Taxonomic novelties in Magnolia-associated pleosporalean fungi in the Kunming Botanical Gardens (Yunnan, China)

This paper represents the first article in a series on Yunnanese microfungi. We herein provide insights into Magnolia species associated with microfungi. All presented data are reported from the Kunming Botanical Gardens. Final conclusions were derived from the morphological examination of specimens coupled with phylogenetic sequence data to better integrate taxa into appropriate taxonomic ranks and infer their relationships. Shearia formosa, the type species of Shearia, lacks type material, and its phylogenetic position accordingly remains unresolved. A fresh collection of Shearia formosa, obtained from Magnolia denudata and M. soulangeana in China, therefore, designated a neotype for stabilizing the application of the species and/or genus name. Phylogenetic analyses of a combined DNA data matrix containing SSU, LSU, RPB2 and TEF loci of representative Pleosporales revealed that the genera Crassiperidium, Longiostiolum and Shearia are a well-defined monophylum. It is recognized as the family Longiostiolaceae and strongly supported by Bayesian and Maximum Likelihood methods. Its members are characterized by immersed to semi-immersed, globose to subglobose ascomata with a central, periphysate ostiole, a peridium composed of rectangular to polygonal cells, cylindrical to clavate asci, broadly fusiform, hyaline to pale brown ascospores, a coelomycetous asexual morph with pycnidial conidiomata, enteroblastic, annellidic, ampulliform, doliiform or cylindrical conidiogenous cells and cylindrical to fusiform, transverse and sometimes laterally distoseptate conidia without a sheath or with a basal lateral sheath. Nigrograna magnoliae sp. nov. is introduced from Magnolia denudata with both asexual and sexual morphs. We observed the asexual morph of Brunneofusispora sinensis from the culture and therefore amended the generic and species descriptions of Brunneofusispora.

Introduction Located in Southwestern China, Yunnan is renowned for harboring one of the botanically richest and most diverse terrestrial regions on Earth [1]. Yunnan represents >50% of China's overall floristic diversity and accounts for 18,000 vascular plant species [2], with high levels of endemism [3]. A highly variable climate and flourishing vegetation facilitate rapid fungal growth, reproduction and speciation. Based on its high level of plant diversity, it is estimated that approximately 104,000 fungal species may exist in Yunnan [4].
Conversely, among these fungal encounters, ascomycetes are being neglected compared to the level of research conducted on basidiomycetes in Yunnan Province. Owing to their abundance in all ecosystems, ascomycetous taxa simply cannot be overlooked in any region. Most taxa of ascomycetes are plant-associated fungi that can be pathogens, endophytes, saprobes or epiphytes across a wide range of hosts in terrestrial as well as aquatic habitats. Microfungi contribute both positively and negatively to human and economic well-being. As pathogens, they pose a threat to agriculture [5], and the rapid identification of potentially problematic species and accurate prediction of their behavior will facilitate the adoption of proper mitigation and phytosanitary measures. Given their ubiquitous nature, additional taxonomic knowledge are prerequisites to understanding the biological and environmental significance of ascomycetes. Supporting this obligation, the CAS Key Laboratory for Plant Biodiversity and Biogeography of East Asia (KLPB) has begun to study plant-based ascomycetes in Yunnan Province. The current study represents the first in a series comprising a taxonomic circumscriptive project of microfungi in Yunnan Province. It accounts for a group of ascomycetes recovered from the twigs of Magnolia denudata and M. soulangeana in the East Garden of Kunming Botanical Garden (Kunming, Panlong District). This garden functions as an ex-situ conservation of plants from Southwest China, particularly focusing on the conservation of endangered, endemic and economically important plant species native to the Yunnan Plateau and the southern Hengduan Mountains [6]. Based on morphology and multi-gene phylogenetic evidences of the collected ascomycetes, we characterized a neotype, a new species and a new host record in the order Pleosporales.

Isolates and specimens
Fresh fungal materials were collected from twigs of Magnolia denudata and M. soulangeana in the East Garden of Kunming Botanical Garden (Yunnan Province, China) during both dry (February) and wet (August) seasons. Kunming Botanical Garden is in the central part of Yunnan Province in the city of Kunming, located in a plateau basin. The wet and dry seasons are distinct in Kunming, and precipitation is concentrated from May to October, accounting for about 85% of the annual precipitation [7]. The dry seasons is from November to April, the rainfall of accounts for only about 15% of total annual rainfall [7]. The Garden is located at 250 7'004.9"-25˚08'054.8"N, 102˚44'015.2"-102˚44'047.3"E at an elevation of 1914-1990 m above sea level, and has an annual average rainfall of 1006.5 mm, an annual average temperature of 14.7˚C and an annual average relative humidity of 73% [6]. The collected specimens were brought to the laboratory in Zip lock plastic bags. Samples were examined with an Olympus SZ61 Series microscope. Single spore isolation was carried out following the method described in [8]. Germinated spores were individually transferred to potato dextrose agar (PDA) plates and grown at 20˚C in the daylight. Isolates including accession numbers of gene sequences are listed in are listed in Suppl. material 1:

Morphological observations
In hand sections of the ascomata/ conidiomata, which were mounted in distilled water, the following characteristics were evaluated: ascomata/ conidiomata diameter, height, colour and shape; width of peridium; height and diameter of ostioles. Length and width (at the widest point) of asci, ascospores, conidiophores and conidia were measured. Images were captured with a Canon EOS 600D digital camera fitted to a Nikon ECLIPSE Ni compound microscope. Measurements were made with the Tarosoft (R) Image Frame Work program and images used for figures processed with Adobe Photoshop CS5 Extended version 10.0 software (Adobe Systems, USA).

DNA extraction, PCR amplifications and sequencing
Mycelia for DNA extraction from each isolate were grown on PDA for 3-4 weeks at 20˚C and total genomic DNA was extracted from approximately 150 ± 50 mg axenic mycelium scraped from the edges of the growing culture. Mycelium was ground to a fine powder with liquid nitrogen and DNA extracted using the Biospin Fungus Genomic DNA Extraction Kit-BSC14S1 (BioFlux, P.R. China) following the instructions of the manufacturer. DNA to be used as template for PCR were stored at 4˚C for use in regular work and duplicated at -20˚C for long term storage.
DNA sequence data was obtained from the partial sequences of three ribosomal and two protein coding genes. The genes, primers, references and PCR protocols are summarized in Table 1. Polymerase chain reaction (PCR) was carried out in a volume of 25 μl which contained 12.5 μl of 2 × Power Taq PCR MasterMix (Bioteke Co., China), 1 μl of each primer (10 μM), 1 μl genomic DNA and 9.5 μl deionized water. The amplified PCR fragments were sent to a commercial sequencing provider (BGI, Ltd Shenzhen, P.R. China). The nucleotide sequence data acquired were deposited in GenBank.

Molecular phylogenetic analyses
Sequencing and sequence alignment. Sequences generated from different primers of the five genes were analysed with other sequences retrieved from GenBank (see S1 Table). Sequences with high similarity indices were determined from a BLAST search to find the closest matches with taxa in Pleosporales, and from recently published data [17,18,19]. The multiple alignments of all consensus sequences, as well as the reference sequences were automatically generated with MAFFT v. 7 (http://mafft.cbrc.jp/alignment/server/index.html; [20,21]), and were improved manually when necessary using BioEdit v. 7.0.5.2 [22].
Phylogenetic analyses. The single-locus data sets were examined for topological incongruence among loci for members of the Pleosporales. The conflict-free alignments were concatenated into a multi-locus alignment that was subjected to maximum-likelihood (ML) and Bayesian (BI) phylogenetic analyses. The evolutionary models for Bayesian analysis and maximum-likelihood were selected independently for each locus using MrModeltest v. 2.3 [23] under the Akaike Information Criterion (AIC) implemented in both PAUP v. 4.0b10. GTR+I+G model is resulted in each locus for Bayesian analysis and maximum-likelihood by AIC in MrModeltest as the best-fit model.
Bayesian analysis was conducted with MrBayes v. 3.1.2 [24] to evaluate Bayesian posterior probabilities (BYPP) [25,26] by Markov Chain Monte Carlo sampling (BMCMC). GTR+I+G was used in the command. Six simultaneous Markov chains were run for 2,000,000 generations and trees were sampled every 200th generation. The distribution of log-likelihood scores was examined to determine stationary phase for each search and to decide if extra runs were required to achieve convergence, using the program Tracer 1.5 [27]. First 20% of generated trees were discarded and remaining 80% of trees were used to calculate posterior probabilities of the majority rule consensus tree. BYPP greater than 0.95 are given above each node (Fig 1).
Maximum likelihood trees were generated using the RAxML-HPC2 on XSEDE (8.2.8) [28,29] in the CIPRES Science Gateway platform [30] using GTR+I+G model of evolution. Maximum likelihood bootstrap values (ML) equal or greater than 70% are given above each node (Fig 1).

Compliance with ethical standards
The authors declare that there is no conflict of interest (financial or non-financial) and agree to the submission of this paper. The authors also confirm that the fieldwork did not involve endangered or protected species and no specific permissions were required for these locations as the land belongs to the Kunming Institute of Botany, which is the first affiliated institution on this paper.

Nomenclature
The electronic version of this article in Portable Document Format (PDF) in a work with an ISSN or ISBN will represent a published work according to the International Code of Nomenclature for algae, fungi, and plants, and hence the new names contained in the electronic publication of a PLOS ONE article are effectively published under that Code from the electronic edition alone, so there is no longer any need to provide printed copies.
In addition, the new name contained in this work has been submitted to Index Fungorum from where they will be made available to the Global Names Index. The unique Index Fungorum number can be resolved and the associated information viewed through any standard web browser by appending the Index Fungorum number contained in this publication to the prefix www.indexfungorum.org/. The online version of this work is archived and available from the following digital repositories: PubMed Central, LOCKSS.

Phylogenetic analysis
The concatenated dataset (SSU, LSU, RPB2 and TEF loci) consisted of 180 strains including the new taxa proposed in this study, with Diatrype disciformis (AFTOL-ID 927), Graphostroma platystoma (CBS 270.87) and Sordaria fimicola (AFTOL-ID 216) as the out-group taxa. Topologies of trees (ML and BI) derived from analyses of single gene dataset were compared and the overall tree topology was congruent to those obtained from the combined dataset.
The RAxML analysis of the combined dataset yielded a best scoring tree (  Table 2. The Bayesian analysis resulted in 10001 trees after 2000000 generations. The first 1000 trees, representing the burn-in phase of the analyses, were discarded, while the remaining 9001 trees were used for calculating posterior probabilities in the majority rule consensus tree. Forty-one families and two suborders (Massarineae, Pleosporineae) in Pleosporales are represented in the phylogenetic tree (Fig 1). The topologies resulting from the ML and BI analyses (Fig 1) are generally congruent with the results reported by [19] and other large-scale phylogenies of Dothideomycetes e.g. [32]. Multigenic (Fig 1) analyses agree to support significantly the monophyletic origin of every interleaved family and the two suborders in Pleosporales. Six newly isolated strains from this study (MFLUCC 20-0016, MFLUCC 20-0017, MFLUCC 20-0018, MFLUCC 20-0019, MFLUCC 20-0020 and MFLUCC 20-0021) constituted three lineages in Clade A, B and C (Fig 1). MFLUCC 20-0020 and MFLUCC 20-0021 grouped within Nigrogranaceae (Clade A, Fig 1) as a well-supported monophyletic clade with 100% ML 1.00 BYPP support (Fig 1). This is closely related to Nigrograna antibiotica (CCF 4378T), N. carollii (CCF 4884) and N. peruviensis (CCF 4485), but the corresponding affiliation is not statistically supported (Subclade A1, Fig 1). MFLUCC 20-0016 grouped in Occultibambusaceae (Clade B,  Fig 1) within Pleosporales, which is regarded as Longiostiolaceae. Notes: Clade C (Fig 1) comprises Crassiperidium, Longiostiolum and Shearia, which are phylogenetically highly supported and belong to Longiostiolaceae in the Pleosporales (Fig 1). [33] recently introduced Longiostiolaceae to accommodate Crassiperidium and Longiostiolum. In this study, Shearia has also proven to be a genus in Longiostiolaceae (Clade C1 , Fig 1), and we hereby amend the family description in order to accommodate its morphological characteristics.
Longiostiolaceae has a close phylogenetic affinity to Cyclothyriellaceae in large-scale multigene phylogenetic analyses of Pleosporales. But they are not grouped in a monophyletic clade, rather forming discrete clades adjacent to each other. The asexual characteristics of this new family distinguish it from Cyclothyriellaceae. Cyclothyriellaceae has 1-celled conidia, whereas Longiostiolaceae has multi-celled conidia. Morphologically, Crassiperidium is most similar to Pseudoasteromassaria [34] in its cylindrical, multi-septate, hyaline conidia. However, their ascospores are different, and phylogenetically, both of them are not closely associated. Shearia shares some morphological resemblances to camarosporium-like and stegonsporiopsis-like taxa by its pycnidial conidiomata, holoblastic annellidic conidiogenous cells and distoseptate, pale pigmented conidia [35]. However, neither camarosporium-like nor stegonsporiopsis-like taxa are phylogenetically closely related to Shearia. Notes: [18] established the genus Crassiperidium to accommodate two new species, C. octosporum and C. quadrisporum from Fagus crenata in Japan. Crassiperidium is characterized by 'globose to depressed globose ascomata with a well-developed ascomatal wall at the sides, clavate asci, broadly fusiform, hyaline ascospores, pycnidial conidiomata, and cylindrical, multiseptate, hyaline conidia produced by annellidic conidiogenous cells' [18]. Phylogenetic analyses of multi-genes show that Crassiperidium is closely affiliated with Cyclothyriellaceae (Pleosporales, Dothideomycetes), but the exact familial placement was uncertain. Morphologically, Crassiperidium is similar to Pseudoasteromassaria by its ascomatal, pycnidial and conidial characteristics [18,34]. Also, both are recorded from the same host plant (Fagus crenata).

Accepted genera in Longiostiolaceae
However, phylogenetically, Pseudoasteromassaria belongs to Latoruaceae and is not closely related to Crassiperidium (Fig 1). In addition, their ascospores and conidiogenous cells differ [18,36]. Notes: Longiostiolum is introduced by [37] as a monotypic genus in the suborder Massarineae with L. tectonae as the type species. The genus is characterized by black, immersed to semi-immersed, uniloculate, globose to subglobose ascostromata with a long central ostiole and phragmosporous ascospores. Longiostiolum was an incertae sedis genus in Pleosporales and [38] listed this in Massarinaceae. In this study, Longiostiolum tectonae grouped sister to Shearia in Longiostiolaceae (Clade C, Fig 1).  [44]. However, the asexual morph of Pleomassaria maxima was reported as Shearia formosa [45]. There are no sequence data available for Pleomassaria maxima (= Splanchnonema maximum) in GenBank, thus it is not possible to check its phylogenetic relationship with other taxa in Pleosporales. With the availability of molecular data, we have included Pleomassaria siparia (CBS 279.74) and Splanchnonema platani (CBS 222.37) in our phylogenetic analyses to represent Pleomassaria and Splanchnonema. However, they are not closely related to each other (Fig 1), whereas CBS 279.74 grouped in Pleomassariaceae and CBS 222.37 grouped in Massarineae.
During this study of microfungi from Magnolia, we recollect Shearia fermosa from Kunming. Below we illustrate and re-describe our new collection as a neo-type. In multi-gene analyses (Fig 1)
Culture characteristics: On PDA, colonies reached up to 40 mm diam after 12 d at 18˚C. Colony dense, circular, slightly raised, surface smooth, with serrate edge, floccose, greenish grey at the center and brown towards margin from the top and reverse dark brown.  Notes: The new fungus was collected from Magnolia denudata in Kunming. It can be labeled a typical Nigrograna taxa based on its ascomata, asci and ascospore characteristics. Phylogenetically it has a close affinity to Nigrograna antibiotica, N. carollii and N. peruviensis, nonetheless this relationship is statistically not supported ( Subclade A1, Fig 1). These three species are reported as endophytes from Peru in the phloem of living Ulmus laevis, on living sapwood of wild Hevea brasiliensis, and on living sapwood of wild Virola sp. respectively [54]. All of these species are known only from their culture characteristics, precluding comparing the morphology of either sexual or asexual morph with our new collection. We were able to isolate both ascospores and conidia from the fruiting structures on the natural host. Both single spore isolation of ascospore and conidia formed identical culture morphologies on PDA (Figs 3p, 3q, 4h and 4i) and the DNA based sequences derived from those cultures were also similar in comparisons. Therefore, we introduce Nigrograna magnoliae sp. nov. providing both asexual and sexual morphs.  [56]. Brunneofusispora was introduced by [57] to accommodate B. sinensis and it was known only from its sexual morph. We collected Brunneofusispora sinensis from Magnolia and in this study, we report the coelomycetous asexual morph from the culture. Hence, we amend Brunneofusispora in order to accommodate its asexual morph. Conidiomata 120-160 μm high, 80-120 μm diam. (� x ¼ 144:1 � 107:4 μm, n = 5), solitary, superficial, globose to sub-globose, unilocular, dark brown. Conidiomata wall composed of thick-walled, very dark brown cells of textura angularis. Conidiophores reduced to conidiogenous cells. Conidiogenous cells 6-7.5 μm long, 2.5-3 μm wide (� x ¼ 6:7 � 2:7 μm, n = 10), enteroblastic, phialidic, ampulliform, doliiform or cylindrical, discrete, indeterminate, hyaline, smooth-walled. Conidia 3-4 μm × 1.9-2.5 μm (� x ¼ 3:6 � 1:99 μm, n = 25), mostly cylindrical or ovoid, 1-celled, hyaline, smooth, guttulate.  Notes: In this study we have acquired DNA from the mycelium of a sexual morph and in multi-gene phylogeny, our novel strain and the Brunneofusispora sinensis group in a monophyletic clade (subclade B1, Fig 1). Even though B1 is not strongly supported, there were only three bp differences (not including gaps) in the comparison of the 525 nucleotides across the ITS regions. Morphological characteristics i.e. asci and ascospores, are not significantly different to each other in shape or dimensions [57]. Therefore, we identify our collection as Brunneofusispora sinensis. We observed its asexual morph from the culture and therefore we amended the generic and species descriptions herein. Nevertheless, it is worthy to remark that we note the ascospores of our new isolate were remaining hyaline at maturity whereas brownish in the holotype of Brunneofusispora sinensis.
The distribution of Magnolia is variable but concentrated particularly around temperate and tropical South East and East Asia. They are treasured around the world as ornamental trees due to their attractive flowers and foliage and are used as timber and medicine by local and international communities. However, 48% of all Magnolia species are endangered and facing habitat loss [61]. The Kunming Botanical Garden functions as an ex-situ conservation for endangered, endemic and economically important plant species native to the Yunnan Plateau and the southern Hengduan Mountains [6]. Micro-fungi on Magnolia have been studied for decades around the world, and so far over 1000 fungal species have been reported on this host [62]. However, only 46 [62]. This study highlights the Kunming Botanical Garden as an ideal locale to conduct diverse research on the micro-fungal occurrences on Magnolia as it is a repository of a great number of Magnolia species.
Supporting information S1 Table. Taxa used in the phylogenetic analyses and their corresponding GenBank numbers. The newly generated sequences are indicated in bold. (DOC)