Introduction

Hebeloma section Theobromina was proposed by Vesterholt (2005) and includes H. theobrominum (the section type), H. erumpens and potentially another taxon that cannot be clearly differentiated from H. erumpens (Eberhardt et al. 2009). In this paper we describe a new species, H. vesterholtii spec. nov., that fits both morphologically in the sect. Theobromina and molecularly in the clade /theobromina. We also provide molecular evidence using ITS sequence data that H. vesterholtii is a taxon distinct from the other members of its section. In addition, we include a morphological key to separate this species from the two other species of sect. Theobromina mentioned above and H. plesiocistum which shares a number of characters with the members of sect. Theobromina.

Materials and methods

ITS sequence data were obtained from all known collections of Hebeloma vesterholtii and all collections of H. theobrominum examined in the context of this study. Sequences were produced by direct sequencing following the protocols detailed in Eberhardt et al. (2009) and references therein. In some of the tracefiles evidence was found for length variation within the template PCR products. These were taken into account by including both length variants as separate sequences into GenBank and into the phylogenetic analyses. Collection details and GenBank accession numbers of new accessions are listed in Table 1. Additional sequences considered in the molecular analyses were EU570163—EU570189, EU693013, EU881922, and EU887517 (GenBank acc. numbers) for which collection details were published by Eberhardt et al. (2009). To avoid confusion stemming from the application of different species concepts we did not include sequences in the analysis for which the corresponding collections have not been morphologically examined within this project. Sequence alignment was carried out using Se-al (vs. 2.0a11; Rambaut 2002). Maximum likelihood analyses were carried out as a heuristic search (100 replicates) in RAxML (vs. 7.0.3; Stamatakis 2006) under the GTRMIX model including bootstrap analyses (500 replicates). FigTree (vs. 1.1.2; Rambaut 2008) was used to visualize the results, relying on njplot (vs. 2.2; Perrière and Gouy 1996) for the correct placement of the bootstrap values.

Table 1 Collection details and GenBank accession no. of new Hebeloma spp. ITS sequences included in the phylogenetic analyses

Details of the morphological analyses are given in Eberhardt et al. (2009). For each collection at least 50 spores were measured in Melzer’s reagent, excluding the apiculus. The maximum length and width of each spore was measured and its Q value (ratio of length to width) calculated. Average length, width and Q value were calculated and recorded alongside the median, standard deviation and 5% and 95% percentiles. The assessment and coding of spore characters follows Vesterholt (2005). The average width of the cheilocystidium apex appears to be an important character in the separation of species within Hebeloma (Vesterholt 2005). It is also important, when determining the average apex width, not to be selective with regard to the cystidia chosen for measurement. To determine the average width at the apex about 100 cheilocystidia were measured on the lamellae edge. For other measurements, at least 20 cheilocystidia, separated from the lamella edge, were measured from each collection. Because of the complex shapes of the cheilocystidia four measurements were made: length, width at apex, width at narrowest point in central region and maximum width in lower half. The measurements are given in this order and an average value was calculated for each of these measurements. The measurements were made in 5% KOH. Unless mentioned otherwise, cited specimens are presently located in the personal collection of H.J. Beker (HJB).

Results and discussion

The sequences of the currently known collections belonging to the species described below as Hebeloma vesterholtii (i.e. all collections on which the description is based) form a well-supported clade in the ITS analyses (Fig. 1). Judging from direct sequence data, evidence of intragenomic variation (ambiguous readings in both strands; overlapping peaks as the result of length variation among amplicons) was found in ten of thirteen specimens, implicating seventeen positions in the alignment of all H. vesterholtii sequences. Only one additional position is variable in between-specimen comparisons, but unambiguous within all sequenced specimens. However, counting only clear mismatches and disregarding the length variation, the intraspecific variation is restricted to two sites, which explains the very short terminal branches in the clade. The observation and extent of the intragenomic ITS variation is not uncommon in Hebeloma (Aanen et al. 2001; Matheny et al. 2007; Eberhardt et al. 2009; unpublished data). Likewise, the clade of the closely related H. theobrominum is well-supported. In contrast, H. erumpens, forms a mixed clade with sequences of the insufficiently known Hebeloma sp. For a more detailed discussion see Eberhardt et al. (2009). These two species and one species complex are European representatives of the /theobromina clade, corresponding to Hebeloma sect. Theobromina. From a morphological point of view H. plesiocistum should also be included in this section. In Fig. 1 it is associated with the /theobromina clade, but in a more inclusive analysis including ITS sequences from all purported Hebeloma species occurring in Europe (not published) it appears in a separate clade. A study including genes other than the ITS is underway.

Fig. 1
figure 1

Best of 100 ML tree of from a heuristic search with 100 replicates based on ITS data; bootstrap support was calculated with 500 replicates. For specimens showing length variation in the ITS, both variants were included in the tree and indicated by the same shade font. Single sequences are in black

Taxonomy

Hebeloma vesterholtii Beker & U. Eberh., sp. nov.—Mycobank MB 515246; Figs. 2, 3, 4 and 5

Figures 2, 3, 4 and 5

Fig. 2
figure 2

Collection BR-MYCO 166528-76 (holotype) of Hebeloma vesterholtii growing on calcareous ground under deciduous trees. Photograph H.J. Beker

Fig. 3
figure 3

Collection BR-MYCO 166528-76 (holotype) of Hebeloma vesterholtii: a. spores ×1600 in Melzer’s Reagent; b. basidium ×500 in 5% KOH; c. cheilocystidia ×500 in 5% KOH; d. cheilocystidia ×1000 in 5% KOH; e. ixocutis ×500 in 5% KOH; f. cheilocystidia on lamella edge ×500 in 5% KOH. Scale bars = 10 µm. Photographs H.J. Beker

Fig. 4
figure 4

Spores and cheilocystidia of collection BR-MYCO 166528-76 (holotype) of Hebeloma vesterholtii, Scale bar = 5 µm

Fig. 5
figure 5

Distribution map for collections of Hebeloma vesterholtii

Etymology—in honour of Jan Vesterholt who has made a significant contribution to our understanding of the genus Hebeloma.

Pileus usque ad 55 mm in diam., convexus vel obtuse umbonatus; superficies sicca vel leviter viscida, interdum hygrophana, haud striata, ochracea vel bruno ochracea, praeter marginem pallidum cum coloribus cremeis vel roseis. Lamellae adnatae vel emarginatae, parce densae, tenues; color saturate cremeus, alutaceus vel brunneo-alutaceus pallidus vel brunneus, demum brunneo-olivaceus. Acies fimbriata. Stipes centralis, interdum attenuatus vel clavatus basin versus, usque ad 75 mm longus et 10 mm crassus, candidus vel alutaceus; superficies sicca, pruinosa vel floccosa. Cortina non visa. Caro cremea vel pallide brunnea; odor compositus (cacao et raphanoideus). Sporae amygdaloideae, fere laeves vel inconspicue ornamentatae, valde dextrinoideae; 7.8–9.5 × 4.7–5.6 µm. Basidia tetrasporica, 21–33 × 5.6–8.4 µm, cylindrica vel clavata. Pleurocystidia non visa. Cheilocystidia subclavata, in parte inferiori incrassata, apex 4.4–6.3 mm latus. Caulocystidia cheilocystidiiformia. Fibulae omnipraesentes.

Basidiomes usually in small scattered groups, but occasionally solitary and rarely caespitose. Pileus up to 55 mm in diam. (rarely up to 85 mm), convex to broadly umbonate; surface dry or slightly viscid, can be hygrophanous but is not striate; cuticle colour fairly uniform across most of the pileus except for a thin paler margin, ochraceous to yellowish brown, occasionally with fawn or cinnamon tones, sometimes even slightly olive at centre while the thin paler margin has cream to pinkish buff tones; pileus margin straight, slightly involute in immature basidiomes, wavy (undulate) when old. Lamellae adnate to emarginate, 4–6 mm deep at the widest point, moderately dense, thin, with 60–90 full length lamellae and usually between 1 and 3 lamellules between lamellae, droplets on the lamellae rarely recorded; colour cream, alutaceous or brown when young, later brownish olive following spore maturity; edge fimbriate, paler than lamella surface. Stipe central, cylindrical but occasionally either attenuate or clavate towards the base, up to 75 mm long and with central width 4–10 mm (rarely up to 20 mm); white or alutaceous, sometimes discolouring from the base upwards, becoming hollow with age; surface dry, pruinose to floccose. Cortina not observed. Flesh cream or pale brown, almost never discolouring when bruised. Smell cacao with a raphanoid component but sometimes no smell recorded, taste mild. Spore print brownish olive. Exsiccata often dark or even blackening particularly on the pileus.

Spores amygaloid, with small apiculus and rounded at the end opposite the apiculus, with a distinct thinning of the spore wall and never any sign of papilla, guttulate with one or occasionally more oily drops, almost smooth to weakly ornamented, with no sign of loosening perispore and strongly dextrinoid (O1/2; P0; D3/4); spore colour under the microscope from pale yellow to pale brown; spore size based on n = 87 spores of the holotype, 5% to 95% percentile range 7.8–9.5 × 4.7–5.6 µm, with median 8.6 × 5.2 µm and av. 8.7 × 5.1 µm with S. D. length 0.54 µm and width 0.29 µm, Q value 5% to 95% percentile range 1.54–1.82, with median 1.70 and av. 1.69 with S. D. 0.09; spore size based on thirteen collections medians 8.2–9.8 × 4.7–5.4 µm and av. 8.2–9.8 × 4.7–5.5 µm with S. D. length 0.45–0.71 µm and width 0.18–0.31 µm, av. Q 1.64–1.79. Basidia cylindrical to clavate and 4–spored, 21–33 × 5.6–8.4 µm, with av. 23–29 × 6.2–8.0 µm. Pleurocystidia not found. Cheilocystidia mostly hourglass–shaped, i.e. clavate, occasionally cylindrical in the upper part, the majority narrower in their central region but then swollen again in the lower half, very rarely cylindrical in their lower half, occasionally with unclamped septa or some median thickening or with apex mucronate and often bent in the centre; width of apex holotype 5% to 95% percentile range 4.4–6.3 µm, with median 5.3 µm and av. 5.3 µm with S.D. 0.64 µm; across thirteen collections median 4.3–7.1 µm and av. 4.4–7.1 µm; n = 20–30 selected cheilocystidia of each of thirteen collections yields a range for the averages of 27–37 × 4.4–7.1 × 3.5–4.9 × 5.4–6.6 µm and 30 × 5.3 × 3.6 × 5.6 µm av. for holotype. Caulocystidia resemble cheilocystidia but tend to be larger. Pileipellis is an ixocutis with an epicutis in the range 120–180 µm, embedded hyphae up to 5 µm broad, often encrusted, hyaline or occasionally pigmented. Cutis orange brown and made up of cylindrical to isodiametric elements. Clamp connections present throughout the fruitbody.

Habitat and distribution—all thirteen collections, on which the above description is based, were made in deciduous woodland and with a single exception all habitats were reported as calcareous. There is a range of deciduous trees reported but Quercus appears to have been present on every occasion. This species appears widespread in Europe with a more southerly distribution and is already known from Denmark, England, Belgium, Italy, Spain and Bulgaria. The most northerly collection is from Denmark at N55.86 degrees.

Collections examined

Belgium, Wallonia, Han-sur-Lesse, Turmont, (N50.1142167;E05.1715000; alt. 153 m) on calcareous soil in a deciduous woodland including Carpinus betulus, Corylus avellana, Fagus sylvatica, Quercus robur, 19 Sep. 2004, H. Beker, holotype BR-MYCO 166528-76, BR, double HJB1000133; loc. cit., 19 Sep. 2004, H. Beker, HJB10117; loc. cit., alt. 150–165 m, H. Beker, 1 Sep. 2005, HJB11195; loc. cit., H. Beker, 1 Sep. 2005, HJB11199; loc. cit., H. Beker, 1 Sep. 2005, HJB11206. Bulgaria, Sarantsi (N42.71;W23.77; alt. 650 m), woodland, Quercus cerris; Quercus frainetto; Quercus sp., 7 Oct. 2008, D. Bojantchev, DB8207/HJB12344. Denmark, Eastern-Jylland: Elbæk Skov, East of Horsens, North of Horsens Fjord (N55.8615800;W9.9374770; alt. 4 m), on calcareous clayey soil in dry woodland, Fagus sylvatica, Quercus robur, 5 Sep. 2008, J. Vesterholt, JV08-132/HJB12321; loc. cit., J. Vesterholt, JV08-133/HJB12331. Italy, Emilia-Romagna, nr Carpegna, (N43.62055;E12.13133; alt. 578) on calcareous soil in a deciduous woodland under Castanea sativa and Quercus spp., 15 Oct. 2004, H. Beker, HJB10339; Sardinia, Mount Limbara, (N40.857;E9.168; alt. 1200 m) on calcareous soil under deciduous trees including Quercus sp., 10 Nov. 2006, C. Hobart, HJB11869. Spain, Pais Vasco, Urdiain, (N42.9059833;W2.1503333; alt. 489 m) on calcareous soil under Quercus rubra, 22 Oct. 2005, H. Beker, HJB11377. United Kingdom, East Sussex, St. Dunstan’s Farm (N50.95421;E0.28947; alt. 88 m) in a deciduous woodland under Quercus sp., 10 Oct. 2004, A. Andrews, HJB10674; Calke Abbey (N52.59975;W1.99604; alt. 148 m), Fagus sylvatica, Quercus sp., 18 Sep. 2008. C. Hobart, HJB12696.

Hebeloma theobrominum Quadr., 1987, Mycotaxon 30: 311

Collections examined

Belgium, Wallonia, Han-sur-Lesse, Turmont, (N50.11;E05.16; alt. 150 m) on calcareous soil in a deciduous woodland including Carpinus betulus, Corylus avellana, Fagus sylvatica, Quercus robur, 5 Sep. 2004, D. Ghyselinck, HJB10063. Denmark, Høstemark Skov, south of Mou (N56.93;E10.21; alt 6 m) on mossy ground under Picea abies, 15 Oct. 2002, J. Vesterholt, JV02-782/HJB10907; Hamborg Skov, Lolland, Fuglsang Storskov (N54.78;E11.31; alt. 7 m) under Fagus sylvatica, 6 Oct. 2000, J. Vesterholt, JV00-435/HJB10909. Estonia, Tagamoisa (N58.4611;E22.0089; alt. 25 m), on calcareous soil in a mixed woodland under Tilia sp., 9 Aug. 2004, H. Beker, HJB10009. Italy, Cala Violina, north west of Grosseto (N42.83;E10.77; alt. 20 m) in a deciduous woodland under Quercus cerris, Q. ilex, Q. suber, 9 Nov. 2006, J. Vesterholt, JV06-1095/HJB11848. Norway, Leikanger, Sanden by Klyvi, (N59.77;E9.92; alt. 7 m) in a deciduous woodland, 10 Sep. 2000, J. Heilmann-Clausen, JV00-314/HJB10908. Spain, Catalunya, Brunyola (N41.9001;E2.6595; alt. 187 m) in a mixed woodland under Quercus sp., 23 Oct 2005, H. Beker, HJB11382. United Kingdom, Westmorland, Roudsea (N54.2356;W3.0294; alt. 15 m) on calcareous soil in a deciduous woodland, 24 Sep. 2002, H. Beker, HJB9293; Derbyshire, Coombsdale (N53.2706;W1.6446; alt. 140 m) on calcareous soil with Helianthemum sp., 28 Oct. 2004, H. Beker, HJB10467; East Sussex, St. Dunstan’s (N50.95;E0.29; alt. 88 m) in a deciduous woodland, 10 Oct. 2004, A. Andrews, HJB10679.

Comment

Hebeloma vesterholtii appears to favour deciduous woodland in calcareous areas. It is likely that it forms ectomycorrhizal relationships with species of Quercus. The combination of almost smooth to very weakly ornamented and strongly dextrinoid spores indicates an affinity to either sect. Velutipes or sect. Theobromina (Vesterholt 2005). The known species in the sect. Velutipes tend to have more ornamented spores and/or differently shaped cystidia. Their cheilocystidia are generally much more gently clavate than those of H. vesterholtii. The very pale small spores and the shape of the cystidia strongly support that H. vesterholtii belongs within sect. Theobromina. The new taxon can be readily distinguished both morphologically and molecularly (see above) from the other members of this section. It is likely that previous collections may have been confused with H. theobrominum. However, H. vesterholtii is distinguishable from the latter species by means of its differently shaped spores and different cap colour (Figs. 2, 6, 7). Based on 10 collections of H. theobrominum we find the cap colour to range from deep cinnamon to red–brown to dark brick as opposed to the lighter and often more yellowish to buff colours of H. vesterholtii. Based on the same 10 collections (and measuring a minimum of 50 spores in each case) the average spore size ranges from 8.4–9.9 × 4.5–5.3 µm and av. Q 1.80–1.95. Based on our collections there is no overlap in the average spore Q, though there is only a small gap between the two ranges. Thus a collection of either species with a misleading average spore Q cannot be ruled out. However, in cases where the average spore Q may give an ambiguous identification, the cap colour should help distinguish between the two species. Figure 1 shows the placement of H. vesterholtii in the /theobromina clade based on ITS data. Table 2 summarizes the main characters of the four species currently included in the section Theobromina.

Fig. 6
figure 6

Collection JV02-782 of Hebeloma theobrominum from Hostemark, Denmark growing in mossy ground under Picea

Fig. 7
figure 7

Comparison of spores of Hebeloma vesterholtii and Hebeloma theobrominum ×1600 in 5% KOH: a. Collection HJB10339 of Hebeloma vesterholtii; b. Collection JV06-1095 of Hebeloma theobrominum. Scale bars  = 10 µm. Photographs H.J. Beker

Table 2 Comparison of critical morphological and ecological characters of species of Hebelona, sect. Theobromina

Key to Hebeloma sect. Theobromina and H. plesiocistum:

  1. 1a

    Collected on calcareous ground in deciduous or mixed woodland 2

  2. b.

    Collected under Cistus 3

  3. 2a

    Cap reddish brown and ave. spore Q > 1.80 H. theobrominum

  4. b.

    Cap yellowish brown and ave. spore Q < 1.80 H. vesterholtii

  5. 3a

    Average cheilocystidium apex > 7 µm and spores very strongly dextrinoid, becoming dark brick to dark reddish–brown H. plesiocistum

  6. b.

    Average cheilocystidium apex < 7 µm and spores strongly dextrinoid, becoming yellow brown to orange–brown or brick H. erumpens