Background and aims – During fieldwork in French Guiana in 2018, two fungal specimens resembling Subulicystidium oberwinkleri were collected. This study aims to clarify species- and genus-level assignment of this material and of S. oberwinkleri.

Material and methods – Corticioid fruiting bodies were examined under a light microscope, and spores from the spore prints were studied morphometrically. DNA sequences of large subunit-coding DNA and internal transcribed spacer were used for maximum likelihood and Bayesian phylogenetic analyses. Furthermore, guanine-cytosine content was calculated for the LSU dataset, and UNITE Species Hypotheses Matching Analysis was performed for newly generated ITS sequences.

Key results – The new genus, Palliocystidium, is introduced in the family Hydnodontaceae, based on the peculiar pattern of cystidial encrustation (crystalline plates of various shapes) and cystidial septation and supported by results of phylogenetic analyses. Within the new genus, the new species P. chlamydatum from French Guiana is described. In addition, Subulicystidium oberwinkleri is transferred to Palliocystidium. The two species can be distinguished by the size of their reniform spores. Both species display high levels of guanine-cytosine content at the scale of the order Trechisporales.

Conclusions – In the newly introduced genus Palliocystidium and genera Subulicystidium and Luellia, there is significant potential for further exploration of species diversity and generic boundaries. Additional intensified fruiting-body-based sampling of taxa and genes is necessary to clarify the relationship of genera within Hydnodontaceae.

Agaricomycetes, corticioid fungi, crystals, cystidia, DNA barcoding, generic relationship, reniform basidiospores

The order Trechisporales K.H.Larss. is one of the early-diverging fungal lineages of the class Agaricomycetes Doweld, phylum Basidiomycota R.T.Moore. Taxonomic studies on Trechisporales not only shed light on global fungal diversity but also contribute to a better understanding of early events in the evolution of mushroom-forming fungi, including the evolution of fruiting body complexity and nutritional strategies. The fruiting bodies of the known representatives of Trechisporales display certain variation in shape and considerable variation in hymenophore configuration (Hibbett et al. 2014; de Meiras-Ottoni et al. 2021). Current knowledge of the nutrition strategies of Trechisporales indicates their basic saprotrophic ability along with an evolving ability to form ectomycorrhiza (Dunham et al. 2007; Rinaldi et al. 2008; Vanegas-León et al. 2019).

With increased attention to Trechisporales since its formal description (Hibbett et al. 2007), the view on the composition of families and genera in the order is rapidly changing. The assumed status of a separate family for the genus Sistotremastrum J.Erikss. and its sister genus Sertulicium Spirin, Volobuev & K.H.Larss. has been extensively debated and recently led to the segregation of the separate order Sistotremastrales L.W.Zhou & S.L.Liu (Larsson 2007; Spirin et al. 2021; Liu et al. 2022a, 2022b). The composition of genera within Trechisporales per se, now all treated within the family Hydnodontaceae Jülich, has been differently considered by different authors (He et al. 2019; Liu et al. 2019; Spirin et al. 2021). Currently, 12 genera are accepted as members of Hydnodontaceae, and for 11 of them, phylogenetic placement is molecularly supported (Liu et al. 2022a). However, the intergeneric relationship within Hydnodontaceae remains poorly understood.

Though not present in the largest genus Trechispora P. Karst., cystidia or cystidia-like hyphae are characteristic for many genera of Hydnodontaceae and often serve as a genus-level morphological marker (Oberwinkler 1965; Parmasto 1968; Liu et al. 2019). Among such genera, Subulicystidium Parmasto is currently one of the richest in terms of the number of known species. In the type of the genus, S. longisporum (Pat.) Parmasto, the most collected species in Europe, as well as most other known members of the genus, the smooth crystalline sheath of cystidia is covered with two chains of bow-tie-shaped crystals, which are seen in the light microscope as four chains of rectangular crystals along the cystidium body (Jülich 1975; Keller 1985). A few species deviate from this pattern by the shape, location, or frequency of individual crystals (e.g. Punugu et al. 1980) but were proven to belong to Subulicystidium based on analysis of nuclear ribosomal DNA (Ordynets et al. 2018, 2020).

On cystidia of one species of Subulicystidium, viz. S. oberwinkleri Ordynets, Riebesehl & K.H.Larss., regular chains of rectangular crystals are absent but instead plate-like to irregular oblong crystals present. Particularly, this species was placed at the most basal node and slightly apart from the rest of Subulicystidium species in nuclear ribosomal DNA-based phylogenetic trees (Ordynets et al. 2018). Another study even placed S. oberwinkleri outside of the otherwise well-supported Subulicystidium clade and thus questioned the generic affiliation of the species (Liu et al. 2019). The holotype of this species further stands out as one of the outliers among Agaricomycotina Doweld regarding guanine-cytosine (GC) content in the large subunit of the nuclear ribosomal DNA. This property of the specimen can bias the inference of its phylogenetic position (Kolařík et al. 2021).

During fieldwork in French Guiana in 2018, two fungal specimens (fruiting bodies) that resembled S. oberwinkleri were collected. Morphological and molecular investigations allowed us to consider this material as a new species in a separate genus within Hydnodontaceae. This paper describes the new species as the type of a new genus, and proposes a new generic position for S. oberwinkleri.

This study introduces the new corticioid fungal genus Palliocystidium based on peculiar cystidium morphology, UNITE species hypotheses matching, and phylogenetic analyses. Within the new genus, we describe the new species P. chlamydatum. We also transfer Subulicystidium oberwinkleri to the new genus.

Being covered by crystalline plates of various shapes and bearing septa, Palliocystidium demonstrates a novel cystidium type in the family Hydnodontaceae. This encrustation pattern deviates from the classical one in the better-known genus Subulicystidium (regular rows of rectangular crystals). Ordynets et al. (2020) commented on P. oberwinkleri (as Subulicystidium) in the context of phylogeny of Subulicystidium that “… a peculiar cystidium encrustation (crystal plates) correlates with an isolated phylogenetic position”. We argue that the species treated here as P. oberwinkleri was known to the mycological community much earlier than from the point of its formal naming in 2018. Peculiar cystidia of this species were illustrated and discussed already by Oberwinkler (1977) and later Boidin and Gilles (1988), and then Maekawa (1998). However, neither of the three studies was prepared to assign their specimens to a separate species and instead labelled them as S. nikau. This solution cannot be justified as the holotype of S. nikau has cystidia with a typical regular Subulicystidium encrustation (Cunningham 1955; Ordynets et al. 2018). The specimens mentioned in all of the above cited studies, except Maekawa (1998), were examined by Ordynets et al. (2018) and the results of the measurements were re-used in morphological analysis of the present study.

The two species that we consider within the genus Palliocystidium share the character of reniform spores but they can be easily distinguished based on the difference in the size of the latter. The separation of P. chlamydatum from S. nikau based solely on spores is more challenging and needs careful measurement of spore length. There is also a difference in the average spore width between the holotype of S. nikau collected on nikau palm (Rhopalostylis sapida H.Wendl. & Drude) in New Zealand (PDD 13816) and the specimen labelled as S. nikau but collected on dead wood on La Réunion (KAS: L1296). Molecular data are available only for Reunionese material, while DNA amplification from the holotype of S. nikau did not succeed (Ordynets et al. 2018). It cannot be excluded that the material from New Zealand and La Réunion represents two different species, and additional fresh fungal specimens and DNA sequences are necessary to explore this hypothesis.

In our study, only for specimens of P. chlamydatum spore measurements are based on spore prints and performed in Melzer’s reagent. For other collections, they are based on sections from fruiting bodies (i.e. potentially containing immature spores) and mounted in potassium hydroxide (potentially causing some swelling in microscopic structures). Therefore, morphometric comparison presented in this study should be treated with caution. Although spores of Basidiomycota do not change in size as dramatically as in some Ascomycota depending on vitality and mounting medium, we admit that future measurements should ideally be based entirely on spore prints and performed in media not potentially causing size or shape change of the spores. The measuring protocol should be comprehensively documented (Matočec 1998; Duhem 2010).

Both ITS-based phylogenetic analyses suggested a sister relationship of P. chlamydatum and Subulicystidium. This clade did not include a single accession of P. oberwinkleri (MH041511, ex-holotype) which was, in turn, placed at a distant node as a sister taxon to Luellia cystidiata. It is very unlikely that two species with almost identical LSU sequences possess strongly deviating ITS sequences. LSU and ITS of the holotype of P. oberwinkleri were amplified from different DNA extractions, and specifically ITS originates from DNA extraction experiment in bulk (Ordynets et al. 2018). Therefore, it is possible that accession MH041511 is a contamination representing Luellia-related species from the soils of Madagascar and Réunion, whose fruiting bodies are yet to be discovered. Nonetheless, the ITS region is highly variable in the two largest genera of Hydnodontaceae, Trechispora, and Subulicystidium, as well as sister order Sistotremastrales, which warns of inferring interspecific and intergeneric relationships based on this marker only (Ordynets et al. 2018; de Meiras-Ottoni et al. 2021; Spirin et al. 2021). Therefore, in interpreting the intergeneric relationship in Hydnodontaceae, we relied to a large extent upon the more conservative 28S DNA region.

Three LSU sequences of Palliocystidium deviate from each other to an extent typical for the congeneric material in fungi (cf. Schoch et al. 2012). However, it was surprising that LSU sequence from Venezuela (ex KHL 11042), apart from 9 IUPAC ambiguity symbols, was identical to the holotype of P. chlamydatum (French Guiana). The ratio of the mean spore length (1.46) and width (1.21) between the two specimens also does not exceed 1.5, which is a rule of thumb for judging the conspecificity of two randomly taken specimens in Agaricomycetes (Parmasto and Parmasto 1987). It is possible that the Venezuelan specimen named P. oberwinkleri in fact represents P. chlamydatum. The geographic proximity of the two collecting sites also speaks in favour of this assumption. Therefore, the Venezuelan specimen is treated as P. cf. oberwinkleri in the current study. The same uncertainty should probably be assigned to another Venezuelan specimen FO 14338, but an earlier attempt to amplify its rDNA was unsuccessful (Ordynets et al. 2018). It remains open whether P. oberwinkleri can occur in South America and what is the true species diversity of the genus Palliocystidium.

Our study confirmed several phylogenetic hypotheses recently put forward for Trechisporales as circumscribed by Larsson (2007). Earlier studies focusing on nc rDNA lacked consistent phylogenetic support for the clade known as the “Sistotremastrum family” (Spirin et al. 2021; Liu et al. 2022b). We could overcome this issue by providing high support for Sistotremastrales in both the maximum likelihood and Bayesian phylogenetic analyses of the 28S data and thus align our results for the order-level relationship with the multigene phylogeny of Liu et al. (2022a). Automated removal of poorly aligned positions from the dataset was probably helpful for this result.

At the Hydnodontaceae family level, different patterns of intergeneric relationship were observed in our and other studies, depending on the depth of taxon and gene sampling. Liu et al. (2022b) were first to place the acystidiate species Pteridomyces galzinii as a sister taxon of Subulicystidium, although the latter was represented by the single species S. tropicum S.H.He & S.L.Liu in their 28S+5.8S data. By considering all 18 Subulicystidium species currently present in public DNA sequence databases, we found P. galzinii as intruding in the Subulicystidium clade and probably contributing to the low branch support for it in the ML analysis of 28S. This pattern did not appear in the 5-gene phylogeny of Liu et al. (2022a), who recovered Subulicystidium as a monophyletic genus. Surprisingly, the latter study also placed Allotrechispora L.W.Zhou & S.L.Liu among cystidiate genera, although the genus morphologically resembles Brevicellicium K.H.Larss. & Hjortstam and Trechispora P.Karst. Our study, and also that of Spirin et al. (2021), refused such a convenient solution proposed by Liu et al. (2019) as genera with prominent cystidia or special encrusted hyphae (Subulicystidium and Fibrodontia, Dextrinocystis and Tubulicium) were segregated from the non-cystidiate ones (Porpomyces, Brevicellicium, and Trechispora incl. Scytinopogon). Further exploration of the intergeneric relationships within Hydnodontaceae and searching for synapomorphic characters in the family remain meaningful. Using longer reads of ribosomal DNA is a worthy improvement for this task, as well as involving additional genetic markers (Liu et al. 2022a). Unfortunately, thin corticioid fruiting bodies of Hydnodontaceae and a general trade-off of DNA quality vs age of the dried corticioid fungal collections prevent the broad application of this approach (Ordynets et al. 2018; Miller et al. 2022).

In light of the inconsistent phylogenetic placement of Palliocystidium and poorly resolved intergeneric relationship within Hydnodontaceae, an additional impediment of the phylogenetic reconstruction should be mentioned. The 28S sequence of the holotype of P. oberwinkleri was exemplified by Kolařík et al. (2021) as having one of the highest GC content among Agaricomycotina Doweld. We could reproduce this result on a smaller taxonomic scale, i.e. orders Trechisporales/Sistotremastrales, and show even the highest GC content of the holotype sequence of P. oberwinkleri in our sampling. A single 28S sequence of P. chlamydatum in this dataset was also among the GC-richest. It should be noted that the accession of P. cf. oberwinkleri did not diverge much in GC content from the median of the dataset. At the species level, deviations in GC content between taxa may indicate differences in lifestyle, i.e. saprotrophy versus some form of symbiosis. Unlike Kolařík et al. (2021), who focused on mycoparasitic Ustilaginomycotina Doweld, it is hard for us to interpret the reasons for GC variation in Hydnodontaceae and in particular in Palliocystidium in the face of limited genetic data and ecological information.

Sequencing environmental samples may help gain additional information about the DNA barcode variation, lifestyle, and taxon distribution. Though likely due to contamination while aiming to sequence DNA from fungal fruiting body, a lineage inhabiting soils of Madagascar and Réunion was elucidated. Without knowing representatives forming fruiting bodies, it would be challenging to integrate this lineage into phylogenetic analyses without a system like UNITE species hypothesis. Currently, 103 (54 non-singleton) species hypotheses are classified within Subulicystidium and 16 (12) within Luellia and mostly lack an associated species epithet in UNITE v.10.0 (Abarenkov et al. 2024).

In the era of the ongoing development of molecular techniques, species identification based on morphological characters remains highly relevant in mycology. We regret that identification keys are missing from many recent publications introducing new species of corticioid fungi. Therefore, we updated the worldwide key to the species of Subulicystidium and complemented it with data on Palliocystidium.

In the newly introduced genus Palliocystidium and genera Subulicystidium and Luellia, there is significant potential for further exploration of species diversity and generic boundaries. Additional intensified fruiting-body-based sampling of taxa and genes is necessary to clarify the relationships of the genera within Hydnodontaceae. Enrichment of molecular analyses with data from high-throughput sequencing of environmental samples can contribute to a better knowledge of species’ lifestyle and distribution.

Miroslav Kolařík is acknowledged for discussion on genomic GC content and Nitaro Maekawa for sharing the literature. The director of Forêt et Risques Naturels of the French forestry office (ONF-DRFRN), and Paris French Museum (Patrinat) are acknowledged for funding the field trip. CNRS and Jennifer Devillechabrolle (ONF, Nourague Reserve manager) are acknowledged for the perfect organization of GG’s stay at the Nourague scientific station in French Guiana. We thank reviewers and editors for commenting on the manuscript, which substantially improved its quality.