Hydnum pallidum Raddi, the Correct Name for H. albidum Peck in the Sense of European Authors and the Recently Described H. reginae Kibby, Liimat. & Niskanen

The systematics of the genus Hydnum have undergone important advances, and many new species have been described with the aid of molecular data. A revision of old names that refer to Hydnum s. str., considering the knowledge now available, might reveal prioritary names of recently described species. This study focuses on the study of names that refer to white Hydnum in Europe, among which earlier synonyms of Hydnum reginae (=Hydnum albidum s. auct. pl. eur.) are potentially found, a species characterized by producing white basidiomata and smaller spores than any other European species. Our revision revealed the existence of three earlier names based on European material, namely H. pallidum Raddi, H. album Fr. and H. heimii Maas Geest. The earliest of those, Hydnum pallidum, is epitypified using material from Tuscany (Italy), from where it was originally described, and hence, it becomes the correct name for H. albidum s. auct. pl. eur. A full description and photographs of H. pallidum are provided, and further comments on other names that refer to white Hydnum based on European material are made.


Introduction
Hydnum comprises ectomycorrhizal fungi that produce pileate and stipitate basidiomata with a spiny hymenophore, along with suburniform basidia of stichic type [1].Its species are edible and popularly known as "hedgehogs" or "wood urchins".For decades, only a handful of species were recognized in the world (in Europe, just H. repandum L., H. rufescens Pers.and H. albidum Peck), but the systematics of Hydnum have undergone important advances since molecular data, especially from the ITS region, became available [2][3][4], and currently 49 species are recognized in the world [5].
Although several new species of Hydnum were described as new recently [2][3][4], it remained evident for several years that some clades lacked a binomial name, and new names had to be coined for those.In this context, the contribution by Niskanen et al. [5] represented an important advance by naming the up-to-then known diversity of Hydnum based on the ITS region-twenty-two new species were described, along with epitypifying Hydnum repandum and H. rufescens.Since this study based the description of several species on a single or very few specimens, morphological characters could not always be revised in depth, and the morphological variations in those taxa are often poorly understood.In this context, species delimitation is far from being straightforward in Hydnum and largely relies on molecular data from the ITS region.Morphological characters that support species discrimination are often not clear-cut, namely basidioma robustness, pileus colour, the attachment of the spines disposition, along with the size and shape of basidiospores [5].Old names that refer to Hydnum have been largely disregarded in recent studies, probably due to the fact that type specimens are lost or the difficulties in recognizing species morphologically.Maas Geesteranus [6] listed and studied in depth many names that need to be re-interpreted in this new context, as potentially prioritary older names could be available for newly described taxa.
This study focuses on the revision of old names that are applicable to what European mycologists have called H. albidum.Peck [7] described H. albidum from North America and stated that it differed from H. repandum in its smaller size, white colour and smaller spores.The first report of H. albidum in Europe was made by Maas Geesteranus [8] (p.495), based also on the material of "a conspicuous whiteness, the spines very much crowded and more slender than in H. repandum, while its spores are smaller and narrower".Subsequent authors followed Maas Geesteranus and applied the name H. albidum to European material with a white colour and small spores.Currently, reports attributed to H. albidum have been made from temperate and Mediterranean Europe, almost always in coniferous or broadleaf forests with calcareous ground: Austria [9], Denmark [10], France [11], Germany [12], Italy [13], Spain [14], Sweden [15], Switzerland [8,16], and the United Kingdom [17] (as Hydnum reginae).All the European material assigned to H. albidum appears to be homogeneous, and the species is considered easily recognized [18] and even striking in the field ( [17] as H. reginae).Previous authors, Olariaga [18] (p.326), adopted the name H. albidum but noted that the basidiospores of the Iberian material were ellipsoid-as opposed to the subglobose spores mentioned by Peck-and that several earlier putative synonyms described in Europe required further study.Swenie et al. [19] epitypified H. albidum after studying extensive material collected in the region of the type locality and providing sequences that considerably differ from the ones generated from European material identified as H. albidum.Once it became evident that H. albidum was being misapplied in Europe, Kibby and Liimatainen [17,20] erected a new species name, H. reginae Kibby, Liimat.& Niskanen, to accommodate collections previously referred to H. albidum in Europe.Being the only species of Hydnum sect.Alba Niskanen & Liimat. in Europe [5], the only other species that can be confused with H. reginae (=H.albidum s. auct.pl.eur.) is H. boreorepandum Niskanen, Liimat.& Niemelä, which differs in having larger spores, always-white spines and being only known from boreal forests in Finland [5] and Sweden (specimen S F-156605 recorded by Ibai Olariaga).Hydnum vesterholtii Olariaga, Grebenc, Salcedo & M.P. Martín and H. ibericum Olariaga, Liimat.& Niskanen also have a pale-coloured pileus but differ macroscopically in having a pale ochre colour, even in young basidiomata [3,5].Thus, H. albidum s. auct.pl.eur. is a species that can be readily distinguished in the field in Europe, even more straightforwardly in the Mediterranean region.
A comprehensive nomenclatural revision, considering all currently available knowledge of Hydnum, revealed the existence of three names, Hydnum pallidum Raddi [21], H. album Fr. [22] and H. heimii Maas Geest.[6], that are synonyms and have priority over H. reginae.Thus, the goals of this study are to propose stable interpretations of those names and to contribute to nomenclatural stability before a name without priority, H. reginae, is widely adopted by the mycological community.

Morphological Studies
Specimens were studied from AGMT, AH, ARAN, BIO, C, GDAC, JA, MA, PI and S herbaria [23] and from the private herbaria of Marco Cartabia and Luis Ballester (LB).Thirty-five specimens, all of them previously identified as H. albidum Peck, were examined.Colour codes are based on the charts of The Royal Horticultural Society [24] and those of dried material on Munsell Color Corp. (Baltimore, MD, USA) [25].Basidiospores were measured from dried material by rehydrating 1-3 spines in KOH 5%.Only released mature spores were measured.Spore statistics are based on measurements of 25 spores from each collection: Lm = mean length, Wm = mean width and Qm = Lm/Wm.The number of spores measured is provided as n = total number/number of specimens (e.g., n = 25/1).Extreme values are given in parentheses.
2.2.DNA Extraction, PCR Amplification, Sequencing and Alignment DNA was extracted from dried material using either the DNeasy Plant Mini Kit (QIAGEN, Hilden, Germany) or the PTB DNA extraction protocol following [26].The primer combinations used were ITS5-ITS4 [27] and LR0R-LR5 for the ITS and LSU [28], respectively.The same primers were employed for sequencing.PCRs were performed using a Master Mix (QIAGEN Multiplex PCR Kit) in a 20 µL volume and then conducted in an Applied Biosystems GeneAmp ® PCR System 9700 and 2720 Thermal Cyclers.The amplification programme used was initial denaturation at 95 • C for 5 min, followed by 35-45 cycles of 95 • C for 45-60 s, 50-55 • C for 50 s, 72 • C for 1 min, followed by a final extension at 72 • C for 10 min.PCR products were visualized in a 1% agarose gel, stained with SYBR Safe DNA Gel Stain (Invitrogen-Thermo Fisher Scientific, Inc., Waltham, MA, USA), and through a UV trans-illuminator.PCR products were sequenced at the Macrogen Spain service.
Sequences were edited and assembled using Sequencher v. 4.1.4(Gene Codes Corp., Ann Arbor, MI, USA) and deposited in GenBank, and additional sequences were downloaded from the GenBank and UNITE databases (Table 1).The dataset was automatically aligned in Aliview v. 1.26 [29] and manually adjusted.The ITS-LSU data matrix was subjected to Maximum Likelihood (ML) and Bayesian analyses.The partitions of the alignment were inferred in ITSx [30] and subsequently analyzed in PartitionFinder v. 2.1.1 [31] using the greedy algorithm [32] in order to assess the best-fit evolution model for each region (ITS1, 5.8S, ITS2 and LSU).

Phylogenetic Analysis
The ML analysis was conducted in IQ-TREE [33], starting from a random tree and setting four partitions (ITS1, 5.8S, ITS2, LSU) under default options.To assess the branch confidence, 1,000 ML bootstrap replicates were conducted using standard bootstrapping.The Bayesian analysis was carried out in MrBayes v. 3.2.7a[34], in the CIPRES Science Gateway [35], using two parallel runs of 4 Metropolis-coupled Markov chain Monte Carlo (MCMCMC) chains for 30 M generations, starting from a random tree, and sampling one tree every 1000 generations from the posterior distribution.Substitution models were sampled across the GTR space during the MCMC simulation [34].Stationarity was assumed when the average standard deviation of split frequencies fell below 0.01.The burn-in fraction was set to discard 0.25 trees from each analysis.To assess the branch confidence, a 50% majority rule consensus tree was computed with the remaining trees using the SUMT command of MrBayes.An ML standard bootstrap (ML-Boot) ≥95 and Bayesian posterior probability (PP) values ≥ 0.95 were considered supported.
The ML analysis of the combined dataset resulted in a single best ML tree of -lnL = 6412.498.The Bayesian analyses reached an average standard deviation of split frequencies of 0.001699 after 30 M generations.A majority rule consensus tree was constructed from the 45,002 trees sampled from the two runs, each consisting of 22,501 trees sampled from the stationary tree distribution (the first 25% discarded as the burn-in) (Figure 1).The sequences of H. pallidum, including the two specimens from Tuscany, form a strongly supported clade (ML-Boot 100, BPP 1).Sistotrema muscicola (AJ606040, AJ606040) was set as an outgroup for the dataset, previously recovered as a sister clade to Hydnum within Cantharellales [36].
The ML analysis of the combined dataset resulted in a single best ML tree of -lnL = 6412.498.The Bayesian analyses reached an average standard deviation of split frequencies of 0.001699 after 30 M generations.A majority rule consensus tree was constructed from the 45 002 trees sampled from the two runs, each consisting of 22 501 trees sampled from the stationary tree distribution (the first 25% discarded as the burn-in) (Figure 1).The sequences of H. pallidum, including the two specimens from Tuscany, form a strongly supported clade (ML-Boot 100, BPP 1).Sistotrema muscicola (AJ606040, AJ606040) was set as an outgroup for the dataset, previously recovered as a sister clade to Hydnum within Cantharellales [36].The sequences of H. pallidum (many of them identified as H. albidum or H. reginae; see Table 1) show a low sequence divergence and are identical except for the lack of a cytosine at position 142 of the alignment in LJU-GIS 1341 and some gaps at the start of the sequences TUF106235 and TUF101782, which might be due to sequencing errors.

Discussion
The predominant white colour and the small spores support Hydnum albidum s. auct.pl.eur. to be accommodated in Hydnum subg.Alba Niskanen & Liimat.[5].Within this group, H. subcremeoalbum Tedersoo, Liimat.& Niskanen, H. zongolicense Garibay-Orijel and H. treui Tedersoo, Liimat.& Niskanen morphologically differ by their more rounded, globose-to-subglobose spores (Qm < 1.1) [5].The rest of the species assigned to Hydnum subg.Alba, H. albidum Peck, H. alboaurantiacum Swenie & Matheny and H. creomeoalbum Liimat.& Niskanen have similar spores compared to H. albidum s. auct.pl.eur., but none of them occur in Europe, and the available molecular data suggest that they are not conspecific with it.Interestingly, according to the extensive material studied, H. albidum s. auct.pl.eur.possesses a characteristic never mentioned for any other Hydnum species, i.e., the presence of rounded, < 1 µm crystals on the hyphae of spine apices [18] (Figure 142c).Nevertheless, more detailed studies on other species of Hydnum subg.Alba are needed to determine whether this characteristic is restricted to H. albidum s. auct.pl.eur.or has been overlooked in other species.All the European material of Hydnum albidum s. auct.pl. is morphologically homogeneous, and available ITS sequences show little divergence, as described above.It has been recorded in association with coniferous or broadleaf ectomycorrhizal hosts [18], and possibly with Helianthemum [17,20], showing a strong preference for calcareous ground.Although red-listed in some countries, such as Norway [41] and Sweden [42], H. albidum s. auct.eur.pl. is a common and widespread species in Mediterranean Europe, especially under evergreen oaks (Quercus ilex, Q. rotundifolia) on calcareous soils [18].

Study of the Name Hydnum Pallidum Raddi
Hydnum pallidum Raddi was validly published as a description was provided [21] (p.353).Hydnum pallidum was described from Boboli Gardens (Florenze, Tuscany) as an entirely white-pale Hydnum with numerous spines.The drawing provided by Raddi (Table 13, Figure 8; erroneously referred to as "Tav.: V. Figure 8") also shows an entirely white, small Hydnum that conforms beyond doubt to H. albidum s. auct.pl.eur.The only other European species that produces nearly completely white basidiomata, H. boreorepandum, is a boreal species [17], and hence its occurrence in the Florence area, characterized by a Mediterranean climate, (Csa type according to Köppen-Geiger's classification [43], is difficult to conceive.Unfortunately, no original specimen of H. pallidum exists at FI and PI, where most of Raddi s material is kept. A more thorough search for information on H. albidum s. auct.pl.eur.shows that it is a widespread edible mushroom in Italy, even included in regional lists of commercialized species [44] and popularly called "Steccherino bianco" [45].A more specific search in Tuscany revealed several bibliographic records [46][47][48] and treatments in popular websites of mushrooms [49,50].Furthermore, we examined seven herbarium specimens (see material examined) from different sites in Tuscany.The morphological examination of those specimens and the sequence data obtained have confirmed that they belong to H. albidum s. auct.pl.All the information gathered shows that H. albidum s. auct.pl. is a widespread and rather common species in Tuscany and fully supports that the species described by Raddi as H. pallidum is conspecific with H. albidum s. auct.eur.In order to be able to check molecular characters and to support our interpretation, we propose here an epitype collected in Tuscany for H. pallidum.

Study of the Name Hydnum Album Fr.
The earlier publication of Hydnum album by Raddi [21] (p.361), not validly published because it is lacking a description (see under excluded names further down), does not affect the legitimacy of H. album Fr. [22].Although Fries included it among the "Hydnorum stipitatorum species mihi dubia" (doubtful species), H. album Fr. was validly published under Art.36.1 [51].Under H. album Fr., Fries [22] included the illustration provided by Micheli [37] for his "Erinaceus, esculentus, albus, crassus", which is represented in Tab 72, Figure 2 (erroneously referred to as "Table 71. Figure 2" by Micheli).The latter author described a white edible mushroom called "Steccherino" or "Dentino bianco", which occurred in several habitats near Florence in October and was sold in the markets at that season.Micheli's illustration shows a hydnoid stipitate basidioma that, considering the white colour mentioned by Micheli, conforms to H. albidum s. auct.eur.In conclusion, there is little doubt that the fungus described by Micheli [37]  Some confusion has existed as to the attribution of the name H. album to Fries or Persoon.As Maas Geesteranus [6] (p. 140) noted, Persoon did not refer to Fries [52] (p.249) or Persoon [53], even though Persoon himself cited the same Micheli plate [37] (Table 72, Figure 2) as Fries under his treatment of H. album.Probably due to the fact that Persoon provided a more detailed opinion on the interpretation of H. album, Maas Geesteranus ascribed H. album to Persoon.Although Persoon did not link his H. album with Fries, both authors included a single element eligible as a type under their treatments, i.e., the Micheli plate cited above.Hydnum album Pers.cannot be considered a later homonym of H. album Fr., as both names would be based on the same type, and Art.53 cannot be applied [51].Thus, we consider Persoon's H. album as a later treatment of H. album Fr. that does not merit recognition as a different name.

Study of the Name Hydnum Heimii Maas Geest
Hydnum heimii was proposed as a new name for Sarcodon abietinus R. Heim, invalidly published by Heim [38] as no Latin diagnosis was provided (Art.39.1 [51]).The diagnostic features noted by Heim [38,39] were the overall pale colour (cream white) and the yellowisholivaceous pileus centre.Remarkably, Heim [38,39] described the spores as 6-7 × 4.8-5 µm.Maas Geesteranus [6] failed to obtain any of the specimens cited by Heim [38] on loan, but was able to check a specimen collected in the type locality in 1953.During the visit by I. Olariaga to the PC herbarium, the type of H. heimii could not be located either, but he was able to examine the specimen collected in 1953, regrettably in a rather poor condition.The specimen contains around 10 dry basidiomata of a dark reddish brown colour.The spore (5.5-7 × 3.5-4.5(5)µm) measurements are very similar to those provided by Heim, and we observed basidia to be 5-6 µm broad.Both the spore and basidia sizes fall within the range of H. albidum s. auct.pl.eur.Considering the cream-white basidioma colour described for H. heimii [38,39] and the microscopic characters mentioned, we conclude that H. heimii is conspecific with what European authors have called H. albidum.The remaining known species in Europe possess larger spores (e.g., 7.1-9.2× 5.9-8.7 µm in H. repandum, Olariaga [3]) and broader basidia (e.g., 8-11 µm in H. repandum [3]) in our experience.The fact that Maas Geesteranus [40] treated H. heimii as a synonym of H. repandum can be explained by the fact that he included within H. repandum material from several continents that is known to belong to several species today.He therefore treated H. repandum as an extremely variable species and saw a continuum in many characters such as in the basidioma size, colour and spore size in his treatment in 1975.Raddi [21] (p.361) listed this name as "Hydnum album Nob.", but did not provide any description for it.Therefore, it is regarded here as a nomen nudum, as Maas Geesteranus [54]  Persoon [55] described H. medium as having a glabrous and pallid pileus, acute thin spines and a short white stipe.He further argued that aculei and the pileus surface were close to H. repandum, whereas the colour of the spines conformed to H. rufescens.The normal size and the pallid pileus-not described as white-suggest this name is most likely to be a synonym of H. repandum, as considered by Maas Geesteranus [6]  The spore measurements provided by Cejp [56] were 7-7.5 × 6-8 µm and therefore do not conform to those of H. pallidum.

Conclusions
A comprehensive nomenclatural revision revealed the existence of three earlier names for the species that has been traditionally called H. albidum in Europe and for which the name H. reginae was coined recently.The adoption of the name H. pallidum for H. albidum s. auct.pl.eur.will bring nomenclatural stability as it is unlikely that any earlier synonym of it exists.We hope this study encourages the in-depth study of other Hydnum names described before the advent of the molecular era, among which further earlier synonyms of the later described species might be found.

Figure 1 .
Figure 1.Best tree of the Maximum Likelihood analysis of selected sequences of Hydnum from ITS/LSU sequence data.Maximum Likelihood standard bootstrap values (ML-Boot)/Bayesian posterior probabilities (PP) are shown on branches, ordered as ML-UFBoot/PP.Thickened branches received support in both analyses (ML-Boot ≥ 95% and/or PP ≥ 0.95).Values are provided for nodes supported at least in one analysis.The epitype of Hydnum pallidum is marked by an asterisk.The H. pallidum clade is highlighted in a blue colour, whereas the American H. albidum is in green.Subgenera are indicated by vertical bars.

Table 1 .
Sequenced specimens used for the phylogenetic analysis of this study, with GenBank or UNITE (UDB) accession numbers for ITS and LSU regions.Newly generated sequences in this study are in bold, and sequences from type specimens are indicated.Original identifications of the sequences are provided between parentheses.
(p. 132) is what has been called Hydnum albidum by European authors.