Taxonomic recognition of some species-level lineages circumscribed in nominal Rhizoplaca subdiscrepans s. lat. (Lecanoraceae, Ascomycota)

Background Rhizoplaca subdiscrepans (Nyl.) R. Sant., a saxicolous, placodioid lichen, is considered to have a worldwide distribution in warm-temperate to boreal-arctic areas in Asia, Europe and North America. However, recent studies have revealed that this species includes five unrecognized species-level lineages—‘subd A, B, C, D and E’. During research focused on the diversity of saxicolous lichens in mountainous areas of southern Poland, some interesting representatives of the genus Rhizoplaca were found. The main aim of our study was to determine the taxonomic status of the collected specimens by means of molecular tools and a comparative analysis of similar herbarium materials. Methods Detailed morphological, anatomical and chemical examinations of reference material from Asia, Europe and North and South America focused primarily on a selected group of lecanoroid taxa with a placodioid thallus. In addition, 21 new generated sequences representing Lecanora pseudomellea, Protoparmeliopsis muralis, Rhizoplaca opiniconensis, R. subdiscrepans s. lat. and R. phaedrophthalma were selected for molecular study using the internal transcribed spacer region (ITS rDNA), together with 95 available GenBank sequences mainly from the genus Rhizoplaca. Results Polish specimens that clustered with members of a potential species-level lineage ‘subd E’ of Rhizoplaca subdiscrepans complex were recovered. Comprehensive analyses of the lichen group led us to the conclusion that lineage ‘subd E’ represents R. subdiscrepans s. str. and that the taxon appears to have a limited geographical distribution and specific habitat preferences. Furthermore, some of the recently defined species candidates within R. subdiscrepans s. lat.—‘subd D’ and ‘subd A’—should be assigned to two previously known species of Rhizoplaca, namely R. opiniconensis (Brodo) Leavitt, Zhao Xin & Lumbsch and R. phaedrophthalma (Poelt) Leavitt, Zhao Xin & Lumbsch, respectively. These two species are characterized by phenotypic features observed as well in analyzed specimens representing lineages ’subd D’ and ’subd A’. Moreover, the representatives of these lineages demonstrate some differences in occupied habitat and geographical range that also correspond with the indicated species. Additionally, it was found that Lecanora pseudomellea B.D. Ryan is a strongly supported monophyletic lineage within Rhizoplaca, and therefore an appropriate new combination for the species is proposed.


INTRODUCTION
The genus Rhizoplaca Zopf belongs to a large family of lichenized fungi, Lecanoraceae. It was first segregated out of the genus Squamaria DC. by Zopf (1905), and recognized later by Choisy (1929), who classified it under a new name Omphalodina M. Choisy. At the same time, Omphalodina chrysoleuca [Rhizoplaca chrysoleuca (Sm.) Zopf] was selected as the type species of the new genus. Later Omphalodina was included in the genus Lecanora as a section within the subgenus Placodium (Poelt, 1958); however, Leuckert, Poelt & Hahnel (1977) proposed this taxon at a generic level once more under its older name Rhizoplaca, on the basis of the following distinguishing features: umbilicate thalli, well-developed upper cortex, thick lower cortex, loose medulla and cupulate hypothecium.
Currently, the genus Rhizoplaca comprises c. 24 species of lichenized fungi representing umbilicate to placodioid growth forms (Zhao et al., 2016). The thalli predominantly produce usnic acid in addition to various other substances. Rhizoplaca species occupy siliceous or weakly to moderately calcareous rock or, rarely, soil, and grow mainly in open, windy, cool and dry areas (Ryan & Nash, 1997). Species occur in the northern hemisphere (with their centre of diversity in Central Asia and western North America), as well as in South America and Antarctica (Leavitt et al., 2016;Zhao et al., 2016). Some of the known species show broad ecological preferences and geographical ranges, e.g., R. chrysoleuca, R. melanophthalma and R. subdiscrepans, whereas others have more restricted habitat requirements and limited distributions, e.g., R. macleanii, R. maheui and R. marginalis (Leavitt et al., 2016).
Although the species was considered to be well-circumscribed, it recently proved to be polyphyletic and several unrecognized species-level lineages have been recovered in nominal R. subdiscrepans s. lat. by Leavitt et al. (2016). As a result of multigene analyses, these authors proposed five candidate species defined as R. subdiscrepans 'subd A, B, C, D, E', considered to be cryptic-species.
Our objective was to explain the taxonomic status of a putative representative of R. subdiscrepans (Nyl.) R. Sant. found during extensive field studies in the foothills foreland of the Sudety Mountains (Poland). Reference herbarium material from Asia, Europe and North and South America was examined, and sequences of gathered specimens were placed in a phylogenetic framework, including available GenBank sequences of placodioid lichens of selected species of the genera Lecanora, Protoparmeliopsis and Rhizoplaca. Based on morphological, chemical and molecular evidence, it was found that some of the recently defined species candidates within R. subdiscrepans s. lat. should be assigned to previously known species of the genus Rhizoplaca.

Taxon sampling
Material used in this study originated from the following herbaria: ASU, CANL (CMN), KRAM, M, MIN, PRA, WRSL and the private herbarium of K. Szczepańska (hb. Szczepańska). Our sampling focused on genera of lecanoroid lichens with placodioid thalli as follows: Lecanora (L. pseudomellea), Protoparmeliopsis (P. garovaglii, P. muralis) and Rhizoplaca (R. chrysoleuca, R. opiniconensis, R. phaedrophthalma, R. subdiscrepans); holotypes of Lecanora pseudomellea, Rhizoplaca opiniconensis and R. phaedrophthalma were also analysed. Unfortunately, we were unable to study the original collection of R. subdiscrepans housed at the PC herbarium since type material at PC is unavailable for loan and we had no opportunity to visit the herbarium. All of the specimens listed in the text were included in the morphological, anatomical and chemical studies, but due to the age and poor condition of some samples, it was not possible to obtain sequences and include these in our phylogenetic analyses. New sequences generated for this study have been deposited in GenBank.
For light microscopy, vertical, free-hand sections of apothecia were cut by a razor blade and mounted in water. Hymenium measurements were made in water and ascospore measurements in 10% KOH (K); the structure and coherency of paraphyses and the solubility of granules in the epihymenium were also tested with K. At least 10 measurements of morphological variables and measurements of 20 spores were made for each sample and their minimum and maximum values calculated.
Chemical examinations included colour reactions and thin-layer chromatography (TLC). Spot test reactions of thalli, apothecial margins and discs were made with KOH, sodium hypochlorite [commercial laundry bleach] (C) and paraphenylenediamine [solution in 95% ethyl alcohol] (PD). TLC analyses were performed in solvent systems A, B and C using the standardized method of Culberson (1972) and following Orange, James & White (2001).
Descriptions of the species are based on our own observations, measurements and TLC analyses of specimens cited in this paper. The terminology used in the descriptions of the species follows Ryan et al. (2004).

Molecular methods: DNA extraction, PCR amplification and DNA sequencing
To infer relationships between species of lichenized fungi studied, the ITS rDNA region, that contains ITS1, 5.8S, ITS2 sequences, was used. Total DNA was extracted from specimens using a CTAB method (Doyle & Doyle, 1987). Dried samples were mechanically disintegrated using Mixer Mill MM400 (Retsch; Haan, Germany). The quality of the DNA was checked by electrophoresis on agarose gel (1%) with Simply Safe staining chemical (Eurx, Gdańsk, Poland). The complete ITS rDNA region was amplified using primers ITS1F (Gardes & Bruns, 1993) and ITS4 (White et al., 1990). The PCR reaction mix included 1U Taq recombinant polymerase (Thermo-Fisher Scientific, Waltham, USA), 10x Taq Buffer, 1 mM MgCl 2 , 0.5 M of each primer, 0.4 mM dNTP and 1 µl DNA template. Amplification cycles were performed with a Veriti Thermal Cycler (Life Technologies, Carlsbad, USA) and involved 8 min at 95 • C, followed by 32 cycles for 45 s at 95 • C and 45 s at 52 • C (annealing), and 1 min at 72 • C, with the a final extension step of 10 min at 72 • C. Amplified PCR products were purified using GeneJet PCR Purification Kit (Thermo-Fisher Scientific, Waltham, USA). This was accomplished at the Laboratory of Molecular Biology (Environmental and Life Sciences) at Wroclaw University. Sequencing of PCR products, post-reaction purification, forward and reverse directions reading were done by sequencing service Genomed (Genomed, Warsaw, Poland), using an ABI 377XL Automated DNA Sequencer (Applied Biosystems, Carlsbad, USA).

Alignment assembly and molecular phylogenetic analyses
The obtained ITS rDNA sequences were assembled and manually edited using Geneious Pro, version 8.0. (Biomatters Ltd). Sequences of Rhizoplaca subdiscrepans (putative name for our collection) together with related species, namely Lecanora pseudomellea, Protoparmeliopsis garovaglii, P. muralis, Rhizoplaca chrysoleuca, R. melanophthalma, R. novomexicana, R. occulta, R. opiniconensis (as Lecanora in GeneBank), R. parilis and R. phaedrophthalma, were included in the analysis. Our final ITS data-set included 21 sequences newly generated for this study and 95 sequences downloaded from GenBank ( Table 1). The final alignment was performed on Geneious Pro using the MAFFT algorithm (Katoh et al., 2005), then re-checked and improved. Ambiguously aligned regions were removed in GBlocks (Castresana, 2000). The nucleotide substitution models were separately searched for in each partition of the ITS region (ITS1, 5.8S, ITS2) to find the best-fitting model using the corrected Akaike information criterion (AICc) as an optimality model criterion in a greedy algorithm search as implemented in PartitionFinder version 1.0.1 (Lanfear et al., 2012). The GTR+G model for ITS1 and ITS2 and K80 for the 5.8S partitions were selected.
The phylogenetic reconstruction was generated using the CIPRES Scientific Gateway (http://www.phylo.org/portal2/) (Miller, Pfeiffer & Schwartz, 2010). Maximum likelihood (ML) bootstrap tree with simultaneous heuristic search was undertaken as implemented in RAxML-HPC2 on XSEDE (Stamatakis, 2006) under the GTRGAMMA substitution model and 1,000 bootstrap resamples. Bayesian inference was carried out using Markov Chain Monte Carlo (MCMC) implemented in MrBayes 3.2.6 on XSEDE (Ronquist et al., 2012). MrBayes was set to two independent parallel runs each initiated with four incrementally heated chains; the run length was settled to 20M generations, and to infer convergence the average standard deviation of split frequencies was printed every 1000th generation discarding the first 50% of the trees sampled as a burn-in fraction. The analyses were stopped after 1M generations when the standard deviation had dropped below 0.01. The resulting phylogenetic trees were visualized in Figtree software (Rambaut, 2014). The alignment of ITS sequences described here is deposited in TreeBASE with number TB2: S26365.

RESULTS
Results of the morphological and chemical studies are presented in the taxonomic section under particular species descriptions. All taxa representing R. subdiscrepans s. lat. appear to be semi-cryptic, with slight morphological variety and some overlapping features. Nevertheless, some noticeable differences are visible in their chemistry, appearance of apothecia and marginal lobes, as well as in the shape and size of ascospores and conidia. The major distinguishing characters of Rhizoplaca subdiscrepans, R. opiniconensis, R. phaedrophthalma, and R. pseudomellea are summarized in Table 2.
The final alignment matrix contained 538 bp and Protoparmeliopsis was selected as the outgroup (Leavitt et al., 2016;Zhao et al., 2016). The topology of the tree was similar to that presented by Leavitt et al. (2016). Our newly generated sequences from Poland, Canada and the USA were resolved within R. subdiscrepans s. lat. as defined by Leavitt et al. (2016) (Fig. 1). The first group of samples (Rhizoplaca subdiscrepans s. str. 21 and 22) were recovered within the clade 'subd E' with high bootstrap support (BS) = 96 Table 1 The species and specimens treated in current study with locality and GenBank accession numbers. Newly generated sequences are in boldface.

Species
Locality GenBank (ITS) Accesion number and posterior probability (PP) = 1 together with specimens from Ukraine and Russia (Rhizoplaca subdiscrepans s. str. 18 and 19 respectively). Sequences of herbarium specimens marked as R. phaedrophthalma were nested within the clade 'subd A' (BS = 80 and PP = 1), together with the sequence of R. phaedrophthalma published by Arup & Grube (2000) and Zhao et al. (2016), while R. opiniconensis (BS = 100 and PP = 1) nested within the clade 'subd D' of Leavitt et al. (2016). Moreover, the sequence of the type collection of R. opiniconensis (Rhizoplaca opiniconensis 10; see Fig. 2A), another sample identified by Irwin Brodo, who described R. opiniconensis as a new taxon (Brodo, 1986), as well as the sequence of R. opiniconensis analysed by Arup & Grube (2000) and Zhao et al. (2016), were placed within the lineage of candidate species 'subd D'. Lecanora pseudomellea, included for the first time in a phylogenetic analysis of Rhizoplaca, formed a strongly supported monophyletic lineage (BS = 100 and PP = 1) within the latter genus.

DISCUSSION
Rhizoplaca subdiscrepans s. lat. was found to be highly polyphyletic by Leavitt et al. (2016), and as a result of multigene analyses, these authors delimited five cryptic species-level lineages within the species complex, defining them as R. subdiscrepans 'subd A, B, C, D and E'. In parallel, another paper was published concerning the generic classification of lecanoroid lichens and including some representatives of the genus Rhizoplaca (Zhao et al., 2016). In the latter authors' phylogenetic analyses, they also took into account several placodioid species previously classified in Lecanora, e.g., L. opiniconensis and L. phaedrophthalma. Among others, they included in their analysis single sequences of the two latter species, published by Arup & Grube (2000) but not treated by Leavitt et al. (2016). As a result, Zhao et al. (2016) found a core group of Rhizoplaca formed a monophyletic group together with the mentioned species and transferred these species along with some others to Rhizoplaca.
Both R. opiniconensis and R. phaedrophthalma were described as new taxa based on morphological and chemical features in the second half of the 20th century. However, it is presently known that morphological characters may not be sufficient for detecting taxa (Bickford et al., 2007;Crespo & Pérez-Ortega, 2009). On the other hand, careful morphological analysis of distinct phylogenetic lineages may lead to the recognition of some previously overlooked characters (Kroken & Taylor, 2001;Del Prado et al., 2007;McCune & Altermann, 2009;Frolov et al., 2016).
Species candidates within R. subdiscrepans s. lat. were considered to be cryptic by Leavitt et al. (2016), but in our morphological analysis, an attempt was made to identify potential diagnostic features for samples representing different clades. In contrast to recognized candidate species within the R. melanophthalma complex, which were highly variable (Leavitt et al., 2011), the morphology of R. subdiscrepans s. lat. representatives seem to be rather similar, especially in the case of clades 'subd E' (conspecific with R. subdiscrepans s. str.) and 'subd D' (conspecific with R. opiniconensis). The most common and characteristic features of all examined samples were a placodioid, polyphyllous, yellow-green thallus with squamulose-bullate centre, visible marginal lobes, and sessile apothecia with pale yellowish to brown, epruinose discs. However, despite their apparently similar morphology, the samples are heterogeneous and vary in the appearance of their apothecia and marginal lobes, as well as in the shape and size of ascospores and conidia. These characters have been shown to be diagnostic in the case of the Parmeliaceae family (Argüello et al., 2007;Divakar et al., 2010).
In our study, R. opiniconensis seems to be characteristic in its arched conidia and thallus colour that becomes distinctly more orange in herbarium material. Whereas the main characteristic feature of R. phaedrophthalma (conspecific with 'subd A') is the morphology of the apothecia, which are strongly convex with reddish-brown discs and an excluded thalline margin. Furthermore, some differences in the latter species are the size of the conidia, which are slightly larger than in the other discussed taxa, and ascospores that are distinctly smaller and subglobose rather than ellipsoid. Additionally, in contrast to other analysed representatives of R. subdiscrepans s. lat., thalli of mature specimens of R. phaedrophthalma are usually poorly placodioid, with weakly developed marginal lobes.
In the above morphological studies, we noted that none of the representatives of R. subdiscrepans s. lat. had orange and pruinose apothecial discs, which are mentioned in the literature as characteristic for R. subdiscrepans (Ryan, 2001). These characters, as well as the greyish-green tint of the upper surface of the thallus, are more appropriately applied to R. chrysoleuca s. lat. than to the R. subdiscrepans complex, an assumption that is consistent with the opinion of Cansaran et al. (2006). However, according to Zheng, Sheng & An (2007), apothecial discs and their pruinosity do not indicate proper phylogenetic relationships among Rhizoplaca species, so this issue requires further research.
With this in mind, we carefully analysed the content of secondary metabolites in the thalli of R. subdiscrepans s. lat. representatives sampled herein. The survey indicated such lichen substances as isousnic, usnic and placodiolic acids, as well as fatty acids and terpenoids, but some differences in their presence in particular recognized species could be observed. Among the secondary metabolites detected by TLC, neither isousnic acid nor fatty acids have been found in the thalli of R. phaedrophthalma, whereas the presence of fatty acids was observed only in specimens of R. subdiscrepans s. str., including those samples from Poland. None of the examined samples of R. subdiscrepans s. lat. contained psoromic, lecanoric or norstictic acids, as mentioned in the literature (Ryan, 2001).
Rhizoplaca subdiscrepans has until recently been considered to be distributed worldwide. However, phylogenetic analyses showed some slight differences in the geographical range of cryptic lineages within this species complex (Leavitt et al., 2016). All of the clades identified by Leavitt et al. (2016) included representatives on the Asian continent. Nonetheless, individuals belonging to clades 'subd A' and 'subd D' also occurred in North America, whereas clade 'subd E' had a European distribution. The results of our study correspond with theirs.
Rhizoplaca opiniconensis (= 'subd D') has been genetically confirmed as occurring in North America and is not found in Europe. The species epithet, however, is reported here for the first time from outside North America, i.e., from East-Central Asia (Altay). It should be noted that based on detailed habitat analysis of available samples it seems this taxon is hygrophytic, preferring mountain habitats in higher elevations, localized in shaded and moist places close to water resources. Rhizoplaca phaedrophthalma, same as R. opiniconensis, has not been noted from Europe, its centre of distribution being located in North America and Western and Central Asia, with the locus classicus in Nepal, which is in accord with the distribution pattern of clade 'subd A' (Leavitt et al., 2016). The species prefers different habitat conditions than R. opiniconensis, since it occurs mostly on siliceous rock in semi-arid areas but also at higher elevations.
Finally, R. subdiscrepans s. str. (= 'subd E') besides Eastern Asia has been recorded in Europe (Poland and Ukraine) and has not been confirmed in North America. This is important in the light of the fact that R. subdiscrepans s. str. has been described from Switzerland in Europe. The distribution pattern of the species supports our concept that it is indeed conspecific with clade 'subd E' (Leavitt et al., 2016). R. subdiscrepans occupies specific habitats, occurring namely in warm, dry and sunny places in lower elevations, usually on volcanic rocks with a southern exposure. It is the only taxon of the discussed group of lichens with such ecological preferences. It is worth noting here that analysed samples of R. subdiscrepans s. str. are very similar in morphology to representatives of R. opiniconensis. Nevertheless, the species seems to be distinguishable based on its different geographical range and habitat preferences. These characters are often mentioned in literature as a supporting species recognition, even when there is a lack of evident phenotypic differences or they are insufficient to circumscribe the species (Crespo et al., 2002;Argüello et al., 2007;Divakar et al., 2010;Onut-Brännström, Johannesson & Tibell, 2018).
The above data can also be compared to the R. melanophthalma species complex presented in Leavitt et al. (2013b). Two of the identified phylogenetic lineages therein had intercontinental distributions, while four were found exclusively in North America. In this case, phylogenetic analyses, as well as geographical distribution, provided the basis for delimitation and description of new taxa (Leavitt et al., 2013a). In addition, R. melanophthalma s. str. was circumscribed and shown to be represented by a clade with the widest geographical range that also includes Europe.
During our study, a few sequences were generated for specimens representing Lecanora pseudomellea B.D. Ryan, a placodioid taxon occurring only in North America. The species forms a strongly supported monophyletic lineage within the Rhizoplaca and was therefore transferred to this genus. Characteristic features of this taxon are an orange-brown to reddish-brown, glossy upper surface, as well as long, sinusoid, darker at the tips, marginal lobes. Compared to other taxa discussed above, R. pseudomellea has a darker colour, longer and much less convex marginal lobes, as well as a distinctly areolate, not squamulose, thallus centre. It has a very variable secondary chemistry and the occurrence of compounds varies widely between samples. The most common substance occurring in the thallus is isousnic acid, but the species may also contain psoromic acid, which is not present in the other taxa of the discussed group.

CONCLUSIONS
Despite the development of modern phylogenetic methods, species delimitation within lichenized fungi is still problematic. Implementation of integrative taxonomy and incorporation of various data, such as genetic, morphological, chemical, geographical and ecological data, usually deliver some resolution. An increasing number of analysed samples can also become more informative with these tools; these often lead to the conclusion that molecularly defined units are semi-cryptic rather than cryptic, as in the case of our study.
In our study based on molecular and phenotypic data, as well as in reference to previously described species, names are proposed for three lineages, the 'subd A, D and E' of R. subdiscrepans s. lat. delimited by Leavitt et al. (2016) and recognized respectively as R. phaedrophthalma, R. opiniconensis (supported by the placement of the type sequence in our phylogeny) and R. subdiscrepans s. str. Furthermore, we suggest transferring Lecanora pseudomellea to the genus Rhizoplaca with a proposal for a new combination-Rhizoplaca pseudomellea.
The geographical conclusion of our survey is that R. subdiscrepans s. str. appears to be mostly a European taxon with a range extended to Western Asia, whereas R. opiniconensis has a broader distribution than previously recorded, as it occurs not only in North America but also in Asia.