More than one fungus in the pepper pot: Integrative taxonomy unmasks hidden species within Myriostoma coliforme (Geastraceae, Basidiomycota)

Since the nineteenth century, Myriostoma has been regarded as a monotypic genus with a widespread distribution in north temperate and subtropical regions. However, on the basis of morphological characters and phylogenetic evidence of DNA sequences of the internal transcribed spacer (ITS) regions and the large subunit nuclear ribosomal RNA gene (LSU), four species are now delimited: M. areolatum comb. & stat. nov., M. calongei sp. nov., M. capillisporum comb. & stat. nov., and M. coliforme. Myriostoma coliforme is typified by selecting a lectotype (iconotype) and a modern sequenced collection as an epitype. The four species can be discriminated by a combination of morphological characters, such as stomatal form, endoperidial surface texture, and basidiospore size and ornamentation.


Introduction
Correct species recognition is an essential requirement for the understanding of systematics, evolution and ecology. Furthermore, it is a prerequisite for population biological studies, reliable Red List assessments and effective conservation action. Recent molecular studies suggest that the magnitude of fungal taxonomic diversity is seriously underestimated [1][2][3]. Basidiomycete taxonomy has been revolutionized by the use of molecular techniques, which have been particularly valuable in revealing component cryptic or semi-cryptic taxa within species complexes or aggregates [4][5][6]. The drawbacks associated with the traditional morphologyonly approach are succinctly expressed by Stielow et al. [7]: "The difficulties in defining characters and their states, and particularly the fact that distinct taxonomists assigned distinct PLOS

Morphological studies
The morphological analyses were performed on specimens, including types, deposited in the Fungal Collection of the Federal University of Rio Grande do Norte (UFRN Herbarium), the collection of fungi of the Real Jardín Botánico of Madrid (MA-Fungi), the cryptogamy collection (PC) at the Herbarium of the Muséum national d'Histoire naturelle (MNHN-Paris), and the Fungarium of the Royal Botanic Gardens, Kew (K), (Table 1). Macromorphological studies were based on 26 exsiccates using a Nikon H600L stereomicroscope coupled with a Nikon DS-Ri camera for image capture. Colour descriptions followed Kornerup and Wanscher [25]. For micromorphological features, such as basidiospores, eucapillitium and exoperidial hyphae, a Nikon Eclipse Ni light microscope (LM) coupled with a Nikon DS-Ri camera was used. Basidiospore measurements were made at 1000× magnification following Sousa et al. [21] and include ornamentation. Scanning electron microscopy (SEM) was used to observe the patterns of ornamentation on basidiospores, eucapillitium and endoperidial surfaces.

DNA extraction, PCR amplification and sequencing
Genomic DNA was extracted from approximately 10 mg of gleba of mature dry basidiomata. The DNeasy TM Plant Mini Kit (Qiagen, Valencia, CA) was used to isolate DNA from UFRN and MA-Fungi specimens, following the manufacturer's instructions with the following modifications: glebal masses were macerated in 1.5 ml tubes with a micropestle before suspension in lysis buffer and again after overnight incubation at 55-60˚C. Both ITS and the 5'-1450-base region of the LSU were analysed using the primer pairs ITS1F/ITS4 [26][27] and LR0R combined with LR7 or LR5 [28][29]  conditions followed Martín and Winka [30]. PCR products were verified on 1% agarose gels (UtraPureTM Invitrogen), purified using ExoSAP-IT1 (USB Corporation, OH, USA) and sequenced bidirectionally in Macrogen Inc. (Seoul, South Korea). DNA from specimens K(M)138625 and K(M)61641 was extracted using an enzymatic digestion-glass fibre filtration protocol in 96-well plate format with a vacuum-manifold as described in Dentinger et al. [31]. PCR amplifications and sequencing were performed following Dentinger and Suz [32]. DNA from the rest of the specimens from the Kew Fungarium was extracted and ITS and LSU regions amplified using Extract-N-Amp (Sigma, Dorset, UK).
The resulting sequences were edited and the consensus sequence was obtained using Sequencher 5.2.4 (Gene Codes Corp., USA). Preliminary identifications were performed through megablast searches comparing the newly-generated sequences with those in GenBank [33]. Sequences were submitted to GenBank under the accession numbers indicated in Table 1.

Sequence alignments and phylogenetic analyses
Both ITS and LSU sequences were aligned separately using Se-Al v. 2.0a11 Carbon [34]. To infer phylogenetic relationships among Myriostoma specimens, homologous sequences retrieved from the EMBL/GenBank/DDBJ databases were included in the alignment [35]. Since Geastrum is the sister genus of Myriostoma [36], two sequences of Geastrum saccatum Fr. were included as outgroup.
Where ambiguities in the alignment occurred, the alignment generating the fewest potentially informative characters was chosen [37]. Alignment gaps were marked "-", unresolved nucleotides and unknown sequences were indicated with "N". Three types of analyses were carried out for ITS and LSU individual alignments and the combined ITS/LSU alignment: maximum parsimony (MP), maximum likelihood (ML), and Bayesian inference. The combined ITS/LSU alignment was submitted to the TreeBASE Number.
In the MP analyses, minimum length Fitch trees were constructed using heuristic searches with tree-bisection-reconnection branch swapping, collapsing branches if maximum length was zero, with the MulTrees option in PAUP Ã 4.0b10 [38], and a default setting to stop the analyses when reaching 100 trees. Gaps were treated as missing data. Nonparametric bootstrap (MPbs) support [39] for each clade, based on 10,000 replicates using the fast stepwise-addition, was tested [40]. The consistency index, CI [41], retention index, RI [42], and rescaled consistency index, RC [42], were obtained. The ML approach was carried out using RAxML [43] in the CIPRES portal (CIPRES Science Gateway v.3.3) assuming a GTR+I+G model as selected by PAUP Ã 4.0b10; MLbs support for each clade, based on 1,000 replicates was tested. The Bayesian analysis [44][45] was performed using MrBayes 3.2 [46], and assuming the general time reversible model [47], including estimation of invariant sites and assuming a discrete gamma distribution with six categories (GTR+I+G), as selected by PAUP Ã 4.0b10. Two independent and simultaneous analyses starting from different random trees were run for 2.000.000 generations with four parallel chains and trees and model scores saved every 100th generation. The default priors in MrBayes were used in the analysis. Every 1.000th generation tree from the two runs was sampled to measure the similarities between them and to determine the level of convergence of the two runs. The potential scale reduction factor (PSRF) was used as a convergence diagnostic and the first 25% of the trees were discarded as burn-in before stationary was reached. The 50% majority-rule consensus tree and the posterior probability (PP) of the nodes were calculated from the remaining trees with MrBayes. A combination of both bootstrap proportion and PP was used to assess the level of confidence for a specific node [2,48]. The phylogenetic trees were visualized using FigTree v. 1.3.1 (http://tree. bio.ed.ac.uk/software/figtree/) and edited with Adobe Illustrator CS3 v. 11.0.2 (Adobe Systems).

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, new names contained in this work have been submitted to MycoBank, from where they will be made available to the Global Names Index. The unique MycoBank number can be resolved and the associated information viewed through any standard web browser by appending the MycoBank number contained in this publication to the prefix at http://www. mycobank.org/MB. The online version of this work is archived and available from the following digital repositories: PubMed Central, LOCKSS and Digital-CSIC.

Results
This study generated 32 new Myriostoma sequences ( Table 1) The LSU dataset included 21 sequences, 15 generated in this study and six obtained from sequence databases. The alignment resulted in 1391 unambiguously aligned nucleotide positions (1302 constant, 20 parsimony-uninformative, and 69 parsimony-informative). The 100 most parsimonious trees gave a length of 100 steps, CI = 0.9300, HI = 0.0700 and RC = 0.9533. The ML tree and the 50% Bayesian majority rule combined consensus tree (not shown) showed essentially the same topology as the parsimony strict consensus tree (not shown). In the three analyses, the sequences from South Africa [K(M)205482, K(M)205483 and K(M) 205540] were sister to the other Myriostoma sequences, although this relationship was weakly supported (MPbs = 52%, MLbs = <50%, PP = 0.65). In the parsimony strict consensus tree, the sequence from Costa Rica appeared as the sister group to those from Argentina and Brazil, but this relationship had very low support (MPbs = 61%); moreover, in the ML and Bayesian analyses, the sequence from Costa Rica was the sister group to the clade formed by sequences from Europe and Hawai'i, a relationship with moderate support (MLbs = 54%, PP = 0.91).
In the ITS/LSU combined dataset there were 2004 unambiguously aligned nucleotide positions (1797 constant, 61 parsimony-uninformative, and 146 parsimony-informative). The 100 most parsimonious trees gave a length of 235 steps, CI = 0.9234, HI = 0,0766, and RC = 0.9474. The ML tree (not shown) and the 50% Bayesian majority rule combined consensus tree (Fig 1)   Fig 1. The 50% majority-rule consensus tree of ITS/LSU nrDNA sequences of Myriostoma species using a Bayesian approach. Two sequences of Geastrum saccatum were used as outgroup. Terminal branches are labelled with appropriate specimen codes and countries of origin. For further specimen details, see Table 1 Furthermore, the K2P pairwise distance of Myriostoma ITS sequences included in Table 1 show high genetic variation between the four species considered ( Table 2). There are clearly defined barcoding gaps within the ITS sequences of Myriostoma such that interspecific variation exceeds intraspecific variation [8]. Based on these results, a new species is described and two varieties are elevated to specific rank. No type material of M. coliforme is known [13,15] and as Persoon [49] referred to Dickson's (not Withering's) name [50], thereby sanctioning it, Dickson's illustration is selected as lectotype and a recently sequenced collection from the same English region (East Anglia) is designated as epitype (see below). Diagnosis. Myriostoma calongei differs from other Myriostoma species mainly by the verrucose endoperidium, with prominent triangular processes (warts 0.13-0.28 mm high). It is closely related to M. capillisporum, but M. calongei has smaller basidiospores (5.9-8.7 μm diam) with less prominent ornamentation (1.0-2.3 μm high).

Discussion
Since the nineteenth century, the genus Myriostoma has been regarded as monotypic. Pegler et al. [14] indicated that M. coliforme is widespread in north temperate and subtropical regions. However, our study reveals that the name M. coliforme has been applied to at least four members of a species complex each of which is well characterized by a combination of morphological characters, of which the stomata, endoperidial surface and spore size and ornamentation are the most important. Consequently, the distribution of M. coliforme in the original sense has been overestimated (IUCN webpage: http://iucn.ekoo.se/iucn/species_view/ 122233/; [22] ; Fig 7). Although further worldwide sampling is clearly required, current DNAbased evidence supports a European and North American range for M. coliforme.
A lack of knowledge about dispersal mechanisms coupled with insufficient molecular data on Neotropical fungi have resulted in speculative interpretations of their biogeographic distribution, especially for saprotrophic taxa such as Myriostoma [20,65]. Recent studies demonstrate that fungal species with a cosmopolitan distribution are the exception [66,67]. In general, most names applied to species with an apparent worldwide distribution probably represent species complexes rather than good species [68]. Based on the conclusion of Kasuya et al. [20] regarding the earthstar Geastrum triplex Jungh., which has a similar, bellows-like, spore dispersal mechanism, Myriostoma dispersal capacity is not expected to be very effective over long distances.
This work opens new perspectives on this striking genus through the application of integrative taxonomy using a combined molecular and morphological approach. Indeed, the production of revised distribution maps of Myriostoma species is an essential prerequisite for ecological studies and robust and reliable IUCN-compliant conservation assessments.