Study of three interesting Amanita species from Thailand: Morphology, multiple-gene phylogeny and toxin analysis

Amanita ballerina and A. brunneitoxicaria spp. nov. are introduced from Thailand. Amanita fuligineoides is also reported for the first time from Thailand, increasing the known distribution of this taxon. Together, those findings support our view that many taxa are yet to be discovered in the region. While both morphological characters and a multiple-gene phylogeny clearly place A. brunneitoxicaria and A. fuligineoides in sect. Phalloideae (Fr.) Quél., the placement of A. ballerina is problematic. On the one hand, the morphology of A. ballerina shows clear affinities with stirps Limbatula of sect. Lepidella. On the other hand, in a multiple-gene phylogeny including taxa of all sections in subg. Lepidella, A. ballerina and two other species, including A. zangii, form a well-supported clade sister to the Phalloideae sensu Bas 1969, which include the lethal “death caps” and “destroying angels”. Together, the A. ballerina-A. zangii clade and the Phalloideae sensu Bas 1969 also form a well-supported clade. We therefore screened for two of the most notorious toxins by HPLC-MS analysis of methanolic extracts from the basidiomata. Interestingly, neither α-amanitin nor phalloidin was found in A. ballerina, whereas Amanita fuligineoides was confirmed to contain both α-amanitin and phalloidin, and A. brunneitoxicaria contained only α-amanitin. Together with unique morphological characteristics, the position in the phylogeny indicates that A. ballerina is either an important link in the evolution of the deadly Amanita sect. Phalloideae species, or a member of a new section also including A. zangii.


Introduction
Amanita sect. Phalloideae (Fr.) Quél., is characterized by a non-appendiculate and non-striate pileus, a persistent partial veil, and a limbate or saccate volva on the bulbous stipe base [1][2][3][4]. The diagnostic microscopic features include amyloid basidiospores and a pileipellis composed of gelatinous filamentous hyphae [1][2][3][4]. The type species of this section is the "Euro-Asian Death Cap", A. phalloides (Fr.) Link., first described from Europe [5]. The sect. Phalloideae currently comprises approximately 61 taxa that have been described worldwide [6,7]. The cyclooligopeptide toxins found in members of sect. Phalloidae, which include amatoxins, phallotoxins, and virotoxins, have been studied for over a decade [8][9][10]. The amatoxins, including α-amanitin and β-amanitin, are considered ten times more deadly than phallotoxins [11]. Members of sect. Phalloideae are consumed after mistaking them for edible species. In Asia, confusion most often occurs with other Amanita spp. in sect. Caesareae Singer ex Singer, which contains a number of edible species [12]. Recently, a multiple gene phylogenetic analysis rigorously examined species in sect. Phalloideae [13,14] and sect. Caesareae [15], with the aim of exploring biogeographic history and plesiomorphies that may help to resolve evolutionary relationships among deeper clades of these two sections. In order to understand the evolutionary relationships in sect. Phalloideae, researchers have begun to critically evaluate the toxins, toxin encoding genes, and biogeography to support the taxonomy [11,13,14,15].
In a recent publication resulting from biodiversity studies in northern Thailand we documented two first records of sect. Phalloideae, A. rimosa P. Zhang & Zhu L. Yang with white basidiomata and A. zangii Zhu L. Yang, T.H. Li & X.L. Wu [16], although the placement of the latter in sect. Phalloideae is controversial because of very atypical morphology [13,16]. In the present paper, we report two additional species in Amanita sect. Phalloideae, and a third species whose taxonomic placement is also problematic, with two of the species new to science The macro-and micro-morphology of those three taxa are described, and photographs and line drawings are provided. Moreover, their phylogenetic affinities are discussed based on multigene phylogenetic analyses (rpb2, β-tubulin, nrITS, and nr5.8S). Finally, we report on the screening of the three species for toxins known to occur in Amanita, including α-amanitin and phalloidin.

Collections
Specimens were collected mainly in forests dominated by Fagaceae (Castanopsis, Lithocarpus, Quercus) and/or Dipterocarpaceae (Dipterocarpus, Shorea) during the rainy season. The rainy season in the north (including Chiang Mai and Chiang Rai provinces) normally occurs from May to beginning of October, whereas the rainy season in the south (including Song Khla Province) occurs from May to December. Fresh specimens were photographed and described, and then dried using a food dehydrator (at ca. 40-50˚C). Tissue samples were taken from fresh basidiomata with aseptic technique and kept in 10% CTAB (cetyl trimethylammonium bromide) for later DNA analyses. The examined specimens were deposited in either one or two of the following herbaria: Mae Fah Luang University, Thailand (MFLU), Botanic Garden Meise, Belgium (BR), and Chiang Mai University, Thailand (SDBR-CMU). Herbarium codes follow Index Herbariorum [17], with the exception of "RET", which is the code adopted for R.E. Tulloss' Herbarium Rooseveltensis Amanitorum, and HKAS, which stands for the Cryptogamic Herbarium of Kunming Institute of Botany. All author citations of species rank not included in the main body of the text are located in Table 1.

Morphological study
Macro-morphology was described from fresh specimens. Color codes are according to Kornerup & Wanscher (1978). The colour change reaction with 10% KOH was tested on the stipe surface of A. ballerina, as is done in other species with white basidiomata [18]. Microscopic features were studied from dried tissue mounted in H 2 O and 5% KOH aqueous solution. Congo red was used for highlighting all tissues, and amyloidity of basidiospores was observed using Melzer reagent. Dimensions of microscopic structures were measured using Image Frame Work (Tarosoft1, Thailand). Basidiospore measurements are accompanied with the following notation: "[n/m/p]", which indicates that n basidiospores were measured from m basidiomata of p collections, with a minimum of 25-50 basidiospores from each basidiome. Size and shape of basidiospores are presented in a form following the description of ranges for biometric variables according to Tulloss [18] (a-) b-c (-d), in which b represents the 5 th percentile, c, the 95 th percentile, while a and d are the lowest and highest extreme values measured, respectively. The average length is indicated as L' whereas W' is the average width. The range of length/width ratio of basidiospores (Q) is provided. In addition to Tulloss' standard format, standard deviationis provided for Q' (the mean of all Q values computed for a single taxon). Faces of Fungi [19], Index Fungorum [20] and MycoBank [21] numbers are provided.

DNA isolation, amplification and sequencing
Specimens were processed for molecular analyses at two core facilities including the Botany Department, University of Wyoming (UW), USA and Botanic Garden Meise, Belgium, using a variety of methodologies for extraction of genomic DNA, PCR and sequencing. Genomic DNA extractions at the University of Wyoming were performed using a CTAB protocol with phenol-chloroform-isoamyl alcohol purification, followed by cleaning with a silica-matrix binding procedure [22]. At Botanic Garden Meise, DNA extractions were performed using a slightly different CTAB protocol. PCR amplification of ITS (nuclear ribosomal internal transcribed spacer) and LSU (large subunit ribosomal DNA) was performed using the primer pairs ITS4/ITS5 or ITS1-F/ITS4, and LR0R/LR5, respectively. Parts of the protein-coding genes βtubulin and rpb2rpb2 (second largest subunit of RNA polymerase II) were amplified using the primer pairs Am-β-tub-F/Am-β-tub-R and Am-6F/Am-7R, respectively [13]. Purified PCR products were then sequenced at the Nucleic Acid Exploration Facility at the University of Wyoming on an ABI 3130 XL DNA analyzer (Applied Biosystems), or at Macrogen Europe (Amsterdam) on an ABI 3730 XL DNA analyzer (Applied Biosystems), using the same primer combinations as for PCR, except for Am-β-tub-F, which was replaced by the shorter primer Am-β-tub-F-Seq (5'-CGGAGCRGGTAACAAYTG-3'). Forward and reverse reads were assembled and edited with Geneious Pro 5.1.7 (Biomatters Ltd., Auckland, New Zealand).

DNA sequence dataset assembly
Seventy-six sequences of collections from Thailand were newly generated for this study and deposited in GenBank (GB) (http://www.ncbi.nlm.nih.gov/; Table 1). Initial BLAST searches (http://blast.ncbi.nlm.nih.gov) of both LSU and ITS1+5.8S+ITS2 sequences were performed to estimate similarity with Amanita sequences already in GB. Additional sequences were identified using phylogenetic inferences and were also retrieved from GB ( Table 1). The quality of the sequences was considered in selecting sequences from GB for use in the analyses. Because some of the species studied here, namely Amanita ballerina sp. nov. and A. zangii, have morphological characters atypical for sect. Phalloideae and/or have been successively placed in different sections in the past, the taxon sampling was performed to cover not only section Phalloideae, but also the three other sections in subgen. Lepidella, namely sect. Lepidella, sect. Validae, and sect. Amidella, as well as five taxa from subgen. Amanita, sect. Amanita selected as outgroup. Because LSU is of restricted phylogenetic utility (see e.g., Tulloss [23] for a brief discussion), it was not included in the phylogenetic analyses.

Phylogenetic analyses
Sequences were initially aligned with MAFFT v.7.0 [24] using the G-INS-i iterative refinement algorithm, with minimal manual adjustment in BioEdit v.7.0.9 (Hall 1999). Introns of proteincoding genes were excluded from the analyses, mainly because they were not included in most of the sequences retrieved from GenBank. For the ITS region, the different loci of the region were identified on the basis of terminal motifs of 18S, 5.8S and 28S loci catalogued [25]. Only the positions corresponding to ITS1, 5.8S, and ITS2 were kept in the alignment. The program Gblocks v0.91b [26] was then used to exclude poorly aligned positions of the ITS alignment with the following parameter settings: minimum number of sequences for a conserved position = 24 (minimum possible); minimum number of sequences for a flank position = 24 (minimum possible); maximum number of contiguous non-conserved positions = 4 bp, minimum block size = 4 bp, and gaps allowed within selected blocks in half of the sequences. Phylogenetic tree inference was performed using both Maximum Likelihood (ML) and Bayesian Inference (BI). The ML analyses were performed using RAxML-HPC2 [27] on the CIPRES Science Gateway [28], with default settings except the number of bootstrap replicates was set to 1,000 for both single-gene and combined gene analyses. Phylogenetic inference was first performed on each single-gene alignment, and, since no significantly supported conflict [with ML Bootstrap Support (BS) ! 70%] was detected, multiple-genes alignments and trees were built. Because ITS sequences were not alignable with reasonable confidence over the whole set of OTUs including sect. Amanita as outgroup, the four-gene alignment (β-tub, rpb2, ITS1+ITS2, and 5.8S) was restricted to sections Phalloideae and Valideae, the latter being used as the outgroup. BI was performed using MrBayes v.3.2.6 [29], with a mixed model partition. The best substitution model was determined for each partition of the data set separately using jModeltest v. 2 [30] on jmodeltest.org, with default parameters. When the best model could not be specified in MrBayes, the next more complex model was used. The selected models were K80 + G for β-tub, GTR + I + G for rpb2, GTR + G for ITS1 and ITS2, and 5.8S. The Bayesian analyses were conducted with 2 runs, each with five simultaneous Markov chains, and trees were summarized every 250 generation. The analyses were stopped after 1,000,000 generations, when the average standard deviation of split frequencies was 0.004999 for the two-gene analysis, and 0.002576 for the four-gene analysis. The burn-in phase (20% and 25% for the two-gene and four-gene analysis, respectively) was estimated by checking the stationarity in the plot generated by the sump command. The remaining trees were used to generate a majority-rule consensus tree and to compute corresponding posterior probabilities. Phylograms were visualized with FigTree ver. 1.3.1 [31].

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 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 http://www. mycobank.org/MB/. The online version of this work is archived and available from the following digital repositories: PubMed Central and LOCKSS.

Toxin analysis
Dried basidiomata of Amanita ballerina (OR1026), A. brunneitoxicaria (BZ2015-01), and A. fuligineoides (BZ2013-41) were analyzed for major toxins including α-amanitin and phalloidin. The basidiomata were pulverized to a fine powder and ca. 100 mg were extracted in methanol for 30 min. in an ultrasonic bath and the resulting liquid filtered through filter paper. The liquid phase of methanol was then evaporated to dryness. The crude extracts were dissolved in methanol and transferred to 4 mL glass vials, and dried under nitrogen, then weighed. Then crude extracts were again re-dissolved in methanol. Crude extracts were filtered by using SPME Strata™-X 33 u Polymeric RP cartridges (Phenomenex, Inc., Aschaffenburg, Germany) and analysed using high performance liquid chromatography (HPLC). Extracts profiles were compared to standards based on mass spectrum in the negative and positive ESI modes, as well as on their characteristic UV/V is and retention time (R t ). Standards of α-amanitin and phalloidin were obtained from Sigma-Aldrich, Germany.

DNA sequence analyses
In this study, each of the single-gene phylogenies showed similar tree topologies (Supporting Information S1-S3 Figs), without supported conflicts. The two-gene (β-tub and rpb2) and four-gene (β-tub, rpb2, ITS1+ITS2, and 5.8S) alignments contained 63 and 46 OTUs, and were 924 and 1,443 sites long, respectively. For the ITS1 + ITS2 alignment, Gblocks retained 375 sites (57% of a total of 653 sites). The topology of the two-gene and four-gene trees obtained in this study are consistent with previously published trees with similar taxon sampling [6,13]. Interestingly, Amanita ballerina sp. nov., A. sp. HKAS77321, and A. zangii formed a clade that is sister to the Phalloideae sensu Bas [1], i.e. the death caps. This clade, however, was highly supported only in the four-gene tree (BS = 94%, PP = 0.97). In the two-gene tree BS was only 38% and the clade was not retrieved in the BI tree, where A. zangii was sister to the lethal amanitas, while A. ballerina and A. sp. HKAS77321 formed a sister clade to the poorly supported (PP = 0.73) clade comprised of A. zangii and the Phalloideae sensu Bas [1]. A. brunneitoxicaria sp. nov. forms a well-supported clade with A. fuligineoides (BS = 100%, PP = 1.0). Initial BLAST searches of both LSU and ITS1+5.8S+ITS2 sequences are given in Table 2. In this study, the tree topologies obtained from ML and BI did not show any supported conflict. Therefore, only the ML trees are shown (Figs 1 and 2 Etymology:'ballerina' refers to the cottony skirt-like partial veil, and downward tapering of the bulb, reminiscent of a ballet dancer silhouette (Fig 3).
Basidiomata (Fig 3A and 3D) small-sized. Pileus 34-42 mm wide, subglobose to hemispheric when young, convex to broadly convex at maturity, sometimes slightly depressed at center, dry, slightly viscid when moist, with surface under remains of general veil minutely floccose at first, becoming felted or chamois-like in age, dull white (1A2), universal veil a thin cottony layer breaking into small adherent areolate squamules or large flaky patches that are dull white to pale yellowish white (2A2); margin plane, non-striate at first then striate at maturity, non-appendiculate; context 2-3 mm thick above stem, soft to slightly hard, dull white (1A2). Lamellae 4-6 mm broad, narrowly sinuate, attachment to stipe, occasionally forking, close, dull white to pale yellowish white (3A2) at maturity; lamellulae of 2-3 lengths, rounded to nearly truncate. Stipe 43-75 × 10-26 mm (not including the bulb), nearly cylindrical, with short decurrent line at apex, bulbous, dull white, covered with fine flocculae; context stuffed, thin, yellowish white or cream (4A2). Bulb compressible, slightly subglobose when young, then elongate to ventricose and tapering downward, marginate up to 13 mm wide, occasionally cleft, dull white to white (2A2). Universal veil on stipe base a volval limb, up to 4 mm high on bulb margin, cottony-felted, white to dull white (2A2), or with patches dull white to dirty white (1A2). Partial veil medial, 25-35 mm below apex of stipe, persistent, cottony, skirt-like, with a ragged thickened edge, striate inside, white to dull white (1A2). Spores white in deposit. Odor and taste not recorded. 10% KOH tested not turned to yellow in dried specimens.
Phylogenetically, A. ballerina does not show a close relationship with either members of the Phalloideae with white basidiomata that contain deadly toxins or with sect. Lepidella. In the multigene phylogenetic tree, Amanita ballerina, A. zangii, and A. sp. HKAS77321 collected from tropical East Asia [13]  Etymology:'brunneitoxicaria' refers to the brown basidiomata and presence of deadly toxins.
Known distribution: Currently only known from southern Thailand. Remarks: Amanita brunneitoxicaria is characterized by its small dark greyish brown basidioma, mouse grey to bronze virgate surface that is darkest at center, greyish brown towards pileus margin, subglobose to slightly elongated bulb with saccate volva, and the wavy edge of the apical partial veil. Morphologically, A. brunneitoxicaria is similar to A. fuliginea, A. fuligineoides, and A. subfuliginea in having small to medium basidiomata with innately virgate pileus surface and 4-spored basidia. Amanita fuliginea is most similar but differs by its globose to subglobose basidiospores. In addition, A. fuliginea was reported to contain both α-amanitin and phalloidin [15], whereas our study of A. brunneitoxicaria revealed the presence of α-amanitin only. A. fuligineoides and A. subfuliginea have a brownish stipe covered with fibrillose squamules in a spiral or concentric pattern, while A. brunneitoxicaria has a stipe that is white to pale greyish brown, and the fine flocculae do not form a spiral or concentric pattern. Amanita griseorosea can be distinguished from A. brunneitoxicaria by its pinkish lamellae, the medial position of the partial veil, as well as 2-spored basidia. According to the phylogenetic analysis, most species with dark brown basidiomata do not seem to be closely related and fall into several different clades.
Habitat Known distribution: Subtropical regions of Central and Southern China [6] and now northern Thailand.
Remarks: The large-sized basidiomata of A. fuligineoides with a grayish brown to fuliginous umber pileus, persistent apical to subapical partial veil, and a subclavate to napiform blub with a firm limbate universal veil are characteristic of this species. Amanita fuligineoides was originally described from forests dominated by Fagaceae in China at elevations of 900-1,200 m. [12]. The Thai collections were found in a mixed Dipterocarpaceae-Fagaceae forest at elevations of 510-950 m, and agree with salient features reported in the protologue, especially with respect to the rather large basidiomata and the presence of globose to subglobose spores. The gray-brown to dark gray pileus of A. fuligineoides also occurs in A. fuliginea, A. griseorosea, and A. subfuliginea [6]. Amanita fuliginea was originally described from Japan but is widely distributed in China and Thailand [3,4,32]. Amanita fuliginea and A. fuligineoides are very similar in morphological features. However, A. fuliginea differs from the Thai collections by having smaller-sized basidiomata and a subglobose bulb. The molecular analysis indicates that they are not closely related. A previous study reported α-amanitin and phalloidin from both A. fuliginea and A. fuligineoides [15] which is in agreement with the toxins detected in the Thai collections of A. fuligineoides. Amanita griseorosea and A. subfuliginea have significantly smaller basidomata. Amanita fuligineoides, despite the similar morphological similarities was not closely related in the molecular analysis to either A. griseorosea or A. subfuliginea.

Toxins
HPLC analysis of the α-amanitin standard showed a mass peak [M+H+H 2 O] + of 919.41 amu in the positive ESI mode and [M+H-H 2 O]of 917.41 amu in the negative ESI mode, characteristic UV/Vis with an absorption maximum at 304 nm, and a retention time (R t ) at 3.3-3.4 mins (Fig 6). The palloidin standard showed a mass peak [M+H+H 2 O] + of 789.40 amu in the positive ESI mode, and [M+H-H 2 O]of 787.39 amu in the negative ESI mode, characteristic UV/Vis with an absorbed maximum at 290 nm, and a retention time (R t ) at 5.1-5.2 mins (Fig  7). Crude extracts of Thai Amanita taxa were compared with these standards which showed major peaks at the R t of the standards with matching mass spectra in negative ESI mode, as well as their characteristic UV/Vis. Neither α-amanitin or phalloidin were observed in A. ballerina (Fig 8A). Several peaks corresponding to amatoxins were detected in the extracts of A. brunneitoxicaria (Fig 8B) and A. fuligineoides (Fig 8C). However, quantification was only accomplished for α-amanitin and phalloidin for which standards were available. In A. fuligineoides, 6.7 mg α-amanitin per gram of dry weight and 1 mg of phalloidin per gram of dry weight were detected. Amanita brunneitoxicaria contained approximately 21.5 mg of α-amanitin per gram of dry weight. A. brunneitoxicaria contained only traces of α-amanitin and no phalloidin was present. Extensive evaluation of the HPLC-MS results tentatively revealed the presence of additional minor peaks that may correspond with other amatoxins. Cycloamanide C (R t 5.7 min; molecular weight 886 Da; Gauhe and Wieland [33]), phallisin (R t 4.3 min; molecular weight 804 Da [34]) and phalloin (R t 5.9; molecular weight 772 Da [35]) were detected in A. fuligineoides, whereas A. brunneitoxicaria yielded only cycloamanide C. These results are based on the mass spectrometric data in comparison with characteristics of the amatoxins reported in the literature and would have to be confirmed by preparative isolation of the compounds, which would afford the availability of large quantities. In any case, the present study only proves the presence of the hazardous toxins in the new species A. brunneitoxicaria and additional analyses would have to be carried out on additional collections of the respective species in order to ascertain their toxicogenic potential.

Discussion
This study, along with other recent publications [16,36], illustrates the high diversity of Amanita species in Thailand, suggesting that more taxa remain to be discovered and documented. First, Amanita ballerina is a new species in an interesting clade that so far contained only a few species. Second, A. brunneitoxicaria clusters with A. fuligineoides, which in a previous publication was the only taxon in the lethal amanitas clade VIII [13]. Third, Amanita brunneitoxicaria is the first new Amanita species to be discovered in southern Thailand. To date, most of the taxonomic research on Amanita in Thailand has been focused on northern Thailand and the discovery of A. brunneitoxicaria suggests that additional study is needed in other parts of the country with different mycorrhizal hosts and ecology. Additional collections of A. zangii [16] were made during this investigation in Chiang Mai province in hill forest associated with Castanopsis spp. and Quercus spp. or Lithocarpus spp. at elevations of 1450 m, suggesting that A. zangii (Fig 1E) is not a rare species in northern Thailand.
So far, studies of Amanita have devoted much attention to species producing toxins, but molecular studies of Amanita toxin genes have been only recently conducted [8][9][10]. Research has begun to examine the toxins as well as phylogenies and evolutionary relationships among toxin-encoding genes [13,15,37,42]. Here, we report on an initial investigation of the presence of toxins in two species of Amanita sect. Phalloideae and one species, A. ballerina, whose taxonomic placement is uncertain, all collected in Thailand. Screening for both α-amanitin and phalloidin showed that neither toxin was present in A. ballerina, as is also the case for A.
sp. HKAS77321 and A. zangii [13]. The possibility of the presence but lack of expression of toxin-encoding genes in those species should be explored.
The position of A. zangii and of stirps Hesleri to which it is morphologically related has been previously discussed [13,16,23]. Previous molecular analyses showed that A. zangii and A. sp. HKAS77321 formed a moderately supported clade sister to a clade comprised of lethal Three interesting Amanita species from Thailand amanitas (Amanita sect. Phalloideae sensu Bas [1]), and the former was interpreted as an early diverging lineage in the Phalloideae [13,16]. In the present four-gene phylogenetic analysis, A. ballerina is a third taxon belonging to that interesting clade. It should be noted, however, that this clade was not supported in the two protein-coding gene analysis (BS = 38% only). In both the two-gene and four-gene analyses in the present study, Amanita ballerina and A. sp. HKAS 77321 formed a strongly supported clade sister to A. zangii. Similarly to A. zangii, the suite of morphological characters in A. ballerina is somewhat in conflict with its placement in the Phalloideae. They are indeed reminiscent of A. limbatula, a rare species known from only a few collections, currently included in stirps Limbatula of sect. Lepidella [7]. A significant difference between A. ballerina and species of stirps Limbatula, however, is the presence of clamp connections in the latter [1]. Unfortunately, no sequence of A. limbatula is available to confirm the phylogenetic affinities of this species, and no specimen in good condition was available to us. Amanita ballerina shares a similar elongate to ventricose bulb morphology and pileipellis structure with A. zangii. A. ballerina has a striate margin when mature, while A. zangii has an appendiculate margin. Neither of these characters conform to the circumscription of sect. Phalloideae sensu Bas [1], which raises questions about placement of these two taxa. The sequence of A. sp. HKAS77321 also falls into the interesting clade with A. ballerina and A. zangii, but unfortunately no morphological information is available for this taxon. There is thus growing evidence that there is a divergent lineage sister to lethal amanitas, which could be interpreted as an early diverging lineage in sect. Phalloideae, or perhaps as a completely new section. Careful morphological and sequence analyses of members of stirps Hesleri and stirps Limbatula is required. Further exploration of Amanita diversity and historical biogeography in South-East Asia and Australia, which seems to be a hotspot of early diverging Amanita lineages. [13,14], is critical and could reveal more members of this clade, and help elucidate morphological and molecular synapomorphies to support or refute the hypothesis of a new section.