Introduction

Freshwater lignicolous hyphomycetes is a dominant and diverse group, which lives on submerged woody debris and decaying tree leaves (Wong et al. 1998; Shearer et al. 2007; Hyde et al. 2016a). Molecular techniques revealed that most species belong to Sordariomycetes and Dothideomycetes (Cai et al. 2008, 2010; Su et al. 2015; Liu et al. 2016; Luo et al. 2016, 2018; Wei et al. 2018) and rarely to Eurotiomycetes (Liu et al. 2015; Dong et al. 2018). The first molecular report of freshwater fungi in Eurotiomycetes was provided by Liu et al. (2015), in which they obtained sequence data for five species of Minimelanolocus R.F. Castañeda & Heredia in Herpotrichiellaceae and revisited the genus at the molecular level. Their phylogenetic analyses showed that Minimelanolocus species formed a monophyletic clade based on LSU and SSU genes, while Minimelanolocus obscurus (Matsush.) R.F. Castañeda & Heredia was distinct from the Minimelanolocus clade and was related with Exophiala J.W. Carmich species based only on ITS gene (Liu et al. 2015). This conclusion was proven by subsequent studies (Hyde et al. 2016b; Tian et al. 2016). Those studies, however, did not include the genus Thysanorea Arzanlou, W. Gams & Crous, whose morphological characters are very similar to those of Minimelanolocus.

The genus Minimelanolocus was introduced by Ruiz et al. (2001) based on ten new combinations segregated from Pseudospiropes M.B. Ellis, which differs from Minimelanolocus by the presence of distoseptate conidia. Minimelanolocus is characterized by cylindrical and unbranched conidiophores, holoblastic and sympodially proliferating conidiogenous cells, and acrogenous, oblong or clavate to fusiform, hyaline to pale brown or brown, septate conidia (Ruiz et al. 2001; Ma et al. 2011; Liu et al. 2015). In total, 33 species are listed under Minimelanolocus in Index Fungorum (2019), in which seven species were reported from submerged wood (Liu et al. 2015; Hyde et al. 2016b; Dong et al. 2018).

The genus Thysanorea, on the other hand, was established to accommodate Periconiella papuana Aptroot (Arzanlou et al. 2007), which was first isolated in Papua New Guinea (Aptroot and Iperen 1998) and later recorded from India (Pratibha and Prabhugaonkar 2015) and Taiwan (Kirschner 2016). According to published illustrations, Kirschner (2016) suggested that Ramichloridium lignicola C.K.M. Tsui, Goh, K.D. Hyde & Hodgkiss and Alysidiopsis lignicola Mercado, Figueras & J. Mena considered to be possible synonyms of the type of species, Th. papuana (Aptroot) Arzanlou, W. Gams & Crous. Thysanorea papuana was first described as having dimorphic conidiophores with up to five levels of branchlets and typically pyriform, septate conidia with a truncate base and darkened hilum (Arzanlou et al. 2007). Based on re-examination of the material and fresh collections from Taiwan, Kirschner (2016) improved the original description and reported that the conidiophore heads containing branches were easily broken off and could be replaced by regeneration. This character was a good morphological marker not only at the species level but also at the genus level of Thysanorea (Kirschner 2016). It was also reported in A. lignicola and R. lignicola but not in the species Th. aquatica W. Dong, H. Zhang & K.D. Hyde (Dong et al. 2018).

During studies of lignicolous freshwater fungi along a north/south latitudinal gradient in the Asian/Australian region (Hyde et al. 2016a), we collected a novel species of Minimelanolocus and a new specimen of Thysanorea papuana. Both taxa are described here and compared with related species. The holotype of Th. aquatica was re-examined based on morphological and molecular data.

Material and methods

Isolation and morphology

Pieces of decaying wood were collected from several small streams in Chiang Rai Province, Thailand, following the procedures described in Kurniawati et al. (2010). The samples were placed in Ziploc plastic bags with moist sterile tissue, taken to the laboratory, and incubated at room temperature (25 °C). After 1–2 weeks, specimens were examined using a stereomicroscope (SMZ-171) to locate fruiting bodies (Taylor and Hyde 2003). Photomicrographs were taken using a Cannon EOS 600D camera attached to a Nikon ECLIPSE Ni compound microscope. The fungal structures were measured using Tarosoft (R) Image Frame Work program, and images were processed using Adobe Photoshop CS6 Extended version 13.0 software (Adobe Systems, USA). Isolations were made from single conidia as described by Chomnunti et al. (2014). Water agar (WA) was used for conidial germination and incubated overnight in an incubator at room temperature. Single germinating conidia were selected and transferred to a new potato dextrose agar (PDA) plate to obtain pure cultures. The colonies were checked every 3 days. Herbarium specimens are deposited in the herbarium of Mae Fah Luang University (MFLU), Chiang Rai, Thailand. Living cultures are deposited in the Mae Fah Luang University Culture Collection (MFLUCC). Facesoffungi and Index Fungorum numbers are registered as described by Jayasiri et al. (2015).

DNA extraction, PCR amplification, and sequencing

Cultures were grown on PDA at 25 °C until enough mycelia were obtained, and a Biospin Fungus Genomic DNA Extraction Kit (Bioer Technology Co., Ltd., Hangzhou, P.R. China) was used to extract total genomic DNA from the fresh mycelia following the manufacturer’s instructions. DNA amplification was performed using the polymerase chain reaction (PCR). Fragments of three loci, LSU, SSU, and ITS, were used for phylogenetic analyses, and the following primer pairs, LROR/LR5, NS4/NS5, and ITS5/ITS4, were used for amplification and sequencing (Vilgalys and Hester 1990; White et al. 1990). Amplifications were performed in a 25 μL reaction containing 9.5 μL ddH2O, 12.5 μL 2× PCR Master Mix, 1 μL of DNA template, and 1 μL of each primer (10 μM). The PCR thermal cycles for amplification of the gene regions followed Su et al. (2015). The PCR products were examined on 1.0% agarose electrophoresis gels stained with ethidium bromide. Sequencing reactions were conducted by Shanghai Sangon Biological Engineering Technology and Services Co., Shanghai, P.R. China.

Phylogenetic analyses

Phylogenetic analyses were conducted to illustrate the phylogeny of Minimelanolocus and Thysanorea following Tian et al. (2016) and Dong et al. (2018). The newly generated sequences together with other sequences obtained from GenBank (Table 1) were initially aligned using MAFFTv.7 (Katoh and Standley 2013) on the online server (http://maffTh.cbrc.jp/alignment/server/) and optimized manually when needed. A maximum likelihood analysis was performed using RAxMLGUI v. 1.3 (Silvestro and Michalak 2011). For both analyses, the optimal ML tree search was conducted with 1000 separate runs using the default algorithm of the program from a random starting tree for each run. The final tree was selected among suboptimal trees from each run by comparing the likelihood scores using the GTR + GAMMA substitution model. Maximum likelihood bootstrap values equal to or greater than 80% are presented as the first set of numbers above the nodes in the resulting ML tree. Bayesian analysis was conducted using MrBayes v. 3.1.2 (Ronquist and Huelsenbeck 2003) to evaluate posterior probabilities (BYPP) (Rannala and Yang 1996) with Markov Chain Monte Carlo sampling (BMCMC). The best-fit model was GTR + I + G for LSU, SYM + I + G for ITS, and HKY + I + G for SSU. Bayesian posterior probabilities equal to or greater than 0.90 are shown above each node of the resulting consensus tree. These trees were viewed in Treeview (Page 1996) and edited further using Adobe Illustrator CS v.5. Newly generated sequences are deposited in GenBank. The alignment was deposited in TreeBASE (https://www.treebase.org/) under the accession number 23791. In this study, we follow the recommendations of Jeewon and Hyde (2016) to determine new taxa and introduce the new species Minimelanolocus nonramosus.

Table 1 Isolates used in this study and their GenBank accession numbers
Full size table

Results

Phylogenetic study

The placement of the two isolates within the family Herpotrichiellaceae was determined based on the phylogenetic analyses of LSU, SSU, and ITS sequences. The maximum likelihood tree included 31 isolates belonging to Herpotrichiellaceae species, and two outgroups Cyphellophora oxyspora (W. Gams) Réblová & Unter. (CBS 698.73) and C. sessilis (de Hoog) Réblová & Unter. (CBS 243.85) with an alignment length of 2349 characters. The best scoring RAxML tree with a final likelihood value of − 9697.482313 was selected to represent the relationships among the taxa and is presented in Fig. 1. For Bayesian analysis, six simultaneous Markov chains were run for 1,171,000 generations, and trees were sampled every 1000th generation and 1171 trees were obtained. The first 293 trees, representing the burn-in phase of the analyses, were discarded, while the remaining 878 trees were used to calculate posterior probabilities in the majority rule consensus tree (the critical value for the topological convergence diagnostic was 0.01).

Fig. 1
figure 1

Maximum likelihood tree generated by combining LSU, SSU, and ITS. Bootstrap support values for maximum likelihood (ML, first value) equal to or greater than 90% are shown above the nodes. The values of the Bayesian posterior probabilities from MCMC analyses (BYPP, second value) equal or higher than 0.90 are shown above the nodes. Hyphen (“-”) indicates a value lower than 90% for ML and a posterior probability lower than 0.90 for Bayesian. Ex-type isolates are in bold, and newly generated sequences are indicated in red. The tree was rooted to Cyphellophora sessilis CBS 243.85 and C. axyspora CBS 698.73 (Cyphellophoraceae)

Full size image

The relationships among various isolates of Th. papuana were determined based on the phylogenetic analyses of ITS sequences and is presented in Fig. 2. The maximum likelihood tree included 6 isolates of Th. papuana, and 8 Minimelanolocus species as outgroups based on the above tree. The best scoring RAxML tree with a final likelihood value of − 1759.781121 was selected (Fig. 2). For the Bayesian analysis, six simultaneous Markov chains were run for 1,540,000 generations, and trees were sampled every 1000th generation and 1540 trees were obtained. The first 385 trees, representing the burn-in phase of the analyses, were discarded, while the remaining 1155 trees were used to calculate posterior probabilities in the majority rule consensus tree.

Fig. 2
figure 2

Maximum likelihood tree generated by ITS gene. Bootstrap support values for maximum likelihood (ML, first value) equal to or greater than 90% are shown above the nodes. The values of the Bayesian posterior probabilities from MCMC analyses (BYPP, second value) equal or higher than 0.90 are shown above nodes. Hyphen (“-”) indicates a value lower than 90% for ML and a posterior probability lower than 0.90 for Bayesian. Ex-type isolates are in bold, and newly generated sequences are indicated in red. The tree was rooted to Minimelanolocus species

Full size image

Taxonomy

Minimelanolocus nonramosus X.D. Yu, G.N. Wang & H. Zhang, sp. nov.

Index Fungorum number: IF 555776; Facesoffungi number: FoF 04784; Fig. 3.

Fig. 3
figure 3

Minimelanolocus nonramosus (holotype, MFLU 17–1736). a, b Colonies on submerged wood. c, d Conidiophores. e–g Conidiogenous cells with pigmented conidiogenous loci. The arrow shows them becoming suddenly slimmer and paler, which probably implies a new percurrent proliferation after breaking off. h–m Conidia. n Germinating conidium. o Colony on PDA (front view). p Colony on PDA (reverse). Scale bars: a, b = 100 μm; c, d = 50 μm, e–g = 10 μm, h–n = 5 μm

Full size image

Holotype: MFLU 17–1736.

Etymology: referring to the unbranched conidiophores

Saprobic on decaying, submerged wood in freshwater. Sexual morph: Undetermined. Asexual morph: Colonies superficial, effuse, scattered, pale brown to dark brown. Mycelium mostly immersed, composed of septate, pale brown, smooth hyphae. Conidiophores macronematous, mononematous, erect, cylindrical, slender, mostly flexuous, brown to dark brown, gradually paler and becoming pale brown to subhyaline at the upper part, unbranched, septate, slightly constricted at the septa, thick-walled, smooth, upside of the stipe sometimes becoming suddenly slimmer and paler, which probably implies a new percurrent proliferation after the break offs, (160–) 270–360 μm long, 3–7 μm wide (\( \overline{x} \) = 290.6 × 5 μm, n = 15). Conidiogenous cells holoblastic, terminal, discrete, sympodially proliferating, pale brown to subhyaline, thin-walled, smooth, subcylindrical, slightly thickened, covered by several pigmented, minutely denticulate conidiogenous loci, 5–23 μm long, 1.5–4.0 μm wide. Conidia acrogenous, cylindrical, rounded at the apex, narrow and truncate at the base with a darkened hilum (ca 0.5–1.0 μm diameter), 1–2(− 3)-septate, slightly or not constricted at the septa with several minute guttula, solitary, hyaline, thin-walled, smooth, 11–17 μm long, 1.5–4.0 μm wide (\( \overline{x} \) = 13.6 × 2.7 μm, n = 20), with a ratio of length/width of approximately 4:1.

Material examined: Thailand, Phang Nga, Thap Put, saprobic on decaying wood submerged in a stream, 1 September 2017, X.D. Yu, V9 (MFLU 17–1736, holotype), ex-type living culture MFLUCC 17–2378.

Culture characteristics: Conidia germinating on PDA; colonies slow growing, reaching 30 mm diameter in 20 days at 25 °C, brown to gray, reverse dark brown, circular, velvety, dry, fairly dense, margin entire.

Notes: The phylogenetic tree of LSU, SSU, and ITS sequence data shows the new collection clusters within the Minimelanolocus sensu lato clade. It fits well with the characters of Minimelanolocus in having mononematous and unbranched conidiophores, terminal conidiogenous cells with sympodial, denticulate conidiogenous loci, and septate conidia (Ruiz et al. 2001; Liu et al. 2015). Therefore, it is described as a new species of Minimelanolocus. Minimelanolocus nonramosus is distinguished from the phylogenetically close species Minimelanolocus obscurus (Matsush.) R.F. Castañeda & Heredia by cylindrical, hyaline, and smaller conidia (\( \overline{x} \) = 13.6 × 2.7 μm), while the latter has clavate, hyaline to pale brown, and larger conidia (\( \overline{x} \) = 21.5 × 4.6 μm). Minimelanolocus aquatilis L.B. Conc., M.F.O. Marques, Gusmão & R.F. Castañeda, which has no DNA sequence available yet, differs from M. nonramosus in having shorter conidiophores (43–78 μm in M. aquatilis vs. 270–360 μm in M. nonramosus) and obclavate, 3–5-septate, verruculose, pale brown to subhyaline, and larger conidia (21–30 × 5–8 μm in M. aquatilis vs. 13.6 × 2.7 μm in M. nonramosus) (Fiuza et al. 2017).

Minimelanolocus is morphologically close to Pseudospiropes in having polyblastic and integrated conidiogenous cells with sympodial percurrent proliferations (Ruiz et al. 2001). However, the conidiogenous loci in Pseudospiropes are broad, protuberant, thickened, and strongly melanized, apparently with several layers, forming a discoid black scar and markedly different from the loci in Minimelanolocus. In addition, Pseudospiropes species have distoseptate conidia, which differ from the conidia of M. nonramosus (Ruiz et al. 2001). The genus Pleurophragmium Costantin resembles Minimelanolocus in having single unbranched conidiophores with sympodial, denticulate conidiogenous cells (Hughes 1958; Ellis 1971). However, conidial color of Pleurophragmium species is dematiaceous except for P. acutum (Grove) M.B. Ellis, P. malaysianum Matsush. and P. naviculiforme Matsush. (D’Souza and Bhat 2012). M. nonramosus differs from these species in having cylindrical, 1–2-septate conidia and conidial size.

The upper part of some conidiophores in M. nonramosus (e in Fig. 3) becomes suddenly slimmer and paler, which seems to be a new percurrent proliferation after the head secedes. However, this phenomenon is not very common. We also observed this phenomenon in the plate of Minimelanolocus melanicus H.Y. Su, Udayanga & K.D. Hyde, which was not described by the authors (Liu et al. 2015).

Thysanorea papuana (Aptroot) Arzanlou, W. Gams & Crous. Stud. Mycol. 58: 80 (2007) Fig. 4.

Fig. 4
figure 4

Thysanorea papuana (form MFLU 17–1657). a, b Colonies on submerged wood. c Conidiophore. d Long regenerative extension on the conidiophore with percurrent proliferation. e Heads and branches breaking off from the conidiophore. f–h Conidiogenous cells with denticulate conidiogenous loci. i–l Conidia. m Germinating conidium. n Colony on PDA (front view). o Colony on PDA (reverse). Scale bars: a, b = 80 μm, c = 30 μm, d–h = 10 μm, i–m = 5 μm

Full size image

≡ Periconiella papuana Aptroot, Nova Hedwigia 67: 491 (1998)

= Ramichloridium lignicola KM. Tsui, Goh, KD. Hyde & Hodgkiss, Cryptog. Mycol. 22: 141 (2001)

? = Alysidiopsis lignicola Mercado, Figueras & J. Mena, Mycotaxon 60: 444 (1996). Facesoffungi Number: FoF02731

= Thysanorea aquatica W. Dong, H. Zhang & K.D. Hyde, Mycol. Progr. 17: 625 (2018), syn. nov. Fig. 5.

Fig. 5
figure 5

Thysanorea papuana (from MFLU 15–2695). a Herbarium specimen. b–e Conidiogenous cells with conidia. The arrow shows where the branched head broke off, and the regenerative extension regrew. f Conidia. Scale bars: b–e = 10 μm, f = 5 μm

Full size image

Material examined: Thailand, Prachuap Khiri Khan, on submerged wood in a stream, 30 July 2015, W. Dong, 29C (MFLU 15–2695), living culture: MFLUCC 15–0966; Thailand, Chiang Rai, on submerged wood in a stream, 4 August 2017, G.N. Wang, G3 (MFLU 17–1657), living culture: MFLUCC 17–2315.

Notes: Molecular evidence supports that our new collection (MFLUCC 17–2315) belongs to Thysanorea papuana (Fig. 2). Morphologically, our collection is similar to the holotype in conidiophores and conidial morphology, but they differ in the levels of branchlets (2–3 in our collection vs. 3–5 in the holotype) and the length of regenerative extension (35–65 μm in our collection vs. 15–30 μm in the holotype) (Arzanlou et al. 2007; Pratibha and Prabhugaonkar 2015; Kirschner 2016). These differences are considered not significant, since Kirschner (2016) stated that more complex branching appeared to be an artifact of prolonged cultivation in Th. papuana.

Thysanorea aquatica (MFLUCC 15–0966) was recently introduced by Dong et al. (2018) based on only two levels of branchlets and hyaline conidia with guttules and constricted septate, which they considered to be distinct enough from Th. papuana. The character of the conidiophores was not clearly described in the original description. We re-examined the herbarium material and re-described the conidiophores as “composed of unbranched stipe and branched head, with stipe sometimes percurrently proliferating below the branched head which is paler than stipe, branched head sometimes breaking off and replaced by a regenerative extension, each branch composed of (1–)2(–6) cells, each cell 4–6 μm long, 1–3 μm wide.” The significant character is that conidiophore heads contain branches which could break off and be replaced by regeneration, a feature consistent with the holotype. In addition, only elder conidia have guttulae and more constricted septa. The others are morphologically identical to the holotype (f in Fig. 5). Therefore, we synonymize Th. aquatica under Th. papuana.

The morphology of Th. papuana is variable (Table 2). Branches were up to four levels in the specimen from Taiwan (Kirschner 2016) and up to three levels in the specimen from India (Pratibha and Prabhugaonkar 2015), while they were up to 5–6 levels from the culture of Arzanlou et al. (2007). Septation ranges from 0 to 3, but are mostly 1. The conidial size is between 4 and 11 in length and 2–4 μm in width. The ITS divergence of 1–2% between the ex-type, MFLUCC 15–0966 and MFLUCC 17–2315 are within the generally accepted 1.5% nucleotide differences in the ITS regions that may be indicative of conspecificity (Jeewon and Hyde 2016).

Table 2 Comparison of different collections of Thysanorea papuana
Full size table

Discussion

The generic delimitation of Thysanorea with one single species Th. papuana was discussed by Kirschner (2016) who suspected that more complex branches appeared to be an artifact of a degenerated culture, because the conidiophore heads found on the natural substrate were not as complex as those in culture. This difference was attributed to prolonged growth of the fungus in culture (Kirschner 2016). Unfortunately, our collection did not sporulate in culture. The conidiophore character “branched head sometimes or often breaks off from the stipe and with a percurrent proliferation” was easily observed in all the collections of Th. papuana. This character is an important morphological marker in Thysanorea at the genus level and morphologically significant to distinguish it from other related hyphomycetous taxa, such as Dactylaria.

The genus Minimelanolocus has a strong resemblance to Thysanorea in having mononematous and cylindrical conidiophores, holoblastic and sympodial proliferating conidiogenous cells with denticulate conidiogenous loci, as well as mostly septate, hyaline to pale brown conidia. The only difference between the two genera is the branched conidiophores in Thysanorea and the unbranched conidiophores in Minimelanolocus (Liu et al. 2015; Kirschner 2016; Dong et al. 2018). In this study, as well as in Dong et al.’s 2018, the type species, Th. papuana, was added to the phylogenetic tree. Both analyses showed that Thysanorea is closely related with Minimelanolocus and nested within a clade corresponding to Minimelanolocus which is revealed to be paraphyletic (Fig. 1). There is no coding gene reported for any Minimelanolocus species. We sequenced the TEF gene for Th. papuana (GenBank No. 392331 for strain MFLUCC15–0966 and GenBank No. 392330 for strain MFLUCC 17–2315) and Minimelanolocus thailandensis (GenBank No. MK392329). The similarity between MFLUCC15–0966 and MFLUCC 17–2315 is 98%, while that between Th. papuana and M. thailandensis is 92%. More sequence data, particularly protein coding genes, together with the recollection and sequencing of M. navicularis (R.F. Castañeda) R.F. Castañeda, the generic type, will help to clarify phylogenetic relationships between these closely related genera. Once its placement is known, the phylogenetic status of both genera will be further clarified and some generic rearrangement might be needed in the future.

Thysanorea was ecologically widespread from Oceania to Asia occurring on dead woody substrates to submerged wood (Table 2). In Thailand, Th. papuana has been reported not only on submerged wood (Dong et al. 2018; this study) but also on terrestrial baits (as Ramichloridium lignicola, Kodsueb et al. 2016). This study increases our understanding of freshwater hyphomycetes in Eurotiomycetes.