MYCOLOGIA 2017, VOL. 109, NO. 4, 568–577 https://doi.org/10.1080/00275514.2017.1369830 Phylogeny and taxonomy of Echinodontium and related genera Shi-Liang Liua, Yan Zhaoa, Yu-Cheng Daia, Karen K. Nakasoneb, and Shuang-Hui Hea aInstitute of Microbiology, Beijing Forestry University, Beijing 100083, China; bCenter for Forest Mycology Research, Northern Research Station, U.S. Forest Service, Madison, Wisconsin 53726-2398 ABSTRACT ARTICLE HISTORY The phylogenetic relationship of eight species of Echinodontium, Laurilia, and Perplexostereum of Received 5 February 2017 Russulales were analyzed based on sequences of the nuc rDNA ITS1-5.8S-ITS2 (ITS [internal Accepted 17 August 2017 – transcribed spacer]) and D1 D2 domains of nuc 28S rDNA (28S). Our results show that KEYWORDS Echinodontium tinctorium, E. ryvardenii, and E. tsugicola represent Echinodontium sensu stricto. Bondarzewiaceae; Based on morphological and phylogenetic evidence, the new genus Echinodontiellum is estab- Echinodontiaceae; lished to accommodate Echinodontium japonicum. Amylostereum, Echinodontium, Echinodontiellum; Laurilia; Echinodontiellum, and Larssoniporia form the Echinodontiaceae clade. The Bondarzewiaceae Lauriliella clade includes Bondarzewia, Heterobasidion, Laurilia, and Lauriliella. The new genus Lauriliella is established for the species initially described as Stereum taxodii and Lauriliella taiwanensis new to science. The monotypic genus Perplexostereum forms a distinct clade. A key to the genera in the Echinodontiaceae and Bondarzewiaceae as well to Perplexostereum is provided. INTRODUCTION Ryvarden and Tutka (2014) proposed a new genus, Species of Echinodontium sensu lato are character- Perplexostereum Ryvarden & S. Tutka, for Stereum ized by conspicuous basidiocarps, dentate to smooth endocrocinum Berk. Like Echinodontium and Laurilia, hymenophores, encrusted cystidia, and ornamented, Perplexostereum develops large, pileate basidiocarps amyloid basidiospores. In Gross’s(1964)monograph and ornamented amyloid basidiospores and inhabits of the Echinodontiaceae, he included six species in gymnosperms but differs in lacking encrusted cystidia. Echinodontium Ellis & Everh. Later, Bernicchia and Tabata et al. (2000) demonstrated that Piga (1998) added a new species, E. ryvardenii Echinodontium and Amylostereum Boidin were phylo- Bernicchia & Piga. Among these species, E. sulcatum genetically related and belonged in the (Burt) H.L. Gross and E. taxodii (Lentz & H.H. Echinodontiaceae. This was confirmed by subsequent McKay) H.L. Gross were also placed in Laurilia studies (Hibbett et al. 2000; Hibbett and Donoghue Pouzar by some mycologists (Pouzar 1959; 2001; Binder and Hibbett 2002; Hibbett and Binder Parmasto 1968; Eriksson and Ryvarden 1976; 2002; Binder et al. 2005; Chen et al. 2016). However, Chamuris 1988; Ginns and Lefebvre 1993;Stalpers Echinodontium was also shown to be closely related to 1996). Bondarzewia Singer and Heterobasidion Bref. in the Although similar in some characters, the type species Bondarzewiaceae/Echinodontiaceae clade by Larsson of Echinodontium and Laurilia can be easily distin- and Larsson (2003) and Miller et al. (2006). In addition, guished in morphology. Echinodontium tinctorium some of these studies showed that Laurilia and (Ellis & Everh.) Ellis & Everh. has pileate to ungulate Echinodontium were closely related and possibly con- basidiocarps, a coarsely dentate hymenophore, brick generic (Hibbett and Donoghue 2001; Hibbett and red–colored context, and a dimitic hyphal system. In Binder 2002; Larsson and Larsson 2003; Binder et al. contrast, Laurilia sulcata (Burt) Pouzar has resupinate 2005; Miller et al. 2006). to effuse-reflexed basidiocarps, smooth to tuberculate We wanted to explore and clarify the phylogenetic hymenophore, beige subiculum, and a trimitic hyphal relationship of Echinodontium, Laurilia, and system (Gross 1964; Eriksson and Ryvarden 1976; Perplexostereum within Russulales by employing and Stalpers 1996). analyzing sequence data of the nuc rDNA ITS1-5.8S- CONTACT Shuang-Hui He [email protected] Shi-Liang Liu and Yan Zhao contributed equally to the paper. Color versions of one or more of the figures in this article can be found online at www.tandfonline.com/umyc. Supplemental data for this article can be accessed on the publisher’s Web site. © 2017 The Mycological Society of America Published online 11 Oct 2017 MYCOLOGIA 569 ITS2 (ITS [internal transcribed spacer]) and D1–D2 cbrc.jp/alignment/server/). The sequences were domains of nuc 28S rDNA (28S). We include morpho- adjusted in BioEdit 7.0.5.3 (Hall 1999) and then con- logical, distributional, and ecological data to character- catenated manually. The concatenated alignments were ize the taxa in these genera. A key to the genera in the deposited at TreeBase (http://treebase.org/treebase- Echinodontiaceae and Bondarzewiaceae as well to the web/home.html; submission ID 21276). taxa discussed herein is provided. Maximum likelihood (ML), maximum parsimony (MP), and Bayesian inference (BI) analyses were per- MATERIALS AND METHODS formed for the data set by using RAxML 7.2.6 (Stamatakis 2006), PAUP* 4.0b10 (Swofford 2002), The specimens and cultures examined are deposited at and MrBayes 3.1.2 (Ronquist and Huelsenbeck the herbaria of Institute of Microbiology, Beijing 2003), respectively. In ML analysis, statistical support Forestry University (BJFC); Center for Forest values (LB) were obtained by using rapid bootstrap- Mycology Research, U.S. Forest Service (CFMR); and ping with 1000 replicates, with default settings used National Museum of Natural Science, Taiwan (TNM). for other parameters. In MP analysis, gaps in the Samples for microscopic examination were mounted in alignments were treated as missing data. Trees were cotton blue, Melzer’s reagent, or 1% phloxine and stu- generated using 100 replicates of random stepwise died at magnifications up to 1000× using a Nikon addition of sequence and tree bisection reconnection Eclipse 80i microscope (Nikon Corporation, Japan). (TBR) branch-swapping algorithm, with all characters Drawings were made with the aid of a drawing tube. given equal weight. Branch supports (PB) for all The following abbreviations are used: L = mean spore parsimony analyses were estimated by performing length, W = mean spore width, Q = L/W ratio, n (a/b) 1000 bootstrap replicates (Felsenstein 1985)witha = number of spores (a) from number of specimens (b), heuristic search of 10 random-addition replicates for KOH = 2% potassium hydroxide. Color codes and each bootstrap replicate. The tree length (TL), con- names follow Kornerup and Wanscher (1978). sistency index (CI), retention index (RI), rescaled The ITS and 28S gene regions were amplified consistency index (RC), and homoplasy index (HI) from cultures or a small piece of herbarium speci- were calculated for each generated tree. For Bayesian mens using a cetyltrimethylammonium bromide inference (BI), best models of evolution were (CTAB) rapid plant genomic DNA extraction kit obtained using MrModeltest 2.2 (Nylander 2004), (Aidlab, Beijing, China). The primers ITS5 and and posterior probabilities (PP) were obtained by ITS4 (White et al. 1990) were employed to amplify Markov chain Monte Carlo sampling in MrBayes the ITS region by using the following cycling pro- 3.1.2 (Ronquist and Huelsenbeck 2003). Four simul- tocol: initial denaturation at 95 C for 4 min, fol- taneous Markov chains were run for 5 million gen- lowedby34cyclesat94Cfor40s,58Cfor45s, erations for the data set, and trees were sampled and 72 C for 1 min, and a final extension of 72 C every 100th generation. The first quarter of the for 10 min. The 28S gene region was amplified with trees, which represented the burn-in phase of the primer pair LR0R and LR7 (http://www.biology. analyses, were discarded, and the remaining trees duke.edu/fungi/mycolab/primers.htm), using the fol- were used to calculate posterior probabilities in the lowing procedure: initial denaturation at 94 C for 1 majority rule consensus tree. min, followed by 34 cycles at 94 C for 30 s, 50 C for 1 min, and 72 C for 1.5 min, and a final extension RESULTS of 72 C for 10 min. DNA sequencing was performed at Beijing Genomics Institute, China, using the same Twelve ITS and 15 28S sequences were generated for primers. Newly generated sequences were deposited this study. The data set contained 69 samples represent- in GenBank (SUPPLEMENTARY TABLE 1). ing 53 ingroup and 2 outgroup taxa The phylogeny of Russulales was inferred from ITS (SUPPLEMENTARY TABLE 1). The data set had an and 28S sequence data. Sequences obtained from aligned length of 2442 characters, of which 840 were GenBank were primarily from Larsson and Larsson parsimony informative. MP analysis yielded 90 parsi- (2003; SUPPLEMENTARY TABLE 1). Sistotrema monious trees (TL = 5372, CI = 0.390, RI = 0.572, RC = brinkmannii (Bres.) J. Erikss. and S. muscicola (Pers.) 0.223, HI = 0.610). The best-fit evolution model for BI S. Lundell were selected as outgroup taxa following was “GTR+I+G.” The average standard deviation of Larsson and Larsson (2003). The ITS and 28S split frequencies was 0.009192. The topologies of trees sequences were aligned separately by using MAFFT 6 obtained from ML, MP, and BI were almost the same. with a LINSI option (Katoh and Toh 2008; http://mafft. Only the ML tree is shown in FIG. 1, with maximum 570 LIU ET AL.: ECHINODONTIUM AND RELATED GENERA Figure 1. Phylogenetic tree inferred from maximum likelihood (ML) analysis based on a concatenated data set of ITS and 28S