Introduction

Initially, the polyporoid genus Trametes Fr. was created by Fries (1835), in his ‘Tribe’ Polyporei to accommodate coriaceous species with poroid hymenophore characterized by a context continuously descending into the hymenial trama. In addition other genera were created based on other structures of the hymenophore: lamellate in Lenzites Fr., or daedalean in Daedalea Fr. for instance. Later, Quélet (1886) refined the systematics of the polypores (Fam. II Polyporei) and separated species with regular pores (Trib.II Polypori, including stipitate species such as Caloporus Quél. or Leucoporus Quél. and sessile or resupinate species: Coriolus Quél. Phellinus Quél. etc.), from species with alveoloid to daedalean pores (Trib. III Daedalei, including Trametes gibbosa, T. suaveolens, and Daedalea Pers., Hexagona Poll. etc.). Lenzites, with lamelloid hymenophore, was extracted from Polyporei and placed in the Agarici close to Pleurotus and allied genera despite of obvious natural affinities with Daedalei.

Later other morphological characteristics were considered relevant for defining new genera from the classical Trametes. For example, Quélet (1886) considered the presence of a tomentum on the abhymenial surface as a distinctive feature for Coriolus.

From this Friesian tradition the type of hymenophore, an easily observable and striking character, was considered the main distinctive feature at the generic level within the polypores. Pilát (in Kavina and Pilát 1936) first doubted its importance and considered hymenophoral morphology to be devoid of real systematic value. Thus, the genus Trametes sensu Pilát encompasses poroid, daedaleoid as well as lamelloid species and genera such as Lenzites or Daedalea, (e.g. T. betulina (L.: Fr.) Pilát; T. quercina (L.: Fr.) Pilát).

On the basis of the context pigmentation, Coriolopsis Murrill 1905 (based on Trametes occidentalis (Klotzsch) Fr., now Coriolopsis polyzona) and Pycnoporus P. Karsten 1881 (based on Trametes cinnabarina (Jacq. : Fr.) Fr.) were respectively created to distinguish trametoid specimens with brown or cinnabarin red color.

Considering as many genera as available, Patouillard (1900) recognized their affinities in his “série des Trametes”, in which he gathered poroid, daedaleoid as well as lamelloid genera. Considering a new character, the mitism of the context, Kotlaba and Pouzar (1957) restricted the genus Trametes to species with a trimitic hyphal system, but like Patouillard they gather in a same “Trametes-group” all genera with di- or trimitic hyphal system and colorless, smooth and inamyloid spores such as Cerrena, Daedalea, Hexagona, Pycnoporus, Trametes etc., whatever the aspect of hymenophore. At last the significance of the wood-rotting types (brown-rot versus white-rot types) was revealed by Nobles (1958) as a distinguishing feature between genera in the Polyporaceae. Thus, the white-rotting abilities become a new feature for the Trametes-group, excluding Daedalea, which causes a brown rot.

Once these characters were identified, controversies developed in their respective importance for generic delimitation. Corner (1989) weakened the value of rot-type, hymenophore configuration and context- colour, and came back to Kavina and Pilát’s (1936) enlarged Trametes concept. Indeed, he gathered all taxa showing trimitic hyphal systems with exception of three genera (Daedalea, Lenzites and Trichaptum) because of the presence of cystidiform ends of the hymenial binding hyphae. In contrast Ryvarden (1991), in a Trametes-group inspired from Kotlaba and Pouzar’s (1957) concept, accepted all white-rot genera such as Coriolopsis and Pycnoporus, with colored hyphal pigments, Lenzites with distinct pointed hyphal ends in the catahymenium and hymenial lamellate surface, and 16 others based on narrow combinations of all the above mentioned characters (Ko and Jung 1999). In addition to the ability to produce a white-rot, all of these genera are characterized by di-trimitic hyphal system, clamped generative hyphae, hyaline, thin-walled, mostly cylindrical, smooth and non amyloid spores with no true hymenial cystidia.

The first molecular analysis on Trametes and related genera, by Hibbett and Donoghue (1995), and Ko and Jung (1999), contributed significantly to understand the phylogenetic structure of the family Polyporaceae, based on mitochondrial small subunit ribosomal DNA. Trimitism and white-rotting were confirmed as common features for all genera in a Trametes-clade within the “core Polyporaceae group”, which matched Ryvarden’s arrangement with only a few exceptions such as Trichaptum, which is related to the Hymenochaetaceae (Hibbett and Donoghue 1995; Ko and Jung 1999). An extensive work by Ko (2000) based on mt SSU rDNA and ITS sequences divided the core Polyporaceae group into 2 subgroups: the first (“A”) which gathers Cryptoporus, Daedaleopsis, Datronia, Funalia (including “Coriolopsis” gallica and “Trametella” trogii), Ganoderma, Lentinus, Microporus, Polyporus and the second (“B”) which gathers Coriolopsis (C. polyzona only), Lenzites, Pycnoporus and Trametes. Recently, Rajchenberg (2011) suggested a morphological and cytological support for a Lenzites-Coriolopsis-Pycnoporus-Trametes group (‘subgroup B’ of Ko 2000) on the basis of a normal nuclear behavior, tetrapolarity, white rot and trimitic hyphal system, consistent with the phylogenetic results described above. Moreover, heterocytic nuclear behavior with bipolar mating system separates Funalia and Cerrena from Trametes and Coriolopsis (David 1967). Although Tomšovský et al. (2006) already recognized a “main Trametes-clade” for a small group of tomentose species better matching the genus Coriolus, the question whether narrowly related genera in the ‘subgroup A’ (Ko 2000), such as Coriolopsis, Coriolus, Lenzites, Pycnoporus, should be recognized as independent monophyletic genera or included in an enlarged genus Trametes remains open. A more detailed analysis was required, taking into account more taxa (especially tropical), for defining a robust generic concept in coherence with morphological, chemical and ecological features.

On the basis of previous molecular data and taking into account all the above considerations and partial discrepancies or still questionable points, our study aimed: i) to examine the phylogenetic placement of species belonging to core Trametes, and to the related Coriolopsis, Lenzites, and Pycnoporus, with a focus on tropical taxa, and ii) to evaluate the taxonomic significance of the morphological characters traditionally used to discriminate the four genera. For these purposes 31 species including 16 tropical taxa were included in our molecular and morphological study. Phylogenetic analyses were performed using sequence data from three nuclear ribosomal regions (internal transcribed spacers ITS1 and ITS2 and 5,8 S gene) and the protein-coding gene RPB2. An analysis of 41 NCBI nuc-ribosomal 28 s LSU sequences is also provided.

Materials & methods

Material studied

A cluster of 50 dikaryotic isolates was used for DNA analyses: taxa and strains studied along with geographical origin and herbarium number are listed in Table 1. Twenty-nine strains were isolated from fresh basidiomes collected in Europe, French Guiana, and French West Indies (Guadeloupe and Martinique) between 2007 and 2010. They are deposited at the Banque de Ressources Fongiques de Marseille (BRFM) belonging to the Centre International de Ressources Microbiennes - Champignons Filamenteux (CIRM-CF). The source exsiccates were deposited at the herbarium LIP (Lille). Twenty-one additional strains were obtained from the culture collections at CBS (Baarn, NL), MUCL (Louvain-la-Neuve, B), and CIRM-CF (Marseille, F). Daedaleopsis tricolor, Hexagonia nitida, H. mimetes and Trametella trogii were used as outgroups (Ko and Jung 1999; Tomšovský et al. 2006).

Table 1 List of Taxa and strains and Genbank accession numbers for RPB2 and ITS
Full size table

Sampling was enlarged with 6 sequences retrieved from GenBank: Trametes elegans JV021237J, T. aff. .junipericola AY684171, T. lactinea GQ982887 and Damm 4703, T. maxima AB158315 and Daedalea microsticta FJ403209 (Table 1).

In addition, 41 nuc-ribosomal 28 s LSU sequences were downloaded from Genbank and were analyzed separately (Table 2).

Table 2 List of Taxa and Genbank accession numbers for nucLSU
Full size table

Collection description

The 29 collections of basidiomes and 2 specimens loaned from MUCL, corresponding to the strains MUCL 38443 Funalia polyzona and MUCL 38649 Trametes socotrana, were described on the basis of macro- and micro- morphological features. Fresh specimens were photographed then air dried. Microscopic features were observed on a Zeiss Axioscop light microscope. All observed elements and structures were described and hand-drawn from radial sections of exsiccata examined in Melzer’s reagent (iodine 0,5 g, potassium iodine 1,5 g, hydrated chloral 20 g, for 22 cm3 of water), 1% Congo red in 10% aqueous ammonium hydroxide and 5% aqueous potassium hydroxide solution (abb. KOH).

DNA extraction, PCR and sequencing

Strains were grown on Malt Agar medium (2% malt extract, 2% Bacto-agar DIFCO) at 25 ° C for 1 week. Genomic DNA was isolated from mycelial powder (40–80 mg) as described by Lomascolo et al. (2002). The primers bRPB2-6 F, bRPB2-7.1R (Matheny 2005), and ITS1, ITS4 primers (White et al. 1990) were used for PCR amplification and sequencing reaction. The ITS1-5.8S rRNA gene-ITS2 and RPB2 were amplified from 50 ng genomic DNA in 50 μl PCR reagent containing 1.5 U Expand™ High Fidelity PCR systems (Roche, France), with a protocol adapted from Lomascolo et al. (2002). Annealing temperatures and extension times were respectively 51°C and 1 min for ITS1/ITS4 amplification; 53°C and 1 min for RPB2 amplification. The PCR products were sequenced by GATC Biotech AG (Konstanz, Germany) or Cogenics (Meylan, France). All the nucleotide sequences were deposited in GenBank under the accession numbers given in Table 1. An additional gene (β-tubulin) was sequenced from a selection of the same strains but phylogenetic analysis gave a weak resolution and is not presented here.

Sequence alignments

Sequences generated in this study and those obtained from GenBank were aligned under Clustal W (Higgins et al. 1994). They were carefully refined by eye on the editor in Mega 4.0 (Tamura et al. 2007). Several insertions in the ITS sequence of Pycnoporus puniceus, and another in the RPB2 sequences of several species in the Trametes-clade (see Discussion) were discarded before analyses.

Phylogenetic analysis

Two methods of phylogenetic analysis were applied i.e. Maximum Likelihood (ML) and Bayesian. ML analysis was performed on the Phylogeny.fr platform with the following parameters : 1) the phylogenetic tree was reconstructed using the maximum likelihood method implemented in the PhyML program (v3.0 aLRT), 2) the default substitution model was selected assuming an estimated proportion of invariant sites (of 0.474) and 4 gamma-distributed rate categories to account for rate heterogeneity across sites, 3) the gamma shape parameter was estimated directly from the data (gamma = 0.470), 4) reliability for internal branch was assessed using the ML bootstrapping method (500 ML bootstrap replicates), 5) transition weighted four times over transversion and log likelihood = −9403,75196. Estimated base frequencies were: f(A) = 0.22636, f(C) = 0.269792, f(G) = 0.26798 and f(T) = 0.23773. Sequence file: phymlla96ToTm4/input.phy.

Bayesian analyses were monitored by software Mr Bayes v3.1 (Ronquist and Huelsenbeck 2003). According to the Bayesian Information Criterium (BIC) score, SYM + G + I and K80 + G (K2P; Kimura 1980) were chosen respectively for combined (ITS + RPB2) and 28S sequences analyses as the optimal substitution model defined by TOPALi v2.5 (Milne et al. 2004). Bayesian analyses were conducted using four Metropolis-coupled Markov chain Monte Carlo (MCMC) with one tree sampled per 100th. The first 5000 trees were excluded of our analyses. For the both Bayesian analysis, potential scale reduction factors (PSRF) were reasonably close to 1.0 for all parameters. Bayesian Posterior Probabilities (Bayesian PP) of each node were obtained with majority rules with all compatible partitions. Whatever the method, gaps were scored as missing and trees were rooted by Midpoint rooting application.

Selection of outgroups

Initial analyses based on ITS sequences (not shown here) confirmed that several species fell outside of the core genus Trametes and of the related genera. Among these, Hexagonia nitida, Daedaleopsis tricolor and Trametella trogii (syn. Funalia trogii; for a comparison between Funalia and Trametella especially based on polarity: see (Pieri and Rivoire 2007) were selected as outgroups since all were shown to belong to the sister “subclade A” of Ko (2000). A strain identified as Trametes mimetes was found from our preliminary analysis to be closely related to Hexagonia nitida, as suggested earlier by Reid (1975), therefore the name Hexagonia mimetes (Wakef.) D.A.Reid is retained here assuming a correct identification of the strain (voucher specimen not seen). This species had not been included in previous phylogenetic works (e.g. Tomšovský et al. 2006), The corresponding sequences were also used as outgroups.

Results of the phylogenetic analysis

Morphological analysis

All 31 collections have been observed, including the type material of Lenzites acutus, Trametes cingulata, T. lactinea, T. menziesii, T. ochroflava, T. sclerodepsis and T. subectypus, in order to confirm field identifications. European specimens from Marcel Bon’s collections (LIP herbarium) were studied to complete morphological observations. For each collection the hymenophoral trama, hymenium, spores, pileus, structure of context, and structure on radial cuts were analyzed. Following various keys of neotropical species of Trametes (Ryvarden et al. 2009; Gomes-Silva et al. 2010; Læssøe and Ryvarden 2010) the KOH reaction was systematically investigated on abhymenial and hymenial surfaces of basidiomes (dry and also fresh specimens when possible).

Morphological analysis of 31 collections for which culture was successful resulted in the identification of 20 species, 10 being strictly tropical taxa (‘Coriolopsis’ polyzona, Pycnoporus sanguineus, ‘Trametes’ elegans, T. lactinea, T. maxima, T. menziesii, T. socotrana and T. villosa (Table 1). Two species collected repeatedly in French Guiana remain unidentified: one showed morphological characters close to those of the paleotropical species T. meyenii (here called ‘Trametes aff. meyenii’: GUY 08-152 and GUY 10-36, LIP), the other could not be compared to any well-defined species (here called ‘Leiotrametes sp.’: GUY 08-20, GUY 08-225, GUY 08-167 and GUY 08-156, LIP).

ITS + RPB2 combined analysis

Compared to separate gene analyses, the combination of ITS and RPB2 sequences produced the best resolved phylogeny and the highest number of strongly supported clades. A combined sequence dataset was thus constructed for 41 strains of Trametes and allied genera (24 being tropical areas, the others from Western Europe). The Bayesian 50% majority rule consensus tree is shown, in which 27 clades receive more than 95% Bayesian PP and 20 received more than 70% ML bootstrap support (Fig. 1). The ML analysis (not shown) was very similar in topology as the Bayesian analysis but differed by a lack of basal resolution for the main clades and revealed no more information.

Fig. 1
figure 1

Phylogenetic reconstruction of the Trametes-clade based on the combined analysis of ITS1-5.8S-ITS and RPB2 (50% majority rule consensus tree). Interpretative features are figured on the right part of the figure: Pil = Pileus structure (letters a-g refer to type structures in Fig. 4); Ha = presence (+ to +/- if disappearing with age) or absence (°) of hairs (tomentum) on pileus; Pig: presence (+) or absence (°) of incrusting pigment (see Fig. 4); K = reaction to 5% KOH (°: none; +: brown; ++: black; p = only on pileipellis); St = presence (+) or absence (°) of a pseudostipe; Hy = morphology of hymenophore (P = poroid, Fig. 5d–f; D = daedaloid, Fig. 5a,c; L = lenzitoid, Fig. 5b right; d = with protruding dissepiments); BL = presence (+) or absence (°) of a “black line” under pileipellis

Full size image

ITS and RPB2 sequences have an alignment of 594 and 697 bp, respectively, including gaps. After removing poorly aligned positions and divergent regions of DNA, ITS and RPB2 sequences had respectively an alignment of 532 bp with 178 variable regions and 131 parsimony informative characters, and 644 bp with 284 variable regions and 254 parsimony informative characters. 5.9% of gaps were maintained in the final combined alignment but were scored as missing.

As shown in Fig. 1, topology of the Bayesian tree is composed of three highly supported clades:

  1. 1)

    A strongly supported (Bayesian PP = 1; ML bootstrap = 100%) group of specimens that were identified as Lenzites elegans sensu Ryvarden and Johansen (1980) (French Guiana, French West Indies, New Caledonia and Cuba).

  2. 2)

    A clade (Bayesian PP = 0.92) of a groups specimens with glabrous upper surface. It comprises three distinct sub-clades:

    • Pycnoporus forms a strongly supported monophyletic group (Bayesian PP = 0.98; ML bootstrap = 0.78);

    • Sister sub-clade of Pycnoporus, moderately supported (Bayesian PP = 0.60), comprising two close species of unclear systematic position: Trametes ljubarskyi (France) and T. cingulata (Southern Africa);

    • Third sub-clade, strongly supported, comprising 3 tropical species, T. menziesii, T. lactinea and an unidentified Guianese species that shows hymenial surface evolving from pored to more or less lamellate pattern while ageing (Bayesian PP = 1; ML bootstrap = 100%).

  3. 3)

    Third clade (Bayesian PP = 0.86) comprising a group of specimens with pubescent to hirsute upper surface. Three distinct sub-clades are identified within this clade:

    • Firstly a strongly supported sub-clade comprising genuine Trametes species (i.e. with strictly poroid hymenophore): Trametes versicolor, T. hirsuta, T. ochracea, T. suaveolens, a chinese species close to T. junipericola, T. socotrana, T. pubescens and T. villosa (Bayesian PP = 1; ML bootstrap = 92%). Most of them excepting T. socotrana and T. villosa are from temperate areas.

    • Second sub-clade formed by a species with radially elongated pore surface (T. gibbosa), a lenzitoid species (‘Lenzites’ betulinus) and a strictly pored tropical species (Coriolopsis polyzona); the position of C. polyzona relative to the T. gibbosa-L. betulinus group is weakly supported (Bayesian PP = 0,58)

    • Third strongly supported (Bayesian PP = 1; ML bootstrap = 0.92) sub-clade grouping 3 tropical species with intermediate hymenophore configuration, Trametes maxima, T. meyenii, and a Guianese species morphologically close to T. meyenii.

  4. 4)

    ‘Lenzites’ warnieri’ comes out as a single branch at the same phylogenetic level as the three main above-mentioned clades.

RBP2 analysis

The alignment of RPB2 sequences revealed an interesting insertion area for some species (Fig. 2): most species of Trametes s.str. (T. maxima, T. meyenii, T. ochracea, T. pubescens, T. versicolor) have a 15-nucleotide long insertion (21-nucleotide long in T. ochracea BRFM632), all of rather similar composition. Trametes gibbosa and ‘Lenzites’ betulinus show a much longer insertion, 51- and 69-nucleotide long respectively. This insertion (not included in the phylogenetic analysis) supports the inclusion of Trametes meyenii and T. maxima in the core Trametes group (a hypothesis already well-supported in the phylogenetic analysis, see above), and a further argument for considering the Trametes gibbosa-Lenzites betulinus clade as closely related. Interestingly enough this insertion is absent from all other lineages and suggests a basal origin of the “third clade” with an internal fast evolution; it might have disappeared in some derived lineages such as Trametes suaveolens or Coriolopsis polyzona, the alternative hypothesis (a multiple origin of this insertion) from an evolutionary point of view being less parsimonious.

Fig. 2
figure 2

Distribution and composition of insert in RPB2 sequences in the Trametes clade; species are disposed according to the ITS + RPB2 phylogeny in Fig. 1

Full size image

28S rLSU analysis

In order to obtain additional information, a 28S rLSU analysis was processed, independently from the former, by using sequences downloaded from GenBank (Fig. 3). A group of 41 reliable sequences of Trametes and allied taxa (incl. 8 tropical species) was considered (Table 2). Most of them have been previously published by Tomšovský et al. (2006), whose species concepts match those adopted here. No rLSU sequence of Lenzites warnieri or T. cingulata is available in public databases. Laetiporus sulphureus, Trametella trogii and T. (Coriolopsis) gallica were used as outgroups (Tomšovský et al. 2006).

Fig. 3
figure 3

Phylogenetic reconstruction of the Trametes-group based on Bayesian analysis of rLSU (50% majority-rule consensus tree). Only the Pycnoporus/Leiotrametes clade including “Trametes” ljubarskyi shows a significant support compared to the ITS + RPB2 phylogeny (Fig. 1)

Full size image

This single-gene analysis using Bayesian methods gives a weak basal support, which does not contribute to a better definition of the clades defined with ITS + RPB2. Nevertheless a good support (Bayesian PP = 0.94) is given to the “second clade” of the former analysis, including Pycnoporus and the Trametes lactinea-group. The displacement of Coriolopsis polyzona, Lenzites betulinus and Trametes elegans e.g., compared to the former analysis, is not supported and cannot be considered as consistent. It is assumed that the 28S rLSU sequences are not pertinent for reconstructing the phylogeny of the Trametes-clade, although easily aligned. The necessity of choosing a very distant outgroup (Laetiporus sulphureus) in order to get a better ML bootstrapping suggests that the resolution power of rLSU is insufficient with the currently available data, as it is for the other gene studied by us (β-tubulin, data not shown). More taxa might partly improve this analysis.

Discussion and new systematic arrangement of the Trametes-clade

General systematics in the Trametes-group

As expected, the close relationships between the genera Pycnoporus, Lenzites, Coriolopsis and Trametes, as previously described by Ko (2000), Garcia-Sandoval et al. (2011) and Rajchenberg (2011) were confirmed. Species such as Hexagonia nitida, Daedaleopsis tricolor, Trametella trogii with binucleate spores and heterocytic nuclear behavior, previously located in a sister clade position (Ko and Jung 1999; Tomšovský et al. 2006), and the newly positioned Hexagonia (Trametes) mimetes, represented convenient outgroups for our study.

However, relationships within the subgroup “B” Trametes-Lenzites-Pycnoporus-Coriolopsis (Ko 2000) of the core polyporoid group remained uncertain. Morphological features defining these four genera such as lamellate or pored hymenophore and colour of the hyphae have not yet proved their worth at the generic level.

By addition of more tropical and rare temperate taxa, such a configuration is no more fully supported by our phylogenetic results, and three (ITS + RPB2 analysis, Fig. 1) well-supported monophyletic lineages can be identified, with some still uncertainly placed outstanding taxa such as Lenzites warnieri for which some molecular data are missing.

Although the basal resolution of the three main clades (1, 2, 3) remains relatively weak, whatever the data sets and analyses, each of them received a good support by the concatenate analysis as well as by the macro- and microcharacters (Fig. 1).

At this stage two possibilities can be considered according to such results: either recognizing an unique genus Trametes, enlarged to encompass the three traditional genera cited above; or, as far as some monophyletic clades can be supported by morphological features, split this clade into different genera, each of them defined by a thorough combination of characters. Morphology supplies strong information where molecular phylogenies provide weak support, and helped us propose a better systematic arrangement. Therefore, we propose separation and delimitation of four distinct genera in the Trametes group (Fig. 1; Table 3):

  1. 1)

    Trametes, corresponding to the species with pubescent to hirsute upper surface, including most temperate species fitting the traditional definition of the genus, in addition to ‘Lenzites’ betulinus and ‘Coriolopsis’ polyzona;

  2. 2)

    Pycnoporus to include species with red basidiomes, blackening with KOH;

  3. 3)

    Artolenzites to include the tropical ‘Lenzites’ elegans;

  4. 4)

    Leiotrametes gen. nov., comprising three tropical species: ‘Trametes’ menziesii, T. lactinea, ‘Leiotrametes sp.’

Table 3 Morphologic characteristics of genera and species groups in the Trametes-group
Full size table

This classification is nevertheless incomplete, since some critical taxa from various tropical parts of the world were not accessible to us and might either add new lineages to the system, or illustrate more continuities between some of the proposed divisions. In the same way two still unplaced lineages not included in previous analyses: ‘Lenzites’ warneri and the ‘Trametes’ ljubarskyi-T. cingulata group, cannot reasonably justify new genera according to their uncertain position in our analyses, nor can they be included in Trametes s.s. because of outstanding morphological features, and will deserve further studies. There are here provisionally maintained in their traditional genera.

Morphological characters in the four branches within the Trametes clade

Structure of upper surface

Aspect and structure of the abhymenial surface is a discriminating morphological feature of major importance at the generic level in the core polyporoid clade, as already shown in Ganoderma (Steyaert 1980; Gottlieb et al. 1999; Moncalvo 2000; Welti and Courtecuisse 2010). In the Trametes group differences in pileus-structure (glabrous or tomentose) have already been described for each species studied here and are considered by Læssøe and Ryvarden (2010) as an essential feature for species recognition; they nevertheless never been used for phylogenetic interpretation. Taking our phylogenetic results, fundamental differences in structure (Fig. 4) and consequently in macroscopic aspect of the basidiome surface, explain the evolutionary history of the groups. Differentiation of hairs (pileus tomentum) is a synapomorphy of our redefined genus Trametes (Fig. 4a–c), without any known exception, although some species are only minutely pubescent when young and become somewhat glabrous whilst ageing (T. gibbosa, T. ochracea, T. suaveolens).

Fig. 4
figure 4

Pileus structures in Trametes and allied species. a: trichoderm with differentiated subpellis, with incrustations (Trametes versicolor); b: idem, without incrustations (T. villosa); c: trichoderm without differentiated subpellis (T. gibbosa); d: intermixed structure without incrustations (Artolenzites elegans); e: idem, with brown pigment in skeletal hyphae (Leiotrametes lactinea); f: idem, with incrustations at hyphal apex (Pycnoporus cinnabarinus); g: idem, with brown intracellular pigment and resinoid matrix (‘Trametes’ ljubarskyi)

Full size image

The most superficial layer in strictly glabrous species shows other architecture specificities (Fig. 4d–g): Thus in Artolenzites (Fig. 4d) and Pycnoporus (Fig. 4f) the pileipellis is made of a single cutis composed of a +/- gelatinized layer of undifferentiated hyphae, whilst in Leiotrametes and Lenzites warnieri (Fig. 4e) superficial hyphae are thick-walled and filled with brown, resinous material. In Trametes ljubarskyi (Fig. 4g) the same kind of hyphae are overlapped by a 150–200 μm thick layer of colourless +/- resinous or mucilaginous substance soluble in KOH. In Trametes cingulata the brownish resinous layer from the accumulation of amorphous resinous material from damaged hyphae reminds one of the upper surface of the laccate Ganoderma species but lacks clavate pileocystidia.

All glabrous species have a dull superficial aspect, except T. ljubarskyi and T. cingulata which have a glossy surface due to the upper resinous layer.

Differentiation of subpellis (“black line”)

The hairy-tomentose species Trametes betulina, T. maxima, T. meyenii, and T. versicolor – and often also T. hirsuta – typically differentiate a dark subpellis (“black line” or BL). When observed under the light microscope, the BL is very refractive and consists of a dense layer of radially arranged hyphae embedded in a mucus partly dissolving in 5% KOH. In Trametes species where the BL is not apparent this structure is not (T. gibbosa, T. suaveolens) or only weakly (T. polyzona, T. socotrana, T. villosa) developed. Contrary to Ryvarden (1991) and Tomšovský et al. (2006) who consider the BL as a characteristic of the whole “Coriolus-subclade” (our core Trametes clade) we failed to systematically observe it in T. hirsuta and never in T. gibbosa, T. ochracea, T. pubescens, or T. polyzona. Thus the BL is not a synapomorphic feature in Trametes and does not support the distinction of a genus or subgenus (such as Coriolus) based on this character (Ryvarden 1991).

Such a differentiated subpellis is absent in glabrous species of the Trametes clade (Pycnoporus, Leiotrametes, Artolenzites, L. warnieri, T. ljubarskyi, T. cingulata). In the same way Trametes species without differentiated subpellis (especially T. gibbosa and T. suaveolens) tend to soon become glabrous whilst ageing.

Parietal crystal pigment

Red to orange parietal crystals located along skeletal hyphae, especially those quite close to the upper surface and hymenophore, is the main feature differentiating Pycnoporus species from those belonging in the genus Leiotrametes and more generally from the glabrous members of the Trametes group, where we never found the pigment. Although these crystals are very quickly soluble in 5% KOH and must be searched for carefully, such a feature is so far relatively significant to justify monophyly of the genus Pycnoporus. However, some species outside this genus also showed relatively similar crystals located at the upper surface level: colorless and turning black with 5% KOH for T. cingulata or blue and incompletely soluble in 5% KOH for T. versicolor.

Hymenophore

Despite its importance in traditional systematics, the phylogenetic analysis does not support a classification based on the type of hymenophore at generic level. All genera (Artolenzites, Trametes, and Leiotrametes) except the exclusively pored Pycnoporus contain some species with lamellate hymenophore. Although the type of hymenophore is usually stable at species level (Fig. 5), its structure is variable within the tropical Artolenzites elegans and even more in Leiotrametes sp. (Fig. 5a–b) according to the specimen (mainly daedalean, mainly lamellate, or a mixed pattern).

Fig. 5
figure 5

Types of hymenophores of Trametes and allied species. a: daedaleoid (Artolenzites elegans); b: poroid (left), daedaleoid (middle) and lenzitoid (right), in three sporocarps of “Leiotrametes sp.”; c: secondarily daedaleoid (L. menziesii); d: poroid with protruding dissepiments (Trametes villosa); e: poroid with angular pores (T. polyzona); f: poroid with round pores (Leiotrametes lactinea). Pictures of S. Welti (b,f), R. Courtecuisse (c,d), P.-A. Moreau (a,e)

Full size image

The origin of daedalean or heteromerous (mixture of rounded and elongate pores) hymenophore seems to species-correlated. On comparing the aspect of mature specimens of T. gibbosa the pores elongate irregularly from the origin. In contrast in L. menziesii young specimens show regular pores, of which only radial dissepiments develop with age to give a secondarily false daedalean or somewhat lenzitoid structure, with the primary septa still visible in the bottom of the alveoli (Fig. 5c). Such development may be correlated to the inclination of the basidiomes on its substrate. When dimidiate and horizontally growing the hymenial surface remains pored, but when growing oblique or erect the continuous geotropic growth of the dissepiments from a regularly pored ground yields an irpicoid (T. maxima or T. villosa; Fig. 5d) or more or less lenzitoid (L. menziesii) aspect.

Presence of a pseudostipe

A distinct and sterile base clearly delimited from the hymenophore, mostly attached to the substrate with a disc is found in various species: Leiotrametes menziesii, the Guianese Leiotrametes sp., Artolenzites elegans and Pycnoporus sanguineus. All species of Trametes known to us are sessile, as well as Leiotrametes lactinea, Lenzites warnieri and T. ljubarskyi (T. cingulata having a contracted basal attachment). Despite great morphological variability within the Trametes group, this character is very stable in all studied collections of the above mentioned taxa.

KOH reaction

Basidiomes were tested in both fresh and dry conditions with 5% KOH, on pileus, context and hymenophore. All species of Pycnoporus showed an immediate black reaction on all surfaces, in addition to T. cingulata (Table 3). The other species studied never showed such significant reactions (either nil to yellow or brownish and never in all parts of the basidiomes). Nevertheless, the upper surface in species belonging to the new genus Leiotrametes turned deep brown or even almost black with 5% KOH, but the colour of the context did not show a strong reactivity and remained pale yellow. Indeed, this KOH reaction was already used to distinguish Leiotrametes lactinea (turning to deep brown) from ‘Trametes’ modesta or T. supermodesta (becoming red to brownish) by Gomes-Silva et al. (2010).

Biogeography Leiotrametes and Artolenzites are common in all tropical areas, some species, such as L. lactinea and A. elegans being apparently pantropical (Neotropics and New Caledonia). Nevertheless L. lactinea has been recently collected by Vlasák and Kout (2011) in Eastern USA (especially Florida) and interpreted as a recent colonization. According to Gilbertson and Ryvarden (1987), A. elegans is common in South Eastern USA. However, since Vlasák and Kout (2011) “were able to find only one specimen of this species in ten year”, such a statement could result from a misidentification with either L. lactinea or T. gibbosa the intr0oductions of which could possibly be recent in the North American continent. Leiotrametes menziesii (= T. menziesii) is so far known from Paleotropical area (Ryvarden and Johansen 1980; Corner 1989) and is reported here from the Neotropics for the first time.

Trametes and Pycnoporus are more widely distributed. Some species are commonly found in Northern temperate or Mediterranean areas, but they also include common tropical species such as T. maxima, T. meyenii, T. villosa, P. sanguineus or P. puniceus. Finally Lenzites warnieri and Trametes ljubarskyi are mainly Mediterranean species.

Taxonomy

Genus Trametes Fr., Fl. Scand.: 339 (1836), emend.

Synonyms : Lenzites Fr., Fl. Scand. : 339 (1836); Coriolus Quél., Enchir. Fung.: 175 (1886); Coriolopsis Murrill, Bull. Torrey Bot. Club 32: 358 (1905).

Type species : Trametes suaveolens Fr. (Murrill 1905).

Species studied: T. betulina (L.: Fr.) Pilát (lectotype of Lenzites), T. gibbosa (Pers.: Fr.) Fr., T. hirsuta (Wulfen: Fr.) Pilát (lectotype of Coriolus), T. junipericola Manjón et al., T. maxima (Mont.) David & Rajchenberg, T. meyenii (Klotzsch) Lloyd, T. ochracea (Pers.: Fr.) Gilbertson & Ryvarden, T. polyzona (Pers.: Fr.) Corner (holotype of Coriolopsis), T. pubescens (Schum.: Fr.) Pilát, T. socotrana Cooke, T. suaveolens (L.: Fr.) Fr., T. versicolor (L.: Fr.) Lloyd and T. villosa (Swartz: Fr.) Kreisel.

Observations: The main feature which could characterize this genus is certainly the pubescent to hirsute upper surface of the pileus in all species (Fig. 4a–c). Although T. suaveolens, T. ochracea and T. gibbosa are characterized by a glabrescent abhymenial surface, they are in fact tomentose at early stages of their development (Fig. 4c). In contrast, species of Artolenzites, Leiotrametes, Pycnoporus, as well as ‘Lenzites’ warnieri and the ‘Trametes’ cingulata-T. ljubarskyi group, all excluded from Trametes in this study, are always glabrous, and the hyphae located at the far edge of the upper surface are bent or adpressed and never protruding (Fig. 4d–h).

As defined here, Trametes encompasses species with various types of hymenophore: typical from circular or angular pores (T. versicolor complex; Ko 2000; Fig. 5d–e) to also radially elongated to lamellate (T. gibbosa - T. betulina group; Tomšovský et al. 2006) or daedaleoid pores (T. maxima and T. meyenii, formerly classified in Cerrena by Hansen 1960 and Sclerodepsis by Ryvarden 1972). These results confirm that hymenophoral structures, although conspicuous and on which traditional systematics was mainly based (Fries 1835; Ryvarden 1991), is of low taxonomic value at generic level. However it represents a relevant morphological character for species delimitation. Moreover, except T. polyzona with strictly poroid hymenial surface, which moderately clusters (Bayesian PP = 0,58; Fig. 1) with T. betulina and T. gibbosa, each type of hymenial surface corresponds to a monophyletic subclade of Trametes.

The Black line is frequent in Trametes but has no taxonomic value at subgeneric level, as it can be found in various subclades (Figs. 1, 4a–b) and shows no correlation with hymenophoral structures. In the T. meyenii subclade all species analyzed herein show a black line. However an ITS sequence of Daedalea microsticta deposited in Genbank clusters with T. meyenii and T. maxima (data not shown); for Ryvarden et al. (2009) Daedalea microsticta is a synonym of T. ochroflava, whose type specimen is glabrous, strictly pored and without black line (personal observation). More precision on this still confused group of species is required.

Trametes polyzona, a species with brown context, was encorporated into Trametes by the mttSSU and ITS rDNA analyses of Ko (2000), who also established a close relationship between T. polyzona, T. gibbosa, T. hirsuta and also T. meyenii (Ko and Jung 1999; Garcia-Sandoval et al. 2011). Consequently the brown color of the skeletal hyphae is not significant in excluding T. polyzona from the genus Trametes we propose. Morphological similarities between T. hirsuta, T. betulina, T. socotrana, T. villosa, T. maxima and T. polyzona, are especially significant regarding the upper surface with hirsute hairs along narrow sulcate zones (Gilbertson and Ryvarden 1987; Ryvarden and Gilbertson 1994). Finally, the effused-reflexed basidiome of T. polyzona is another characteristic of the genus Trametes, in contrast to the other clades mostly characterized by pseudostipe or contracted basis (Fig. 1).

Once compared morphological characters with phylogenetical results, we can deduce that the major characteristic distinguishing Trametes from the other genera of the core Trametes-clade is the pilose upper surface. As quoted above, Quélet (1886) was the first to emphasize the significance of this feature by separating the genus Coriolus from Trametes which was restricted to glabrous species (or supposed so, erroneously for T. gibbosa and T. suaveolens). However, as soon as Trametes suaveolens (type species of the genus Trametes, unless one of its less representative members) is considered congeneric with the type species of Coriolus (C. versicolor), the genus Trametes in this clade, Trametes keep priority on Coriolus.

Genus Pycnoporus P. Karst., Rev. Mycol. (Toulouse) 3(9):18 (1881)

Type species: Polyporus cinnabarinus Jacq.:Fr.

Species studied: Pycnoporus cinnabarinus (Jacq. : Fr.) P. Karsten, and, P. sanguineus (L.: Fr.) Murrill.

Observations: In a large phylogenic study of Pycnoporus, Lesage-Meessen et al. (2011) clearly separated four species of Pycnoporus and defined the genetic intraspecific variability of each according to geographic distribution.

Monophyly of this genus is strongly supported by both of the phylogenetic methods (Bayesian PP = 0,98; ML bootstrap = 78%). This is correlated with the presence of red, extracellular pigments soluble in 5% KOH, a relevant morphological character at generic level (Fig. 4f). In addition, black KOH reaction on all parts of the basidiomes clearly separates Pycnoporus from Trametes (Ryvarden and Johansen 1980)

Genus Leiotrametes Welti & Courtec., gen. nov.

Mycobank MB 563399

Basidiomata lignatilia, annua vel perennia, coriacea, sessilia vel pseudostipitata nonnunquam basi discoidea, dimidiata usque ad fere circularia; contextus albidus usque ad cremeum, homogeneus; superficies hymenialis porata ad aspectum labyrinthiforme vel lenzitoideum vertens sive ex incremento radiali dissepimentorum sive ex porrectione radiali pororum; superficies superior semper glabra, zonis concentricis angustis interdum tantum marginalibus; frequens proventus excrescentiarum verrucosarum in basi superioris partis pilei. Structura tramae trimitica; hyphae generativae fibulatae; hyphae skeleticae incolores usque ad pallide flavas, aliquot repletae pigmento resinoideo specialiter sub zonis concentricis coloratis pileipellis. Pigmenta parietalia nulla. Basidiosporae cylindratae, incolores, laeves, nec amyloideae nec cyanophilae. Cystidia hymenialia nulla. Saprotropha, in ligno mortuo Angiospermarum; caries alba. Distributio pantropicalis.

Holotypus hic designatus : Polyporus lactineus Berk., Ann. Nat. Hist. 10: 373 (1843)

Species studied: Leiotrametes lactinea (Berk.) Welti & Courtec. comb. nov. (basionym: Polyporus lactineus Berk., Ann. Nat. Hist. 10: 373, 1843; Mycobank MB 563400), L. menziesii (Berk.) Welti & Courtec. comb. nov. (basionym : Polyporus menziesii Berk., Ann. Nat. Hist. 10: 378, 1843; Mycobank MB 563401) & Leiotrametes sp.

Observations: in all our phylogenetic analyses (Figs. 1 and 3) this group of three tropical species separates from all other clades with strong support; the Bayesian analysis includes it in the “second clade” and suggests a sister position to the Pycnoporus + ‘Trametes’ cingulata-T. ljubarksyi lineage.

Leiotrametes lactinea was recently documented from South America by Ryvarden (2000) and from Southern USA by Vlasák and Kout (2011); descriptions of this material match the morphological characteristics of our numerous Guianese and Caribbean collections. Our collections of L. menziesii are the first reported from the Neotropics and their morphological features match those of Polyporus menziesii as described by Ryvarden and Johansen (1980) and our personal observations (isotype – K).

The third species here mentioned as ‘Leiotrametes sp.’ from French Guiana does not match any species known to us nor described in the literature. Nevertheless hymenial surface of this species could evoke the temperate Daedalea quercina (L.: Fr.) Fr., a phylogenetically unrelated species producing a brown rot (also showing other morphological discrepancies). Since Daedalea quercina was mentioned by Patouillard (in Duss 1903) after a collection by Duss in Guadeloupe and taking into account its unlikely occurrence in the Carribean (see Courtecuisse and Welti 2011) it is possible that Duss’s material represents this still undescribed Leiotrametes sp.

The main characteristic separating Leiotrametes from Trametes and Pycnoporus is the glabrous upper surface, the lack of black line under the pileipellis and of parietal crystals (red in Pycnoporus, colorless in T. cingulata and blue in T. versicolor). Another interesting character is the brown resinous substance filling the lumen of the skeletal hyphae in the pileipellis, particularly those concentrated in the narrow grayish concentric zones (Fig. 4e). They were also found in some species of Trametes: T. gibbosa and T. villosa.

A comparable resinous content also appears in T. cingulata and T. ljubarskyi but differs by its conspicuous accumulation in uppermost level inducing cellular walls rupture (Fig. 4g) and so generating a glossy and brown, surface. ‘Lenzites’ warnieri, of still unsolved phylogenetic position, also showed similar resinous hyphae; nevertheless, they appear less abundant in the upper surface level and did not show resinous accumulation at the surface (Fig. 4e).

‘Trametes’ cingulata and ‘Trametes’ ljubarskyi

The position of Trametes cingulata and T. ljubarskyi has already been shown to be ambiguous according to our study. However the Bayesian analyses on ITS + RPB2 (Fig. 1) and to a lesser degree on 28S rLSU, suggest a sister-clade relationship between both species and Pycnoporus. As a support to this hypothesis we detected crystals darkening in 5% KOH under the upper surface of T. cingulata. Furthermore, the orange-brown, dry basidiomes of this species, as well as its tendancy to turn blackish with 5% KOH 5%, at a lower degree the characteristic of Pycnoporus species (red basidiomes and KOH reaction).

So far a close relationship between Trametes ljubarskyi and T. cingulata has never been mentioned, probably because of their distinguishing morphological characteristics (Ryvarden and Johansen 1980; Gilbertson and Ryvarden 1986; Ryvarden and Gilbertson 1994). The Asian and African distribution of T. cingulata versus Europe for T. ljubarskyi; the thickness of the basidiomes: 2–10 mm. for T. cingulata versus 30 mm. for T. ljubarskyi; the pore pattern and dissepiments: round and regular, 4–6 per mm., fairly thick dissepiments for T. cingulata versus circular to angular, 3–4 per mm., thin dissepiments for T. ljubarskyi and strikingly different upper surface: frequently concentrically sulcate and whitish to ochraceous becoming sooty black spreading from the base for T. cingulata versus azonate and whitish to ochraceous becoming pale grayish brown in spots for T. ljubarskyi. Furthermore, according to our own observations, basidiomes of T. ljubarskyi are paler than those of the isotype of T. cingulata which is rather red brown.

Nevertheless these 2 species share several common features: somewhat broadly ellipsoid basidiospores (a very unusual character in this group) with similar sizes strictly pored hymenial surface remaining so during development of the basidiomes and glabrous and somewhat glossy upper surface.

‘Lenzites’ warnieri

As mentioned above, ‘Lenzites’ warnieri creates a unique branch according to the topology of the Bayesian tree. This unresolved phylogenetic position is reflected in the fact that the species possesses many morphological features from other genera, and this ultimately would place L. warnieri in a separate genus.

This Mediterranean species is always glabrous and dull, with strictly lamellate hymenial surface (character in common with T. betulina), without parietal crystals on the hyphae (Artolenzites, Leiotrametes). L. warnieri shows superficial skeletal hyphae filled with a brown resinous content not accumulating at the apex (Fig. 4e) and its abhymenial surface turns deep brown with 5% KOH. These 2 features also characterize species of the genus Leiotrametes.

This supports one of the striking points emphasized in this study: there is no correlation at all between type of hymenial surface and phylogenetic position of a species within the Trametes-group. The lamellate Lenzites warnieri, Artolenzites elegans and T. betulina are not monophyletic and show no close relationship. Lenzites is therefore discarded.

Unfortunately, because of the absence or very weak development of the hymenium in most of our specimens, we cannot rule about the taxonomic significance of the hymenial sword-like pseudo-cystidia previously mentioned for T. betulina, T. gibbosa and L. warnieri (Ryvarden and Gilbertson 1993; Tomšovský et al. 2006). For the same reason, the basidiospores could not be properly analyzed in these species. Nevertheless, while Pieri and Rivoire (2007) revealed that pseudocystidia were not found in T. gibbosa, Ryvarden and Johansen (1980) mentioned the presence of such pseudocystidia in the hymenium of Lenzites acutus (with pored to lamellate hymenial surface) and Trametes cubensis (strictly pored) (Gilbertson and Ryvarden 1987; Ryvarden and Johansen 1980). Thus, further investigations should be undertaken to evaluate the relevance of pseudo-cystidia at generic level.

Although Ko (2000) showed recently on the basis of ITS sequences that Daedaleopsis flavida (Lév.) A. Roy & A. Mitra clustered with Pycnoporus, Ryvarden and Johansen (1980) considered this taxon in the synonymy of L. acutus, a species closely related by several morphologic similarities to L. warnieri (Gilbertson and Ryvarden 1987). Morphologic description (Ryvarden and Johansen 1980) and molecular results of L. acutus remind us of our Guianese species named Leiotrametes sp. but thorough comparison of both species finally reveals no real morphological similarities.

Genus Artolenzites Falck, Hausschwammforsh 3: 37 (1909)

Type species: Daedalea repanda Pers. (= A. elegans (Spreng.: Fr.) Teixeira)

Species studied: Artolenzites elegans (Spreng.: Fr.) Teixeira, Rev. Brasil. Bot. 9(1):43 (1986).

Observations: So far only one species is recognized in this genus, with an abundant synonymy (Ryvarden and Johansen 1980). However, we noted several morphological and genetic differences between our collections from New Caledonia and French West Indies, and do not exclude that the type species of the genus - Daedalea repanda Pers., originally from New Guinea (Gaudichaud-Beaupré 1827) might be different from L. elegans from Guadeloupe (Fries 1821). Further comparisons within this cosmopolitan and polymorphic species are required.

The morphology of specimens in this clade matches those formerly described by Vlasák and Kout (2011) and Ryvarden and Johansen (1980). All basidiomes are white to cream-coloured, glabrous, of large size, spathulate to reniform with acute margin, sometimes with stipe-like base attached to the substrate with a disc. The hymenophore is narrowly daedaleoid to lamellate (Fig. 5a). All possess hyphal pegs.

As already stated above the hymenial surface cannot be considered as a separating character at generic level so that Ryvarden (1991) was right on this very point in considering Artolenzites as a taxonomic synonym of Trametes. However, since molecular results clearly separate T. elegans from the core Trametes, the type of abhymenial surface turns out to present the main feature for distinguishing Artolenzites from Trametes.

Thus, the aspect and structure of the upper surface are much more significant than the hymenial pattern to separate the genera from the Trametes group.

Finally, Artolenzites is distinguished from the other glabrous genera (Pycnoporus, Leiotrametes, ‘Lenzites’ warnieri and the T.cingulata-T. ljubarskyi clade) by lack of both resinous accumulation in the upper surface skeletal hyphae and parietal crystals (Fig. 4d).

Key to genera of the Trametes group (see Table 3)

1. Upper surface pubescent to hirsute...........genus Trametes

1. Upper surface glabrous................................................2

2. Basidiome red, incrusting pigment present as orange-red parietal crystals soluble in 5% KOH ..........genus Pycnoporus

2. Basidiome not red, lacking red incrusting pigment......3

3. No part of basidiome darkening in 5% KOH .....................................................................genus Artolenzites

3. Entire basidiome or at least upper surface darkening to black or dark brown with 5% KOH .....................................4

4. Entire basidiome initially orange-brown becoming black with 5% KOH. Upper surface glossy, hymenophore strictly pored….......................……....Trametes cingulata

4. Only upper surface or context becoming deep brown with 5% KOH. Superficial layer of pileipellis with numerous skeletal hyphae filled with brown resinous contents...........................................................................5

5. Upper surface glossy. Hymenial surface strictly pored, context staining brown with 5% KOH.........Trametes ljubarskyi

5. Upper surface dull. Hymenial surface pored to lamellate, upper surface staining brown with 5% KOH.....................6

6. Temperate to Mediterranean species. Hymenophore strictly lamellate. Basidiome never pseudostipitate, lacking narrow, coloured, concentric zones on the abhymenial surface.......................................................Lenzites warnieri

6. Tropical species. Hymenophore pored or daedalean to lamellate. Basidiome sometimes pseudostipitate, with mostly numerous and narrow grayish or brownish, concentric zones on abhymenial surface...............................genus Leiotrametes