Polypores and genus concepts in Phanerochaetaceae ( Polyporales , Basidiomycota )

We explored whether DNA-phylogeny-based and morphology-based genus concepts can be reconciled in the basidiomycete family Phanerochaetaceae. Our results show that macromorphology of fruiting bodies and hymenophore construction do not reflect monophyletic groups. However, by integrating micromorphology and re-defining genera, harmonization of DNA phylogeny and morphological genus concepts is possible in most cases. In the case of one genus (Phlebiopsis), our genetic markers could not resolve genus limits satisfactorily and a clear morphological definition could not be identified. We combine extended species sampling, microscopic studies of fruiting bodies and phylogenetic analyses of ITS, nLSU and rpb1 to revise genus concepts. Three new polypore genera are ascribed to the Phanerochaetaceae: Oxychaete gen. nov. (type Oxyporus cervinogilvus), Phanerina gen. nov. (type Ceriporia mellea), and Riopa (including Ceriporia metamorphosa and Riopa pudens sp. nov.). Phlebiopsis is extended to include Dentocorticium pilatii, further species of Hjortstamia and the monotypic polypore genus Castanoporus. The polypore Ceriporia inflata is combined into Phanerochaete. The identity of the type species of the genus Riopa, R. davidii, has been misinterpreted in the current literature. The species has been included in Ceriporia as a species of its own or placed in synonymy with Ceriporia camaresiana. The effort to properly define R. davidii forced us to study Ceriporia more widely. In the process we identified five closely related Ceriporia species that belong to the true Ceriporia clade (Irpicaceae). We describe those species here, and introduce the Ceriporia pierii group. We also select a lectotype and an epitype for Riopa metamorphosa and neotypes for Sporotrichum aurantiacum and S. aurantium, the type species of the anamorphic genus Sporotrichum, and recommend that teleomorphic Riopa is conserved against it. Copyright Otto Miettinen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. MycoKeys 17: 1–46 (2016) doi: 10.3897/mycokeys.17.10153 http://mycokeys.pensoft.net A peer-reviewed open-access journal


Key words
Systematics, taxonomy, morphology, anamorphic fungi introduction Fruiting bodies are the most visible and easily studied element of the life cycle of macrofungi. Fruiting body morphology, including overall shape and construction of the spore-producing surface (hymenophore in basidiomycetes), was adopted early on as the guiding principle of fungal classification. This practical, but artificial, system has been largely replaced by a more natural, phylogenetic classification based on molecular characters (Hibbett et al. 2007, McLaughlin and Spatafora 2014.
At higher levels, there is rampant convergence and parallelism in the evolution of fruiting body and hymenophore types, possibly with a general trend towards evolution of more complex types. For instance, some orders of basidiomycetes only contain simple, effused fruiting bodies (e.g. Atheliales, Corticiales), while others are dominated by more complex forms (e.g. Agaricales, Gloeophyllales). Nevertheless, fruiting body morphology and hymenophore type remain significant for classification of fungi, particularly at very low taxonomic levels (e.g. within genera). The separate research traditions of specialists on morphological groups such as agarics, corticioid fungi and polypores have hindered comparisons of morphologically distinct yet closely related taxa. Otherwise well implemented studies for instance in polypore systematics sometimes neglect closely related corticioid fungi (Li and Cui 2013, Chen et al. 2015. A number of studies have shown that hymenophore types classified separately may actually belong to the same genus. Examples include Hyphodontia/Xylodon (Langer 1994, Resupinatus (Thorn et al. 2005), Schizophyllum (Nakasone 1996), Sidera (Miettinen and Larsson 2011), Steccherinum , and Trechispora (Larsson 1994, Birkebak et al. 2013). In the present study we explore whether phylogenetic genus-level classification and hymenophore type based classification can be united into a coherent system in the family Phanerochaetaceae. Larsson (2007) suggested the adoption of Phanerochaetaceae for a clade of corticioid fungi around the genus Phanerochaete. A more comprehensive sampling of the Polyporales by Binder et al. (2013) suggests that Phanerochaetaceae is indeed a wellsupported subclade of the large phlebioid clade, with the polypore genus Bjerkandera as the sister clade to the rest of the family. The family, as well as others mentioned in this paper, will also be adopted in the forthcoming treatment of Polyporales systematics by Justo et al. (in preparation). Aside from Bjerkandera, all the members of the Phanerochaetaceae identified in previous analyses have been corticioid or hydnoid fungi, most of them simple septate and monomitic, with the exception of Hapalopilus, a polypore genus with clamped hyphae. Here we describe two new polypore genera for the family (Figure 1). The corticioid members of the Phanerochaetaceae have been popular subjects of phylogenetic research, which has resulted in revision of genus concepts within the family. Greslebin et al. (2004) created the new genus Rhizochaete for pigmented Phanerochaetelike taxa in a separate clade within the Phanerochaetaceae. Wu et al. (2010) produced an extended phylogeny of the Phanerochaetaceae, extending the genera Hjortstamia and Phlebiopsis. The most comprehensive phylogenetic treatment until now, produced by Floudas and Hibbett (2015), resulted in creation of Phaeophlebiopsis for Phlebia-like taxa that are phylogenetically separated from the similar Phlebiopsis species, and moved a species of Hjortstamia to Phlebiopsis. Chikowski et al. (2016) extended the genus Rhizochaete further, including species with inconspicuous, poorly differentiated cystidia.
To better understand the morphological variation and evolution within the Phanerochaetaceae, we have incorporated new species -polypores and corticioid fungito the datasets published by earlier authors. With this new data we provide an updated phylogeny of the family, and revise species concepts therein.
We compiled three datasets for phylogenetic analyses: 1. LSU-dataset of the phlebioid clade (Irpicaceae, Meruliaceae, Phanerochaetaceae) based on nuclear ITS and LSU sequences, with 122 specimens. Of these, 100 had ITS and 118 nLSU sequence available. Total alignment length after manually removing unalignable characters was 1799 bp with 474 (26%) parsimony informative characters. The tree was rooted with Phlebia radiata (Meruliaceae). 2. Rpb1-dataset for Phanerochaetaceae based on rpb1, ITS and nLSU sequences with 34 species, all containing all three genetic markers. Total alignment length after re-moving unalignable characters was 3064 bp with 672 (22%) parsimony informative characters. The tree was rooted with Bjerkandera adusta. 3. Hapalopilus dataset with 16 ITS sequences, with a total alignment length 593 bp and 20 (3%) parsimony informative characters. The tree was rooted with H. percoctus (described in this paper).
We used MrBayes 3.2 (Ronquist et al. 2012) for inferring Bayesian consensus trees for the three datasets. The LSU and rpb1 datasets were partitioned as follows: ITS1 and ITS2 in one partition, 5.8S and LSU in another, and rpb1 separately. The nucleotide substitution model GTR+I+G was used for all partitions except Hapalopilus ITS, for which GTR was used. Models were chosen based on AIC scoring produced in jmodeltest (Darriba et al. 2012). Bayesian analyses were run with eight chains in three parallel runs, temp=0.1. LSU dataset was run for 10 (LSU dataset), 2 (rpb1) and 4 (Hapalopilus) million generations sampling every 2000 generations. All runs converged to below 0.01 average standard deviation of split frequencies. A burn-in of 25% was used before computing the consensus tree.
In parallel with the Bayesian analyses, we used RAxML 8.1.3 (Stamatakis 2014) for maximum likelihood inference and bootstrapping, partitioned similarly as in Bayesian analysis but using the GTR+G substitution model for all datasets. The tree with the highest likelihood from 100 individual runs was selected, and bootstrap values were calculated from 1000 repetitions. All the phylogenetic analyses were done at the CSC -IT Center for Science (https://www.csc.fi) multi-core computing environment. The resulting phylograms were pre-edited in FigTree 1.4.2 (Rambaut 2014) and processed further in CorelDRAW X6. Since the Bayesian and maximum likelihood analyses had similar topologies in all well-supported and relevant nodes, we report here only the Bayesian results amended with bootstrap support values from the maximum likelihood analyses. The alignments and phylograms are available in TreeBase (http://purl.org/ phylo/treebase/phylows/study/TB2:S19710).

Microscopy
We used a Leica DMLB microscope with optional phase contrast illumination for microscopic observations. Basic mountant was Cotton Blue (CB, Merck 1275) made in lactic acid, but we also used Melzer's reagent (IKI), 5% KOH, and Cresyl Blue (CRB, Merck 1280). Sketches were made using a drawing tube with the exception of spores that were drawn with free hand after a real measured spore. The sketches were then imported to CorelDRAW X6 and converted to vector graphics. Spore statistics were produced with R version 3.0.2 (R Core Team 2013).
In microscopic descriptions, the following abbreviations are used: L -mean spore length, W -mean spore width, Q -L/W ratio. Entry CB+ means cyanophily, CBacyanophily; IKI-means neither amyloid nor dextrinoid reaction. While reporting pore and spore measurements, the whole range is given in parentheses; 90% range excluding 5% extreme values from both ends of variation is given without parentheses; in case the values are identical, parentheses are omitted. For basidial and hyphal width measurements, the 20% tails are in parentheses.

Results
Our phylogenetic analyses support the division of the phlebioid clade into three lineages in line with previous research (Binder et al. 2013, Floudas andHibbett 2015): Meruliaceae, Irpicaceae (Byssomerulius clade in the sense of Larsson 2007) and Phanerochaetaceae ( Figure 2). In the analyses of our LSU dataset (ITS+nLSU), the Phanerochaetaceae receives excellent support (posterior probability=1, bootstrap sup-port=98%) and the Irpicaceae good to moderate support (pp=0.97, bs=59%), while the tree was rooted within the Meruliaceae (Phlebia radiata).
The Bjerkandera clade contains three genera: pileate polypores in the genus Bjerkandera, the effused corticioid genus Terana, and Porostereum spp. with smooth hymenophore and caps. All known species in these genera have clamped septa.
The Phanerochaete clade contains numerous corticioid species as well as five species of polypores: Ceriporia inflata, Oxychaete cervinogilva (=Oxyporus cervinogilvus), Phanerina mellea (=Ceriporia mellea), Riopa metamorphosa (=Ceriporia metamorphosa), and Riopa pudens. This clade contains only simple-septate species with one exception (Phanerochaete krikophora nom. prov.), whereas clamped and simple-septate species are intermixed in other parts of the Phanerochaetaceae. To create monophyletic genera, we have two options: a wide, morphologically heterogeneous Phanerochaete that includes a number of different-looking polypores, or three polypore genera in addition to a more homogenous Phanerochaete. We have opted to use three polypore genera: Oxychaete, Phanerina and Riopa. Even after this, a polypore species, Ceriporia inflata with incomplete pores, is nested within Phanerochaete, where it is closely related and microscopically very similar to spiny species. Nevertheless, this arrangements allows us to stick largely with morphologically identifiable genera (Tables 1 and 2).
Even though somewhat different from Phanerochaete, the polypore species in the Phanerochaete clade have an uncharacteristically simple hyphal structure for a polypore. They have no hyphal pegs or cystidioles. The subhymenial structure is loose, reminding a cymoid corymb in botanical terms (see Figs 7-9). In contrast, a typical polypore subhymenium is more difficult to study, hyphae are tightly interwoven, less clearly oriented and more irregular. Pores of Phanerochaete clade polypores are shallow and in many species irregular. Basidiocarps are relatively thin. All cystidia are hymenial, and no cystidia of tramal origin typical for many cystidioid polypores (such as Rigidoporus) are present.
The Donkia clade is a sister to the Phanerochaete clade, and contains the genera Donkia, Hyphodermella and Pirex as well as some species ascribed to Phlebia sensu lato. It includes smooth to hydnoid, pileate to effused species, many of which have clamped septa and are also otherwise morphologically quite different from Phanerochaete.
The Phlebiopsis clade contains a wide variety of different fruiting body types: pileate polypores with clamped septa (Hapalopilus), a resupinate polypore with simple septa (Phlebiopsis castanea or Castanoporus castaneus), phlebioid taxa with tight, simple- septate fruiting bodies and encrusted cystidia (Phlebiopsis), and loose rhizomorphic fruiting bodies (Rhizochaete). The internal structure of the clade is poorly resolved in the LSU dataset ( Figure 2). The rpb1 dataset ( Figure 3) includes too few species to be of much help either at this point. Three clades are well supported -Hapalopilus, Phaeophlebiopsis and Phlebiopsis -but Rhizochaete is poly-or paraphyletic. Further species sampling and genes may help the situation, but in our experience poor resolution of nrDNA markers in Polyporales often persists in expanded datasets.   No intuitively pleasing genus arrangement seems to be in reach for the Phlebiopsis clade. Based on our LSU dataset, the only well supported options for including all species in monophyletic genera would be either one genus for the whole clade (for which Hapalopilus has priority), or 10-13 separate genera, most of them new and monotypic. Neither is a satisfactory solution, and we have therefore taken a pragmatic stand and chosen a strict concept of Hapalopilus as a polypore genus and expanded the genus Phlebiopsis to include Castanoporus, leaving classification for the rest of the clade unresolved.
Thus defined, Hapalopilus is a small genus, currently with four polypore species (Figure 4). The rest of the species currently accepted in Hapalopilus (11 species), with different pigmentation and denser fruiting body consistency (cf. Aurantiporus croceus), do not belong to Phanerochaetaceae but rather to Meruliaceae ( Figure 2) and probably also other families. The expanded concept makes Phlebiopsis variable in terms of fruiting body morphology: smooth and effused (Phlebiopsis), poroid effused (Castanoporus), and stereoid, pileate species with smooth hymenophore (Hjortstamia). Microscopically the genus is rather uniform but not distinguishable from Phaeophlebiopsis, so for now we have had to abandon a strictly morphological genus concept for this species group.
The genus Riopa described by Reid (1969) has been considered a taxonomic synonym of Ceriporia, typified by C. viridans (Irpicaceae, Figure 2). This conclusion arises from an incorrect interpretation of the identity of the type species of the genus, R. davidii, as Ceriporia camaresiana (Ryvarden 1991, Bernicchia 2005. Our study of the type specimen shows that R. davidii is instead a synonym of Ceriporia metamorphosa (=Riopa metamorphosa, Phanerochaetaceae). The species called Ceriporia davidii (=Riopa davidii) by Pieri and Rivoire (1997) turns out to be an undescribed member of the Ceriporia clade in the Irpicaceae. The new species, named here as C. pierii, and four other newly described species form a well-supported group within the Ceriporia clade ( Figure 2).

Discussion
In our treatment, Phanerochaetaceae contains 14 genera, half of them with poroid species. We expect further sampling to result in more polypores and polypore genera for the family. Even so, corticioid species and genera will likely dominate Phanerochaetaceae.
Our taxonomic revision has managed to retain morphological genus concepts within Phanerochaetaceae, although this has required creation of three new genera for polypores. We show that natural genera (Phanerochaete, Phlebiopsis) contain a wide variety of hymenophore types -poroid, hydnoid and smooth -and can be best defined with a combination of microscopic characters of fruiting bodies. However, in one case (the Phlebiopsis clade, genus Rhizochaete in particular) no morphologically unique, phylogenetically justified genera could be defined, and we have felt the need to adopt an interim, partial classification arrangement.
Our results mirror those of Miettinen et al. (2012), whose similar treatment of Steccherinaceae identified genera (Antrodiella, Metuloidea, Steccherinum) each with variable hymenophore types (poroid, hydnoid or smooth). Like us, they found it generally possible to integrate phylogenetic information and morphological genera, but also identified one clade (Steccherinum), for which no morphologically satisfactory genus arrangement was in reach.
These studies reinforce the view that genera of macrofungi may contain species with widely variable fruiting body morphology. It seems that morphological genus concepts do have a future, but in many cases only when based on a wide set of microscopic characters. Finally, in a small minority of cases, it appears that morphologically unique genera of macrofungi may not be feasible.
Any taxonomist working with DNA sequences has the advantage of comparing their taxa with publically available sequences regardless of morphology of the source. We encourage a broad-minded approach outside traditional morphological conventions in taxonomic studies. When studying genus limits in particular, sampling and taxonomic treatment should be extended to include all the taxa with similar micromorphology and DNA sequences.
What factors gave rise to the diversity of fruiting body types in Phanerochaetaceae? We believe that ecological specialization is the major factor in driving fruiting body evolution within the family. For instance, rhizomorphic species with pellicular, simple fruiting bodies in Phanerochaete and Rhizochaete prefer decaying wood in advanced stages of decomposition and seem to colonize suitable substrates by growing through soil vegetatively. Their closest relatives in Phanerochaete and Phlebiopsis with denser fruiting bodies occur more frequently on recently fallen logs or even still attached branches. Most poroid, hydnoid and stereoid Phanerochaetaceae with relatively complex fruiting bodies produce them in earlier stages of wood decomposition, living trees or drier microclimatic conditions (Bjerkandera, Donkia, Oxychaete, Phlebiopsis castanea, Pirex, Porostereum, Riopa metamorphosa, Terana).
We see here a pattern where simple, ephemeral, rhizomorphic fruiting bodies belong mainly to species growing in soil and very decayed wood, whereas more persistent, complex and denser fruiting bodies tend to belong to species inhabiting living or recently dead trees. Species specialized in colonizing quickly consumed substrates such as rotten pieces of wood in soil are probably better off producing short-lived, simple fruiting bodies. Species using more concentrated and longer-term energy sources, such as recently fallen logs, can invest in more complex or longer-living fruiting bodies. Yet Phanerochaetaceae includes no species with long-lived perennial fruiting bodies, and it might be that the genetic make-up of species in the family sets limits to evolution of fruiting body forms.

Type species. Castanoporus castaneus (Lloyd) Ryvarden
Remarks. This monotypic genus contains one conifer-dwelling resupinate polypore species from East Asia. With its simple-septate hyphae, monomitic and dense structure (in basal layer) with thick-walled hyphae, middle-sized spores and subulate, encrusted cystidia the species brings into mind Phlebiopsis under the microscope. For a more detailed description see Nuñez and Ryvarden (2000).
Phylogenetically the species comes close to Phlebiopsis flavidoalba and P. pilatii. Together those three species form a sister clade to core Phlebiopsis, typified by P. gigantea (Figures 2 and 3). For now the most practical solution is to include Castanoporus in Phlebiopsis (see discussion under Phlebiopsis). Hjortstam (1987) listed Castanoporus castaneus under Phlebiopsis in his check-list of corticioid fungi, but made no formal combination. If Phlebiopsis would be defined more strictly, then Castanoporus could be put in use.
The genus Cystidiophorus has been described for Castanoporus castaneus, but for nomenclatural reasons described below we think Castanoporus should prevail against Cystidiophorus. Bondartsev and Ljubarsky (1963) described the monotypic genus Cystidiophorus with the species C. merulioideus as the type. Unfortunately, they did not indicate a type specimen for the species, which makes the species name invalid, and also rendered the genus invalid (Melbourne Code Art. 40; the cut-off year for type indication is 1958). Later, Imazeki (Imazeki and Hongo 1965) made the combination Cystidiophorus castaneus based on Merulius castaneus Lloyd, mentioning C. castaneus and C. merulioideus as synonyms. This combination does not qualify as a validation of Bondartsev and Ljubarsky's genus name, because Imazeki did not provide reference to the genus description, which is clearly separate from the species description in the original paper (Art. 38.1). In such a case, the genus could be considered valid with the condition that no previously described species is mentioned (Art. 38.5a), but this is not the case as Imazeki mentions Lloyd's species. Thus, we follow Ryvarden (1991) and regard Castanoporus as the correct name for this genus. Ginns (1969) lectotypified C. castaneus and gave a description of the type, which agrees well with our concept of the species as well as that of Imazeki's and Bondartsev's. Also Maas Geesteranus (1974) studied the lectotype from BPI. Zmitrovich et al. (2006) combined C. castaneus in Australohydnum. We do not have material of Australohydnum from Australia (the type locality of the type species) or any sequences, but judging from the type of cystidia and hyphal structure we think it is unlikely (but possible) that Australohydnum belongs to Phlebiopsis as delineated here (see Oxychaete for further notes on Australohydnum). If Phlebiopsis were to be split, Castanoporus and Australohydnum would probably both persist being morphologically quite distinct. Description. Pileate to resupinate polypores with soft to cottony corky, ochre to pink basidiocarps. Hyphal structure monomitic, clamps always present, generative hyphae slightly thick-walled, 2-5.5 µm in diameter, CB−, IKI−, KOH−, covered with granular, golden yellow pigment that dissolves in KOH turning purple. Cystidia absent. Hymenial cells relatively long, 12-25×4.2-5.5 µm. Spores ellipsoid to subcylindrical, thin-walled, 3-5×2-3.2 µm.
Remarks. Altogether 36 species have been combined to Hapalopilus, most of them bright-colored, soft polypores with a monomitic, clamped hyphal system. The genus type H. nidulans belongs to the Phanerochaetaceae as shown by us ( Figure 2) and previous work (Binder et al. 2005, Binder et al. 2013, Floudas and Hibbett 2015. Other species traditionally referred to this genus (H. alborubescens, H. croceus, H. ochraceolateritius etc.) belong to other lineages of the Polyporales (Niemelä et al. 2012, Dvořák et al. 2014, and their phylogeny and taxonomy will be revisited on further occasion.
Here we include four species in Hapalopilus in the strict sense, three of which are new to the genus. According to our data, Hapalopilus rutilans is a holarctic species, H. eupatorii and H. ribicola are found in Europe, and H. percoctus is so far only known from the type locality in Botswana. These species are morphologically very similar, and thus Hapalopilus as a genus is morphologically easy to characterize. The purple KOH reaction of Hapalopilus is shared by its pigmented, corticioid relatives in Rhizochaete (Wu et al. 2010, Chikowski et al. 2016. Unlike other Phanerochaetaceae polypore genera recognized here, Hapalopilus has a typical polypore subhymenium of sinuous, tightly packed, interwoven hyphae instead of the loose corymb type seen in Oxychaete, Phanerina, Phanerochaete and Riopa. Also Phlebiopsis species (including Castanoporus) have an interwoven subhymenium.
Morphological, ecological and geographic data of Hapalopilus species are summarized in Table 3. Remarks. H. eupatorii has completely resupinate, thin basidiocarps on dead herbaceous stems (Arctium, Eupatorium, and Reynoutria). It has been recorded once on thin fallen branches of Robinia in a thicket of Reynoutria. Karsten (1884)  France as Physisporus eupatorii, but it long remained an enigma for mycologists (Lowe 1956, Donk 1974. Recently it was reported from England as Ceriporiopsis herbicola (Fortey and Ryvarden 2007) and Germany as H. nidulans f. resupinata (Dämmrich 2014).
Remarks. Similar to Hapalopilus rutilans with pileate basidiocarps. Microscopically otherwise identical, but H. percoctus has clearly wider spores and tramal hyphae ( Table 3). The spore dimensions come close to H. eupatorii, which has larger pores, effused basidiocarps and grows usually on woody herbs. Its tramal hyphae are also narrower. Hapalopilus percoctus is the only species in the genus known to us from the Southern Hemisphere.  Remarks. This species was described by Karsten (1881) based on the sole collection from Finland. It had usually been regarded as a form of H. rutilans (Lowe 1956). However, our data show that specimens growing on Ribes spp. in North Europe are distinct from H. rutilans and phylogenetically closer to H. eupatorii. All specimens of H. ribicola studied by us are from Finland, from branches of both wild and cultivated Ribes spp. The species is evidently widely distributed and just overlooked.  (Bondartsev 1953, Gilbertson and Ryvarden 1986, Bernicchia 2005, Ryvarden and Melo 2014, but also the latter name has been in use (Murrill 1904, Donk 1974, Niemelä 2005. Hapalopilus rutilans is an older name than H. nidulans, and since both were sanctioned by Fries, the former has priority (ICBN Melbourne code art. 15.4).

Hapalopilus rutilans (Pers.) Murrill
Neither of the names has been typified. Persoon's original publication includes a rather uninformative painting of the fungus, probably Hapalopilus rutilans or Inonotus sensu lato. The original description of H. nidulans is similarly scanty. No material suitable for lectotypification remains of either species, so we have chosen to designate neotypes for both species to fix the nomenclature: H. rutilans based on a French specimen from oak in accordance to the protologue (Persoon 1798) as Persoon got material mainly from Germany and France, and H. nidulans based on a Finnish specimen, since Fries (1821) based his description on his own collection from neighboring Sweden. Ryvarden (1991) attempted to designate a lectotype for H. nidulans. We dispute his typification, since he used an illustration in Bulliard's publication from 1791 as the type, whereas Fries's original work does not refer to Bulliard. The fact that Fries later (1836-1838) referred to Bulliard doesn't make the drawing available for lectotypification: only the original material is valid under the code (ICBN Melbourne art. 9.2, 9.12).
Remarks. Other hydnoid and poroid genera with simple-septate hyphae and encrusted, thick-walled cystidia include Australohydnum, Phlebiopsis, Flavodon and Irpex. The latter two are phylogenetically distantly related to Oxychaete, and they possess dimitic hyphal structure quite different from the loose monomitic structure of Oxychaete. Phlebiopsis is phylogenetically distinct from Oxychaete (Figure 2), and its hyphal structure is more compact, even agglutinated (basal layer). Hyphae are also winding and covered with abundant brownish encrustation, which is lacking in Oxychaete. Cystidia are tramal in origin (as opposed to hymenial in Oxychaete). Due to the hyphal structure the basidiocarp is tougher and not board-like when cut as in Oxychaete.
Australohydnum is a more difficult case to decide on since there are no good references on the microscopic characters of the type species, Hydnum griseofuscescens Reichardt from Australia. Descriptions vary so much that it is possible that many species and even genera have been recognized as Australohydnum dregeanum (Berk.) Hjortstam & Ryvarden and its supposed synonyms (Jülich 1978, Hjortstam and Ryvarden 1989, Gilbertson and Adaskaveg 1993, Melo and Hjortstam 2002, Zmitrovich et al. 2006. Sometimes the structure is monomitic, sometimes dimitic; cystidia may be subulate or obtuse; basidiocarps may be resupinate with smooth hymenophore or hydnoid with caps. Reid (1955Reid ( , 1963 refers directly to Australian material and the type, and provides an illustration (under Irpex vellereus). His A. griseofuscescens is a pileate, hydnoid species with violaceous brownish basidiocarps, very thick-walled, simple-septate hyphae 4-9 µm in diameter, and abundant long, obtuse, poorly differentiated cystidia with tramal origin and fine apical encrustation. Reid states that the hyphal structure is monomitic, but has also drawn long aseptate hyphae. Spores are ellipsoid, medium-sized. The description and illustrations provided by Melo and Hjortstam (2002) from Portugal are very similar to those of Reid, and agree largely with an Indian specimen we have studied.
Hymenium dominated by basidioles and cystidia, cells with constrictions especially in older basidiocarps. Basidia cylindrical to narrowly clavate, collapsing upon spore release and difficult to spot, with 4 sterigmata. Cystidioles absent.
Ecology. Apparently prefers small-diameter dead wood of angiosperms. According to the description, the type was collected in a wet, shady forest in Javanese mountains. Australian collections we have seen are from drier localities (monsoon forest and city park).
Remarks. This monotypic genus comes close to Riopa both morphologically and phylogenetically, though the two do not seem to form a monophyletic group ( Figure  2). Morphological differences are summarized in Table 2. Description. Basidiocarp resupinate, yellow, ranging from yellowish cream to brownish yellow, 1-10×1-5 cm patches, 1(2) mm thick. Consistency fragile when dry. Pores shallow, somewhat irregular, splitting and eventually may turn dentate, 2-4 per mm, larger when split. Subiculum cream-colored, a bit lighter than pore surface, pellicular, cottony under the lens, 0.1-0.3 mm. Margin thinning out, smooth areas of several millimeters similar to tube bottoms may be present.
Distribution. Described from Sri Lanka. We can confirm it from East Africa (Tanzania, Kenya), Japan (Okinawa), and Indonesia (New Guinea). Sequences of Chinese specimens are also available in the INSDC.
Ecology. Grows on dead dicot trees, both standing and fallen, often in sun-exposed habitats. Remarks. East Asian, East African and New Guinean specimens have neither ITS sequence differences nor morphological differences, so we feel it is safe to assume that the type from Sri Lanka belongs to the same species. Morphologically the type specimen agrees very well with other material. Its spores are a little larger on average than in other specimens studied, but considering the large variability in size and shape of spores this is best interpreted as normal variance within species.

Remarks.
We have chosen to apply the genus name Phanerochaete for most of the Phanerochaete clade, excluding the three polypore genera Oxychaete, Phanerina and Riopa ( Figure 2). Morphologically, species in the Phanerochaete clade share microscopic characters such as simple-septate, relatively simple, loose hyphal structure, mid-sized hymenial cells, mid-sized straight cylindrical to narrow ellipsoid spores, and cystidia of subhymenial origin (Table 1 and 2). However, cystidia are rare and poorly differentiated or absent in three of the polypores (in the genera Phanerina and Riopa), and spores are distinctly curved in two species (Riopa). The third newly introduced polypore genus Oxychaete with its encrusted cystidia and large spores produces pileate and poroid basidiocarps. With the inclusion of these species, the genus Phanerochaete would become difficult to define morphologically.
Ceriporia inflata described by Jia and Cui (2012) belongs to Phanerochaetaceae with P. raduloides as the closest relative ( Figure 2). The hymenophore of C. inflata is composed of irregular pores with lacerate mouths, and that of P. raduloides of irregular teeth. Also Ceriporia jianxiensis (no sequence available) described in the same paper as Ceriporia inflata may be closely related. Their identity against P. capitata and P. aculeata along with other species in the P. raduloides group should be checked.
For now we consider Ceriporia inflata a species of Phanerochaete. Splitting the hydnoidporoid Phanerochaete of this group into a separate genus (possibly Phanerodontia Hjortstam) would make it necessary to split Phanerochaete into many small genera and would place morphologically very similar corticioid species into separate genera. For this reason we strongly prefer a wide concept of Phanerochaete that includes the hydnoid and poroid members, which are microscopically very similar to Phanerochaete sensu typi. See Tables 1 and 2 for characterization of the genus against similar genera in the Phanerochaetaceae. Hjortstam and Ryvarden (2010) described Phanericium and Phanerodontia for a few species placed traditionally in Phanerochaete. Their Phanerodontia includes four taxa with smooth to hydnoid hymenophores. Phanerodontia is probably a taxonomic synonym of Phanerochaete. Although the type, P. dentata, has not been sequenced, two other members of the genus have (P. chrysosporium and P. magnoliae). They clearly belong to Phanerochaete, and according to the rpb1 dataset to the same subclade within the genus with smooth to poroid members (Figure 3). Phanerodontia dentata does not closely resemble any polypore genus discussed here (except Phanerochaete) with its combination of thin-walled tubular cystidia, long basidia, thick-walled subicular hyphae and ellipsoid spores.
Phanericium is a monotypic genus, and the type P. subquercinum is characterized by hydnoid, effused fruiting bodies, absence of cystidia, hyphae of even width throughout the fruiting body and broad ellipsoid spores. This set of characters does not closely match taxa discussed in detail in this paper, and more detailed study is needed to conclude whether the genus belongs to Phaerochaetaceae.

Phlebiopsis Jülich
Persoonia 10: 137 (1978 Remarks. Phlebiopsis is typified by P. gigantea, a phlebioid species with agglutinated lower subiculum, well-developed basal layer/upper subiculum, thick-walled, simpleseptate hyphae and thick-walled, conical, encrusted cystidia (lamprocystidia). Our wider concept of Phlebiopsis dilutes this set of characters, but lamprocystidia, interwoven subhymenium and tightly built subiculum remain as important characters for genus delimitation against similar genera of the Phanerochaetaceae (Table 1). Hjortstamia crassa has been shown to be a close relative of Phlebiopsis, and has been included in that genus (Floudas and Hibbett 2015). We agree with this conclusion. The type species of Hjortstamia (H. friesii) has not been sequenced, but it is very similar to H. crassa. Thus Hjortstamia should for now be considered as a taxonomic synonym of Phlebiopsis. In addition to the above-mentioned Hjortstamia spp., a third similar species, H. papyrina, is combined to Phlebiopsis on morphological grounds.
The two main differences that have been emphasized to separate Hjortstamia from Phlebiopsis are reflexed basidiocarps and the loose subiculum of the former as opposed to the dense, agglutinated subiculum and totally effused basidiocarps of the latter. A closer look reveals that the difference is not as striking as often described. Whereas the genus type of Hjortstamia -H. friesii -and its close relative H. papyrina are distinctly pileate, basidiocarps of Hjortstamia crassa are much of the time fully resupinate or caps are small. Hjortstamia crassa also has an agglutinated upper subiculum or basal layer similar to agglutinated Phlebiopsis structures, as depicted by Wu and Chen (1992). Hjortstamia friesii has a tight (though not agglutinated) subicular layer composed of parallel hyphae as well (Hjortstam andRyvarden 1989, Boidin andGilles 2002). Subicular/cystidial hyphae of the above-mentioned species are strikingly similar, thick-walled, straight, stiff and sparsely septate.
A loose subiculum or pileate fruiting bodies do not seem to be useful characters separating Hjortstamia from Phlebiopsis, since loose and agglutinated species are widely intermixed phylogenetically within Phlebiopsis sensu lato (Figure 2). Hjortstamia crassa for instance is more closely related to the type species of Phlebiopsis than is Phlebiopsis flavidoalba with a very dense structure and effused fruiting bodies.
Sequences made available by Wu et al. (2010) include Phanerochaete brunneocystidiata and Phanerochaete laxa. The former is based on a paratype and the latter on the holotype. Wu combined the species in Hjortstamia due to sequence similarity to H. crassa. We haven't seen authentic material, but according to original descriptions, they seem to share basic Phlebiopsis characters except that no agglutinated layer was described (Wu 2000(Wu , 2004. Some Phlebiopsis species may turn out to belong to the Hapalopilus-Rhizochaete subclade instead of the Phlebiopsis subclade. For instance Phlebiopsis roumeguerei is nested within Phaeophlebiopsis as defined by Floudas and Hibbett (2015). More indepth research is needed to settle genus classification for Rhizochaete and Phaeophlebiopsis-like taxa. Description. White, resupinate polypores with shallow pores, 2-5 per mm. Hyphal structure monomitic, clamps absent. Hyphae thin-to slightly thick-walled, similar throughout the basidiocarp, hyphae not swollen, wider (3-5 µm in diameter) in subiculum, a bit narrower in trama (2.8-3.5 µm). Hymenial branching corymb-like. Thin-walled, poorly differentiated hymenial cystidia and conidia in one species. Spores curved cylindrical, sausage-like, thin-walled, mid-sized (4.5-6.5×2-3 µm).
Remarks. Reid (1969) Pieri and Rivoire (1997) regarded Riopa davidii and Ceriporia camaresiana as separate species, and made the combination Ceriporia davidii. Their concept of the species was mixed, as can be seen already from the spore variation they report. Their specimens from mainland France did seem to represent a species of Ceriporia separate from C. camaresiana, and consequently Ceriporia davidii was adopted by Bernicchia (2005) and Ryvarden and Melo (2014).
We studied the type of Riopa davidii, and it turned out to be a more recent synonym for Ceriporia metamorphosa (Fuckel) Ryvarden & Gilb. After studying the French material of Ceriporia davidii collected by B. Rivoire, we could also conclude that Ceriporia davidii sensu Pieri and Rivoire (1997) needs to be described with a new name (Ceriporia pierii). Ceriporia pierii and also C. camaresiana belong to the Ceriporia clade and are only distantly related to Riopa (Figure 2). Description. Basidiocarp resupinate, white, cream or straw-colored, consistency fragile when dry. Forms patches of a few cm that can fuse to extensive basidiocarps, up to 2(-3) mm thick. Pores rounded angular, soon splitting and then irregular and sinuous, mouths smooth, 2-3(4) per mm, up to 2 mm wide when split. Subiculum very thin, arachnoid to pellicular, white to cream, often lighter than pores. Margin thinning out, usually no sterile margin.
Anamorph known as Sporotrichum aurantiacum Link present or absent. Most but not all basidiocarps produce at least conidia in subiculum. When the anamorphic stage is well developed, it appears as an orange mass of conidia similar in shape to Haplotrichum aureum, in conjunction with basidiocarps or separately. Microscopically composed of thick-walled, ellipsoid to constricted conidia (8.2-12.2×5.2-7.8 µm, n=36/3) born singly as apical parts of slightly to clearly thick-walled, partly encrusted hyphae, (3.2)3.6-4.5(7.2) µm in diameter, walls ≤1.5 µm. The conidia and hyphae are yellow, the plasma of the conidia stains in CB, and the walls are CB− to CB(+) and slightly dextrinoid. In KOH the conidia stain pinkish red in masses. Wakefield (1952) proved in the lab that the polypore and conidial stages belong to the same organism.
Ecology. Grows preferably on rotten oak trunks. We have seen it on Eucalyptus and Salix caprea, also reported on Castanea, Juglands and Malus (Bourdot and Galzin 1928, Ryvarden and Gilbertson 1993, Pieri and Rivoire 1997. Remarks. Fuckel's herbarium is in Wiesbaden (WIES), but its material is not available for loan. A duplicate of an original Fuckel specimen in Stockholm is chosen as the lectotype here. It represents an almost completely destroyed anamorphic stage. For practical reasons we also select an epitype from the Czech Republic.
Ecology. Grows on fairly rotten angiosperm wood. The type comes from low-land rainforest.
Remarks. The species lacks any distinct characters. Cream-colored basidiocarp with non-inflated hyphae and corymb-subhymenium help to distinguish this species from Ceriporia spp. It is similar to Phanerochaete inflata and Ceriporia jianxiensis, but differs in having long-celled, narrower subicular hyphae (mostly <5 µm in diameter). The relatively small cylindrical curved spores exclude Oxyporus spp. and Emmia spp. Except for the smaller pores and the lack of cystidia and a conidial stage it is very similar to Riopa metamorphosa.
To formally settle the names Sporotrichum, S. aureus and S. aurantiacum we need to designate neotypes for the two species in question. In line with Stalper's interpretation, we designate here the collection Vlasák 0511/15 (H 7008577) as the neotype of S. aureum Link, and collection Spirin 2456 (H 7029505) as the neotype of S. aurantiacum.
This makes Sporotrichum an older name available for Riopa under the ICBN Melbourne code article 59.1. However, adoption of Sporotrichum, traditionally a very heterogeneous set of anamorphs, for a small genus of polypores would only create confusion. Stalpers (1984) described the genus as a "litterbag" of conidiogenous fungi, and accepted only three species. According to him the teleomorphs of those three species are in separate genera (Laetiporus, Phanerochaete and Pycnoporellus/Riopa) that we now know are phylogenetically distinct. Although the type species Riopa produces an anamorph, we have seen no conidia in the other species of the genus (R. pudens). In this situation it is better to coin Riopa, a name without identity problems, for this polypore genus. We suggest conservation of the teleomorphic name Riopa D. A. Reid 1968 over the anamorphic Sporotrichum Link 1809. Ceriporia pierii and four closely related species described below seem to form a subclade of the large Ceriporia -Leptoporus clade (Figure 2). In morphological terms, the C. pierii group encompasses species with pale colored (white, pale pink or pale ochraceous), minutely rhizomorphic basidiocarps (Figure 10), and cylindrical to ellipsoid basidiospores normally exceeding 2 µm in width. In addition, fan-like crystal aggregations occur among hyphae (Figure 11g), and subicular hyphae are considerably wider than tramal and subhymenial ones. The latter feature is not unique for the C. pierii group but is found for instance in the genus type C. viridans and its closest relatives. The C. viridans group is not very closely related to C. pierii and its sibling species (Figure 2), although morphological differences are very subtle. In the Ceriporia viridans complex the basidiospores are curved and mostly cylindrical, less than 2 µm in width (except C. excelsa), and hyphae possess more or less thickened walls (hyphal walls are thin in the C. pierii group). The C. purpurea and C. spissa species complexes have much brighter, red-colored basidiocarps, cylindrical spores, and hyphae of more or less equal diameter throughout the basidiocarp.

Key to genera of Phanerochaetaceae
Morphologically species in the C. pierii group are very similar to each other, pore and spore characters being the most useful for identification (Table 4). ITS sequence differences are clear, 3.2-10.6% between species. Below is a general description for species in this group.
Description. Basidiocarps annual, resupinate, very thin (below 1 mm), 1-20 cm wide. Sterile margin byssoid, white to cream-colored, producing thin, white rhizomorphs (in all species but not all specimens). Pore surface pale-colored (white-yellow-pale ochraceous), pores shallow, uneven, angular, partly fusing together and even irpicoid, 2-6 per mm. Dissepiments mostly thin, wavy to dentate. Subiculum byssoid, white, very thin (up to 0.1 mm). Hyphal system monomitic, simple-septate. Subicular hyphae thin-to moderately thickwalled, branched at sharp angles, producing abundant H-like connections, always wider than Remarks. Ceriporia humilis produces rather large basidiocarps with rhizomorphs at the marginal area or in the substrate. The type specimen was collected from a fallen oak log in Nizhny Novgorod Region, European part of Russia. Another, much older collection derives from Helsinki, Finland (HFR009978, a fallen log of Acer platanoides). One sequence of C. viridans in the INSDC from Shanxi, China belongs to C. humilis (KC182775, Dai 7642) showing that the species is present in East Asia, too. Ceriporia humilis has the narrowest spores in the whole species complex. Etymology. Named after mpur, the people and language spoken around the type locality.
Remarks. Ceriporia mpurii is very similar to C. humilis (see above), differing in slightly darker color of the basidiocarps and a bit rounder spores. Moreover, hyphae in older parts of tubes are densely arranged and glued together, while they are loosely arranged in C. humilis. Ceriporia mpurii is known so far from its type locality in New Guinea.
Etymology. Named after Max Pieri, who with Bernard Rivoire first discovered this species.
Remarks. Ceriporia pierii is introduced here to encompass C. davidii sensu Pieri and Rivoire (1997). Pieri and Rivoire identified C. camaresiana (Bourdot & Galzin) Bondartsev & Singer as the most similar species to C. pierii, but our data show that the two are not closely related (Figure 2). Basidiospores of C. camaresiana are clearly curved, mostly bean-shaped and longer, 5.26×2.74 µm (Table 5). Moreover, the hy-table 5. Spore measurement statistics of polypores. Bold-face values are composite statistics for species. L = average of spore length, W = average of spore width, Q = L/W, and n = number of spores measured. The whole range is given in parentheses; 90% range excluding 5% extreme values from both ends of variation is given without parentheses; in case the values are identical, parentheses are omitted.
Description. Basidiocarps 0.2-0.5 mm thick, up to 20 cm in the widest dimension. Sterile margin up to 3 mm wide. Pore surface yellowish, in dry specimens pale to dirty ochraceous, in a few portions with pinkish hues, pores 3-4 per mm. Subicular hyphae subparallel, 5-13.6 µm in diameter, some inflated. Tramal hyphae 2.6-4 µm Remarks. Ceriporia sordescens is a close relative of C. pierii differing by its ochraceous colors and smaller spores. We have studied one morphologically very similar specimen to C. sordescens from Ontario, Canada identified (incorrectly in our view) as Poria griseoalba by R.F. Cain (H ex TRTC 33465). It may represent yet another species in the C. pierii group, differing from C. sordescens mainly by its smaller pores 4-5 per mm, and longer, thick cylindrical spores 4.2-5.1×2-2.3 µm (n=30), L=4.54, W=2.15, Q=2.12. Poria griseoalba (Peck) Saccardo was described from Osceola, New York (Peck 1885) as having small-pored, grayish white basidiocarps, and Lowe (1966) placed it among the synonyms of Poria rhodella Fr. (= Ceriporia viridans s. lato). Even if Lowe's species concept was probably wider than today, Poria griseoalba belongs in the vicinity of C. viridans and is clearly not conspecific with C. sordescens.

Specimens examined
We studied specimens from herbaria H, O, K and LY, as well as specimens from the personal herbarium of Josef Vlasák (JV). Type specimens of species described here are omitted since their specimen information is found in the descriptions. Sequenced specimens are marked with an asterisk (*).

Five new species of Graphidaceae introduction
Phenotypical characters, such as morphology of the thallus and ascomata and anatomy of the ascomata as well as secondary chemistry have traditionally guided species delimitation in lichenized ascomycetes. However, especially crustose lichens often exhibit only few traits and without independent markers, such as DNA sequence data, it is often difficult to assess whether variation is due to genetic differences or plasticity. Indeed, recent phylogenetic studies suggest high amounts of homoplasy in phenotypical characters used to delimit taxa in lichenized fungi (Grube et al. 2004;Tehler and Irestedt 2007;Schmitt et al. 2009;Rivas Plata and Lumbsch 2011;Lumbsch et al. 2014a). Hence, molecular data have greatly increased our ability to identify distinct lineages, including the detection of numerous cryptic lineages (Crespo and Lumbsch 2010;Lumbsch and Leavitt 2011;Leavitt et al. 2015). While numerous foliose and fruticose lichen groups have been studied in some detail, especially in the diverse Parmeliaceae, our knowledge on species delimitation in crustose lichens is still in its infancy. However, among predominantly crustose families, Graphidaceae is now relatively well known. Graphidaceae constitutes the largest family of crustose tropical lichens with about 2100 accepted species (Rivas Plata et al. 2012;Lücking et al. 2013;Cáceres et al. 2014;Van den Broeck et al. 2014;Lumbsch et al. 2014b;Kraichak et al. 2014). The family has its center of distribution in the tropics, but also occurs in temperate regions with a smaller number of species, in some cases even extending towards the Sub-Antarctic region. The family is most common, however, in the tropics where its species occur often on bark, but can also be found on rocks, wood or soil and sometimes on leaves. Recently, the first author started a project on the diversity of thelotremoid Graphidaceae in East Asia . Thelotremoid Graphidaceae have rounded ascomata (formerly placed in Thelotremataceae), in contrast to species with lirellate ascomata. The group is still relatively poorly known in Thailand and generally in south-east Asia, but preliminary studies have provided important baseline data for the distribution of species and have indicated that numerous additional species can be expected in Thailand (Homchantara and Coppins 2002;Sutjaritturakan and Kalb 2015). Molecular data have been used to identify distinct lineages in this group of lichenized fungi and subsequent re-analysis of phenotypical characters often allowed identification of morphological or chemical traits to separate those species Mangold et al. 2014;Poengsungnoen et al. 2014;Medeiros et al. 2016). This paper employs molecular, morphological and chemical data to identify six distinct lineages of thelotremoid lichens from Thailand and to describe them as species new to science. Based on our limited sampling of thelotremoid Graphidaceae from other regions of southeast Asia, we expect the new species described here from Thailand to occur in other countries of the region.

Material and methods
This study is mainly based on new collections made by the first two authors deposited in F and MSUT. Sections of thalli and apothecia were cut using a razor blade and examined in water, a solution of KOH, and Lugol's solution using a ZEISS Axioscope 2 plus compound microscope. Chromatography (HPTLC) was performed with standard solvent systems A and C (Culberson 1972;Arup et al. 1993).
We performed two different phylogenetic analyses: 1) sequences of six samples of the genus Glaucotrema were aligned with two outgroup taxa (Leptotrema wightii, Reimnitzia santensis) and 2) sequences of 35 samples of Ocellularia s. str. were aligned with O. cavata as outgroup. Selection of samples was done using Blast searches and included best hits to ensure that all similar sequences were included. In addition sequences of morphologically similar species were added to the data set. Sequences of mtSSU rDNA, nuLSU rDNA, and the protein-coding RPB2 gene were used for this study. Voucher information and Genbank numbers are listed in Table 1. DNA isolation, PCR, and direct cycle sequencing conditions were described previously .
For the phylogenetic analyses, the alignment of the nucleotide sequences for each dataset was performed separately using Geneious version 8.0.3 (Drummond et al. 2014) and manually inspected for removal of any ambiguous characters. We then performed a maximum likelihood analysis, using RAxML-HPC Blackbox version 8.2.8 (Stamatakis 2006) with the default rapid hill-climbing algorithm and the GTRGAM-MA model of nucleotide substitution. The analysis was carried out on the online server CIPRES science Gateway version 3.3 (Miller et al. 2010) with a total of 1,000 pseudoreplicates to assess the rapid bootstrap value support. A bootstrap support value of 70 and above was considered a strong support for a clade. The resulting bipartitioned trees were visualized with the program FigTree version 1.4.2 (Rambaut 2012).

Phylogenetic analysis
The final alignment of the combined data set for the Glaucotrema analysis consisted of 802 unambiguously aligned nucleotide positions for mtSSU, 865 for nuLSU, and for 985 RPB2. The final alignment of the dataset for the Ocellularia taxa consisted of 787 unambiguously aligned nucleotide positions for mtSSU, 879 for nuLSU, and for 913 RPB2. As the topologies of the single locus phylogenies for these two datasets did not show any conflicts, they were analyzed in a concatenated matrix.
In the Glaucotrema tree (Fig. 1), the Thai material formed an unsupported sistergroup relationship with G. glaucophaenum, and G. stegoboloides. The latter two species were not separated in our analysis but were supported as different species in a broader table 1. Genbank numbers and voucher information of specimens used in this study. For author names see Index Fungorum (http://www.indexfungorum.org). Missing data are indicated by [-]. analysis in Kraichak et al. (2014) with more samples, in which the Thai material was also included and supported as distinct species. In the Ocellularia tree (Fig. 2), O. aff. ascidioidea from Thailand did not form a monophyletic group with O. ascidioidea from New Caledonia but an unsupported sister-group relationship with Thai material of O. exigua, similar to the analysis by Kraichak et al. (2014). Subsequent morphological re-analysis revealed that the Thai samples previously identified as O. aff. ascidioidea are identical to O. krathingensis described from Thailand (Homchantara and Coppins 2002). As already indicated by Kraichak et al. (2014), Ocellularia aff. fumosa from Thailand did not cluster with O. fumosa but appeared closely related to O. natashae and O. thryptica. The latter differs in having a clear hymenium and containing protocetraric acid, whereas O. natashae has longer ascospores and contains the hirtifructic acid chemosyndrome (Hale 1973;Rivas Plata and Lücking 2013). The close phylogenetic relationship of these three taxa, which are not only phenotypically disparate but also have distinct geographic distributions, suggests that the loci here used may be of limited use for species delimitation in recently evolved complexes, which has already been discussed for mtSSU by Kraichak et al. (2014). Two samples, included as spec. nov. 8 in Kraichak et al. (2014), formed an unsupported sister-group relationship with O. albocincta, a species that differs morphologically (see below) and so the Thai material is described as a new species (O. siamensis) below. Nine samples included as spec. nov. 6 in Kraichak et al. (2014) from Thailand clustered together, related to O. diacida, which is readily distinguished by the presence of the hirtifructic acid chemosyndrome. The species is described new to science below as O. phatamensis. A single specimen, included as spec. nov. 7 in Kraichak et al. (2014), is also related to O. diacida but differs -among other characters -by the absence of secondary metabolites. Diagnosis. Characterized within the genus by having submuriform ascospores. Etymology. The specific epithet refers to the country where the type specimen was collected.
Distribution and ecology. The new species was found in northeastern Thailand, growing on bark in a dry evergreen forest. It is known only from the type locality.
Secondary chemistry. No substances detected by TLC. Distribution and ecology. The new species was collected in northeastern Thailand, growing on bark in a dry evergreen forest. It is known only from the type locality.
Remarks. Similar in ascospore size, lack of secondary metabolites and only apically carbonized exciple to O. krathingensis but differing in having a whitish gray, rimose thallus with ascomata in verrucae and surrounded by a black ring, reminiscent of O. wirthii . The latter species is readily distinguished by having a broader, carbonized columella and the presence of the psoromic acid chemosyndrome. The species would key out at alternative 60 in the Ocellularia key for Thailand (Sutjaritturakan & Kalb 2015).
Diagnosis. Differing from the similar O. krathingensis in having an a grayish, thick and rimose thallus.
Etymology. The specific epithet refers to the name of the Pha Tam National Park in Ubon Ratchathani Province, Thailand.
Secondary chemistry. No substances detected by TLC. Distribution and ecology. The new species was collected in northeastern Thailand, growing on bark in a dry evergreen forest. It is known only from the type locality.
Remarks. The new species is similar to O. krathingensis in having an apically carbonized exciple and columella, transversely septate, amyloid ascospores, and lacking secondary metabolites, but differs in having a grayish and thicker thallus (Homchantara and Coppins 2002). Another similar species is O. klinhomii, but differs in lacking a dark apothecial rim and the ascomata are not immersed in verrucae. Molecular data support the distinction of these two species (Fig. 2). Another similar and related species is O. diacida, which is readily distinguished by the presence of the hirtifructic acid chemosyndrome. The species would key out at alternative 60 in the Ocellularia key for Thailand (Sutjaritturakan and Kalb 2015).

Diagnosis.
Differing from O. fumosa in having ascospores with rounded ends. Etymology. The specific epithet refers to the ascospore shape with rounded ends and to the similarity with O. fumosa.
Secondary chemistry. No compounds detectable by TLC. Distribution and ecology. The new species was collected in northeastern Thailand, growing on bark in a dry evergreen forest. It is known only from the type locality.
Remarks. Similar to O. fumosa, but differing in having rounded ends of the ascospores instead of acute ones in O. fumosa. Molecular data support the distinction of the species (Fig. 2). Characters to separate the related O. natashae and O. thryptica are discussed above. The species would key out at alternative 23 in the Ocellularia key for Thailand (Sutjaritturakan and Kalb 2015).