 8-spored. Ascospores 5.5-9 × 1.5-2 µm (average 7.3 × 1.7 µm, n = 12), oblong elliptic, ends rounded, flattened on one side, straight to slightly curved, 0-septate, hyaline. Notes. The two Chilean species differ macroscopically, B. magellanicus having noticeably thinner stipes than B. osornoensis. The only known collection of B. magellanicus is from Magellanic subpolar forest in Patagonia that is dominated by Nothofagus betuloides. It is possible that this Brahmaculus species is restricted to these sub Antarctic cold southern forests but more specimens are needed to determine the range of the species.  Etymology. From the type locality. Historically important as a gold mining area (where T.H. Chinn, the great-great grandfather of the collector of the type specimen, prospected for gold in the 1880's), the name also reflects the deep golden colour of the apothecia of this fungus.
Notes. Brahmaculus moonlighticus has a stipe that consistently has 4 distinct branches near the apex. The other species have several separate hymenial cups, but these are held on very short branches arising from across the apex of the stipe, the margins of these cups superficially forming a more or less continuous layer. Etymology. Refers to the type locality, Volcan Osorno. Diagnosis. Phylogenetically distinct from other known Brahmaculus spp., apothecia 3-6 × 1-2.5 mm, paraphyses taper slightly to rounded apex, ascospores 6.5-10(-11) × 1.5-2 µm (average 8.3 × 2 µm).
Notes. The two Chilean species differ macroscopically, Brahmaculus osornoensis having noticeably broader stipes than B. magellanicus and slightly longer ascospores. B. osornoensis is known only from Nothofagus dombeyi forest in northern Patagonia on Volcan Osorno in the Vicente Perez Rosales National Park. It is possible that this Brahmaculus species is restricted to the wetter and warmer forests in northern Patagonia, but more specimens are needed to determine the range of the species.   Etymology. Named after the late Jillian ("Jill") Mary Packham whose assiduous collecting activities detected the type collection.
Notes. Brahmaculus packhamiae is macroscopically and microscopically similar to the Chilean B. magellanicus, both species having relatively long and narrow stipes. Notes with the specimen, indicate that when fresh the ascomata "seem to be attached to roots".  Etymology. Refers to the host substrate of the known specimens. Diagnosis. Phylogenetically a Chlorociboria, differs in developing on dead leaves rather than wood and in the asci being 4-spored when mature.
Notes. The substrate in both specimens was partly rotted leaves. It is possible that this fungus has a broader host range as most host-specialised, leaf-inhabiting Leotiomycetes are found on recently fallen leaves of their preferred host. Cultures are slow growing (on PDA, 9 mm after 8 weeks) with sparse mycelium and pale brownish pigmentation, remaining sterile. Etymology. Refers to the country of origin, in contrast to Argentina and the morphologically similar C. argentinensis, with which C. novae-zelandiae was previously confused.
Diagnosis. Similar to Chlorociboria argentinensis in having small, allantoid ascospores and lacking tomentum hyphae, but phylogenetically distinct and with smaller ascospores and narrower asci. Subsequent DNA sequencing of specimens from Argentina identified as C. argentinensis, showed that the New Zealand species is phylogenetically distinct. Morphologically, the two species are similar, both with an excipulum comprising highly gelatinous textura intricata, the apothecia lacking hair-like tomentum hyphae, and with small, allantoid ascospores. The New Zealand species has somewhat smaller ascospores (average 7.0 × 1.5 µm versus 9.9 × 1.9 µm) and narrower asci (3.5-4.5 µm versus 4-5.5 µm) compared with the Argentinian specimens recognised here as C. argentinensis. The Argentinian specimens match closely the description of Dixon (1975). Etymology. Refers to the host substrate of the holotype. Diagnosis. Phylogenetically a Chlorociboria, developing on fallen leaves rather than wood, differs from Chlorociboria metrosideri in having flexuous, coiled hairs and lanceolate paraphyses.
Additional specimens examined. New Zealand - Taupo  Notes. Chlorociboria solandri is micromorphologically distinctive in having scattered, large, lanceolate paraphyses, short-cylindric to more or less globose, thick-walled excipular cells, excipular tissue reacting either blue or red to Melzer's reagent, and coiling, rough-walled hairs. Known from two specimens from Fuscopora solandri leaves. A third specimen in poor condition, PDD 92925, could be the same species; it is morphologically similar but has longer hairs than the other two specimens. Cultures on agar are very slow growing (10 cm after 4 weeks), have little aerial mycelium and pale yellow brown pigments, remaining sterile. Etymology. From subtilis (delicate) referring to the stature of the apothecia. Diagnosis. Blue-green apothecia on blue-green stained fallen, partly decomposed leaves, hairs on receptacle rough-walled, somewhat flexuous, ascospores filiform, 45-55 × 1 µm. Description. Apothecia erumpent from blue-green stained leaf tissue. Apothecia less than 1 mm diam., cupulate with short, broad stipe, receptacle pale blue-green with tangled, white hairs, especially near the edge of the cup. Hairs 55-75 × 3-4 µm, somewhat flexuous, wall roughened. Apothecia in vertical section with ectal excipulum up to 30 µm wide, cells 6-10 µm diam., short-cylindric to square, walls thick, cells arranged in rows with a high angle to the receptacle surface; medullary excipulum poorly developed, two or three rows of narrow-cylindric cells, walls encrusted with blue-green material; stipe with thick-walled textura intricata. Paraphyses 1.5-2 µm diam., slightly wider towards the apex, often branched in the upper 20-30 µm, extending 15 µm beyond asci. Asci 85-105 µm × 5.5-6.5 µm cylindric, tapering gradually to small, truncate apex, wall thickened at apex, amyloid pore in inner half of wall, reaction most intense on inner edge of wall, pore appears more or less U-shaped, sloping outwards slightly through the wall, 8-spored, spores confined to the upper 60-100 µm of ascus, crozier present. Ascospores 45-55 × 1 µm, filiform, straight, 0-septate, hyaline. Notes. Most specimens are on fallen leaves of Dracophyllum spp., but the host range may be more extensive. A specimen on Fuscopora cliffortioides (PDD 55523) is morphologically similar, but perhaps with longer ascospores.

Discussion
The Brahmaculus species described here are so morphologically and ecologically divergent from Chlorociboria that they must be placed in their own genus. All four new species are members of a well-supported monophyletic lineage within Chlorociboriaceae (Fig.  2). However, in both the multigene and ITS analyses (Figs 1, 2) the Brahmaculus clade makes Chlorociboria, as currently understood in a morphological sense, paraphyletic. If alternative generic limits were to be drawn to recognise only monophyletic genera within Chlorociboriaceae, it is unclear how these genera could be distinguished morphologically. The type of Chlorociboria (C. aeruginosa) sits within the main Chlorociboria clade, and hence the name Chlorociboria will remain attached to the bulk of the species so far described in the genus. However, further sampling of Chlorociboria, including of species lacking green pigments (see below) is required before redrawing generic limits, especially in regard to the distinguishing morphological characters of the main Chlorociboria clade in relation to the phylogenetically differentiated species C. halonata and C. aeruginella.
The multi-gene phylogeny places Chlorociboriaceae in an isolated position near the base of Helotiales. Earlier analyses had suggested a relationship between Chlorociboriaceae and Cyttariaceae (Peterson and Pfister (2010a). The multiple genes newly available from a Cyttaria nigra specimen (PDD 117571) allowed Cyttariaceae to be treated in the multi-gene analysis. This showed that although Cyttariaceae was similar to Chlorociboriaceae in having an isolated position near the base of Helotiales, no particular phylogenetic relationship was found between the families. Cyttariales is treated here as a synonym of Helotiales.
Direct observations of the mycelium at the base of the stipes of several of the Brahmaculus spp. suggests a biotrophic relationship with either the roots of Nothofagaceae (possibly as root endophytes), or the mycorrhizal fungi associated with those roots (possibly as parasites). Johnston and Park (2005) noted a possible ecological relationship between wood rotting basidiomycetes and some of the wood-inhabiting Chlorociboria spp.
Not all of the specimens accepted here as Chlorociboria develop green pigment on their substrate. These include C. glauca and two of the newly described species from New Zealand (C. metrosideri and C. solandri). Both of these newly described species develop on fallen leaves, they have whitish rather than green apothecia, form no green pigment on their substrate, but have an excipular structure and the short, rough-walled, hair-like elements typical of several of the New Zealand representatives of the genus. The third newly named species from New Zealand, Chlorociboria subtilis, also develops on fallen leaves, but both the apothecia and the adjacent parts of the leaf have a blue-green pigment. The apothecial hairs of this species are better developed than those of most Chlorociboria species. Fungi morphologically similar to C. subtilis occur on fallen leaves in both eastern Australia (e.g. PDD 117581) and southern South America (unpubl. data) but they are not named here as only small specimens, and no DNA sequences, are available for these fungi.
Most known Chlorociboria species develop on green-stained, fallen wood. It is likely that there are other unrecognised Chlorociboria species, placed in other genera because they lack green pigment and have substrates apart from wood, the visually obvious features historically regarded as characteristic of Chlorociboria. Their true phylogenetic relationship may be revealed only when DNA sequence data becomes available for them, unless an alternative set of morphological features is discovered that is found to be characteristic of the Chlorociboriaceae clade. Huhtinen et al. (2010) discuss other seemingly ecologically or morphologically atypical Chlorociboria spp. from Europe. If these are shown to be Chlorociboria phylogenetically, they may be key to discovering phylogenetically informative morphological characters for the genus and family.

Conclusions
The phylogenetic breadth of Chlorociboriaceae is becoming better understood with the identification of Brahmaculus as a distinct lineage. For Chlorociboria, recognis-ing that not all species form apothecia on green-stained wood is an important step in characterising the genus and family both morphologically and phylogenetically, and in resolving more accurately its geographic distribution globally.