Introduction

Along the Andes, dispersed woodlands of Polylepis (Rosaceae) constitute a common component of the treeline. Polylepis woodlands are characterized by being dominated either mostly or exclusively by representatives of this genus and by being found in areas of difficult access, including rocky outcrops and mountain slopes (Fjeldså and Kessler 1996; Kessler 2006). These woodlands are important habitats for plant and animal species, including several endangered habitat-specialist bird species (Fjeldså and Kessler 1996) and constitute important sources of firewood, fodder, and medicinal plants for local indigenous communities (Fjeldså and Kessler 1996; Domic et al. 2014; Hurtado et al. 2018). Polylepis woodlands are experiencing a rapid decline due to unsuitable land use practices, habitat destruction, and ongoing climate change (Navarro et al. 2005).

Polylepis tarapacana is a species that includes small trees and shrubs distributed along the semiarid Andean highlands from Peru to Argentina and Chile (Kessler 1995). The species possess morphological (Simpson 1979; Kessler 1995) and physiological adaptations (Rada et al. 2001; Azócar et al. 2007; González et al. 2007) to tolerate frost and desiccation due to extreme environmental conditions, including high solar irradiance, night frosts, and high diurnal temperature variation. In Nevado Sajama in Bolivia, P. tarapacana forms one of the world’s highest-altitude woodlands, reaching elevations up to 5200 m a.s.l. (Jordan 1980). The species is currently categorized as “Vulnerable” in Bolivia due to firewood extraction, habitat loss, and climate change (MMAyA 2012; Cuyckens et al. 2016). A potentially pathogenic fungus, Leptosphaeria polylepidis, constitutes an emerging threat to the permanence of remaining P. tarapacana woodlands, as its infection has been attributed to the mortality of several individuals in the Sajama National Park (Coca-Morante 2012). The fungus, originally described from two collections made in 2002 in the Sajama National Park, forms stromatic black knots on the branches of infected individuals of P. tarapacana (Macía et al. 2005). Recent evaluations have shown that the infection is widespread in protected areas particularly of the southeast and northeast slopes of Nevado Sajama. However, actions to control the spread of the infection have not been implemented due to a lack of knowledge regarding the life-cycle and pathogenicity of L. polylepidis.

The genus Leptosphaeria, with the type species Leptosphaeria doliolum, resides in the Leptosphaeriaceae within Pleosporales (Zhang et al. 2012; de Gruyter et al. 2013; Ariyawansa et al. 2015). It contains hundreds of described species that in recent molecular analyses were proven to belong to different genera, including Brunneosphaerella, Heterospora, Neoleptosphaeria, Paraleptosphaeria, Plenodomus, Pseudoleptosphaeria, or Subplenodomus (Crous et al. 2009; de Gruyter et al. 2013; Ariyawansa et al. 2015). The generic placement of the majority of Leptosphaeria species is, however, still uncertain, and their rearrangements into the correct genus need molecular phylogenetic analyses, ideally using freshly collected materials. When describing Leptosphaeria polylepidis, Macía et al. (2005) provided a nuc rDNA ITS1-5.8S-ITS2 (ITS) sequence from the holotype material. Comparing this sequence with sequences available at that time in GenBank, they found that L. polylepidis was most similar to Leptosphaeria dryadis, a parasite of Dryas octopetala in arctic and alpine regions of the Northern Hemisphere. Leptosphaeria dryadis is currently placed in the genus Paraleptosphaeria (de Gruyter et al. 2013). However, it is not clear if this is also the correct genus for L. polylepidis because the ITS sequence generated by Macía et al. (2005) was not included in any of the subsequent studies dealing with Leptosphaeria-like species; many new sequences of Leptosphaeria-like species were generated in recent years that could disprove close relationships between L. polylepidis and Paraleptosphaeria dryadis, and sequences from more than one DNA locus can better resolve generic boundaries within Leptosphaeriaceae (Ariyawansa et al. 2015).

Therefore, it is the aim of this study to resolve the phylogenetic and generic placement of Leptosphaeria polylepidis by employing integrative morphological and molecular analyses using sequences from multiple DNA loci. During the analyses of freshly collected material, we unexpectedly found that pseudothecia of L. polylepidis were overgrown by conidiomata that were not reported in the original description of the fungus. It was not clear if these conidiomata represented an asexual state of L. polylepidis or, alternatively, belonged to a mycoparasitic fungus. Therefore, the second aim of this study was to answer this question by applying morphological and molecular analyses using nuc rDNA ITS and LSU sequences.

Materials and methods

Specimen sampling and documentation

This study is based on newly collected material of Leptosphaeria polylepidis. Four specimens were collected in the Sajama National Park in Bolivia, which is the type locality and the only known place of occurrence of L. polylepidis. One of them was partly covered with conidiomata. This specimen was used for detailed morphological/anatomical and molecular analyses. The voucher material is preserved at the Herbario Nacional de Bolivia (LPB) and at the fungal herbarium of the W. Szafer Institute of Botany, Polish Academy of Sciences, Kraków (KRAM F). Attempts to obtain cultures failed.

Morphological analyses

The morphology was examined using standard stereo and light microscopy (Nikon SMZ 800, Nikon Eclipse 80i DIC). Thin sections of ascomata and conidiomata were made manually using a razor blade or with the aid of a freezing microtome Thermo Scientific Microm HM430 equipped with BFS-MP freezing stage and BFS-3MP controller. The sectioned material was examined in distilled water, 10% solution of potassium hydroxide (KOH), or lactophenol cotton blue (LPCB). The amyloidity of fungal structures was tested using Lugol’s solution (IKI), or a combination of first KOH and then IKI (KOH/IKI). All measurements were made in distilled water.

DNA isolation, PCR, and sequencing

Genomic DNA was extracted in three separate extraction rounds, namely one extraction from 10 pseudothecial hymenia of Leptosphaeria polylepidis, one extraction from conidial mass obtained from five conidiomata, and one extraction from 15 entire conidiomata of unknown fungus, using the DNeasy™ Plant Mini Kit or QIAamp DNA Investigator Kit (Qiagen, Germany), following the manufacturer’s instructions. For Leptosphaeria polylepidis, a total of four genetic markers were amplified, namely nuc rDNA 18S (SSU), nuc rDNA ITS1-5.8S-ITS2 (ITS), nuc rDNA 28S (LSU), and a fragment of the translation elongation factor 1-alpha gene (TEF1). For unknown conidiomatal fungus, two genetic markers were amplified, namely ITS and LSU. The following primer pairs were used for amplification: NS1-NS24 (White et al. 1990; Gargas and Taylor 1992), ITS1F-ITS4 (White et al. 1990; Gardes and Bruns 1993), LROR-LR7 (Vilgalys and Hester 1990; Rehner and Samuels 1994), and EF1983F-EF12218R (Rehner and Buckley 2005) for SSU, ITS, LSU, and TEF1, respectively. Polymerase chain reactions (PCRs) were performed as follows: for Leptosphaeria polylepidis, 2 μl of DNA extract was used for SSU, ITS, and LSU and 5 μl for TEF1, while for unknown conidiomatal fungus, 3 μl of DNA extract was used for ITS and LSU in a total volume of 25 μl PCR reactions; the rest of components were added according to Flakus et al. (2019). Cycling conditions were performed as reported by Rodriguez-Flakus and Printzen (2014) for the nuclear markers and Rehner and Buckley (2005) for TEF1 with modification of the initial denaturalization to 4 min and final extension to 5 min (M. Mardones, pers. comm.). PCR products were visualized on agarose gels and later purified using ExoSap. The PCR amplicons were sequenced by Macrogen Europe B.V. (Amsterdam, the Netherlands).

Phylogenetic analyses

Newly generated sequences were trimmed, assembled, and edited using Geneious Pro, version 8.0.5 (Biomatters Inc.). BLAST searches in GenBank (Altschul et al. 1997) were performed to find sequences of most closely related species, revealing affinities of sequences from Leptosphaeria polylepidis to sequences of the Leptosphaeriaceae and sequences obtained from unknown conidiomatal fungus to sequences of the Dictyosporiaceae. Accordingly, two separate datasets were assembled to resolve the phylogenetic placement of each fungus.

To resolve the phylogenetic position of Leptosphaeria polylepidis, a four-gene dataset (SSU+ITS+LSU+TEF1) was assembled that included sequences generated in this study and sequences from 45 specimens of closely related members of the Leptosphaeriaceae and a member of the Cucurbitariaceae used as an outgroup (selected from Zhang et al. 2012 and Tibpromma et al. 2017; see Table 1). Each single gene dataset was aligned separately using the MAFFT algorithm (Katoh et al. 2005) as implemented on the GUIDANCE web-server (Penn et al. 2010); to remove poorly or ambiguously aligned uncertain columns, a default cut-off score of 0.93 in all single gene alignments were chosen. The single-gene datasets were concatenated to a final alignment using Geneious Pro and consisted of 1023 bp (SSU), 452 bp (ITS), 1326 bp (LSU), and 915 bp (TEF1). Subsequently, the analyses were performed in the CIPRES Scientific gateway portal (http://www.phylo.org/portal2) (Miller et al. 2010). The optimal partitioning scheme and substitution models for each single-gene alignment were inferred by PartitionFinder 2 (Lanfear et al. 2016). A single partition for SSU and LSU; three partitions for ITS1, 5.8S, and ITS2 regions; and three for each of the codon positions of TEF1 were found. Substitution models selected under the greedy search algorithm and Akaike information criterion (Lanfear et al. 2016) were TVM+G+I for SSU, ITS2, and LSU; JC+I for 5.8S; and GTR+G for ITS1 and used as priors for Bayesian phylogenetic inference (BI) analyses. Maximum likelihood (ML) and bootstrap support (BS) analyses were performed on each locus separately and concatenated data set by using RAxML-HPC2 on XSEDE with a rapid BS algorithm (Stamatakis 2014), 1000 replicates, and GTRGAMMA substitution model. Bayesian inference (BI) analyses were carried out using Markov Chain Monte Carlo as implemented in MrBayes on XSEDE (Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003) and were based on 10 million generations, sampling every 1000th tree, two independent parallel runs of four incrementally heated chains (0.15) and discarding the first 0.5 of the sampled trees.

Table 1 Species used for inferring the phylogeny of the Leptosphaeriaceae (and a member of the Cucurbitariaceae used as an outgroup) with details of their host, country of origin, strain or voucher information, GenBank accession numbers (ITS, LSU, SSU, TEF1) and references
Full size table

To resolve phylogenetic position of the conidiomatal fungus, a two-gene dataset (ITS+LSU) was assembled, which contained sequences generated in this study and sequences from 28 related members of the Dictyosporiaceae and two sequences of Periconia igniaria used as an outgroup (selected from Iturrieta-González et al. 2018 and Yang et al. 2018; see Table 2). Phylogenetic analyses were performed as described above and were based on 820 bp (LSU) and 412 bp (ITS) alignments. A single partition for LSU (GTR+I+G) and three partitions for ITS (GTR+G for ITS1, SYM+I+G for 5.8S, and JC+I+G for ITS2) were selected based on PartitionFinder 2 (Lanfear et al. 2016). The ML and BI analyses were performed with similar parameters as described above, but 20 million generations in four independent parallel runs were sampled and six gradually heated chains used. Relationships receiving bootstrap support (ML-BP) above 70% and 0.95 as Bayesian posterior probability (PP) were considered strongly supported. The resulting trees from both phylogenetic analyses were built in Figtree v.1.4.2 (http//tree.bio.ed.ac.uk/software/figtree).

Table 2 Species used for inferring phylogeny of the Dictyosporiaceae (and Periconia igniaria used as an outgroup) with details of their substrate or host, country of origin, strain or voucher information, GenBank accession numbers (ITS, LSU), and references
Full size table

Results

Phylogenetic analyses

The phylogenetic analyses of the four-locus (SSU+ITS+LSU+TEF1) dataset comprising representatives of the Leptosphaeriaceae (and a member of the Cucurbitariaceae used as an outgroup) resulted in overall similar tree topologies obtained from BI and ML analyses. The Heterospora clade nested together with Subplenodomus in the ML analysis but was sister to the remaining genera of the Leptosphaeriaceae in BI analysis. In general, high bootstrap support values above 70% (ML–BP) and posterior probabilities above 0.95 (PP) were retrieved and supported monophyletic lineages representing distinct genera as in previous studies (de Gruyter et al. 2013; Ariyawansa et al. 2015; Tibpromma et al. 2017). The Bayesian tree is displayed in Fig. 1. The newly generated Leptosphaeria polylepidis sequences clustered with ex-holotype sequence of L. polylepidis (ML–BP = 100%, PP = 1) in their well-supported Paraleptosphaeria lineage (ML–BP = 78%, PP = 1); relatedness of L. polylepidis and Paraleptosphaeria dryadis was weakly supported (ML–BP = 60%). The ITS sequence from the newly generated material and holotype of L. polylepidis differed in 6 bp. The SSU, LSU, and TEF1 sequences were not available from holotype material.

Fig. 1
figure 1

Bayesian inference consensus tree inferred from a four-locus dataset (SSU+ITS+LSU+TEF1) of the Leptosphaeriaceae clade. Numbers above bold branches indicate bootstrap support values ≥ 70% from maximum likelihood and posterior probabilities ≥ 0.95 from Bayesian inference analyses. Cucurbitaria berberidis is used as an outgroup

Full size image

The phylogenetic analyses of the two-locus (ITS+LSU) dataset, including representatives of the Dictyosporiaceae and two sequences of Periconia igniaria used as an outgroup, resulted in similar tree topologies. Dictyosporium grouped in a single clade together with Aquadictyospora in ML analysis but was sister to the Pseudodictyosporium lineage in the BI analysis. The Bayesian consensus tree is displayed in Fig. 2. The phylogenetic tree shows 11 monophyletic lineages of the Dictyosporiaceae, corresponding to 10 known genera and one unknown lineage, the latter of which consisted of the conidiomatal fungus from Leptosphaeria polylepidis. Results were roughly consistent with previous phylogenetic results (e.g., Iturrieta-González et al. 2018; Yang et al. 2018), except for some differences in branches with low statistical support values mainly at the backbone of the phylogenetic tree. The sequences of the conidiomatal fungus from two independent extraction rounds were identical and of high quality, excluding the possibility of DNA isolation from contaminating fungi, and were nested within the currently circumscribed family Dictyosporiaceae (ML–BP = 100%, PP = 1). They formed a sister group to the members of the genus Pseudocoleophoma (ML–BP = 89%, PP = 1). The phylogenetic relationships of the conidiomatal fungus and the Pseudocoleophoma lineage with their relatives in the Dictyosporiaceae remain unresolved, due the lack of support in our molecular analyses.

Fig. 2
figure 2

Bayesian inference consensus tree inferred from two-locus dataset (ITS+LSU) of the Dictyosporiaceae clade. Numbers above bold branches indicate bootstrap support values ≥ 70% from maximum likelihood and posterior probabilities ≥ 0.9 from Bayesian inference analyses. Periconia igniaria is used as an outgroup

Full size image

Taxonomy

Paraleptosphaeria polylepidis (M.J. Macía, M.E. Palm & M.P. Martín) Piątek, Flakus & Rodr. Flakus, comb. nov. (Fig. 3)

Fig. 3
figure 3

Paraleptosphaeria polylepidis (KRAM F-59659). a, b, d Habit of the crowded ascomata on twig of Polylepis tarapacana. c Longitudinal section through ascomata and stroma. e Paraplectenchymatous peridium from longitudinal section. f Hamathecium composed of thin interascal filaments. g Bitunicate asci with 5–8 ascospores inside (left mounted in water, right mounted in KOH/IKI). h Ascospores mounted in water. Scale bars: a–d 500 μm, e 25 μm, f, h 10 μm, g 50 μm

Full size image

MycoBank no. MB832195

Basionym: Leptosphaeria polylepidis M.J. Macía, M.E. Palm & M.P. Martín, Mycotaxon 93: 402 (2005)

Ascomata pseudothecial, developed on stroma. Stroma 1–3 mm high, 2–7 mm wide, black, surface with visible individual pseudothecia, usually with base growing inward and intermixed with host tissue. Pseudothecia dark brown to black, matt, rough, 300–700 μm high, 300–600 μm wide, globose to subglobose with base usually protruding into ca. 200–1000 μm high stipes that fuse below into stroma; papillae short or lacking. Peridium 60–150 μm wide, pale to dark brown (KOH+ greenish grey), consisting of a single stratum, paraplectenchymatous, textura angularis, composed of several layers of isodiametric to slightly elongate thin-walled cells, 15–25 × 9–15 μm. Hamathecium persistent, composed of relatively thin, 1.5–4 μm wide, septate, branched and strongly anastomosed pseudoparaphyses. Asci 120–250 × 20–45 μm, 5–8-spored, bitunicate, cylindrical-clavate, slightly curved, with a short stipe, apically widened, with an ocular chamber, IKI–, KOH/IKI–; endosacus KOH/IKI+ orange. Ascospores 45–60 × (14–)20–23 μm, pale brown, broadly ellipsoidal, narrower at the top, constricted at septa, uniformly thin-walled, without perispore or gelatinous coat, (2–)3(–4)-septate. Asexual state unknown.

Specimen examined (used for morphology and molecular analyses): See host fungus in the type of Sajamaea mycophila.

Additional specimens examined: Bolivia, Oruro, Sajama National Park (=Parque Nacional Sajama), W slope of Sajama’s volcano, Quebrada Huaytana, 18°11′ S, 68°51′ W, elev. ca. 4000–4200 m, on Polylepis tarapacana, 1 Feb. 2016, A. Domic (three collections: AD1, AD2, AD5; all in LPB).

Sajamaea Flakus, Piątek & Rodr. Flakus, gen. nov.

MycoBank no. MB832196

Etymology: Referring to the place of occurrence of the new genus – slopes of Nevado Sajama, the highest mountain in Bolivia.

Description: Mycoparasitic. Conidiomata pycnidial, uniloculate to multi-loculate, subglobose to irregular in shape, pale brown. Peridium composed of several layers of isodiametric to slightly elongate cells in the form of textura angularis. Conidiophores hyaline, reduced to phialidic conidiogenous cells. Conidia pigmented (pale brown), broadly ellipsoidal, aseptate, smooth, thin-walled, guttulate. Generic type: Sajamaea mycophila Flakus, M. Piątek & Rodr. Flakus.

Sajamaea mycophila Flakus, Piątek & Rodr. Flakus, sp. nov. (Fig. 4)

Fig. 4
figure 4

Sajamaea mycophila (KRAM F-59659). a–c Habit of pale brown conidiomata growing on ascomata of Paraleptosphaeria polylepidis. d–e Longitudinal section of uniloculate conidioma erumpent through the outermost layer of host peridium (e mounted in LPCB). f Longitudinal section of multi-loculate conidioma showing layers of conidiogenous cells (hyaline areas) and conidial masses (brown areas). g Section through paraplectenchymatous peridium showing conidiogenous cells mounted in LPCB. h Conidiogenous cells and young conidia mounted in LPCB. i Conidia. Scale bars: a 1000 μm, b–c 500 μm, d–f 50 μm, g–i 10 μm

Full size image

MycoBank no. MB832197

Etymology: The epithet name mycophila refers to the occurrence of this fungus on other fungus.

Description: Mycoparasitic on Paraleptosphaeria polylepidis, causing moderate damages of host ascomata. Conidiomata pycnidial, as pale brown galls on host surface, 150–300 μm wide, 120–300 μm high, first erumpent through the outermost layer of host peridium, later almost sessile on pseudothecial clusters of host, uniloculate to multi-loculate, solitary to aggregated, subglobose to irregular in shape, pale brown; better seen when wet. Peridium 20–30 μm wide, pale to dark brown, composed of about 3–8 layers of cells with walls up to 2 μm wide, cells isodiametric to slightly elongate, textura angularis. Conidiophores reduced to conidiogenous cells. Conidiogenous cells 5–11 μm high, 4–8 μm wide, enteroblastic, phialidic, smooth-walled, hyaline, with a small collar, densely outlining inner surfaces of pycnidia. Conidia 9–13 × 5.5–7.5 μm, pale brown, broadly ellipsoidal, sometimes slightly narrower at one point, with rounded ends, aseptate, smooth, thin-walled, guttulate. Sexual state unknown.

Specimen examined: Bolivia, Oruro, Sajama National Park (=Parque Nacional Sajama), W slope of Nevado Sajama, closely located to the permanent monitoring plot “Cohiri,” 19K 512102, 8002927, elev. ca. 4425 m, parasitic on Paraleptosphaeria polylepidis growing on Polylepis tarapacana, 15 Oct. 2016, A.N. Palabral-Aguilera (APA-2999), C. López & M.I. Gómez (LPB 0003512 – holotype, KRAM F-59659 – isotype).

Discussion

Molecular phylogenetic analyses, based on the four-gene dataset, placed Leptosphaeria polylepidis in a well-supported clade consisting of members of the genus Paraleptosphaeria, including the type species Pa. nitschkei. The genus Paraleptosphaeria has been described by de Gruyter et al. (2013) for Pa. dryadis, Pa. macrospora, Pa. nitschkei, Pa. orobanches, and Pa. praetermissa that form a distinct monophyletic lineage among Leptosphaeriaceae within Pleosporales. Paraleptosphaeria is morphologically similar but clearly divergent genetically from the Leptosphaeria lineage. Separateness of Paraleptosphaeria from Leptosphaeria has been confirmed by subsequent molecular studies (Ariyawansa et al. 2015; Chen et al. 2015; Liu et al. 2015; Wanasinghe et al. 2016; Tibpromma et al. 2017); additionally, two novel species, Pa. padi and Pa. rubi, have been described in the genus (Ariyawansa et al. 2015; Tibpromma et al. 2017) after its original establishment. Therefore, prior to this study, seven species were accepted in Paraleptosphaeria.

The generic concept of Paraleptosphaeria includes species having immersed, solitary or aggregated, thick-walled, ostiolate and unilocular pseudothecia, bitunicate, 8-spored asci, and fusiform, 5–8-septate, hyaline to yellow-brownish ascospores. The asexual state, when present, is coelomycetous consisting of pycnidial, unilocular conidiomata, phialidic conidiogeneous cells, and oblong or ellipsoidal, aseptate, hyaline conidia (de Gruyter et al. 2013). Leptosphaeria polylepidis generally fits well to this concept, except that its pseudothecia are superficial and developed on stroma. However, this character may be variable in the genus, and, for example, Liu et al. (2015) reported superficial pseudothecia in their material of Paraleptosphaeria nitschkei. Therefore, based on molecular evidence, Leptosphaeria polylepidis is transferred to Paraleptosphaeria (as Pa. polylepidis) in this study.

Other than genetically divergent, Paraleptosphaeria may be ecologically different from Leptosphaeria by having other host specialization. Five of the eight Paraleptosphaeria species occur on Rosaceae, and three species on other host families, Asteraceae, Orobanchaceae, and Polygonaceae, while Leptosphaeria species tend to inhabit hosts in very diverse families but not Rosaceae (de Gruyter et al. 2013; Ariyawansa et al. 2015; Dayarathne et al. 2015; Liu et al. 2015). The molecular phylogenetic analyses showed that Paraleptosphaeria polylepidis was weakly supported as sister species of Pa. dryadis supporting earlier finding by Macía et al. (2005). These authors pointed out that Pa. polylepidis and Pa. dryadis occur both at high elevations. Their host plants, Polylepis tarapacana and Dryas octopetala (Rosaceae), are adapted to grow in cold ecosystems. According to our molecular analyses, the branch leading to Pa. polylepidis is very long, which may suggest an old evolutionary origin of this lineage. It is, however, also possible that undiscovered closely related Paraleptosphaeria species may still exist in poorly studied South American countries. The weakly supported sister-group relationship of Pa. polylepidis and Pa. dryadis and the circumstance that their hosts are phylogenetically unrelated (Potter et al. 2007) suggests that they evolved from different ancestral species.

The assignment of conidiomata found on the pseudothecia of Paraleptosphaeria polylepidis to a new phylogenetic lineage of the Dictyosporiaceae was unexpected as we earlier hypothesized having encountered the asexual state of Pa. polylepidis. The sequences obtained from the fungus formed a strongly supported sister clade to members of the genus Pseudocoleophoma but could not be classified in Pseudocoleophoma on the basis of different morphological characteristics. The genus Pseudocoleophoma has been described for two species, Ps. calamagrostidis (generic type) and Ps. polygonicola, that produced sexual states on the natural substrates and asexual states in cultures (Tanaka et al. 2015). Two more species have been described later, Ps. typhicola forming only asexual state on the natural substrate (Hyde et al. 2016) and Ps. bauhiniae producing both asexual and sexual states on the natural substrate (Jayasiri et al. 2019). While the fungus found on the pseudothecia of Pa. polylepidis lacks a sexual state and forms uniloculate to multi-loculate pycnidia and pale brown, broadly ellipsoidal conidia, the species of Pseudocoleophoma usually produce a sexual state and pycnidia of the asexual state are uniloculate and the cylindrical conidia are hyaline. It also differs ecologically by parasitism on another fungus, while all Pseudocoleophoma species are saprobic on dead plant materials. Though we did not check mycelial interactions in culture experiments, the fungus forms pycnidia erumpent through the outermost layer of host peridium and, therefore, is likely mycoparasitic (at least necrotrophic as defined by Sun et al. 2019). Furthermore, the molecular phylogenetic analyses indicated genetic distances between this fungus and Pseudocoleophoma comparable to distances between other Dictyosporiaceae clades, assigned to distinct genera (Boonmee et al. 2016). No morphologicaly similar coelomycete genus was found in Sutton (1980) and in the recent literature; therefore, a new genus and species names, Sajamaea and S. mycophila, are introduced for this fungus. The family Dictyosporiaceae contains mostly asexual species, while sexual species are very rare, being represented only by Dictyosporium meiosporum, D. sexualis, Gregarithecium curvisporum, Pseudocoleophoma calamagrostidis, P. polygonicola, and P. bauhiniae (Tanaka et al. 2015; Boonmee et al. 2016; Jayasiri et al. 2019). The asexual species are mostly hyphomycetous, often producing characteristic cheirosporous conidia (Tanaka et al. 2015; Boonmee et al. 2016; Iturrieta-González et al. 2018; Yang et al. 2018), but Sajamaea and Pseudocoleophoma have coelomycetous asexual states, which supports their close phylogenetic relationship.

The parasitism of Sajamaea mycophila suggests that this species may be potentially used as a biopesticide that could prevent the development of Paraleptosphaeria polylepidis and consequently protect Polylepis tarapacana woodlands.

Additional emphases are thus required to re-collect S. mycophila and aim at culturing it in order to unravel its mycoparasitic and biopesticidal potentials.