A total of 37 specimens of Trametes were collected in three different macroclimatic zones and different forests of Benin (Fig. 1A, C) from July to September in 2017 (Olou et al. 2019) and in 2018 (another series of surveys). Small pieces of fresh fruit bodies were placed in plastic bags half-filled with silica gel for DNA extraction. The rest of fruit bodies were air- or oven-dried at 45–50 °C for 1–2 days depending on the consistency of the fruit body. The dried fruit bodies were then preserved in plastic bags for morphological investigation. Specimens are deposited at the mycological herbaria of the University of Parakou (UNIPAR; Thiers 2019) and the University of Kassel (KAS).

DNA extraction. Genomic DNA of all specimens classified into nine morphotypes was extracted using the microwave DNA extraction method (Dörnte and Kües 2013) or the NucleoSpin Plant II DNA extraction kit (Macherey, Nagel, Germany).

Amplifications and sequencing. The extracted genomic DNA was amplified targeting two nuclear ribosomal DNA (nrDNA) regions, internal transcribed spacer (ITS) and ribosomal large subunit-coding DNA (28S rRNA) for all specimens. Additionally, three protein-coding genes, RNA polymerase II largest subunit (RPB1), RNA polymerase II second largest subunit (RPB2), and translation elongation factor 1-alpha (TEF1) were amplified for specimens forming part of the T. elegans species complex and specimens of Trametes sp. The amplification of the 5.8S rRNA gene region, including ITS were performed in Mastercycler nexus gradient (Eppendorf, Germany), using the primer pair ITS-1F/ITS4 (White et al. 1990; Gardes and Bruns 1993). The Polymerase Chain Reaction (PCR) procedure for the ITS region, was as follows: initial denaturation at 95 °C for 3 min, followed by 35 cycles at 95 °C for 30 s, 52 °C for 30 s and 68 °C for 1 min, and a final extension at 68 °C for 3 min. Amplifications of LSU and three protein-coding genes were performed in 96-well TGradient Thermocycler (Biometra, Göttingen, Germany). PCR procedure for amplifying partial LSU rDNA using the primer pair LR0R/LR5 (Vilgalys and Hester 1990) approximately 964 bp differed to the ITS only by the annealing temperature (55 °C instead of 52 °C) and increased cycle extension time (90 s per cycle). The primer pairs EF1-983F/EF1-1567R (Rehner and Buckley 2005), RPB1-Af/RPB1-Cr (Stiller and Hall 1997; Matheny et al. 2002), and RPB2-b6F/RPB2-b7.1R (Liu et al. 1999; Matheny 2005) were used to amplify approximately 500 bp of TEF1, 1000 bp of RPB1, and 800 bp of RPB2. To amplify the protein-coding genes RPB1 and RPB2, the touchdown PCR protocol following Justo and Hibbett (2011) was used. PCR products were checked on 1% agarose gel stained with GelRed fluorescence dye (Biotium, Hayward, California, USA) in the Transilluminator Biometra Ti5 equipped with BioDocAnalyze software (Biometra GmbH, Göttingen, Germany). They were further cleaned up with QIAquick PCR Purification Kit according to manufacturer’s instructions (QIAGEN GmbH, Hilden, Germany). Thereafter, Sanger sequenced at GATC Biotech in Germany.

At least one sequence per specimen was generated for each morphotype except for the morphotype named Trametes aff. versicolor (Fig. 2N; Suppl. material 1). All newly generated sequences composed of 25 ITS, 20 LSU, two RPB1, four RPB2, and three TEF1 were deposited in GenBank (for accession numbers, see Table 1).

Figure 2. 

Macromorphology of Trametes species in Benin and specimen numbers in parentheses. A Trametes cingulata B hymenophore of Trametes cingulata (10) C Trametes flavida D hymenophore of Trametes flavida (05) E Trametes lactinea F hymenophore of Trametes lactinea (01) G Trametes palisotii H hymenophore of Trametes palisotii (04) I Trametes parvispora J hymenophore of Trametes parvispora (04) K Trametes polyzona L hymenophore of Trametes polyzona (06) M Trametes sanguinea (04) N Trametes aff. versicolor (01) O Trametes socotrana P hymenophore of Trametes socotrana (02). Scale bar corresponds to 1cm except in E, F where it corresponds to 2 cm.

Sequence alignment and phylogenetic analyses. To place all the 25 generated ITS sequences of specimens of Trametes spp. in a phylogenetic context, we aligned them in addition to 66 ITS sequences retrieved from GenBank (Benson et al. 2011). Further, 48 LSU sequences were aligned with 20 LSU sequences generated here. For the T. elegans species complex, seven newly generated sequences of protein-coding genes were aligned in addition to sequences used by Carlson et al. (2014). Each marker was aligned separately using MAFFT version 7, with the algorithm L-INS-i (Katoh et al. 2017) and standard settings as default. The resulting multiple species alignments were slightly adjusted and trimmed at both ends a bit from incomplete sequences in Geneious 5.6.7 (Kearse et al. 2012). Eight different datasets were assembled for the phylogenetic analyses: (i) ITS dataset with 91 sequences of Trametes spp., (ii) combined ITS-LSU dataset with 91 sequences Trametes spp., (iii) combined RPB1-RPB2 dataset with 23 sequences of Trametes spp., (iv) ITS dataset with 17 sequences of T. elegans species complex, (v) RPB1 dataset with ten sequences of the T. elegans species complex, (vi) RPB2 dataset with 12 sequences of T. elegans species complex, (vii) TEF1 dataset with 14 sequences of T. elegans species complex, and (viii) combined dataset of four genes (ITS, RPB1, RPB2, TEF1) of T. elegans species complex. The combined datasets were concatenated using Geneious 5.6.7 (Kearse et al. 2012). For the phylogenetic analyses, the partitioning of the combined datasets of Trametes spp. was considered. Lopharia cinerascens (Schwein.) G. Cunn., and Dentocorticium sulphurellum (Peck) M.J. Larsen & Gilb., were chosen as the outgroup in all datasets (Justo and Hibbett 2011). Two phylogenetic tree inference methods, Maximum likelihood (ML) and Bayesian (BY) were performed in each dataset. The ML of all datasets were performed using RAxML 8.2.10 (Stamatakis 2014) and the BY of all individual genes and combined dataset of T. elegans species complex were performed using MrBayes 3.2.6 (Ronquist et al. 2012) at the Cipres Science Gateway V.3.3. (Miller et al. 2010). The BY of the partitioned datasets of Trametes spp. were run independently using MrBayes 3.2.7 (Ronquist et al. 2012). The parameters in BY inference were set as follows: lset applyto = (all), nst = 6, rates = invgamma, ngammacat = 4, sampling frequency = 1000, and the command “unlink” was used to unlink parameters across characters on partitioned datasets. Two independent Markov Chain Monte Carlo (MCMC) processes were run, each in 4 chains, for 5 million generations, and 0.2 fraction were discarded as burn-in. The Phylogenetic Tree Summarization (SumTrees) program within DendroPy 4.3.0. (Sukumaran and Holder 2010) was used to build the consensus tree with branch supports (posterior probabilities). Further, by using IQ-Tree (Trifinopoulos et al. 2016), we assigned the bootstrap values (BS) of ML to the consensus tree of BY. The resulting phylogenetic trees were inspected in FigTree v. 1.4.2 (Rambaut 2014). All sequence alignments and phylogenetic trees generated in the study were deposited in TreeBASE: http://purl.org/phylo/treebase/phylows/study/TB2:S24354. The topologies of the consensus trees obtained from BY are presented in all figures throughout the document. Posterior probabilities (PP) and bootstrap values (BS) on or below branches as followed (PP/BS).

Macro-morphological descriptions were based on fresh and dried herbarium specimens. Microstructures are described using dried herbarium specimens. Fine sections through the basidiomata were prepared for observation using a razor blade under a stereomicroscope Leica EZ4 and mounted in 5% aqueous solution of potassium hydroxide (KOH) mixed with 1% aqueous solution of Phloxine. Melzer’s reagent (to test for dextrinoid or amyloid reactions), Cotton Blue (to test for cyanophilic reaction) were used and then examined at a magnification of 1000× using a Leica DM500 light microscope. Measurements were done with the software “Makroaufmaßprogramm” from Jens Rüdigs (https://ruedig.de/tmp/messprogramm.htm) and analysed with the software “Smaff” version 3.2 (Wilk 2012). In total, 135 basidiospores were measured from the sequenced specimen OAB0022 and additional examined specimen OAB0268. The basidiospore size is given as length and width of the spore. As measurements we present the mean with standard deviation and minimum and maximum values in parentheses (see below). The length (L), arithmetic average of all spore lengths, and the width (W), arithmetic average of all spore widths, were calculated. In addition, the ratio of length/width (Q) was calculated.

All alignments and phylogenetic trees generated in this study are available in TreeBASE under this link: http://purl.org/phylo/treebase/phylows/study/TB2:S24354. Newly generated sequences are available in GenBank, and the accession numbers are given in Table 1. Alignments, phylogenetic trees, and accession numbers of newly generated sequences will be public after the paper is published. Collected specimens are available at the mycological herbarium of the University of Parakou (UNIPAR). The new species was registered in mycoBank, and the registration number is given in the taxonomy section of this paper.

a.s.l. above sea level

BS Bootstrap values

BY Bayesian

ITS Internal Transcribed Spacer

KAS Mycological herbarium of the University of Kassel

L Length

LSU Large Subunit

MCMC Markov chain Monte Carlo

ML Maximum likelihood

nrDNA nuclear ribosomal DNA

PP Posterior probabilities

Q Length to width ratio

RPB1 RNA polymerase II largest subunit

RPB2 RNA polymerase II second largest subunit

TEF1 Translation elongation factor 1-alpha

UNIPAR Mycological herbarium of the University of Parakou

ITS dataset. The 25 ITS sequences obtained from Trametes spp. from Benin clustered in eight distinct clades (Suppl. material 2). All sequences of Trametes spp. from Benin fell into the monophyletic corresponding clades except the clade of Trametes lactinea (Berk.) Sacc., which, besides sequences of T. lactinea, accommodated also sequences of Trametes cubensis (Mont.) Sacc. with a very high support (BP = 1.00/BS = 100). Sequences of specimens of Trametes sp. (OAB0022 and OAB0023) from Benin formed a separated and well-supported clade within the Trametes clade (BP = 0.73/BS = 66).

ITS-LSU dataset. Results of ML and of BY show higher congruency, higher support values, and a higher number of resolved nodes than the results obtained with ITS data only. As evident by the ITS dataset, the sequence of T. lactinea from Benin clustered in addition to other sequences of T. lactinea retrieved from GenBank with sequences of T. cubensis with high support (BP = 1.00/BS = 92). Like in the analysis of the ITS dataset, sequences of Trametes sp. from Benin formed a distinct clade (Fig. 3). The two sequences of the new species of Trametes from Benin clustered in a distinct lineage within the Trametes clade (Figs 2I, J; 4). The clade of the T. elegans species complex is presented in the section below.

Figure 3. 

ML phylogeny of Trametes spp. based on combined ITS-LSU dataset. Branch support values given as PP/BS. All clades where newly generated sequences clustered are highlighted in grey and bars with names are given beside for more readability. Taxon names are followed by voucher or stain number and country of origin.

The phylogenetic trees generated from individual gene regions ITS, RPB1, RPB2, and TEF1 (Suppl. material 3) and the combined datasets (Fig. 5) show similar results for phylogenetic relationships within the T. elegans species complex. Four distinct and well-supported clades were evident in all datasets. The clade highlighted in grey (Fig. 5; Suppl. material 3) is distinct from all other clades within T. elegans species complex and highly supported in all individual gene and combined dataset. This clade contains only sequences of T. elegans from Benin and Cameroon.

Figure 4. 

Crossection of the hymenium at the base of a pore of Trametes parvispora. Basidiospores, hyphae, basidia, basidioles, and a hyphal peg are showing. The box (lower left corner) shows the location (small rectangle) of the line drawing in the cross-section of the hymenophore. Scale bar = 10 μm

Figure 5. 

ML phylogeny of Trametes elegans species complex based on combined dataset of four-gene regions (ITS, RPB1, RPB2, TEF1). Branch support values given as PP/BS. Sequences of T. elegans from tropical Africa investigated in this study are highlighted in grey.

In Benin, seven species of Trametes were previously reported (Olou et al. 2019). By the present, study two additional species, namely T. lactinea and Trametes aff. versicolor (Fig. 2E, F, N), were recorded in addition to previous species. Thus, to our knowledge, nine species of Trametes are currently known for Benin. Of these nine species, only two species, T. elegans and T. sanguinea, were reported in Benin until 2002 (Yorou and De Kesel 2002). The remaining seven species, namely T. cingulata, T. flavida, T. lactinea, T. parvispora, T. polyzona, T. socotrana, and Trametes aff. versicolor, were recorded between 2017 and 2018. Given this history, it is most likely that more species will be found. Nonetheless, this number is significant when compared to the total diversity of 9–14 species of Trametes reported for Europe (Ryvarden and Gilbertson 1994; Ryvarden and Melo 2014). Further studies are needed to document the overall diversity of species of Trametes in Benin.

To place specimens of Trametes spp. from Benin in a larger phylogenetic context, we generated sequences of several genes. Generated sequences were placed into the phylogeny of the genus Trametes as established by Justo and Hibbett (2011). Eight distinct clades corresponding to eight different species were obtained from these sequences.

Our phylogenetic analyses from ITS and combined ITS-LSU datasets reveal sequence similarities and taxonomic misplacement within the clades of T. flavida and T. lactinea (Fig. 3; Suppl. material 2). The clade of T. flavida accommodated, in addition to sequences of T. flavida, sequences of Trametes sp. from French Guiana which is known as Leiotrametes sp. (Welti et al. 2012). This species was proposed as a new species by Welti et al. (2012). Here, Trametes sp. clustered together with T. flavida with high support in the ITS dataset (PP = 0.84/BS = 89) and the combined ITS-LSU datasets (PP = 0.98/BS = 99). Both species share also high morphological similarity (Welti et al. 2012; Fig. 2C, D) and a tropical distribution. We therefore suggest that Trametes sp. from French Guiana should not be considered as a new species but should be referred to as T. flavida. In addition to the T. flavida clade, our phylogenetic analyses showed that the T. lactinea clade contains not only sequences of T. lactinea, but also sequences of T. cubensis with high support in the ITS and ITS-LSU datasets (Fig. 3; Suppl. material 2). This result is similar to previous phylogenetic analyses on Trametes using the ITS marker (Justo and Hibbett 2011; Carlson et al. 2014). Trametes lactinea and T. cubensis are still valid names and both species share quite similar morphological characters. They are characterized by an applanate, broadly attached to dimidiate, white to cream basidiomata and a white to cream pore surface (Ryvarden and Johansen 1980; Gilbertson and Ryvarden 1987). Nevertheless, although both species are sharing quite similar morphological characters, they also differ in some characters. Trametes. cubensis is characterized by an annual basidioma, small pores, almost invisible to the naked eye, 5–7 per mm, and cylindrical basidiospores 7–9 × 3–3.5 μm (Gilbertson and Ryvarden 1987), while T. lactinea has an annual to perennial basidioma and large pores, which are visible to the naked eye, mostly 1.5–2 per mm, but can reach up to 3–4 (5) per mm in some specimens with cylindrical-ellipsoid basidiospores 4–7.5 × 2.2–3 μm (Ryvarden and Johansen 1980). Our specimen of T. lactinea (Fig. 2E, F) matches the morphological description of T. lactinea with 3–4 pores per mm, but we did not observe any spore despite numerous attempts. Thus, considering the result of our phylogenetic analyses, absence of spores in our T. lactinea specimen, and the high morphological similarity between species within Trametes (Gilbertson and Ryvarden 1987), we cannot reasonably distinguish T. lactinea from T. cubensis. Further morphological, chemotaxonomic, and molecular studies integrating proteins coding genes (e.g. RPB1, RPB2, and TEF1) are therefore needed to confirm whether T. lactinea and T. cubensis refer to the same species.

Previously the phylogenetic resolution of T. cingulata was problematic due to low sequence availability. Here we generated a total of 17 de novo sequences and show that T. cingulata appears as a monophyletic group within Trametes with high support in ITS and combined ITS-LSU datasets respectively (PP = 1.00/BS = 97) and (PP = 1.00/BS = 100) (Fig. 3; Suppl. material 2). Thus, contrary to the uncertain position of T. cingulata within the genus Trametes (Welti et al. 2012), our results revealed that the latter does not belong to Trametes sensu stricto in the sense of Justo and Hibbett (2011) and Welti et al. (2012) (Fig. 3; Suppl. material 2) but rather to Trametes sensu lato.

The specimens from Benin identified as members of the T. elegans species complex correspond to the morphological descriptions of T. elegans by Gilbertson and Ryvarden (1987) and Ryvarden and Johansen (1980). The clades evident in all datasets within the T. elegans complex (Figs 3, 5; Suppl. material 2, 3) represent three clades previously attributed to three different species by Carlson et al. (2014), and a new clade highlighted in grey (Fig. 5; Suppl. material 3) represents specimens of T. elegans from Benin and Cameroon (Tropical Africa). This new clade contains only sequences of T. elegans from Benin and Cameroon due to the non-publication of most T. elegans sequences from tropical Africa (Olusegun 2015; Awala and Oyetayo 2016; Ueitele et al. 2018). Thus, prior to this study, only sequences of T. elegans from Cameroon and Gabon are available in GenBank for Africa. However, the sequences of T. elegans from Gabon (GenBank accession number: KY449397, KY449398) were not considered because they fell outside the T. elegans species complex and were instead closely related to T. lactinea. We, therefore, excluded these sequences from our analyses. All in all, since the sequences of T. elegans from tropical Africa investigated in this study are demarcated from sequences of T. elegans s. str., the adoption of another correct name for specimens of T. elegans from this area is necessary.

Specimens belonging to the T. elegans species complex have been reported in the past for tropical African countries (Ryvarden and Johansen 1980), with the first name applied to such specimens being Daedalea amanitoides P. Beauv., which was based on a specimen from Nigeria (cited as kingdom of Oware) (Palisot-Beauvois 1804). The morphological characteristics evident in the very short description and illustration of a fruiting body of D. amanitoides match the characteristics of the specimens examined in this study. However, for reasons that we ignore, Fries (1821) replaced this name (D. amanitoides) by the name Daedalea palisotii Fr., which is sanctioned and therefore must be used. The combination Trametes palisotii (Fr.) Imazeki (Imazeki 1952) is available and must be used for African specimens known previously as T. elegans (Fig. 5).

The sequences belonging to the new species named T. parvispora form a distinct and well-supported clade in the ITS and the combined ITS-LSU datasets (Fig. 3; Suppl. material 2). This species forms a sister clade with the still formally undescribed Trametes sp. (KT896651) from Finland. However, unlike T. parvispora where fruiting bodies were available for morphological characterization (Fig. 2I, J), the Finnish specimen was isolated as mycelium from the bark beetle Ips typographus L. (Linnakoski et al. 2016). Thus, anatomical and morphological comparisons are currently not possible. Furthermore, both sequences of T. parvispora share a clade with Trametes meyenii (Klotzsch) Lloyd. This clade was confirmed by phylogenetic analyses including two additional markers RPB1 and RPB2 (Suppl. material 4). Trametes meyenii has hispid and cream-yellow pilei, irpicoid and white to ochraceous hymenophore, pores 1–3 per mm, 4.5–6 × 2–2.5 μm basidiospores (Zmitrovich et al. 2012), whereas T. parvispora has glabrous and whitish pilei, a daedaleoid and white hymenophore, 3.2–4.6 × 2.1–2.8 μm basidiospores, and the presence of regular hyphal pegs (Figs 2I, J, 4). These morphological differences confirm that T. parvispora and T. meyenii are distinct species as shown by the phylogenetic analyses (Fig. 3; Suppl. material 2, 4). However, some species lacking DNA sequences, namely Trametes barbulata Corner, Trametes daedaleoides Corner, and Trametes rugosituba Corner (Corner 1989; Hattori 2005; Hattori and Sotome 2013), share with T. parvispora a quite similar spore size range. But the latter species differs from each other species by the combination of macro- and microscopic characteristics outlined above. Thus, the rare anatomic features of the regular hyphal pegs and the small size of the basidiospores together with the phylogenetic placement within the Trametes clade, provide enough evidence for T. parvispora as a distinct new species.