The genus Hebeloma in the Rocky Mountain Alpine Zone

Abstract Numerous taxa of Hebeloma have been reported in association with Salix, Dryas, and Betula in arctic-alpine habitats. However, species are notoriously difficult to delineate because morphological features overlap, and previously there was little reliable molecular data available. Recent progress in ITS-sequencing within the genus, coupled with an extensive database of parametrically described collections, now allows comparisons between species and their distributions. Here we report 16 species of Hebeloma from the Rocky Mountain alpine zone from some of the lowest latitudes (latitude 36°–45°N) and highest elevations (3000–4000 m) for arctic-alpine fungi in the northern hemisphere. Twelve of these species have been reported from arctic-alpine habitats in Europe and Greenland and are now molecularly confirmed from the Middle and Southern Rockies, greatly expanding their distribution. These are: Hebelomaalpinum, H.aurantioumbrinum, H.dunense, H.hiemale, H.marginatulum, H.mesophaeum, H.nigellum, H.oreophilum, H.subconcolor, H.spetsbergense, H.vaccinum, and H.velutipes. Hebelomahygrophilum is known from subalpine habitats in Europe, but was never recorded in arctic-alpine ecology. Three species recorded from the Rockies, but as yet not reported from Europe, are H.alpinicola, H.avellaneum, and H.excedens. The last two have never previously been reported from an arctic-alpine habitat. For all three of these species, the holotypes have been studied morphologically and molecularly, and have been incorporated into the analysis.


Introduction
The alpine is defined as the life zone above treeline on high mountain tops and this biome constitutes 3% of the earth's land (Körner 1999). In northern latitudes, it is characterized by low, open vegetation and a climate dominated by cold temperatures (Chapin and Körner 1995). Diurnal temperature fluctuations and periodic strong winds during the short growing season affect both plant development and basidiome production. Ectomycorrhizal fungi are critical to the survival of alpine woody plants such as Salix, Dryas, Betula, and non-woody plants such as Persicaria (Bistorta) and Kobresia in the alpine zone (Cripps and Eddington 2005). The most diverse ectomycorrhizal fungal genera in the Northern Hemisphere alpine are Cortinarius, Inocybe, Hebeloma, Laccaria, Entoloma, Lactarius and Russula (Gardes and Dahlberg 1996;Cripps and Horak 2008).
The Rocky Mountain alpine exists as islands on high mountain tops and plateaus separated by vast forests and grasslands. The middle and southern Rockies span some of the lowest latitudes (36°-45° N) and highest elevations (3000-4000 m) known for northern hemisphere alpine. Yet, species of Inocybe and Lactarius from the Rocky Mountain alpine zone have been found to be conspecific with those occurring in arctic and alpine habitats in the Alps, Pyrenees, Norway, Sweden, Finland, Svalbard, and Greenland through molecular matching of ITS (internally transcribed spacer) sequences (Cripps et al. 2010;Larsson et al. 2014; ; . The genus Hebeloma is common in arctic and alpine habitats, but species are poorly known. It is phylogenetically placed in the Hymenogastraceae Vittad. (Matheny et al. 2006) and is characterized by smooth to roughened brown spores that lack a visible germ pore, distinct cheilocystidia, an absence (usually) of pleurocystidia, and an ixocutis resulting in a smooth viscid pileus which is often two-colored (usually darker in the center). Distinctive odors, typically of radish or raw potato described as raphanoid are often present (Vesterholt 2005). However, not all species exhibit all features and character states overlap. Although most experienced mycologists will normally be able to identify a mushroom as a Hebeloma with relative ease, taxa are notoriously difficult to delineate at the species level because of variable morphological features and, until recently, a lack of reliable reference literature and a lack of confirmed DNA reference sequences of type materials. While the recent monograph by Beker et al. (2016) provides a great deal of reference material, this was centered on the Hebeloma of Europe; overlap between the European and American continents is currently being studied.
Numerous taxa of Hebeloma have been reported in association with Persicaria, Betula, Salix, and Dryas from arctic-alpine habitats including those in the Alps (Favre 1955;Bon 1986;Bruchet 1974;Debaud et al. 1981;Kühner and Lamoure 1986;Senn-Irlet 1990;Senn-Irlet 1993;Jamoni 2008;Graf 1994;Brunner et al. 2017), Iceland (Eyjolfsdottir 2009), Scandinavia (Vesterholt 2005(Vesterholt , 2008Knudsen and Vesterholt 2008), Svalbard (Hutinen 1987;Ohenoja 1971;Gulden and Torkelsen 1996;Beker et al. 2018), Pyrenees (Corriol 2008), and the Carpathians (Eberhardt et al. 2015b). In North America, there are reports from Greenland (Lange 1957;Borgen 2006;Borgen et al. 2006), Canada (Ohenoja andOhenoja 1993, 2010), Alaska (Miller 1998), and the Rocky Mountains (Miller and Evenson 2001;Cripps and Horak 2008;Beker et al. 2010). A table comparing the occurrence of species in various arctic and alpine locations was presented in Beker et al. (2018); this table indicates 10 species from the Rocky Mountains. Beker and co-workers (2016) list 25 species occurring in arctic or alpine habitats, 14 of which appear (almost) restricted to these habitats; others also occur in a variety of habitats from subalpine or boreal with coniferous and hardwood trees right down to sand dunes where they grow with dwarf Salix. The veiled species of Hebeloma in Western North America have been treated in a monograph by Smith et al. (1983), but few (if any) of their collections are from above treeline, although many are from high elevations in the Rocky Mountains. While recent work on the genus Hebeloma in Europe now provides a basis for comparison of morphological and molecular data for a significant number of species and make possible comparisons of distribution patterns (Vesterholt 2005;Beker et al. 2016), much more work is needed before we will have a complete picture of the different species that occur on the different continents and their distribution across those continents. Here we delineate 16 species of veiled and unveiled Hebeloma primarily with Salix from the Rocky Mountain alpine zone. Thirteen of these taxa were described in detail in Beker et al. (2016) but three species described here were not included in that discussion of European Hebeloma. These three species (H. alpinicola A.H. Sm., Evenson & Mitchel, H. avellaneum Kauffman, and H. excedens (Peck) Sacc.), whose holotypes have been studied morphologically and molecularly, are described within this paper and their relationship with other Hebeloma species is explored.
As demonstrated in Beker et al. (2016), morphological differences do exist between species and although separation between species does need careful work, in almost all cases a morphological analysis may be used for determination of species and in some cases morphology is even better suited for species delimitation than the data of the five loci applied. Here we have carried out a morphological analysis to determine species and have found no conflict between our morphological placement and that provided by our molecular analysis based on ITS data. Tree and network building methods have been applied to demonstrate the taxonomic placement of the Rocky Mountains collections in relation to type specimens and confirmed collections of species treated by Beker et al. (2016). For the three species not treated in Beker et al. (2016) we include type sequences from American types. We do not provide lists of synonyms in the species descriptions, because we have not yet re-evaluated all species described outside Europe and any list that we could give would be provisional. Where we deem it necessary, synonyms are mentioned in species discussions. Species names and their synonyms from Europe have been treated to great detail by Beker et al. (2016).
A great majority of the encountered species was shown to be paraphyletic and part of species complexes by Beker et al. (2016) and previous works (Eberhardt et al. 2015aGrilli et al. 2016). In the course of the studies for this work we found that the same is true for two species (H. alpinicola and H. excedens) not treated by Beker et al. (2016). We have chosen to illustrate the problems of species recognition and delimitation based on ITS data by showing networks for taxa treated by Eberhardt et al. (2015aEberhardt et al. ( , 2016 and Grilli et al. (2016), i.e. members of the H. sects. Denudata and Velutipes; and in addition to trees for members of H. sect. Hebeloma. The ITS region of members of these species complexes often differs only by a small number of base pairs between species, and comparable differences occur within species. Additionally, species often do not form monophyla within these complexes.
Median-Joining Networks have been recommended for inferring intraspecific phylogenies (i.e. Bandelt et al. 1999). Pruned quasi-median networks (Ayling and Brown 2008) are a tool to visualize DNA sequence variation when evolution has not necessarily been treelike. No assumptions are made as to which evolutionary mechanisms (i.e. hybridization, recombination, etc.) have been responsible for the observed variation. In the networks, observed sequence variants are shown as circles and the size of each circle represents the number of times the respective sequence variant has been observed. Two circles connected by an unsegmented line differ in 1 bp. So-called quasimedians, a kind of placeholder for unobserved sequence variants, are placed between observed sequence variants that each differ from the quasi-median by 1 bp. The number of segments to a line represents the number of base pair changes between two sequence variants or a sequence variant and a quasi-median. A pruning mechanism is applied to reduce the complexity of the networks while depicting at least one shortest path between all pairs of sequence variants (Ayling and Brown 2008).
Ideally, we would have been able to present networks of haplotypes. What we here refer to as 'ITS variants' are sequencing results of dikaryotic material; in many cases, the sequences do not seem to correspond to a single haplotype. Although the ITS exists in multiple copies in the genome, it has been shown to behave like a dikaryotic locus in Hebeloma (Aanen et al. 2001) and other fungi (i.e. Schnabel et al. 2005;Hughes et al. 2013). Even good quality reads of ITS and other nuclear loci of many Hebeloma species contain one or several ambiguous positions and/or indications of indels, which we consider as evidence of variation between haplotypes of the same locus. Here, the level of variation was such that attempts to phase all ITS data into haplotypes (Flot et al. 2006;Flot 2010) were aborted and each collection is represented by a single ITS variant, i.e. the consensus sequence of both 'haplotypes'.

Study sites
Our primary study sites are in the Middle-Northern and Southern Floristic zones of the Rocky Mountains that extend from Montana to Colorado (Fig. 1); the phytogeography is described in Cripps and Horak (2008) and further site details are in  and Osmundson et al. (2005). Primary collecting sites include the Beartooth Plateau (latitude 45° N, elevation 3000-3500 m) in Montana and Wyoming, and the Front Range, Sawatch Range, and San Juan Mountains in Colorado (latitude 36°-38° N, elevation 3600-4000 m). Ectomycorrhizal vascular plants include Salix reticulata, S. arctica, S. rotundifolia, S. cascadensis, S. planifolia, S. glauca, Betula glandulosa (= B. nana), Dryas octopetala, Persicaria vivipara, and Kobresia mysuroides (Cripps and Eddington 2005). While our study was focused on areas of tundra above the tree line, occasionally small Picea shrubs also occurred and it was not possible to unambiguously specify the mycorrhizal partner.

Collections and morphological descriptions
Basidiomes were collected from late July through August, which constitutes the field season, from 1980 to 2017. Most collections were described in fresh condition, photographed, and dried on a dehydrator overnight. Dehydrated material was deposited in the MONT herbarium (Montana State University), ETH (Zurich, Switzerland), DBG (Denver Botanic Gardens), and/or the HJB private herbarium. Microscopic examination of dried material was done in 5% KOH to measure spores, cystidia, basidia, and other important features and in Melzer's solution to assess dextrinoid reactions follow- ing Beker et al. (2016) and Vesterholt (2005). Within the species descriptions below we conform to spore descriptions based on spore ornamentation measures (O1-O4), spore dextrinoidity measures (D0-D3) and perispore loosening measure (P0-P3), as described in Beker et al. (2016). Similarly, cheilocystidia measurements include length, maximum width near the apex, minimum width in the median part of the cystidium and maximum width in the basal part of the cystidium. No distinction is made in the spore measurements for spores from two-and four-spored basidia. Measurements for the two types of spores are given separately in the Suppl. material 1. Exsiccate were also described. Unless otherwise mentioned, the species descriptions given are based on the collections from the Rocky Mountains cited here.

Molecular analyses
ITS sequence data from the 115 Hebeloma collections from the Rocky Mountains (which is referred to as the RM dataset), 221 reference sequences including some type sequences from Europe (referred to as the FE (Fungi Europaei) dataset, see  and 10 type collections of species described from the US, pertinent to the RM collections, were generated using a variety of protocols (Eberhardt 2012;Eberhardt et al. 2016). Newly generated sequences were submitted to GenBank (acc. no. MK280985-MK281025, MK286558-MK286561, and MK305906-MK305939).
The DNA of old material was extracted using the Gentra Puregene kit (Qiagen, Hilden, Germany), modifying the procedure that is described in the manual (version 2014) for yeasts, generally replacing any pipetting of DNA-containing fluids by pouring (see Eberhardt et al. 2016). A small amount of basidiome material was crushed in a TissueLyser II (Qiagen), suspended in 300 µl suspension solution plus 1.5 µl lytic enzyme for 30 min at 37 °C. The samples were centrifuged for 5 min at 8000 rpm and the supernatant poured out. Lysis was done in 300 µl of Cell Lysis Solution, the samples mixed by vortexing and incubated overnight at 37 °C, followed by 1 h at 65 °C. Samples were cooled to room temperature and 100 µl Protein Precipitation Solution added. Prior to centrifugation (maximum speed, 5 min), the samples were placed in the freezer for 10-15 min. Each sample was then poured into a prepared tube with 300 µl absolute isopropanol and 1 µl of glycogen (Life Technologies, Darmstadt; diluted 1:1 with ultrapure water). After mixing by repeatedly inverting for 1 min, the DNA was precipitated overnight to several days in the fridge. The pellets were washed in 300 µl 70% ethanol, air-dried for 30 min and re-desolved in 50 µl DNA Hydration Solution. The purified DNA was re-desolved by heating the samples for one hour at 65 °C and keeping them overnight at room temperature. DNA extracts were diluted for PCR as required. ITS1 and ITS2 were amplified separately in 35-40 cycles of PCR (30 s denaturation at 95 °C, 45 s annealing at 55 °C, and 60 s elongation at 72 °C) with 1.25 U/25 µl MyTaq Red (Bioline, Luckenwalde, Germany), using the primer pairs ITS1F/ITS2 and 58SF/ ITS4 (White et al. 1990;Gardes and Bruns 1993;Tedersoo et al. 2013 [who erroneously ascribed the primer 58SF (3' -ATG CAT CGA TGA AGA ACG C -5' to Martin and Rygiewicz 2005]). Sequencing was carried out at LGC (Berlin, Germany).
Taxonomic assignment to section and species cluster was done via BLAST searches against the collections analyzed in depth by Beker et al. (2016), the FE dataset, in Geneious R10 (version 10.2.3,Biolmatters,Auckland,NZ). To illustrate the taxonomic placement of the RM collections, eight alignments were assembled using Mafft online with the G-INS-I option (Katoh et al. 2017), breaking up the large number of sequences into manageable datasets based on BLAST results. Alignments include RM and FE representatives of the target species, i.e. species occurring in the Rockies, relevant types for non-European species, and (where applicable) FE sequences of taxa that cannot be unambiguously distinguished from the target taxa, i.e. neither target species nor sister species forming monophyla in the ITS analyses of Beker et al. (2016) for arctic-alpine species. For better readability, non-arctic-alpine sister species clearly distinct from the target species were excluded from the final analyses. Also, for better readability, the number of European representatives of the included species was restricted to 10 (if available) or, for species present in the RM dataset in more than 10 collections, matching (if possible) the number of collections of the RM dataset. An exception was made for H. velutipes, for which 20 sequences were included because of the known high intraspecific diversity of this species. For each included species, the selection of included representatives from Beker et al. (2016) was random, but only considering sequences with high quality reads. For illustrating the placement of H. avellaneum, not included in Beker et al. (2016), a small alignment was assembled representing all species accepted by Beker et al. (2016) in H. sect. Naviculospora. For tree analyses, outgroup sequences were added; selection of outgroup taxa followed Beker et al. (2016). Details are given in Table 1 for the sequences of Rockies collections, in Table 2 for other American collections, the majority types, and in Suppl. material 1 for FE data (Supplementary Data). Alignments were viewed and reformatted using AliView version 1.24 (Larsson 2014) (Miller et al. 2010), with the GTRGAMMA option, five searches for the best ML tree, using the MRE option to limit the number of fast bootstrap replicates. Trees were visualized using FigTree version 1.4.2 (Rambaut 2014). Pruned quasi-median network analyses were carried out in SplitsTree (version 4.14.6, Huson and Bryant 2006) using the default settings apart from activating the 'scale nodes by taxa' and 'subdivide edges' options. Nodes representing different classes of sequences (differentiated by species and origin, RM versus FE) were replaced in Adobe Illustrator CS6 by pie charts of corresponding diameters, showing the relative numbers of sequences for each class.
Distances between sequences were calculated in PAUP* (Swofford 2003), as the total number of differences of standard data, disabling the default 'equate' scheme for sequence data. By doing this, ambiguity reads like i.e. 'y' are not equated with the corresponding bases, here 'c' and 't'. Missing data were recoded as '?'; gaps were treated as standard characters. In addition, differences in PAUP* 'standard DNA/RNA absolute' differences with default settings (equating scheme in place; gaps treated as missing data) are given in square brackets. For those who wish to convert absolute to relative distances, alignment length was between 698-722 bp.

Results and general discussion
Species recognition is often not easy in Hebeloma, and although species can normally be identified by morphology alone, species are delimited by a combination of morphology, multi-locus molecular data and ecology. In some sections (H. sects. Denudata and Velutipes) the efforts of Aanen and co-workers (i.e. Aanen and Kuyper 1999, 2004, Aanen and Kuyper 2004 also gave some evidence with regard to the limits of biological species. As described earlier (Eberhardt et al. 2015a, 2015b, Grilli et al. 2016, species definitions based on several lines of evidence may share ITS or other loci' haplotypes, presumably as a result of incomplete linage sorting, hybridization or other population processes. The molecular distance between some species is so small that we assume that not all groups we recognize as species had sufficient time to reach monophyly in all loci. Thus, we do not necessarily expect species to form monophyla in ITS trees. In spite of this, and this is visualized by the networks, certain haplotypes or combination of haploypes (as in dikarya, here referred to as "variants") is normally characteristic for a single species and occurs only rarely in sister species. Therefore, in spite of its lack of resolution in phylogentic trees, BLAST searches against an ITS database of well identified collections very often retrieve the correct species name in relation to other lines of evidence. We are not aware of a single locus that can differentiate between all species of Hebeloma. In particular in H. sect. Hebeloma, the search for a locus that is more powerful in recognizing species than the loci used by Beker et al. (2016), namely ITS, RPB2, Tef1a, and variable regions of the mitochondrial SSU, is still ongoing. We are at the beginning of our research into the Hebeloma funga of America and all of our conclusions rest heavily on our insights into Hebeloma of Europe and there on the available material. For some species, for example H. velutipes, we have hundreds of collections to choose from, while for other species, like H. pubescens we have only a few specimens. As our research goes on and more data becomes available, we will revisit and if necessary rectify the conclusions drawn here. Sixteen species of Hebeloma were identified morphologically among the collections from the Rocky Mountains alpine zone. The molecular analysis carried out supported the morphological analysis. A key is given below. In all, 115 collections and 10 relevant types from North America were sequenced successfully for the ITS region (Tables 1, 2). Figure 2 shows the taxonomic positions of the treated species (complexes) mapped on the ITS tree of Beker et al. (2016). Of the 16 species collected in the Rockies, three were not treated by Beker et al. (2016), namely H. alpinicola, H. avellaneum and H. excedens. These species were named based on type studies. Figure 3 shows that H. avellaneum is a member of H. sect. Naviculospora and forms a monophylum. The only other species encountered in the Rocky Mountains that is clearly distinct in the ITS region is H. hiemale Eberhardt et al. 2016;Fig. 4B). For all other species, several taxa were included in a single network (Figs 4A, 4C, 4D, 5, 6). The networks show that there are usually only a small number of unambiguous base pair differences between members of the same species, irrespective of their origin, even though some parts of some networks (H. aurantioumbrinum, H. marginatulum) are exclusively of RM origin. While ITS trees were published for H. sects. Denudata and Velutipes (Eberhardt et al. 2015aGrilli et al. 2016), this is not the case for H. sect. Hebeloma. Therefore ITS ML trees, rooted with H. grandisporum Beker, U. Eberh. & A. Ronikier, are shown in Figure 6. Details, including base pair (bp) differences between species, are discussed in the Taxonomy section. Beker et al. (2016) showed that in a number of Hebeloma species clusters or complexes, morphology is better suited for species distinction and delimitation than molecular data. The majority of the species encountered in the Rocky Mountains belong  (Beker et al. 2018).
In the Taxonomy part, minute levels of sequence variation are discussed. We do that against the background of multilocus analyses presented by Beker et al. (2016) and other works, indicating in which cases the ITS is wanting for species differentiation. Thus, even though ITS differences between species may be slight or not constant, and even considering that morphological distinction in some cases relies on minute differences, the combination of morphology, ecology, and ITS data provides a reliable set of information for species assignment.
Based on previous studies, delimitation of most species is now well understood (Eberhardt et al. 2015aBeker et al. 2016;Grilli et al. 2016), and consequently we did not consider it necessary to include all species discussed as morphologically similar in the same molecular analysis. Our aim has been to show what information, even in the case when it is sparse, is contained in ITS data.
We have made an effort to combine sequence analyses based on different subsets of data and displaying different levels of complexity in the visualization. We have considered several different methods for analyzing ITS sequence data: ML trees, pruned quasi-median networks, and base pair difference counts between aligned se- quences. Sometimes, the relationship between sequences and species may appear differently between trees, networks and difference counts. In the ML analyses, gaps are treated as missing data and ambiguous reads are equated. The networks are based on clean base pair exchanges and gaps; polymorphic positions with two states, i.e. positions with ambiguous codes are treated as missing data. Owing to the complexity of networks displaying this kind of information in full, such networks are, as far as we are aware, used for data verification rather than for data analysis (Bandelt and Dürr 2007;Brandstätter et al. 2007). For the direct sequence difference counts, all kinds of differences were counted equally, thus giving the maximum number of differences plus giving absolute DNA differences in square brackets, which do not count gaps and polymorphic positions as different. Whereas ML trees pruned quasi-median networks and absolute DNA differences are prone to omitting observed intragenomic and thus intraspecific variation, total distance counts are overestimates. In spite of that, we have decided to present these values here, because they could influence species identificaton.
Key to Hebeloma species of the Rocky Mountain Alpine Zone Description. Cortina not observed. Pileus 10-11 mm in diameter, convex, buff to brownish with a hoary coating, rather unicolor, smooth, shiny, tacky; margin turned down, a bit crenulate, faintly striate; edges white. Lamellae adnexed, L = 38 plus lamellulae, buff to milk coffee. Stipe 10 × 3 mm, equal, cream, finely floccose at apex and fibrillose for length, delicate. Context cream. Odor not apparent, but previously noted as raphanoid. Exsiccate: very tiny, brown, not shiny, lamellae not blackening.
Rocky Discussion. Beker and co-workers Eberhardt et al. 2016; including ML ITS analyses) showed that H. vaccinum can be recognized by its ITS region from all species apart from H. cavipes Huijsman, which differs in morphology and ecology. The RM H. vaccinum collection fits in with the diversity found within the species (Fig. 4D)  This species is usually described as larger (13-40 mm) than the Rocky Mountain specimens described here. Microscopically, the species has spores that are strongly dextrinoid (D3) with a frequently loosening perispore. The spores and cheilocystidia characteristics (swollen at the apex and at the base but constricted in the middle part) put it in H. sect. Denudata, subsect. Clepsydroida. Hebeloma vaccinum is known to occur in low elevation dunes and woodlands with Salix; it is widespread in Northern Europe. Other arctic-alpine collections are from the European Alps, the Carpathians in Slovakia, and Greenland, always with Salix species Eberhardt et al. 2015b). It could be recognized in the Rocky Mountains by its association with dwarf Salix, small size, lack of a veil, and distinct spores and cystidia; compare with H. aurantioumbrinum. Description. Cortina absent. Pileus small, 10-20 mm in diameter, convex, slightly conic-convex, appearing smooth, greasy, not hygrophanous, cream, then buff, pinkish buff, orange brown, can be lighter towards margin but not clearly two-toned, somewhat hoary; margin weakly involute, possibly crenate with a white rim. Lamellae deeply indented, deeply sinuate-arcuate, rather distant, L = 25-40 plus lamellulae, cream, then buff, pinkish buff, milk coffee; edges fimbriate, white but graying, drops visible. Stipe 15-28 × 2-3 mm, equal, bit curved, dingy whitish cream but darkening at base to watery brown (in CLC3093), floccose/pruinose for top third and smoothfibrous below. Context dingy whitish. Odor faint or raphanoid. Exsiccate: pileus buff, lamellae brown; stipe very thin, whitish.
Rocky Discussion. Beker and co-workers Eberhardt et al. 2015a) showed that H. aurantioumbrinum cannot be distinguished from the non-arctic-alpine H. helodes J. Favre based on ITS sequencing, but it can be separated from all other members of H. sect. Denudata. An ITS tree is given in Eberhardt et al. (2015a). The RM dataset includes more collections of H. aurantioumbrinum (15) than the FE dataset (7). Therefore, it is not surprising that the molecular diversity of the RM sequences is higher than that of the FE dataset (Fig. 4C). There are 0-6 [0] bp differences among the FE sequences of H. aurantioumbrinum, 0-9 [0-3] bp differences among the sequences of RM H. aurantioumbrinum and 2-11 [0-3] bp differences between H. aurantioumbrinum and H. helodes. Morphologically, H. aurantioumbrinum and H. helodes are quite different and can be easily separated, for example H. helodes always has a distinct thickening of the cheilocystidium wall at the apex, a feature that is absent in H. aurantioumbrinum. Further, they occur in very different habitats; H. helodes has never, to our knowledge, been confirmed in arctic-alpine habitats.
Hebeloma aurantioumbrinum may have been confused with H. pusillum J.E. Lange, although H. pusillum has much more slender basidiomes that are distinctly two-toned. Hebeloma aurantioumbrinum is squatter and rarely two-toned. Additionally, we are not aware of any confirmed records of H. pusillum in arctic-alpine habitats. Both these species, without any veil (beyond the primordial stage) and with clavate-stiptate cheilocystidia, belong to the Crustuliniformia subsection of section Denudata. This subsection contains many small species that are arctic-alpine specialists that occur with Salix, and these species have only recently been split out and described (Eberhardt et al. 2015a). Collections of H. aurantioumbrinum have been confirmed from a number of arctic and alpine habitats, including Greenland, Iceland, Scandinavia, and Svalbard ). In the Rockies, this species can be recognized by its alpine habitat, association with willows (primarily S. planifolia), small size, lack of veil, and pinkish buff to orange brown uniformly colored pileus often with a white, crenate margin. Etymology. concolor for the similar coloration of pileus and stipe, which is not a consistent feature.
Rocky Mountain ecology. Two collections reported under willow at alpine elevations of 4000 m in Colorado; noted as cespitose to gregarious. This small species has a grayish cast not found in other taxa in sections Denudata and Velutipes that we report from the Rocky Mountains; also, the lamellae are well separated and few in number. It should be compared to the other non-veiled, small species such as H. aurantioumbrinum and H. vaccinum. Hebeloma velutipes has a different coloration and is larger with many more full length lamellae. Hebeloma subconcolor is known from arctic and alpine locations in the European Alps, Greenland, Iceland and Scandinavia (Beker et al. , 2018. Figures 4B, 10, 23(4) Etymology. From hiemalis, winter or wintry, presumably to denote the production of basidiomes in colder seasons or habitats Description. Cortina absent. Pileus 15-35 mm in diameter, slightly conic-convex or domed-convex, smooth, greasy, pinkish buff, yellowish buff, to pale cream at the margin, with uniform coloration, somewhat hoary, with or without a white rim a few mm wide at margin; margin turned down or rolled in, then wavy. Lamellae narrowly attached, emarginate, somewhat crowded, L = 48-60 plus lamellulae, white to pale milk coffee, pale brown, wood brown; edges white floccose, with drops of liquid. Stipe 20-45 × 5-12 mm, equal, slightly clavate towards the base, whitish cream, totally pruinose (big floccules) for most of length and smoother below. Context white to watery cream, firm. Odor raphanoid, faint. Exsiccate: pileus yellowish brown, not distinctly two-toned; lamellae brown with white edges; stipe white and slimmer than for H. alpinum.
Rocky Mountain ecology. In the alpine zone with dwarf willows, Dryas and Betula, confirmed from Colorado, Montana, and Wyoming.  Discussion. An ITS tree including H. hiemale is given by Eberhardt et al. (2016); the respective network is shown in Figure 4B. The RM dataset includes ITS sequences from 22 collections. These were matched by the same number of sequences from the FE dataset. Hebeloma hiemale ITS sequences were shown to form a well-supported monophylum in ML results presented in earlier studies Eberhardt et al. 2016). Beker et al. (2010) showed that it is a species with a relatively high number of different ITS variants. The disparity between variants is mostly caused by gaps and SNPs (singlenucleotide polymorphisms). The number of differences between any pair of sequences of the presented H. hiemale data set is 0-9 [0-2] bp, within the RM sequences 0-8 [0] bp. This species is widespread across Europe occurring from the subalpine to the alpine, in lowland dunes, shrublands, gardens, and parks; it occurs with a wide array of deciduous and coniferous trees and this includes a number of willow species, including dwarf Salix. Confirmed arctic-alpine reports include those from Canada, Greenland, Iceland, Scandinavia, and Svalbard with Salix herbacea and S. polaris as well as Dryas and Persicaria ). Here it is confirmed with S. reticulata. Hebeloma hiemale has rarely been reported from North America in either subalpine or alpine habitats (Beker et al. 2010), but many collections previously labeled H. alpinum are now confirmed as H. hiemale.
This species looks like a small version of Hebeloma crustuliniforme but usually has more color in the pileus, particularly at the center. It has cheilocystidia that are generally swollen in the lower half, giving an hourglass appearance. The spores are verrucose, more warty than those of H. alpinum, but less so than the spores of H. vaccinum. There was some ambiguity around the delineation of H. hiemale, which was ultimately resolved with selection of an epitype (Beker et al. 2010;Eberhardt et al. 2015a). Figures 3, 11, 23(5) Etymology. For the color of hazelnuts, such as Corylus avellana.
Rocky Mountain ecology. Cespitose, or clustered, in low alpine krummholz with conifers and willows. Both collections we have studied are from Colorado.
Rocky Discussion. Based on ITS data, H. avellaneum is monophyletic, but unsupported by bootstrap values (Fig. 3). In terms of phylogeny, its closest relative is H. catalaunicum Beker, U. Eberh., Grilli & Vila, a Mediterranean species. It is also close to H. naviculosporum Heykoop, G. Moreno & Esteve-Rav. and H. nanum Velen. All three species appear to associate with Pinaceae ). The identification of H. avellaneum is supported by type studies. The fourth collection used in Fig. 3 is from Canada (Newfoundland) and has been presented by Voitk et al. (2016) as "Hebeloma sp. sect. Naviculospora".
Based on our studies of this taxon and of the habitats where it has been collected, we strongly suspect that this species is typically associated with conifers in temperate to subalpine or subarctic habitats. The holotype was collected in a temperate rainforest within the Olympic Peninsula in western Washington state. The often pruinose pileus with distinctive orange tones is indicative of H. sect. Naviculospora. These specimens were found in the low alpine where conifers are possible, and indeed Picea was noted for one collection, but only willows for the other. In the low alpine of the Rocky Mountains, the species might be confused with H. alpinum, H. velutipes, or H. hiemale because of its robust habit and lack of veil, however there are more orange color tones of the pileus; the spores are smaller and more dextrinoid than one would expect for H. alpinum and H. hiemale. Etymology. velutinus, for the velvety appearance of the stipe surface.
Rocky Mountain alpine ecology. In alpine situations, mostly reported with Dryas octopetala and also with Salix in Montana and Colorado. Discussion. Grilli and co-workers (2016) showed that in ITS ML analyses H. velutipes falls into three unsupported clusters, i.e. one with H. incarnatulum, one with H. leucosarx, and one with H. subconcolor. The latter is discussed above; the former two species do not occur in the kind of habitats sampled in the Rocky Mountains Grilli et al. 2016). Hebeloma velutipes cannot be distinguished from these three species based on ITS, but it is distinct from all other species treated in Beker et al. (2016). The reason for the intraspecific variation observed in H. velutipes has already been shown by Aanen et al. (2001), namely that H. velutipes possesses ITS alleles that differ greatly. In the Rocky Mountains, representatives of two of the clusters were found, the H. leucosarx cluster and the H. subconcolor cluster, and the collections from Montana fall into the first of these clusters while those from Colorado fall in the latter cluster. Accordingly, the number of differences are between 2-23 [0-5] bp; seven pairs with 2-6 [0-1] bp differences and seven pairs with 20-23 [2-5] bp differences. Looking at all included collections, the overall figure hardly changes (1-23 bp), although the collections randomly selected from the FE dataset include representatives of all three clusters (Fig. 5). To date we have not observed any morphological or ecological differences between members of the different clusters. The geographical differentiation of the RM representatives of H. velutipes is possibly a sampling artifact. This species displays the characteristic features of H. sect. Velutipes, i.e. the absence of a veil, presence of a velutinate stipe, and rather strongly dextrinoid spores (reaction can take a while), as well as the gently clavate cheilocystidia. It is known to be common and widely distributed in Europe at lower elevations primarily with deciduous trees but also with coniferous hosts. There are a number of arctic and alpine records, particularly from Svalbard with Dryas octopetala and Salix polaris , and it has been previously reported from the North American alpine zone (Beker et al. 2010). This species produces relatively large basidiomes for the genus in the alpine; but because of its pale coloration and lack of a veil, young specimens may have been incorrectly identified as H. alpinum or H. hiemale, which are typically smaller. Phylogenetically H. velutipes is not close to these two species but, as mentioned, is related to H. subconcolor, which is smaller with fewer lamellae, grayer coloration and is also reported from the Rocky Mountain alpine zone. Interestingly, almost all Rocky Mountain specimens of H. velutipes were found with Dryas, which might help with field recognition, in addition to its robust stature, and stout white stipe. Figures 4A, 13, 23 (7) Etymology. alpinum from the alpine.
Rocky Mountain ecology. Information is based on one collection from Montana, with mixed dwarf and shrub Salix species.
Discussion. The only confirmed report we have for this species from the Rocky Mountains relies on a single collection of a few specimens found near Cooke City, Montana at an elevation of 3000 m with dwarf and shrub Salix species. In the network Fig. 4A, this single RM representative of H. alpinum appears rather distant from its European counterparts, which are clustered at one of the centers of the network, i.e. the biggest circle, of the H. alpinum complex. An ITS tree including the H. alpinum complex is given in Eberhardt et al. (2015a). Although this collection appears molecularly quite far removed from its conspecifics, 6-10 [1-2] bp, the total distance is largely due to a 5 bp indel repeating a sequence motif generally present in members of the H. alpinum complex. Thus, the molecular results do not argue against this being H. alpinum. This species is quite variable molecularly as well as morphologically (see the discussion of the alpinum-complex in Beker et al. 2016). The spores of this collection are on the lower end of the range for this taxon, as given in Beker et al. 2016, but still comfortably within the range.
Hebeloma alpinum has been reported previously in North America from the Rocky Mountain alpine zone (Cripps and Horak 2008) and Alaska (Miller 1998), however, most sightings were not molecularly confirmed. There are three records from the Canadian Arctic collected in 1971 and 1974 (Ohenoja and Ohenoja 2010), which have been confirmed molecularly (Beker et al. 2018). Ten collections at the Denver Botanic Garden, originally labeled H. alpinum, are now molecularly confirmed as H. hiemale (see comments for this species).
Favre originally described this species from the Swiss Alps as Hebeloma crustuliniforme var. alpinum Favre (Favre 1955) and Bruchet (1970) elevated it to species level. Hebeloma alpinum appears confined to arctic-alpine habitats and has been reported from such regions of the European Alps, Carpathians, Pyrenees, Greenland, Iceland, Scandinavia, Svalbard, and Switzerland, primarily with Salix reticulata, S. polaris, S. retusa, and Dryas octopetala as well as Persicaria ). The species is in H. sect. Denudata, subsect. Crustuliniformia because of the lack of a veil, the clavatestipitate shape of the cheilocystidia and molecular data (Eberhardt et al. 2015a). As a relatively robust alpine species, it should be compared to H. hiemale and H. velutipes; the latter has a robust floccose white stipe.

Hebeloma section Hebeloma
We will address this next section in two parts, again following the outline of the key: first those that have ellipsoid indextrinoid spores (H. alpinicola, H. dunense, H. excedens, H. marginatulum, and H. mesophaeum), also referred to as the H. mesophaeum complex and secondly those with amagdaliform spores that are rather strongly dextrinoid (H. hygrophilum, H. nigellum, H. oreophilum, and H. spetsbergense), also referred to as the H. nigellum complex.
Hebeloma section Hebeloma, Part one: cortina present, spores ellipsoid, not dextrinoid Description. Cortina present, remnants distinctly present in some. Pileus 15-40(-50) mm in diameter, slightly conic-convex, domed convex, irregular, sometimes with a flat center that can even be dished, smooth or rough due to velipellus, shiny, strongly canescent, underneath dark brown, dark chestnut, to dark caramel color, mostly uniform but two-toned in some and then lighter at margin (more hoary, dingy whitish, or ochraceous in one), with a fine white border around the pileus perimeter a few mm in from margin, not hygrophanous; margin turned down or in, rather persistently so, and then covered with copious veil, often irregular, wavy, fragile. In one collection, the cuticle is rather thick and rubbery. Lamellae deeply emarginate and squared off, some pulling away, somewhat broad, L = 30-40 plus lamellulae, cream, then pinkish buff, darkening to medium coffee brown; edges fimbriate. Stipe 20-40(-45) mm × 2-6(-10) mm, equal, undulating or not, pale buff (some with possible yellow tint), and dark (up to black) at base, pruinose at apex, longitudinally fibrous lower, with a few longitudinal fibrils. Context dingy whitish, some with yellowish tones and dark at base. Odor raphanoid or sourish, sometimes faint. Exsiccate: pileus pale brown to dark brown, some obviously canescent; lamellae medium brown; stipe buff or ocher, darker at base.

Rocky
This taxon was first described as H. versipelle var. marginatulum by Favre (1955) from the alpine region of the Swiss Alps and was later raised to species level by Bruchet (1970). It is now considered to be restricted to arctic and alpine habitats primarily with dwarf willows (Beker et al. , 2018. Confirmed records show it to be present in these habitats in Canada, Greenland, Iceland, Scandinavia, Svalbard as well as the European Alps and the Carpathians and Rocky Mountains (Eberhardt et al. 2015b;Beker et al. 2016). Vesterholt (2005) described H. polare as a darker brown closely related species, but this has been synonymized with H. marginatulum . The Rocky Mountain specimens are also mostly uniformly dark brown with a canescent sheen.
Collections from the alpine that are very hoary and dark brown have been misinterpreted as H. bruchetii Bon (Miller and Evenson 2001) before molecular techniques; H. bruchetii, first described as an alpine species, has now been synonymized with H. mesophaeum and should have smaller spores. Hebeloma marginatulum is mentioned as a subalpine species (in Idaho) by Smith et al. (1983) who described two varieties (var. fallax, var. proximum) from the subalpine in Colorado. Smith's spore descriptions (dextrinoid with sharp ends) for his varieties may not fit this species, but the authors recognize that these varieties of H. marginatulum, and indeed other closely related species, need more study in North America. This species is in H. sect. Hebeloma because of basidiomes with a cortina and the ventricose cheilocystidia together with the non-dextrinoid, or barely dextrinoid, spores that are primarily elliptical; within this group, it has an arctic-alpine habitat and relatively large spores (greater than 10 × 6 µm). Figures 6B, 15, 23(9) Etymology. alpini-and cola, meaning dweller, to emphasise its alpine habitat, although this taxon is not found exclusively in such habitats.

Hebeloma alpinicola A.H. Sm., V.S. Evenson & Mitchel, Veiled species of Hebeloma in the western United States (Ann Arbor): 48 (1983)
Description. Cortina present. Pileus robust, fleshy, 20-40 mm in diameter, irregular convex, somewhat domed or not, reddish brown center with grayish tones, outwards ocher and lighter towards margin (buff not white), not particularly two-toned, with hoary canescent coating that dries shiny; margin turned in at first, and then turned down. Lamellae narrowly attached, slight emarginate, or with a tooth, or pulling away, somewhat broad, milk coffee, L = 36-44; edges white floccose. Stipe 30-40 × 5-10 mm, equal, straight or not, whitish and pruinose at apex, dingy ocher and longitudinally fibrillose and striate in lower part, base sometimes encased in sand or earth. Context dingy whitish, darker below, and flesh staining brown; stipe solid or slightly hollow. Odor raphanoid. Exsiccate: pileus and stipe medium ochraceous brown; lamellae dark brown; stipe base encased in soil in the large collection (CLC1577).
Rocky Mountain ecology. Collected from two different sites, one in Montana, the second in Colorado. The first site is a mixture of Dryas, Salix planifolia and S. reticulata, with some Persicaria present. The second site is a low alpine zone with dwarf willows. In both cases the growth habit was gregarious, sometimes in rings, sometimes cespitose, but not completely joined.
Rocky Other specimens examined. See Table 2. Discussion. Figure 6B shows H. alpinicola as paraphyletic and closely related but not mixed with species from the H. mesophaeum complex other than H. has not yet been fully tested in multilocus analyses, we consider its distinctive morphology combined with the ITS evidence to be sufficient to assign the four RM collections to this species.
This taxon, with its small ellipsoid, indextrinoid spores and ventricose cheilocystidia is a member of H. sect. Hebeloma. Morphologically it is closely related to H. excedens and H. mesophaeum. It is generally more robust than these two species, espe-cially the stipe, and the pileus is not as two-toned. Colorado collections were described as having gray tones. While further work is needed to decide whether this really is a species distinct from the other two, the molecular evidence coupled with the morphological evidence suggest this to be the case. We have studied a number of collections, from a variety of habitats within North America that all appear to represent this taxon. Hebeloma chapmaniae, H. littenii, H. nigromaculatum, H. perigoense, and H. subargillaceum were all published by Smith et al. (1983) in the same publication that featured H. alpinicola; the replacement name H. smithii is later (Quadraccia 1987). Although there is some molecular variation between these seven collections, it is very small and we see insufficient evidence to separate these species. We have selected the name Hebeloma alpinicola on the grounds that although not all collections are strictly alpine, the majority are at least subalpine. Description. Cortina present. Pileus 10-28 mm in diameter, convex, slightly conic-convex, with or without a slight umbo (one papillate), or almost applanate, some sunken in center, smooth, greasy, pale pinkish buff at first, becoming caramel color in center, outwards remaining pale, with a hoary coating, some flecks of white in outer part, mostly appearing pale unicolor; margin turned in or down, covered with white veil tissue or not. Lamellae emarginate to subdecurrent, or pulling away, variable, L = 25-48 plus lamellulae, a bit distant, cream buff to pinkish buff at first, then milk coffee; edges white fimbriate. Stipe 20-50 × 2-6 mm, equal or narrowing a bit at base, dingy whitish buff in top part, sometimes pruinose and base darkening to golden color then blackish brown (not always obvious unless cut open), with fibrils on lower part and/or a few 'patches of tissue'. Context dingy white, watery buff, dark at base, sometimes splitting, often hollow when mature; tough in base. Odor faintly raphanoid or absent. Exsiccate: mostly pale; pileus buff or more ochraceous buff, center a bit caramel or not; lamellae pale light ocher; stipe buff, not obviously darker at base.
Rocky Mountain ecology. In the alpine zone of the San Juan Mountains, with dwarf willows S. reticulata and S. arctica, and shrub willow S. planifolia, some in moss or near streams.
Rocky Discussion. Based on ITS data, Hebeloma dunense is phylogenetically not clearly distinguishable, but neither is it molecularly identical to other members of the H. mesophaeum complex (Fig. 6B). The intraspecific variation is 0-10 [0-2] bp (17 sequences), within the RM dataset (5 sequences), 1-7 [0-1] bp. The exclusively RM circle in Fig. 6B is a result of the data selection; this corresponds to ITS variants that do occur in the FE dataset, but did not come up in the random selection of sequences for this species. For the Rocky Mountain collections, so far, H. dunense has been found more often with dwarf willows S. arctica, S. reticulata, and shrub willow S. planifolia in contrast to H. mesophaeum and H. excedens, which were more often with S. glauca. Originally described from low-elevation dunes with Salix, this species has been more recently recognized in arctic and alpine habitats and from Canada, Greenland, Svalbard, the European Alps, and the Carpathians Beker et al. 2018;Eberhardt et al. 2015b).
Rocky Mountain specimens of H. dunense are pale, often with narrow subdecurrent lamellae; the cortina can be scant or absent, some cheilocystidia have dense yellow contents, and the spores, which are ellipsoid and distinctly but not strongly ornamented, are slightly larger than those of H. mesophaeum and H. excedens.  (11) Etymology. From Greek meso, in the middle, and phaeus, dark-colored. Persoon (1872) particularly mentioned the peculiar reddish brown pileus center "disco rufofusco peculiaris" which is characteristic of this taxon.
Basidiospores yellow brown, elliptical, a few slightly ovoid, no big apiculus, not guttulate, looks almost smooth even under high magnification (O1), not or only very slightly dextrinoid (D0, D1), and no perispore loosening (P0), 8-10.5(-11) × 5-6.5 µm, on average 9.7 × 5.8 µm, Q = 1.66. Basidia 20-30 × 6-9 µm, clavate, four-spored mostly. Pleurocystidia absent. Cheilocystidia cylindrical in the upper part and slightly swollen to more swollen at the base, rarely fully cylindrical, 30-55 µm long × 4-7 µm at apex, 4-7 µm in middle, and 6-9.5(-10.5) µm wide at base, with occasional thickening of the apical wall, some septate. Epicutis thickness 60-350 µm, with some encrusted hyphae.  Discussion. Only two collections from the RM dataset turned out to be H. mesophaeum that differ in their ITS region by 7 [2] bp (Fig. 6B) Previously Hebeloma bruchetii Bon was one of the most commonly reported species from arctic and alpine areas, but it has now been synonymized with and folded into H. mesophaeum . Hebeloma mesophaeum has relatively small elliptical spores that are smooth to slightly rough and not dextrinoid. Hebeloma mesophaeum is a widespread species reported in almost all arctic and alpine habitats, as well as from subalpine, boreal, and lower elevation habitats with a wide variety of hosts . Also, many varieties have been described in North America (Smith et al. 1983) and in Europe (Vesterholt 2005). Some of the European taxa have been synonymized by the authors  and it remains to check the 12 North American varieties delineated by Smith et al. (1983).
Description. Cortina present. Pileus 10-25 mm in diameter, shallow convex, campanulate, then almost applanate, slight umbo or not, viscid or greasy, medium cocoa brown to orange caramel in center and pale brown on most of the pileus, with ences, the intraspecific variation of the H. excedens sample (RM + type = 7 sequences) is 0-8 [0-1] bp, whereas the variation in the sample between H. excedens and H. mesophaeum is 2-11 [0-4] bp. In terms of absolute differences, the type of H. excedens is 5-8 [0-1] bp different from other collections referred to this species, but as Fig. 6B shows it is not strongly differentiated from other members of H. excedens, if ambiguous positions are treated as missing data as in networks or equated to their constituting bases as in the ML tree. In terms of absolute differences, the type of H. excedens is 5-11 [0-3] bp away from the H. mesophaeum sequences of the sample. Thus, within the limited support ITS data can give in this case, we do consider the species identification of the RM H. excedens collections as molecularly supported. Until the question of the distinctness and delimitation of this species can be clarified, we prefer to treat it as an independent taxon.
Hebeloma pubescens Beker & U. Eberh. is another species from the H. mesophaeum complex that might occur in the sampled habitats of the Rocky Mountains and is close to H. excedens in Fig. 6B. Based on a small sample (3 sequences available for H. pubescens; 7 sequences for H. excedens), the species vary 5-10 [1-3] bp in their ITS region.
Hebeloma excedens was first described by North American mycologist C.H. Peck; the species, with its lageniform to ventricose cheilocystidia and small elliptical, almost smooth, indextrinoid spores belongs to H. sect. Hebeloma. It is closely allied with Hebeloma mesophaeum, with which we believe it has often been confused. Separating these two taxa morphologically is rather difficult, but it does appear that the pileus of H. excedens may be more evenly colored, less yellow brown, less brown in the center, and it was originally described as having a cuticle that extended beyond the lamellae.
The stipe surface appears to have fibrils arranged in zones, in contrast to that of H. mesophaeum. However, further work is required before we can have confidence that these characters are consistently different.
We have examined a number of collections from North America that are morphologically and molecularly consistent with this taxon. Based on these studies it would appear that Hebeloma excedens is widespread across North America and occurs in a wide variety of habitats. Etymology. From oreophilus, mountain loving to emphasize its presence in alpine habitats.
Description. Cortina present. Pileus 15-30 mm in diameter, convex, hemispherical, not umbonate, smooth, dry or greasy, medium brown, bay brown, reddish brown, dark black brown, with white to cream rim of fibrillose veil remnants at margin, with hoary coating; margin even or weakly scalloped. Thick waxy pellicle mentioned in one collection. Lamellae emarginate, subdistant, L = 40-50 plus lamellulae, cream at first then milk coffee color, pinkish cinnamon; margin floccose, white. Stipe 15-60 × 3-8 mm, equal or slightly enlarged at base, a bit curved or undulating, whitich, tan, brown, in top part and darkening to blackish brown at base, pruinose in top half and fibrous below, with patches of fibrils. Context watery buff with yellow tint, and blackish brown in base, stipe hollow. Odor raphanoid. Exsiccate pale brown all over, not dark.
Rocky Discussion. Hebeloma oreophilum is a member of the H. nigellum complex that cannot always be distinguished from H. nigellum based on ITS data (Fig. 6A). In terms of differences, the H. oreophilum sequences from the sample (9 RM, 10 FE) differ by 0-9 [0-3] bp; 0-8 [0-1] bp within the RM sample. Most similar to H. oreophilum is H. clavulipes, which in this sample differs by 1-11 [0-3] bp. The two species do not share the same habitats. The differences between species sharing the same habitats (H. nigellum and H. spetsbergense) are 3-10 [0-5] bp. Morphologically, the easiest way to separate H. oreophilum from H. hygrophilum and H. nigellum is by the number of full length lamellae, always at least 40 for H. oreophilum and less than 36 for the others. Hebeloma clavulipes is not known from arctic-alpine habitats and has spores with an average width at most 6.6 µm while the average spore width for H. oreophilum is on average at least 6.8 µm. Hebeloma oreophilum has a persisting cortina and the lageniform/ ventricose cheilocystidia of H. sect. Hebeloma. This species was first described from the western Carpathians (Slovakia) with Salix reticulata, S. retusa, or Dryas octopetala on calcareous soil (Eberhardt et al. 2015b). It has since been reported from Canada, Greenland, Scandinavia, Svalbard, and the Rocky Mountains Beker et al. 2018). Figures 6A, 20, 23 (14) Etymology. hygrophilus, because it is often found in moist, wet, boggy ground.
Rocky Discussion. Figure 6A supports Beker et al. (2016) in that H. hygrophilum is paraphyletic in relation to the other members of the H. mesophaeum complex based on the ITS sequence, although some genotypes seem to be restricted to this species. The four H. hygrophilum representatives from the Rocky Mountains differ by 2-20 [0-2] bp in their ITS, whereas the intraspecific variation of H. hygrophilum within the sample is 1-22 [0-3] bp (14 sequences). Responsible for the high distance values is sample CLC1476 (HJB15297), which differs from all other conspecifics by 15-22 [0-1] bp and from all sequences of the ingroup by 14-22 [0-2] bp, while all other H. hygrophilum samples differ by only 1-9 [0-2] bp from each other. The morphologically closest taxon occurring in the Rocky Mountains is H. nigellum which differs by 3-10 [0-5] (14-21 [0-2]) bp. The values in round brackets are for CLC1476. An unusually high number of SNP positions in CLC1476 is responsible for the large total differences. However, sequences with numerous SNP positions occur occasionally in Hebeloma and are normally reproducable .
Hebeloma hygrophilum was first described from the Pyrenees in non-alpine habitats above 1250 m (Poumarat and Corriol 2009) and it is known in boreal habitats from northern Europe ). Thus it is typically in subalpine or subarctic habitats. It appears to have been found mostly with Salix and usually in wet areas with moss, typically Sphagnum. Here we report it for the first time in the alpine habitat (with S. planifolia); at least one collection was found in Sphagnum moss. It is molecularly close to H. clavulipes, H. nigellum and H. oreophilum (see below). When found in the alpine, it could be confused with H. nigellum, which is morphologically very similar. However, the spore width of H. nigellum is reported typically with an average over 7 µm, while that for H. hygrophilum is reported with an average of less than 7 µm; to add confusion, both appear to have occasional very large spores likely from two-spored basidia. not only in arctic-alpine habitats, and is reported from alpine and arctic habitats in Canada, Greenland, Iceland, Svalbard and the European Alps (Beker et al. , 2018. In molecular and morphological features it is close to H. hygrophilum (which normally associates with Salix in non-arctic-alpine habitats). Hebeloma kuehneri Bruchet, a commonly reported arctic-alpine species, was described in the same paper as H. nigellum with the main differentiation being that the former has more brownish coloration and the latter more blackish tones (Bruchet 1970); a distinction that could not be supported by other lines of evidence. The holotype of H. kuehneri was lost, however, and a new lectotype (selected from the paratypes) has been established LY BR66-15); it is sequenced and is a molecular match to H. nigellum. We here follow Beker et al. (2016) in selecting the name H. nigellum over H. kuehneri for this species. Figures 6A, 22 Etymology. Originally found in Svalbard.

Conclusions
The 16 species of Hebeloma we report from the Rocky Mountain alpine zone are from some of the lowest latitudes (latitude 36°-45° N) and highest elevations (3000-4000 m) for arctic-alpine fungi in the northern hemisphere. Twelve of these species have been reported from arctic-alpine habitats in Europe and Greenland, and are now molecularly confirmed from the middle and southern Rockies, greatly expanding their distributions. These are: Hebeloma alpinum, H. aurantioumbrinum, H. dunense, H. hiemale, H. marginatulum, H. mesophaeum, H. nigellum, H. oreophilum, H. spetsbergense, H. subconcolor, H. vaccinum, and H. velutipes. Hebeloma hygrophilum is known from subalpine habitats in Europe, but has never been recorded in arctic-alpine ecology. Interestingly, hosts can overlap or vary among continents and while Rocky Mountain collections are primarily with S. arctica, S. reticulata, S. glauca, S. planifolia, and Dryas octopetala, those from other continents were with these plants or additionally with S. herbacea, S. polaris, S. retusa, Persicaria vivipara, and Helianthemum sp. Eberhardt et al. 2015b).
Three species, not known from Europe, have never previously been reported from a true arctic or alpine habitat; they are H. alpinicola, H. avellaneum, and H. excedens. All are species first reported as growing with Pinaceae in North America (Peck 1872; Kauffman and Smith 1933;Smith et al. 1983;Hesler unpublished manuscript). We note that the H. avellaneum collections described above are from the low alpine and conifers (and conifers are noted in some original descriptions); we do suspect that the ectomycorrhizal association is indeed with Pinaceae. The Rockies H. excedens collections were all reported with Salix in the alpine, yet the holotype was with pine in New York state. This species, like H. dunense, H. mesophaeum, and H. nigellum, appears not to be confined to alpine and arctic habitats. Similarly, H. alpinicola appears to be found with a variety of hosts in both alpine and subalpine habitats.
The Rocky Mountain alpine exists as islands on high mountain tops and plateaus far from the arctic and alpine areas of other mountain ranges. While the recent trend, due to molecular analysis, has been to discover differences between European and North American taxa given the same names, in the alpine the reverse appears to be true. Of the ectomycorrhizal genera, a majority of Inocybe, Lactarius, and Cortinarius species from the Rocky Mountain alpine zone have been found to be conspecific with those occurring in arctic and alpine habitats in the European Alps, Pyrenees, Scandinavia, Svalbard, and Greenland through molecular matching of ITS sequences (Cripps et al. 2010;Larsson et al. 2014;. Only a few alpine species of Agaricales and Russulales are so far considered endemic to the Rocky Mountain alpine including Laccaria pseudomontana Osmundson, C.L. Cripps & G.M. Muell. (Osmundson et al. 2005) and Lactarius pallidomarginatus .
The distributions of various ectomycorrhizal plant hosts in the Rocky Mountains alpine have been shaped by glaciation, topography, parent rock, and climate. Glaciation during the quaternary allowed mixing at the glacial forefronts, interspersed with glacial retreat and withdrawal of cold-adapted plants to mountain tops, which include dwarf Salix and Dryas (Birks 2008). Tertiary connections have also been suggested (Webber 2003). A view from the North Pole shows Arctic areas as more contiguous than generally considered, and corridors during interglacial periods stretched from the Rockies to the Arctic and Siberia allowing migration and genetic mixing.
Alpine areas, like the arctic, are known to be sensitive to climate change. Greening of these areas is primarily due to shrub encroachment (Tape et al. 2012), and this involves ectomycorrhizal host plants; consequently, ectomycorrhizal fungi communities are likely to change with the loss or gain of different hosts (Geml et al. 2015;Morgado et al. 2015).