Phylogeny and diversity of Haploporus (Polyporaceae, Basidiomycota)

Abstract Four species of Haploporus, H.angustisporus, H.crassus, H.gilbertsonii and H.microsporus are described as new and H.pirongia is proposed as a new combination, based on morphological characteristics and molecular phylogenetic analyses inferred from internal transcribed spacer (ITS) and large subunit nuclear ribosomal RNA gene (nLSU) sequences. Haploporusangustisporus, H.crassus and H.microsporus occur in China, H.gilbertsonii occurs in the USA, and the distribution of H.pirongia is extended from New Zealand to Australia. Haploporusangustisporus is characterized by the distinct narrow oblong basidiospores measuring 10.5–13.5 × 3.9–5 µm. Haploporuscrassus is characterized by the presence of ventricose cystidioles occasionally with a simple septum, dissepimental hyphae usually with a simple septum, unique thick-walled basidia and distinctly wide oblong basidiospores measuring 13.5–16.5 × 7.5–9.5 µm. Haploporusgilbertsonii is characterized by its large pores (2–3 per mm), a dimitic hyphal structure with non-dextrinoid skeletal hyphae and wide oblong basidiospores measuring 12–15 × 6–8 µm. Haploporusmicrosporus is characterized by distinctly small pores (7–9 per mm), the presence of dendrohyphidia, and distinctly small ellipsoid basidiospores measuring 5.3–6.7 × 3–4.1 µm. Haploporuspirongia is proposed as a new combination. Haploporusamarus is shown to be a synonym of H.odorus and Pachykytosporawasseri is considered a synonym of H.subtrameteus.

During a study on taxonomy of Polyporaceae, several specimens of Haploporus from USA, Australia and China were studied. After morphological examinations and phylogenetic analysis of ITS and nLSU sequences, four new species were confirmed to be members of the Haploporus lineage. In this paper, we describe and illustrate these new species. In addition, Poria pirongia G. Cunn. was originally described from New Zealand (Cunningham 1947), and treated as a synonym of Poria papyracea (Schwein.) Cooke (= Haploporus papyraceus (Schwein.) Y.C.Dai&Niemelä (Cunningham 1965, Lowe 1966and Buchanan and Ryvarden 1988) is shown to represent an independent species, based on new specimens and both morphology and phylogenetic evidences. Therefore, a new combination (H. pirongia) is proposed.

Morphological studies
Sections were studied microscopically according to Dai (2010) at magnifications ≤1000× using a Nikon Eclipse 80i microscope with phase contrast illumination. Drawings were made with the aid of a drawing tube. Microscopic features, measurements, and drawings were made from sections stained with Cotton Blue and Melzer's reagent. Spores were measured from sections cut from the tubes. To present spore size variation, the 5% of measurements excluded from each end of the range are given in parentheses. Basidiospore spine lengths were not included in the measurements. Abbreviations include: IKI = Melzer's reagent, IKI-= negative in Melzer's reagent, KOH = 5% potassium hydroxide, CB = Cotton Blue, CB+ = cyanophilous, L = mean spore length (arithmetic average of all spores), W = mean spore width (arithmetic average of all spores), Q = the L/W ratio, and n = number of spores measured / from given number of specimens. Color terms follow Petersen (1996). Herbarium abbreviations follow Thiers (2018).

Molecular study and phylogenetic analysis
A CTAB rapid plant genome extraction kit (Aidlab Biotechnologies, Beijing) was used to obtain PCR products from dried specimens, according to the manufacturer's instructions with some modifications (Cao et al. 2012, Zhao et al. 2013. The DNA was amplified with the primers: ITS5 and ITS4 for ITS (White et al. 1990), and LR0R and LR7 (http://www.biology.duke.edu/fungi/mycolab/primers.htm) for nLSU (Vilgalys and Hester 1990). The PCR procedure for ITS was as follows: initial denaturation at 95 °C for 3 min, followed by 34 cycles at 94 °C for 40 s, 54 °C for 45 s and 72 °C for 1 min, and a final extension of 72 °C for 10 min. The PCR procedure for nLSU was as follows: initial denaturation at 94 °C for 1 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 of 72 °C for 10 min. The PCR products were purified and sequenced at the Beijing Genomics Institute, China with the same primers.
Maximum parsimony (MP) and Bayesian inference (BI) were employed to perform phylogenetic analysis of the two aligned datasets. The two phylogenetic analysis algorithms generated nearly identical topologies for each dataset, and, thus only the topology from the MP analysis is presented along with statistical values from the MP and BI algorithms. Most parsimonious phylogenies were inferred from the ITS + nLSU, and their combinability was evaluated with the incongruence length difference (ILD) test (Farris et al. 1994) implemented in PAUP* 4.0b10 (Swofford 2002), under a heuristic search and 1000 homogeneity replicates giving a P value of 1.000, much greater than 0.01, which means there is no discrepancy among the two loci in reconstructing phylogenetic trees. Phylogenetic analysis approaches followed Zhao et al. (2015). The tree construction procedure was performed in PAUP* version 4.0b10 (Swofford 2002). All characters were equally weighted, and gaps were treated as missing data. Trees were inferred using the heuristic search option with TBR branch swapping and 1000 random sequence additions. Max-trees were set to 5000, branches of zero length were collapsed and all parsimonious trees were saved. Clade robustness was assessed using a  (Felsenstein 1985). Descriptive tree statistics tree length (TL), consistency index (CI), retention index (RI), rescaled consistency index (RC), and homoplasy index (HI) were calculated for each maximum parsimonious tree (MPT) generated. jModeltest v.2.17 (Darriba et al. 2012) was used to determine the best-fit evolution model of the combined dataset for Bayesian inference (BI). The Bayesian inference (BI) was conducted with MrBayes 3.2.6 (Ronquist et al. 2012) in two independent runs, each of which had four chains for 10 million generations and started from random trees. Trees were sampled every 1000 th generation. The first 25% of sampled trees were discarded as burn-in, whereas other trees were used to construct a 50 % majority consensus tree and for calculating Bayesian posterior probabilities (BPPs).
Phylogenetic trees were visualized using Treeview (Page 1996). Nodes that received Bootstrap support ≥50% and Bayesian posterior probabilities (BPP) ≥0.90 are considered as significantly supported.

Molecular phylogeny
The combined ITS and 28S dataset included sequences from 46 fungal collections representing 21 species. The dataset had an aligned length of 2054 characters, of which 1399 characters are constant, 98 are variable and parsimony-uninformative, and 557 are parsimony-informative. MP analysis yielded 4 equally parsimonious trees (TL = 1370, CI = 0. 639, RI = 0.870, RC = 0.556, HI = 0.361). The best model for the combined ITS and 28S sequences dataset estimated and applied in the BI was GTR+I+G. BI resulted in a similar topology with an average standard deviation of split frequencies = 0.004515 to MP analysis, and thus only the MP tree is provided. Both BT values (≥50%) and BPPs (≥0.90) are shown at the nodes (Fig. 1). The ITS-based phylogenies included ITS sequences from 47 fungal collections representing 21 species. The dataset had an aligned length of 711 characters, of which 317 characters are constant, 54 are variable and parsimony-uninformative, and 340 are parsimony-informative. MP analysis yielded 4 equally parsimonious trees (TL = 927, CI = 0. 653, RI = 0.888, RC = 0.580, HI = 0.347). The best model for the ITS sequences dataset estimated and applied in the BI was GTR+I+G. BI resulted in a similar topology with an average standard deviation of split frequencies = 0.005040 to MP analysis, and thus only the MP tree is provided. Both BT values (≥50%) and BPPs (≥0.90) are shown at the nodes (Fig. 2).
In both 28S+ITS-and ITS-based phylogenies (Figs. 1-2), five new well-supported lineages were identified. Among them three well-supported terminal clades and two isolated branches (100% MP and 1.00 BI).   Diagnosis. Differs from other Haploporus species by the combination of its resupinate habit, a dimitic hyphal structure with dextrinoid skeletal hyphae, the absence of dendrohyphidia, and distinct narrow oblong basidiospores measuring 10-13.5 × 4-5 µm.
Etymology. Crassus (Lat.): referring to the species having wide basidiospores. Fruitbody. Basidiocarps annual, resupinate, adnate, soft corky when fresh, become corky and cracked upon drying, without odor or taste when fresh, up to 35 cm long, 3 cm wide and 1 mm thick at center. Pore surface white to cream when fresh, becoming buff-yellow upon drying; sterile margin indistinct, very narrow to almost lacking; pores round, 3-5 per mm; dissepiments thin, mostly entire, sometimes lacerate. Subiculum cream, corky, thin, about 0.1 mm thick. Tubes light buff, corky, about 0.9 mm long.
Haploporus pirongia is related to H. odorus, but the latter has a perennial and pileate basidiocarp with strong anise odor, ovoid basidiospores and lacks cystidioles (Niemelä 1971). Haploporus pirongia resembles H. thindii and H. subpapyraceus by sharing resupinate basidiocarps with approximately the same pore size. However, Haploporus thindii has a dimitic hyphal structure, lacks cystidioles, and has a distribution in subtropical India and valley of Tibet of China (Natarajan andKolandavelu 1993, Dai et al. 2007). Moreover, H. subpapyraceus has ellipsoid basidiospores (9-12 × 5.5-8 µm, Shen et al. 2016). Gilbertson and Ryvarden (1987) reported Haploporus tuberculosus (as Pachykytospora tuberculosa) from the USA, but only in a small region of southern Arizona where it should be "quite common on oaks, especially in Chiricahua Mountains". Locally, we have collected in this region only H. gilbertsonii and believe that, in most cases, this species was mistaken for H. tuberculosus in Arizona. The presence of H. tuberculosus in America is questionable.