INTRODUCTION

Foliicolous lichens are widespread in the tropics and extraordinarily diverse (Santesson 1952; Lücking 2001, 2008). Their small size and their tendency to occur mixed with other foliicolous lichens and non-lichenized fungi in minute communities, and in some cases their growth under the host plant cuticle, has rendered a re-appraisal of the relationships of these fungi by molecular phylogenetic methods technically difficult. Yet, some chiefly foliicolous lineages such as Gomphillaceae, Pilocarpaceae, Porinaceae, and the genus Chroodiscus (Graphidaceae) have been partly studied phylogenetically (Lücking et al. 2004; Andersen and Ekman 2005; Baloch and Grube 2006, 2009; Papong et al. 2009).

The order Strigulales in class Dothideomycetes was established by Lücking et al. (in Hyde et al. 2013) for the single family Strigulaceae, in an updated phylogeny which showed that the family was separate from other clades recognized as orders. Four genera (Flavobathelium, Phyllobathelium, Phyllocratera, Strigula) were accepted in the family at that time, but with the inclusion of the poorly known, monospecific genus Oletheriostrigula (Huhndorf and Harris 1996), that number has grown to five (Jaklitsch et al. 2016; Lücking et al. 2017). Members of this lineage are mostly found on leaves, with the most speciose genus being Strigula, and more rarely on bark and rocks, mostly in tropical to subtropical habitats, with very few species extending into temperate regions (Harris 1995; Roux et al. 2004; Aptroot and Moon 2014). Strigulales have been analysed molecularly in various studies (Nelsen et al. 2009, 2011a), particularly focusing on species delimitation in foliicolous representatives of Strigula in tropical Asia (Jayalal et al. 2013; Krishnamurthy and Subramanya 2016; Jiang et al. 2016, 2017a, 2017b; Krishnamurthy and Kumar 2017).

As part of a phylogenetic revision of species of Strigula s.lat. and similar taxa, we came across a novel lineage resembling species of Porina but clustering with Strigulales in Dothideomycetes. The diagnostic feature of this new lineage is the thin-walled ascus apex, an unusual feature for Dothideomycetes. Given that Strigulales and the related families Acrospermaceae, Dyfrolomycetaceae, and Kirschsteiniotheliaceae all have bitunicate asci with a thin but distinct tholus and ocular chamber (Huhndorf and Harris 1996; Lücking 2008; Hyde et al. 2013), we recognize this new taxon as a new genus (Tenuitholiascus) in a new family (Tenuitholiascaceae). Initially we also considered introducing a new order, but since families with different ascus types have been shown to be closely related and included in single orders in other instances, such as Lecanorales in Lecanoromycetes (Miadlikowska et al. 2014), we include the new family in the order Strigulales, related to the orders Acrospermales (Minter et al. 2007), Dyfrolomycetales (Hyde et al. 2013), and Kirschsteiniotheliales (Hernández-Restrepo et al. 2017).

MATERIAL AND METHODS

Material examined

The interesting specimens regarding the new family were all collected from Hainan province in China. All are preserved in the Fungarium-Lichenum of the Institute of Microbiology, Chinese Academy of Sciences (HMAS-L).

METHODS

Phenotypic analyses

A LEICA M125 dissecting microscope (Leica Microsystems, Singapore) was used for the morphological studies, and a Zeiss Axioscope2 (Carl Zeiss, Göttingen) compound microscope for the anatomical studies and measuring the size. Photographs were obtained with an AxioCam MRc5 connected to a Zeiss Imager A2-M2 microscope (Carl Zeiss, Göttingen) for microscopic features. Thin-layer chromatography (TLC) (Orange et al. 2001) was employed for the detection of lichen substances. The features of the ascus apex were revealed using Lugol’s solution without pre-treatment by KOH (Baral 1987).

Genotypic analyses

DNA extraction and PCR amplification

All newly collected fresh specimens were subjected to DNA extraction (Table 1), for which a modified CTAB method (Rogers and Bendich 1988) was used. DNA, suspended in ddH2O, was amplified by the polymerase chain reaction (PCR). Partial nuclear ribosomal small subunit SSU sequences were amplified and sequenced using combinations of the primers SF5-CAATTGGAGGGCAAGTCTGG and SR5-CCAAGAGATCCGTTGTTGAAAG (in this study). A portion of the fungal nuclear ribosomal large subunit LSU was amplified and sequenced using primers ITS3 (White et al. 1990) and LR72-TACTACCACCAAGATCTGCAC. Partial TEF1-α sequences were generated using the primers TEF1a-983 F (Rehner and Buckley 2005) and TEF1a-1567R-HTL (Nelsen et al. 2011a). The second largest subunit of RNA polymerase II, RPB2, was amplified and sequenced using primers: fRPB2-5F-GAYGAYMGWGATCAYTTYGG and fRPB2-7cR- CCCATRGCTTGYTTRCCCAT (Liu et al. 1999).

Table 1 Sequences for molecular phylogenetic analysis in Strigulales. The new sequences generated were in bold
Full size table

Photosymbionts of selected specimens were also analyzed phenotypically and molecularly besides phenotype (Table 2). ITS nrDNA sequences of the algal partners were amplified and sequenced using the primers nr-SSU-1780-59 and nr-LSU-0012-39 (Piercey-Normore and Depriest 2001).

Table 2 Sequences of photobiont Trentepohliaceae. The new sequences generated were in bold
Full size table

PCR reactions were carried out in 25 μl reaction volumes and the components used were: 2 μl total DNA, 1 μl each primer (10 μM), 12.5 μl 2 × Taq MasterMix, 8.5 μl ddH2O. Amplification was performed using a Biometra T-Gradient thermal cycler. Cycling parameters for LSU, ITS and SSU were set to an initial denaturation at 95 °C for 5 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 54 °C for 30 s, extension at 72 °C for 1 min, and a final extension at 72 °C for 10 min. PCR amplifications of TEF1-α were initiated with a 2 min denaturation at 94 °C. The annealing temperature in the first amplification cycle was 66 °C, which was subsequently incrementally reduced by 1 °C per cycle over the next 9 cycles. An additional 30 amplification cycles were then performed, each consisting of 30 s denaturation at 94 °C, a 30 s annealing step at 56 °C, and a 1 min extension at 72 °C, concluding with a 10 min incubation at 72 °C (Rehner and Buckley 2005). The PCR conditions of RPB2 included: initial denaturation at 95 °C for 5 min; 35 cycles of 1 min at 95 °C, 2 min at 50 °C, an increase of 1 °C/5 s to 72 °C, and 2 min at 72 °C; and a 10 min incubation at 72 °C (Liu et al. 1999). PCR products were checked on 0.8% agarose electrophoresis gels stained with ethidium bromide and then sent to the sequencing facilities of Majorbiology (Changping district, Beijing, China), for sequencing.

Sequence alignment and phylogenetic analyses

For the mycobionts, sequences generated from different primers (Table 1) were analyzed with others obtained from the GenBank (Table 3). To determine the exact placement of the new lineage, a three-locus (SSU, LSU, and TEF-α) dataset was compiled in which sequences were aligned with those retrieved from GenBank covering the main groups of the class Dothideomycetes. In total, 109 ingroup taxa were used together with four outgroup taxa representing Arthoniomycetes. Further, a four-locus (SSU, LSU, TEF-α, and RPB2) dataset was also analysed with those retrieved from GenBank covering Dothideomycetes and Lecanoromycetes of the phylum Ascomycota, with three outgroup taxa representing Basidiomycota (Table 3). For the datasets, we only used specimens with the highest number of available markers. Each partition of LSU, SSU, TEF1-α, and RPB2 was aligned independently and then the alignments were concatenated for multi-locus analyses. Generated ITS sequences of algal partners were aligned with eight samples of Trentepohliaceae from GenBank (Table 2). All sequences were aligned with MAFFT v7.402 (Katoh and Toh 2010).

Table 3 Other sequences retrieved from GenBank for phylogenetic analysis
Full size table

An ML tree involving 1000 pseudoreplicates was generated by IQ-TREE v1.6.6 (Nguyen et al. 2015) using the shared set of 3 or 4 genes. For this analysis, the best-fit substitution model was selected using ModelFinder (Kalyaanamoorthy et al. 2017), which identifies optimal model of sequence evolution (SE) by combining substitution models (e.g. GTR) with flexible rate heterogeneity across sites model. By allowing the tree topology to vary during the search for an optimal model of SE, ModelFinder reduces the chance of entrapment in local optima during model selection. GTR + F + I + G4 was selected as our best model.

The Bayesian analyses were performed in MrBayes (Ronquist et al. 2012) assuming a general time reversible model including estimation of invariant sites and a discrete gamma distribution with six rate categories (GTR + I + G) for the single-genes and the combined analyses. A run with 5.0 million generations to ensure the average standard deviation of split frequencies lower than 0.01 and employing 20 simultaneous chains was executed. Posterior probabilities above 90% and bootstrap support above 50% are considered significant supports. Every method of analysis for the single-genes and the combined analysis resulted in basically the same tree.

Phylogenetic trees were drawn using FigTree v1.4.2 (Rambaut 2012). The alignments and trees were deposited in TreeBase (http://treebase.org).

RESULTS

Phenotypic analyses

In the new lineage discovered on Hainan island, the thallus was supracuticular and easily separated from the leaf surface and had a Phycopeltis-like photobiont (Fig. 1). The asci were bitunicate in structure, but the ascus apex differed from that of Strigula in lacking a thickened tholus and ocular chamber; instead, the ascus apex had an inconspicuous non-amyloid dome. The ascospores were oblong, 3-septate, with thin septa and walls, and colorless (Figs. 1, 2). This taxon was therefore considered different from any of the five genera currently recognized within Strigulales: from all five genera in the ascus type, from most species of Strigula (except the S. phyllogena group) in the supracuticular growth, from Flavobathelium, which produces similar ascospores, in the general habit, with exposed perithecia, from Phyllocratera, which looks superficially similar, in the very different ascospores, from Phyllobathelium in general habit (exposed perithecia) and ascospores, and from Oletheriostrigula in the lichenized habit and the ascospore type (Huhndorf and Harris 1996; Lücking 2008). These differences, together with the deviating ascus type, not only rendered the new lineage different from Strigulaceae but also implied an unknown position within Ascomycota, with some features also pointing to genera such as Porina in Lecanoromycetes (see below).

Fig. 1
figure 1

Tenuitholiascus porinoides (HMAS–L0139638). a Thallus. b The Phycopeltis algal partner. c Perithecia in vertical section. d Ascus (HMAS–L0139639). e Ascus (HMAS–L0139640). f Ascus apex (HMAS–L0141346). g Ascus with iodine reaction (HMAS–L0139638). h Ascus with iodine reaction (HMAS–L0141348). i Ascospores (HMAS–L0139639). j Ascospores (HMAS–L0139638). Scale bars: a = 300 μm, b = 10 μm, c = 20 μm, d, i, j = 10 μm, e–h = 5 μm

Full size image
Fig. 2
figure 2

Strigula nitidula (HMAS–L0139358): a–b Ascus. c Ascus showing iodine reaction in Lugol’s solution. Strigula cf. smaragdula GD2015025–5 (HMAS–L0138067): d Ascus showing iodine reaction in Lugol’s solution. Tenuitholiascus porinoides (e–g HMAS–L0139638; h–j HMAS–L0141348) e Ascus. f–g Ascus with iodine reaction in Lugol’s solution. h Ascus. i–j Ascus with iodine reaction in Lugol’s solution. White arrows indicate the ocular chamber, and black arrows indicate the non-amyloid dome. Scale bars: a–c = 5 μm, d = 10 μm, e–j = 5 μm

Full size image

Genotypic analyses

The dataset, including 19 LSU sequences, 16 SSU sequences, 19 TEF1-α sequences, and 19 RPB2 sequences newly generated for this study, was complemented with other sequences from different classes retrieved from GenBank (Table 3).

For the concatenated analysis of the three selected markers, SSU, LSU, and TEF1-α, the individual datasets did not show supported conflicts, and so the three loci were combined. The resulting tree showed the new lineage in a well-supported sister group relationship with Strigulaceae (Fig. 3), with the following groups forming further external lineages in a supported clade: Acrospermales, Dyfromycetales, Monoblastiales, and Kirschsteiniotheliales (Fig. 3). The relationships between these lineages were not supported, except for Acrospermales and Dyfromycetales forming a strongly supported clade. This appears to be the first study that places the lichenized Monoblastiales rather close to Strigulales, which is notable as both clades share important characters and have been considered closely related or even belonging in the same family in the past (e.g. Harris 1975; Lücking 2008).

Fig. 3
figure 3

Phylogenetic tree constructed from Bayesian analyses in Dothideomycetes based on three gene (SSU, LSU, TEF1-α) sequences with 3001 bp. Bayesian inference posterior probabilities above 90% (left) and Maximum likelihood bootstrap probabilities above 50% (right) are shown at nodes (B–PP / ML–BP). The families and orders including lichenized taxa are marked with *. The tree was rooted to Arthoniomycetes spp.

Full size image

To assess placement of the new lineage within Dothideomycetes, a dataset consisting of four loci (SSU, LSU, TEF1-α, and RPB2) was also constructed and analysed (Additional file 1). It is evident that the new lineage is a member of the class Dothideomycetes, rather than Lecanoromycetes. Also in this analysis, the specimens of the new genus formed a separate clade supported sister to the known genera of Strigulales.

In the ITS tree of the analysed photobionts, the photobiont of the new lineage clustered with algae identified as Phycopeltis, with Cephaleuros and Trentepohlia forming separate branches with high support (Fig. 4), supporting our phenotypic assessment of the photobiont as Phycopeltis.

Fig. 4
figure 4

Phylogenetic tree constructed from Bayesian analyses based on ITS of photobionts of Trentepohliaceae. Bayesian posterior probabilities (B–PP) > 90%, and Maximum likelihood bootstrap proportions >50% are shown at nodes (B–PP / ML–BP). The new sequences generated in this study are in bold, and algal partners from Tenuitholiascus porinoides are marked with *

Full size image

DISCUSSION

The phenotype and molecular data demonstrate that the material from Hainan island represents a previously unknown lineage of foliicolous lichens which merits the status of a new genus and family. The lineage is closely related to Strigulaceae but differs clearly in the ascus type. Phenotypically, the new taxon bears some resemblance with members of Strigulaceae, but also with some unrelated lichenized lineages in the Lecanoromycetes and Eurotiomycetes. Thus, the general habit with supracuticular growth, a Phycopeltis photobiont, and exposed perithecia resembles that of the Strigula phyllogena group (Lücking 2008) and of Phyllocratera (Lücking and Sérusiaux 2013), whereas the ascospores are similar to those of Flavobathelium (Lücking 2008). The comparatively thin-walled asci, together with the unbranched paraphyses and the oblong, thin-walled, 3-septate ascospores would place the new taxon close to Porina, in particularly the black-fruited species also recognized in the genera Pseudosagedia and/or Trichothelium (Harris 1995; Lücking 2008; Lücking et al. 2017; Sobreira et al. 2018). Among Porina s.lat., the most similar foliicolous species is P. chrysophora (Santesson 1952), which agrees in the black, hemispherical perithecia and the 3-septate ascospores, but differs in the dispersed thallus, the absence of a basally expanding perithecial wall, and the much smaller ascospores. Some foliicolous species of Phylloblastia (Eurotiomycetes: Verrucariales) previously classified in the genus Pocsia (Lücking 2008) are also superficially similar and may produce 3-septate ascospores; however, they are easily set apart by the lack of paraphyses, the apically thick-walled asci, and the different photobiont usually consisting of more rounded cells in irregular arrangement.

The apical ascus structure in all previously recognized members of Strigulales is the so-called Strigula-type, characterized by structurally bitunicate asci with a short tholus and small ocular chamber (Fig. 2a–d). The asci of the new lineage are similar to those of Strigulales in being bitunicate, but differ in their apical structure, in that the asci have an inconspicuous, non-amyloid dome lacking an ocular chamber (Fig. 2e–j). In some stages of development, the inner wall layer becomes gradually thinner and makes the asci appear unitunicate, similar to the genus Porina (Fig. 2h, j).

Although the overall features of the new genus show affinities to other genera currently included in Strigulaceae, the difference in ascus structure is more fundamental and merits recognition of the new taxon at a rank higher than genus. We even pondered the possibility establishing a separate order, but felt this level too high considering that likely many other lineages in this assemblage await discovery. Even foliicolous lichens remain much understudied, as illustrated by the fact that the new genus was quite abundant in the type locality and was collected multiple times during a single day. Also, there are other examples in which taxa with deviating ascus types are classified within a single order, such as Baeomycetales and Lecanorales within Lecanoromycetes (Lumbsch et al. 2007; Miadlikowska et al. 2014).

Our three-locus based analyses (Fig. 3) provided an important insight into the phylogenetic adscription of the order Strigulales amongst Dothideomycetes, the members of which are generally characterized by thin interascal filaments. Our phylogeny largely matches that of Hyde et al. (2013) in the close relationship of Strigulaceae with Acrospermaceae (Acrospermales), Dyfrolomycetaceae (Dyfrolomycetales), and Kirschsteiniotheliaceae (Kirschsteiniotheliales). All four families clustered with strong support in a single clade, which also included the new lineage close to Strigulaceae and the Monoblastiaceae (Monoblastiales). The latter comes as surprise, as this family had not been recovered as not closely related in other analyses (Nelsen et al. 2009, 2011a; Hyde et al. 2013). All families have different morphologies and life styles, Tenuitholiascaceae being closest to Strigulaceae in these aspects but differing from all other lineages in the ascus type. Ascospores in the new lineage are most similar to those of Strigulaceae (Flavobathelium); Arcospermaceae have filiform spores, Kirschsteiniotheliaceae 1-septare but brown spores, and Dyfrolomycetaceae muriform spores (somewhat similar to Phyllobathelium but in shape more similar to those of Strigula). We therefore propose to recognize the new genus under a new family within Strigulales.

TAXONOMY

Tenuitholiascaceae S.H. Jiang, Lücking & J.C. Wei, fam. nov. — Fungal Names FN570578;

Type: Tenuitholiascus S.H. Jiang, Lücking & J.C. Wei.

Diagnosis: Distinguished from Strigulacaeae in the structure of the ascus apex, which lacks the short tholus and ocular chamber characteristic of Strigula-type asci. In the nuLSU alignment (Additional file 2), the following positions are consistently diagnostic at family level: 113 (A vs. G), 143 (A vs. G), 166 (C vs. G), 168 (G vs. T), 207–208 (AT vs. GC), 313 (T vs. C), 363–364 (TC vs. GT), 377 (A vs. G), 385 (T vs. C), 397 (A vs. C), 449 (T vs. G), 484–485 (TG vs. CC), 505 (T vs. C).

Tenuitholiascus S.H. Jiang, Lücking & J.C. Wei, gen. nov. — Fungal Names FN570581;

Etymology. From the Latin tenuis- (slender), the Latin tholus (dome), and the Latin ascus (tube, bag), conveys an important feature of the apically thin-walled asci.

Type: Tenuitholiascus porinoides S.H. Jiang, Lücking & J.C. Wei.

Diagnosis: The only genus of the family, distinguished from Strigulaceae in ascus structure (see above), from the Strigula phyllogena group also in the 3-septate, oblong ascospores, from Phyllocratera in the small, 3-septate ascospores, and from Flavobathelium in the external habit with exposed perithecia.

Tenuitholiascus porinoides S.H. Jiang, Lücking & J.C. Wei, sp. nov. — Fungal Names FN570580; Fig. 1

Etymology. The specific epithet conveys the similarity with the genus Porina, although not related to the latter.

Type: China: Hainan: Changjiang county, Bawangling National Nature Reserve, 19°07′07″N, 109°09′12″E, alt. 700 m, on living leaves, 4 Sept. 2017, S.H. Jiang HN20171851 (HMAS–L0139638 – holotype).

Description: Thallus supracuticular, easily separated from the leaf surface, continuous, smooth, pale green, 3–12 mm diam, 30–52.5 μm thick. Algal partner: Phycopeltis, cells rectangular, 8–14 × 3–5 μm, composed of anastomosing filaments lying in one layer and forming regular radial plates or irregular nets. Ascomata perithecia, globose, scattered or clustered, exposed but covered by thin thallus layer up to the ostiole, central part wart-shaped, sometimes basal part broadly spreading to form horizontal plate, 0.25–0.5 mm diam and 80–150 μm high, greyish black. Involucrellum carbonized, black, 55–125 μm thick. Exciple dense, prosoplectenchymatous, 10–12.5 μm thick, colourless to brown. Interascal filaments: unbranched or simply branched, thin. Asci bitunicate in structure, apex with a non-amyloid rounded, sometimes appearing almost unitunicate in some developmental stages, due to the gradually thinner inner walls (Fig. 1f; Fig. 2h, j), clavate to cylindrical, 75–90 × 10–12.5 μm, I–, KI–, 8-spored. Ascospores fusiform, 3-septate, colourless, 25–30 × 6–8 μm. Pycnidia common, wart-shaped, immersed to erumpent, 0.05–0.1 mm diam, black. Conidia (microconidia) fusiform, hyaline, non-septate, 4–5 × 1.5–2 μm.

Alga partner. The trentepohlioid genera Cephaleuros, Phycopeltis and Trentepohlia have been reported from Strigulaceae (Lücking 2008; Nelsen et al. 2011b). In addition to morphology, four newly generated ITS sequences of the photobiont were aligned with selected Trentepohliaceae from GenBank; the selected sequences of Cephaleuros, Phycopeltis and Trentepohlia formed separate branches each (Fig. 4), and the photobiont of the new lineage clustered with Phycopeltis.

Chemistry: No substances detected by TLC.

Ecology and distribution: At present, the new species is known only from the type locality (Hainan island) in China, where it grows on leaves in wet tropical forest.

Remarks: For similarities and differences of the new species with other taxa in Strigulaceae and the unrelated genera Porina and Pocsia, see above.

Other specimens examined: CHINA: Hainan: Changjiang county, Bawangling National Nature Reserve, 19°07′07″N, 109°09′12″E, alt. 700 m, on living leaves, 4 September 2017, S.H. Jiang HN20171719 (HMAS–L0141342), HN20171740 (HMAS–L0141344), HN20171808 (HMAS–L0141348), HN20171820 (HMAS–L0141343), HN20171826 (HMAS–L0141349), HN20171844 (HMAS–L0141346), HN20171845 (HMAS–L0139639), HN20171850 (HMAS–L0141345), HN20171857 (HMAS–L0139640), HN20171875 (HMAS–L0141347).

CONCLUSIONS

Molecular data of some foliicolous lichens collected in Hainan island revealed a new lineage morphologically similar to Porina but phylogenetically related to Strigulaceae, differing from the latter in ascus type, which merits the status of a new genus (Tenuitholiascus) and family (Tenuitholiascaceae) within Strigulales. The extent of lack of exploration on tropical foliicolous lichens in Asia is indicated by the new genus having been collected 11 times on a single day. Indeed, the discovery of this previously unsuspected lineage is an example of how little we know. It should still be stressed that the effort to complete the inventory of tropical lichens should be made in the future.